Maura Mitrushina, Kyle B. Boone, Jill Razani, Louis F. D'Elia - Handbook of Normative Data For Neuropsychological Assessment (2005, Oxford University Press, USA)

When Handbook of Normative Data for
Neuropsychological Assessment was originally

published in 1999. it was the first book to pro-
vide neuropsychologists with summaries and
critiques of normative data for neuropsycho-
logical tests. The second edition, which has
been revised and updated throughout, 1.1resents
data for ~6 commonly used neuropsychological
tests, including: Trailmaking. Color Trails,
Stroop Color Word Interference, Auditory
Consonant Trigrams. Paced Auditory Serial
Addition, Ruff ~ & 7. Digit Vigilance. Boston
Naming, Verbal Fluency. Rey-Osterrieth
Complex Figure, Hooper Visual Organization.
Visual Form Discrimination, Judgment of Line
Orientation. Ruff Figural Fluency. Design
Fluency. Tactual Performance, Wechsler
Memory Scale-Revised, Rey Auditory-Verbal
Learning. Hopkins Verbal Learning.
WHO/UClA Auditory Verbal Learning, Benton
Visual Retention, Finger Tapping, Grip
Strength (Dynamometer). Grooved Pegboard.
Category. and Wisconsin Card Sorting tests. In
addition. California Verbal Learning (CVLT and
CVLT- II). CERAD ListLearning, and Selective
Reminding Tests, as well as the newest versions
of the Wechsler Memory Scale (WMS-Ill and
WMS-IIIA). are reviewed.
Locator tables guide the reader to the sets of
normative data that are best suited to each indi-
vidual case. depending on the demographic
characteristics of the patient. and highlight the
advantages associated with using data for com-
parative purposes. Those using the book have
the option of reading the authors' critical
review of the normative data for a particular
test, or simply turning to the appropriate data
locator table for a quick reference to the rele-
vant data tables in the Appendices.
The second edition includes reviews of 15
new tests. The way the data are presented has
been changed to make the book easier to use.
Meta-analysis tables of predicted values for
different ages (and education. where relevant)
are included for nine tests that have a sufficient
number of homogeneous datasets.
No other reference offers such an effective
framework for the critical evaluation of norma-
tive data for neuropsychological tests. Like the
first. the second edition will be welcomed by
practitioners, researchers. teachers, and grad-
uate students as a unique and valuable contri-
bution to the practice of neuropsychology.
Maura Mitrushina, Ph.D., is Professor of
Psychology at California State University.
Northridge, and Associate Clinical Professor of
Psychiatry at UClA School of Medicine. She is
an ABPP/ABCN diplomate and maintains a
clinical and forensic practice in Encino,
California. Her research interests include cog-
nitive correlates of normal aging and differen-
tial diagnosis of dementia, as well as factors
influencing rates of recovery after traumatic
brain injury.
Kyle B. Boone, Ph.D., is Professor-in-

Residence of Psychiatry at UClA School of
Medicine, and Director of Neuropsychological
Services and Training at Harbor- UClA Medical
Center. She is an ABPP/ABCN diplomate and
maintains a clinical and forensic practice in
Torrance, California. She has conducted re-
search on the development and validation of
techniques to identify noncredible cognitive
performance, and on the effects of demograph-
ic factors and medical and psychological ill-
nesses on neuropsychological test performance.
Jill Razani, Ph.D., is an Assistant Professor of

Psychology at California State University,
Northridge, and a licensed clinical psychologist
in the state of California. In the past, she has
conducted research on cognitive aspects of
aging and neurodegenerative disorders.
Presently, she has an active program of
research examining issues related to multicul-
tural and cross-cultural neuropsychology, as
well as the relationship between cognitive
functioning and activities of daily living in
patients with dementia.
Louis F. D'Elia, Ph.D., is Assistant Clinical

Professor of Psychiatry, and former Co-
Director of the Neuropsychology Assessment
Laboratory at the University of California, Los
Angeles, School of Medicine. He remains
active in the training, supervision, and men-
taring of UClA Postdoctoral Neuropsychology
Fellows in his work with them in his private
practice in Pasadena, California.
jACKET DESIGN: E\'E SIEGEL
OXFORD
UNIVERSITY PRESS
www.oup.com
PRAISE FOR THE FIR T EDITIO
''Should neuropsychologists purchase this volume? The answer is an unqualified yes. The book is a very
valuable asset to any neurop~ ·chology collection. This reviewer wholeheartedly recommends it for pur-
chase; the tables alone justify the pnce .... The authors are due a great deal of credit for gathering
together material that most of us would understand as a multi-year project. In examining this book in
even a cur orv way. the prospective buver will see that the effort needed to bring it to fruition is humbling ..
-Kenneth M Adams. PhD. in]oumalofClinical and Experimental Neurops_rcholog.r
"Overall, Mitrushina et al. have made a substantial contribution with their text. and it nicely complements
other thorough overviews of neuropsychology authored by Lezak or Spreen and Strauss. It is concise. time-
ly, comprehensive, and cogent, and it holds great utility for the practice of clinical neuropsychology.... Let
us hope they continue this good work as additional data emerge ...
-Michael R. Basso, PhD, in Neuropsychiatry, Neuropsychology. and Behavioral Neurology
" ... a valuable and well-written addition to the literature that should find its way onto the reference
shelves of practicing neuropsychologists. The book will be a useful educational tool. ... There IS a lot to
be gained from consulting this book. In readability, utility, and practicality. it goes way beyond the
norms."
-Russell M. Bauer. PhD, infoumal of the International Neuropsychological Society
90000
9 780195 169300
ISBN 0-19-516930-1
Handbook of Normative Data
for Neuropsychological Assessment
OXFORD
UNIVERSITY PRESS
Oxford University Press, Inc., publishes works that further

Oxford University's objective of excellence
in research, scholarship, and education.
Oxford New York

Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Copyright© 2005 by Maura Mitrushina, Kyle B. Boone, Jill Razani, and Louis F. D'Elia
Published by Oxford University Press, Inc.
198 Madison Avenue, New York, New York 10016
www.oup.com
Oxford is a registered trademark of Oxford University Press
All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying. recording, or otherwise,
without the prior permission of Oxford University Press.
Library of Congress Cataloging-in-Publication Data

Handbook of normative data for neuropsychological assessment I Maura Mitrushina ... [et al.].- 2nd ed.
p. ; em. Includes bibliographical references and indexes.
ISBN-13 978-0-19-516930-0
ISBN 0-19-516930-1
1. Neuropsychological tests-Handbooks, manuals, etc. 2. Reference values
(Medicine)-Handbooks, manuals, etc.
[DNLM: 1. Neuropsychological Tests. 2. Reference Values. WL 141 H23654 2005]
RC386.6.N48M58 2005
616.8'0475-dc22 2004054724
9 8 7 6 5 4 3 2 1
Printed in the United States of America

on acid-free paper
With admiration and gratitude,
we dedicate this book to
those professionals
whose normative research efforts made
this volume possible.
Preface
The Handbook of Nonnative Data for Neu- Osterrieth Complex Figure, Hooper Visual
ropsychological Assessment is our attempt to Organization, Visual Form Discrimination,
provide ready access to neuropsychological Judgment of Line Orientation, Ruff Figural
normative data and to evaluate their strengths Fluency, Design Fluency, Tactual Perfor-
and weaknesses. Because the interpretation of mance, Wechsler Memory Scale (WMS-R,
test scores profoundly affects the quality and WMS-111, WMS-IIIA), Rey Auditory-Verbal
utility of neuropsychological reports and re- Learning, California Verbal Learning, Hopkins
search, we felt that a critical compendium Verbal Learning, WHO-UCLA Auditory Ver-
containing most of the available normative bal Learning, CERAD List-Learning, Selec-
data for commonly used tests was essential. tive Reminding, Benton Visual Retention,
Before this book's publication, only those Finger Tapping, Grip Strength (Dynamome-
lucky individuals with the time or staff to ter), Grooved Pegboard, Category, and Wis-
conduct exhaustive library searches or with consin Card Sorting tests.
extensive professional subscription lists could
hope to be aware of more than a few norma-
tive reports for any specific test.
ORGANIZATION OF THE BOOK
Although several books cover the intricacies
of administration and scoring procedures for The book contains 25 chapters. The basic
neuropsychological tests and a few contain concepts of normative neuropsychology are
some normative data, no previous volume has addressed in the first three chapters. The first
been exclusively devoted to the presentation chapter provides an introduction to the prac-
and discussion of existing normative data for tice and philosophy of neuropsychology as a
specific neuropsychological tests or provided a clinical discipline. The second chapter ex-
framework for judging studies that report plores the interface of neuropsychology with
normative data. other professional/clinical disciplines and re-
This handbook was written to help guide visits critical issues in neuropsychology. The
the busy clinician, researcher, and graduate third chapter provides an overview of statisti-
student to the utility of commonly used neu- cal methods and the use of statistical and
ropsychological tests and to the normative methodological concepts in neuropsychology,
data accompanied by critical reviews for history and applications of meta-analysis in
comparison purposes for most of the tests clinical practice, and description of proce-
described in this book. The following tests dures for the use of meta-analysis in this book.
have been described: Trailmaking, Color The remaining 22 test chapters review and
Trails, Stroop Color Word Interference, Au- present the normative data for specific neu-
ditory Consonant Trigrams, Paced Auditory ropsychological tests, which are derived from
Serial Addition, Ruff 2&7, Digit Vigilance, articles and other communications reporting
Boston Naming, Verbal Fluency, Rey- results of normative and clinical comparison
viii PREFACE
studies. These chapters begin with a brief must be reported. Depending on the test admin-
ovetview of the history, utility, and psycho- istered, other important variables may include
metric properties of the test under discussion, gender, ethnicity/culture, and hand preference.
which indicates whether there are different
versions of the test and/or varying administra- Procedural variables address such issues as:
tion procedures. If more than one version of a
test exists, the differences in content, adminis- 'What version of the test was adminis-
tration, and scoring are described. We pur- tered?"
posely avoided an exhaustive review of the "How was the test administered?"
history and psychometric properties of the tests "How was the test scored?"
because this information is readily available "Did the data reported include mean and
in other Oxford publications, specifically Lezak standard deviation scores?''
et al. (2004) and Spreen and Strauss (1998).
The next part of the test chapters is a sum- The next section of each of these chapters
mary of the findings from research that has summarizes the status of the normative data
examined the influence of demographic vari- for the test and answers the questions:
ables (e.g., age, education, intellectual level,
gender, ethnicity/culture, handedness) and ad- "How many studies are out there?''
ministration procedures on test performance. 'Which versions of the test have been the
The findings from this review highlight the most frequently administered?''
critical variables needed to evaluate the nor- 'What demographic characteristics have
mative reports for the test. These critical vari- been the most frequently studied?''
ables are broken down into two categories: (1)
subject variables and (2) procedural variables. The next section presents critiques of the
studies, with the strengths and considerations
Subject variables address such issues as: regarding the use of each normative report
discussed in some depth. Data tables are
"How broad are the utilized age group presented in the appendix corresponding to
ranges in data reporting?" each chapter. Each appendix starts with the
data locator table for that chapter, which
Optimally, studies report data across rather discrete summarizes the subject and procedural vari-
age groups (e.g., 20-24, 25-29, 30-34, 35--39, 40- ables for each study reviewed in the text, or-
44, 45-49, 50-54 years) rather than across one all- ganized in ascending chronological order. The
inclusive range (e.g., 20-54 years). table quickly highlights the most appropriate
normative data, given the demographic char-
'What is the education and/or IQ of the acteristics of the patient under study, as well
study participants?" as the test administration and scoring criteria
employed. The locator table also indicates
Because education and IQ may have a dramatic the page number on which an extensive crit-
impact on test performance, it is important to in- ical review of the study can be found in the
clude this information so that data that closely text of the chapter and directs the reader to
match the education and/or IQ of the patient under the corresponding data tables in the appendix.
study can be used. Therefore, readers have the option of reading
the critical review of the normative data for a
'What was the sample size in each of the particular test or simply using the data locator
reported age or age/education categories?" table to rapidly identify the appropriate data
"Is the sample from which data were col- set for quick test interpretation.
lected well described?" Several test chapters also include summa-
ries of results of the meta-analyses which were
For instance, the age of the subjects and the used to derive the predicted scores for dif-
country where the study was conducted always ferent age groups. The tables of predicted
PREFACE ix
scores with education or gender correction tests, a patient's IQ and/or educational char-
(where appropriate) are presented in the acteristics should closely match the demo-
corresponding appendices, along with de- graphics of the normative comparison sample.
scriptive statistics for the aggregate sample, Optimally, normative data are reported by age/
significance tests, and scatterplots depicting education or age!IQ categories (i.e., perfor-
dispersion of the data points around the re- mance of those aged 20-25 years with 12 years
gression line. of education, performance of those aged 20-
The test chapters conclude with a summary 25 years with 13-15 years of education, per-
and suggestions for future research to improve formance of those aged 20-25 years with 16
the database for the test. years of education, etc.). Sample size is also
critical because small sample size within any
of the comparison categories (i.e., age, age/
education) can undermine the stability of the
HOW TO BEST USE THE BOOK
normative data and reduce confidence in score
The process of selecting the inost appropriate interpretation. For some tests, gender and
normative report for interpretive purposes handedness must be considered. Ideally, the
involves determining the "best fit" between a administration and scoring procedures used to
patient's demographic characteristics (e.g., assess the patient should be identical to those
age, years of education, IQ, handedness) and used to collect the normative comparison data.
the demographic characteristics of the study If the data locator table suggests that more
sample. It is also critical to insure that the than one study could be appropriately used,
version of the test administered is the same as then the reader is especially advised to read the
that used to collect the normative data. Like- critical reviews of the studies closely to help
wise, it is critical that the scoring procedures determine whether one data set is more ap-
are identical. propriate than others. Close inspection of the
As a general policy, before seeing a patient, details of the studies often leads to clear-cut
we typically determine which normative data conclusions. If the data from different studies
we are going to use to interpret his or her yield contradictory values, the reader is advised
performance. This way we do not discover to consult the table of meta-analytically pre-
after a patient has gone home that the only dicted values (when available) to aid in these-
reference data available utilized a different lection of the appropriate normative data set. If
administration and/or scoring protocol from normative data for a certain demographic group
the version we used. Such "discoveries" un- cannot be found in the studies reviewed, with
dermine confidence in test score interpreta- proper caution (see Chapter 3), the expected
tion. Fortunately, however, the vast majority value for that group can be extrapolated based
of normative reports use standard adminis- on the table of predicted values or can be
tration and scoring procedures. computed based on the regression equation
If the data have already been collected, an provided with the table. However, we strongly
important variable to screen for initially is discourage the use of predicted values when the
country of origin. If the patient was born and/ actual data sets are available.
or educated in the United States, then the
most appropriate comparison data should have
been collected from individuals born and/or
educated in the United States. Another critical HISTORY OF THIS PROJECT
variable is age. A patient's test scores must be
The Beginnings
compared to those of age peers because per-
formance on most neuropsychological tests The idea for this book originally grew out of
changes as a function of age. Educational level the frustration that was experienced by Lou
and/or IQ are also important variables. Be- D'Elia in his attempts to locate appropriate
cause they can have a tremendous impact normative data during the early years of
on performance on most neuropsychological his postdoctoral training. This frustration is
X PREFACE
familiar to anyone who has used normative reference book?'' Fortunately, Lou found two
data and was practicing before 1990. Back in student colleagues in the same training pro-
the "old days," it was fairly typical fQr prac- gram who shared his concern: Kyle Boone and
ticing professionals to have access to, at most, Allen D. Brandon. Lou, Kyle, and Allen ea-
one or two sets of normative data for any gerly returned to the library to collect the data
particular neuropsychological test. More often necessary to produce a reference book. Soon,
than not, graduate students and postdoctoral however, they discovered why no such volume
fellows and trainees were handed a m~ual of existed. It is hard to imagine now, but as re-
norms to be used in the clinic or laboratory. cently as the late 1980s and early 1990s, the
These "lab manuals" containing tables; of nor- majority of neuropsychology-related profes-
mative data were passed from mentorjto trai- sional journals still had not been referenced in
nee {and vice versa) as if they were t&e Holy databases. No subject category for "Norms" or
Grail. Early in his training, Lou beg~ to ask "Normative Data" was listed in the key refer-
"Where did these data come from?"l Some- ence indices such as Index Medicus or Psy-
times a graduate student, postdocto~ fellow, chological Abstracts. As a result, most of the
or faculty member would "discover" a pew set research papers were located by going through
of norms for a particular test and a neW table the various journals article by article. Gather-
would magically appear in the lab 1panual. ing the necessary information proved to be a
Applying the new reference data to .patient very large task, not one that we would rec-
scores often yielded wildly different percentile ommend to a postdoctoral fellow at the be-
performance interpretations from those based ginning of his or her career. Yet, that is exactly
on the "standard" norms. This sent Loo to the what they did. Hindsight is 20/20!
UCLA Biomedical Ubrary to search for the Allen Brandon withdrew from the project
source of the data and to unearth the original upon completing his postdoctoral fellowship.
research articles. Often, as he read the ·article, Private practice called. Only Lou and Kyle
he discovered to his horror that the data had remained. However, for Lou and Kyle, free
been collected from individuals not educated time seemed to evaporate as they pursued de-
in the United States, that the sample size was veloping professional careers and attended to
extremely small (i.e., n < 10), or worse ~t. that their ever-increasing family activities and ob-
the data were generated from a differ~t ver- ligations. The project slowly moved forward.
sion of the test. If the same version of the test Finding and cataloguing the articles, then an-
was being used, often the normative data had alyzing them using the templates required
been collected by a nonstandard administra- much more work than they had imagined.
tion and/or scoring procedure. It was only after Then, about 1994, Maura Mitrushina joined
a thorough examination of how the J~tudies the project, and thanks to her considerable
were carried out-in terms of test ~ enthusiasm and efforts the first edition of the
tration, scoring, and demographic ch~acter book was 6nally completed.
istics of the study participants-that one could
begin to unravel the reasons why the: use of
The Second Edition-Changes and Updates
one set of normative data yielded a ~erent
interpretation than use of another. Now, 6 years later, we are glad to have on
Those trips to the library resulted in the board a new member of the team, the young
first article to summarize the availab.e nor- and vibrant Jill Razani. We invited her to
mative data for any neuropsychological test: participate in the preparation of the second
'Wechsler Memory Scale: A Critical AP)>raisal edition in order to share responsibilities for
of the Normative Studies" {D'Elia :et al., writing new chapters with reviews of addi-
1989). It was during the preparation ~f this tional commonly used tests in response to the
article that our basic template for analyzing wishes of our audience. This was the only way
normative reports was developed. to keep our sanity, attend to our families and
Lou's next question was 'Why has ~o one jobs, and have a semblance of "normal life"
gathered all this information together into a while working on the second edition.
PREFACE xi
The new tests reviewed in the second edi- studies for such analyses. The limitations of
tion include Paced Auditory Serial Addition, such predicted norms were highlighted.
Ruff 2&7, Digit Vigilance, VISual Form Dis-
crimination, Judgment of Line Orientation,
Ruff Figural Fluency, Design Fluency, WMS-
FUTURE DIRECTIONS
IIIA, California Verbal Learning, Hopkins
Verbal Learning, WHO-UCLA Auditory Ver- The handbook is as up-to-date as we could make
bal Learning, CERAD List-Learning, Selec- it. We intend to update the handbook every few
tive Reminding, Benton VISual Retention, and years; and with subsequent editions, it will be
Wisconsin Card Sorting tests. expanded to include additional tests frequently
The chapters in the first edition have been used by neuropsychologists. We have already
updated and revised. Information on meth- made a step in this direction with the second
odological issues, new versions and new ap- edition. Almost all of the tests in this book
proaches to the tests, and their clinical utility continue to appear on lists of the most popular
has been added. Studies published after 1998 tests in neuropsychology. We also managed to
that are based on well-defined, intact samples sneak in some information regarding a couple of
were reviewed. Outdated information, data published tests that were developed in our
on diagnosed clinical groups, and chapters laboratory that seem to be gaining popularity
describing tests that are not in wide use were elsewhere (i.e., Color Trails Test, WHO-UClA
removed. Auditory Verbal Learning Test).
The format of data presentation has been We hope this book finds its place on the
changed. Learning from our mistakes with the desks of professionals performing or reviewing
first edition (data tables are not exactly placed neuropsychological assessments. We also hope
in the text of their description, as we originally it will be welcomed by teachers of assessment
envisioned!), we removed all data tables from and psychological statistics and helpful to
the text and placed them in the appendices. graduate students learning to interpret test
We hope that this change will make it easier to scores. Our goal is to help bolster confidence
locate the needed tables. in the basis for clinical judgments and to
In response to the wishes of the readers of strengthen the credibility of research and
our first edition, we synthesized the data in clinical findings.
meta-analytic tables of predicted values with
supporting statistics for those chapters that Los Angeles M.M., KB.B., J.R., L.F.D.
have sufficient number and homogeneity of . California
Acknowledgments
We extend our deepest gratitude to all the Alfred Marohl, Gayle Marsh, James Marsh,
authors whose normative and clinical com- Joan McConnell, Susan McPherson, Fernando
parison research is reviewed in this book. Melendez, John Meyers, Eric Miller, Robin
Without their work, this book would not have Morris, Hector Myers, Narine Nazari, Linda
been possible. This volume is not intended to Nelson, Tina Noriega, Lara Orchanian, Eliza-
disparage the work of any author as we beth Pacheo, Daniel Parks, Nikki Passanante,
strongly believe that each author has made an Helen Paull, Eileen Pearlman, Marcel Ponton,
important contribution to our overall knowl- Stephen Rebello, Matt Reinhard, Mark Ri-
edge through their research efforts. chardson, Linda Ringer, Marcela Rivera, Ed-
Over the years, several people have helped die Rozenblat, Michael Salmone, Manuela
us with the preparation of the first and second Saul, Robert Sbordone, Jeffrey Schaeffer,
editions of this book. Their help took many Karen Schiltz, David Schretlen, Amanda
forms, including everything from typing tables Schrey, Ola Seines, Glenn Smith, Fabrizio
and checking the accuracy of references to Starace, Norton Stein, Tony Strickland, Do-
providing us with materials to be included in nald Stuss, Donald Trahan, Craig Uchiyama,
the book and simple moral support. We offer Doug Umetsu, Harry Van der Vlugt, Wilfred
each one our heartfelt thanks for every kind- Van Gorp, Valdis Volkovskis, Travis White,
ness and courtesy extended to us along the Jane Williams, Bennett Williamson, Lome
way: Lidia Artiola i Fortuny, Jean Avezac, Yeudall, Betty Young, and Miguel Zavala.
Eyzzz Baccarrdi, Julian Bach, Robert Bom- We express endless gratitude to Courtney
stein, Virdette Brumm, Debora Burnison, Sheen, who organized and coordinated the
Robert Butler, Flo Comes, Lou Costa, Michele preparation of tables for the second edition.
Croisier, Jeffrey Cummings, Janine Czametz- We thank Linda Fidell and Ingram Olkin
ki, Doug Danaher, Dean Dellis, Jack Demick, for their advice on the design and statistical
Lois Desmond, Carl Dodrill, Linda Dukma- treatment of the meta-analyses.
jian, Katharine Earhart, Robert Elliot, Kadi- We are indebted to Xiao Chen and the
mah Elson, Gwenn Evans, Bee Fletcher, UCLA ATS Statistical Consulting Group for
Travis Fogel, David Forney, Jennifer Forrest, their advice and support, ranging from pro-
Paula Fuld, Stephen Ganzell, Ismelda Gon- viding ample literature resources on applica-
zalez, Patricia Gross, Adrienne Gundry, Tif- tions of Stata in meta-analyses to invaluable
fany Harris, Lany Herrera, Charles Hinkin, help with the set-up of command files and
Stacey Horowitz, Robert Ivnik, Lissy Jarvik, interpretation of results of the analyses.
Irene Kassorla, Ellen Kester, Glen Larrabee, Special thanks go to Muriel Lezak and
Asenath LaRue, Stanislav Levin, James Loong, Edith Kaplan, who have been a constant
Enrique Lopez, Christine LoPresti, Anahit source of encouragement and support from
Magzanyan, Mario Maj, Lawrence Majovski, the very beginning of the project.
xiv ACKNOWLEDGMENTS
We extend our gratitude to Paul Satz, who Sincere thanks to our editors Jeff House,
fostered in three of the authors appreciation Fiona Stevens, and ancy Wolitzer, who e
for the complexity and excitement of the field support throughout has been continuous and
of neuropsychology. enthusiastic.
The contribution of Dale Sherman to the Finally, we thank our families: M.M. thanks
methodological accuracy of the first edition Masha, Sasha, and Kaley for their endless
qualifies him for a spot in heaven. patience and understanding; K.B.B. thanks
We also extend special thanks to Allen Rodney, Galen, and Fletcher; J.R. thanks her
Brandon, who was an early collaborator on the parents and family, especially Bill, Rl10nda,
first edition. Allen, your early efforts and great and Mike; L.F.D. thanks his parent and
enthusiasm were deeply appreciated. family, especially Michael D. Salazar, for their
Dr. D'Elia offers his admiration and ap- constant encouragement and support.
preciation to his three coauthors, whose ef-
forts brought this project to completion. M.M., K.B.B., J.R. L.F.D.
Contents
I. BACKGROUND
1. Introduction, 3
Test-Taking Environment, 6
Test Norms, 7
Tests, 9
Standard and Experimental, 9
When Is a Test Considered Experimental?, 10
What Determines Whether a Test Is Considered "Standard?'', 11
2. Use of Methodological Concepts in Neuropsychology Practice, 12

Interface of Neuropsychology with Other Clinical Disciplines, 12
Applications of Neuropsychological Evaluation, 13
Different Levels of Data Integration in Neuropsychology Practice, 15
Judgment and Decision Making in Clinical Neuropsychology, 17
Strategies in Test Selection, 17
Normative References and Interpretation of Clinical Data, 18
Alternative Methods for Interpretation of Clinical Data, 22
Factors Influencing Performance on Neuropsychological Tests, 27
Effort and Motivation, 27
Issues in Cross-Cultural and Multicultural Neuropsychological Assessment, 28
Final Caveats, 30
Data Inclusion in Neuropsychological Reports, 31
3. Statistical and Psychometric Issues, 33

Measurement and Interpretation of Numerical Values, 33
Standardization of Raw Scores, 35
Standard Scores and Normal Distribution, 36
Interpretation of Infrequent (Outlying) Scores, 38
Interpretation of Scores That Are Not Normally Distributed, 38
Psychometric Properties of Tests, 39
Reliability, 39
Methods of Estimating Test Reliability, 39
Standard Error of Measurement, 40
Validity, 41
Decision Theory, 42
Base Rates, 42
XV
xvi CONTENTS
Selection Ratio, 43
Incremental Validity, 43
Cutoffs and Diagnostic Acctiracy of a Test or Interpretive Strategy, 44
Synthesis of Results of Differen~ Studies in a Meta-Analysis, 45
Historical Overview and the Raticinale for Using Meta-Analysis in This Book, 45
Application of Meta-Analysis in Quucal Practice, 46
Advantages, 46
Sources of Bias, 46
Selection of Studies and Procedures for Meta-Analyses Presented in 11lis Book, 47
Uterature Search and Selection ci Studies, 47
Procedures Used in the Analyses, 48
Data Editing, 48
Regression, 50
Prediction, 51
Standard Deviations, 51
Testing Model Fit and Parameter'Specilications, 52
Effect of Demographic Variables, ; 54
Comments on the Applicability oP;the Meta-Analyses Presented in This Book, 55
I
II. TESTS OF ATTENTION AND f::ONCENTRATION: VISUAL AND AUDITORY
I
4. Trailmaking Test, 59
Brief History of the Test, 59
Contributions of Cognitive Mechatlisms and Physical Layout Differences to Performance on
Parts A and B, 60
Utility of the Derived Measures, Which Are Based on Differences in Performance Times for
Parts A and B, 61
Utility of the Error Analysis, 62
Utility of the Cutoffs for lmpairm~nt, 63
Effect of the Order of Presentatioa and Practice Time, Practice Effect, and Alternate Versions
oftheTMT, 64
Culture-Specific Sets of Normativ~ Data and Cultural Adaptations for the TMT, 65
Modified Versions of the TMT, fti
Relationship Between TMT PerfQnnance and Demographic Factors, 67
Method for Evaluating the No~tive Reports, 70
Summary of the Status of the Norms, 71
Summaries of the Studies, 72
Results of the Meta-Analyses of t\le Trailmaking Test Data, 96
Conclusions, 98
5. Color Trails Test, 99

Relationship Between CTT Performance and Demographic Factors, 101
Method for Evaluating the NonnJtive Reports, 102
Summary of the Status of the NofiJls, 103
Conclusions, 106
6. Stroop Test, 1oa

Current Administration Procedures, 110
CONTENTS xvii
Relationship Between Stroop Test Perfonnance and Demographic Factors, 112

Method for Evaluating the Nonnative Reports, 114
Summary of the Status of the Nonns, 115
Results of the Meta-Analyses of the Stroop Test Data, 132
Conclusions, 133
7. Auditory Consonant Trigrams, 134

Administration Procedures, 134
Psychometric Properties, 135
Relationship Between ACT Perfonnance, Demographic Factors,
and Vascular Status, 135
Conclusions, 140
8. Paced Auditory Serial Addition Test, 141

Modifications and Alternate Formats of the PASAT, 142
Psychometric Properties of the Test, 143
Relationship Between PASAT Perfonnance and Demographic Factors, 143
Conclusions, 158
9. Cancellation Tests, 160

Brief History of the Tests, 160
Ruff 2&7 Selective Attention Test, 160
Brief Overview of the Ruff 2&7, 160
Psychometric Properties of the Ruff 2&7, 161
Relationship Between Ruff 2&7 Performance and Demographic Factors, 162
Digit Vigilance Test, 162
Brief Overview of the DVf, 162
Psychometric Properties of the DVf, 163
Relationship Between DVf Performance and Demographic Factors, 163
Conclusions, 170
Ill. LANGUAGE
10. Boston Naming Test, 173
Studies Using BNT Error Quality Analyses, 174
Current Views on the Mechanisms Underlying Confrontation Naming Deficits, 176
xviii CONTENTS
Modifications and Short Versions bf the BNT, 177

Cultural Adaptations and Culture~pecific Normative Data for the BNT, 178
Relationship Between BNT Perf()rmance and Demographic Factors, 180
Method Jor Evaluating the Nonqative Reports, 182
Results of the Meta-Analyses of the Boston Naming Test Data, 197
Conclusions, 199
11. Verbal Fluency Test, 200

Psychometric Properties of the Ttft, 202
Cognitive Mechanisms Underlying Word Generation, 202
Biochemical and Anatomical Cort;lates and Effect of Brain Pathology
on Verbal Fluency, 203 ·
Assessment of Verbal Fluency in JPifferent Languages, 205
Relationship Between VFT Perfopnance and Demographic Factors, 206
Method for Evaluating the No~tive Reports, 208
Summary of the Status of the Nc;ms. 209
Results of the Meta-Analyses of ~e Verbal Fluency Data, 235
Conclusions, 237
IV. PERCEPTUAL ORGANIZATIQN: VISUOSPATIAL AND TACTILE

12. Rey-Osterrieth Complex Figure, 241
Administration Procedures, 241
Alternate Versions, 242
Scoring Systems, 243
Reliability, 248
Clinical Utility, 249 I
Culture-Specific Studies and Nomfative Data for the ROCF, 251
Relationship Between ROCF Performance and Demographic Factors, 251
Method for Evaluating the Norm.tive Reports, 253
Summary of the Status of the Noims, 254
Results of the Meta-Analyses of ~e ROCF Data, 269
Conclusions, 270 ·
13. Hooper Visual Organization ~-est, 272

Construct Validity, 273
Relationship Between HVOT Ped>rmance and Demographic Factors, 274
Method for Evaluating the Norm~tive Reports, 274
Summary of the Status of the No~s, 275
Conclusions, 277
CONTENTS xix
14. Visual Form Discrimination Test, 278

Relationship Between VFDT Perfonnance and Demographic Factors, 280
Conclusions, 282
15. Judgment of Line Orientation, 284

Alternate Brief Forms of the JLO, 286
Relationship Between JW Perfonnance and Demographic Factors, 286
Conclusions, 296
16. Design Fluency Tests, 298

Brief History of the Tests, 298
Psychometric Properties of the Design Fluency Tests, 300
Ruff Figural Fluency Test, 300
Design Fluency Test Oones-Gotman!Milner Vemon), 300
Relationship Between Design Fluency Perfonnance and
Demographic Factors, 301
Conclusions, 310
17. Tactual Performance Test, 312

Psychometric Properties of the TPT, 314
Relationship Between TPT Perfonnance and Demographic Factors, 314
Conclusions, 333
V. VERBAL AND VISUAL LEARNING AND MEMORY

18. Wechsler Memory Scale (WMS-R, WMS-111, and WMS-IIIA), 337
Relationship Between Test Perfonnance and Demographic Factors, 344
Conclusions, 355
XX CONTENTS
19. List-Learning Tests, 357

Rey Auditory-Verbal Learning Test, 357
Variability in Administration of the Rey AVLT, 357
Functioning of Different Memory Mechanisms, as Assessed by the Rey AVLT, 359
Practice Effect and Alternate Fonns of the Rey AVLT, 361
Assessment of Auditory Verbal Learning with the Rey AVLT in Different Languages
and Cultures, 362
California Verbal Learning Test-Second Edition, 362
Structure of the CVLT-11 and Description of the Nonnative Data Provided in the Test
Manual, 362
Alternate and Short Fonns of the CVLT-11, 363
Review of the Recent Literature on the CVLT and CVLT-11, 363
Effect of Semantic Organization on Recoil, 363
Anatomical Correlates, 364
Assessment of Learning and Memory in Traumatic Brain Injury, 365
Assessment of Serial Position Effect in Dementias, 366
Repeated Administration and Practice Effects, 366
Assessment of Effort with the CVLT, 367
Use of the CVLT in Other Languages and Cultures, 367
Adaptations and Alternate Versions of the CVLT, 367
Hopkins Verbal Learning Test, 368
WHO-UCLA Auditory Verbal Learning Test, 369
CERAD List-Learning Test, 370
Selective Reminding Test, 370
Other Verbal and Nonverbal List-Learning Tests, 371
Relationship Between List-Learning Test Perfonnance and Demographic
Factors, 372
Results of the Meta-Analyses of the Rey AVLT Data, 391
Conclusions, 392
20. Benton Visual Retention Test, 394

Relationship Between BVRT Perfonnance and Demographic Factors, 398
Conclusions, 416
VI. MOTOR FUNCTIONS

21. Finger Tapping Test, 419
Relationship Between FIT Perfonnance and Demographic Factors, 421
CONTENTS xxi
Results of the Meta-Analyses of the Finger Tapping Test Data, 441

Conclusions, 442
22. Grip Strength Test (Hand Dynamometer), 444

Relationship Between Hand Dynamometer Performance and Demographic
Factors, 445
Method for Evaluating the Normative Reports, 445
Results of the Meta-Analyses of the Hand Dynamometer Test Data, 457
Conclusions, 458
23. Grooved Pegboard Test, 459

Relationship Between GPT Performance and Demographic Factors, 460
Results of the Meta-Analyses of the GPT Data, 470
Conclusions, 471
VII. CONCEPT FORMATION AND REASONING

24. Category Test, 475
Alternate Formats, 477
Relationship Between Category Test Performance and Demographic
Factors, 480
Results of the Meta-Analyses of the Category Test Data, 494
Conclusions, 495
25. Wisconsin Card Sorting Test, 496

Anatomical Correlates and Effect of Brain Pathology on the WCST, 498
Brief Overview of Clinical Findings Using the WCST, 499
Modifications and Alternate Formats of the WCST, 503
Relationship Between WCST Performance and Demographic Factors, 508
Conclusions, 531
References, 533
xxii CONTENTS
Appendices
1. Where to Buy the Tests, 611
2a. Subject Instructions for ACT According to Boone et al. (1990) and Boone (1999), 613
2b. Auditory Consonant Trigrains (Boone et al., 1990; Boone, 1999), 614
2c. Subject Instructions for ACT According to Stuss et al. (1987, 1988), 615
2d. Auditory Consonant Trigrapts (Stuss et al., 1987, 1988), 616
3. WHO-UCLA Auditory Ve~al Learning Test: Instructions and Test Forms, 618
4. Locator and Data Tables fqr the Trailmaking Test (TMT), 623
4m. Meta-Analysis Tables for tJie Trailmaking Test (TMT), 648
5. Locator and Data Tables f~ the Color Trails Test, 657
6. Locator and Data Tables £ the Stroop Test, 661
6m. Meta-Analysis Tables for Stroop Test (Golden Version, Interference
Version), 680
7. Locator and Data Tables£ Auditory Consonant Trigrams, 684
8. Locator and Data Tables £ the Paced Auditory Serial Addition Test, 689
9. Locator and Data Tables £ the Cancellation Tests, 705
10. Locator and Data Tables£ the Boston Naming Test (BNT), 709
10m. Meta-Analysis Tables for t}t Boston Naming Test (BNT), 724
11. Locator and Data Tables ~<f the Verbal Fluency Test, 728
11m. Meta-Analysis Tables for ~ Verbal Fluency Test, 760
12. Locator and Data Tables f<f the Rey-Osterrieth Complex Figure (ROCF), 767
12m. Meta-Analysis Tables for ; Rey-Osterrieth Complex Figure (ROCF), 782
13. Locator and Data Tables £ the Hooper Visual Organization Test (HVOT), 793
14. Locator and Data Tables£ the Visual Form Discrimination Test, 796
15. Locator and Data Tables ~<f the Judgment of Line Orientation Test, 798
16. Locator and Data Tables ~<l the Design Fluency Tests, 805
17. Locator and Data Tables for the Tactual Performance Test, 812
18. Locator and Data Tables fot the Wechsler Memory Scale (WMS-R, WMS-111, and
WMS-IIIA), 828 i
19. Locator and Data Tables fot the List-Learning Tests, 837
19m. Meta-Analysis Tables forth~ Rey Auditory-Verbal Learning Test (Rey AVLT), 869
20. Locator and Data Tables for the Benton Visual Retention Test, 885
21. Locator and Data Tables for the Finger Tapping Test (FTT), 903
21m. Meta-Analysis Tables for th~ Finger Tapping Test (FTT), 921
22. Locator and Data Tables for the Grip Strength Test (Hand Dynamometer), 934
22m. Meta-Analysis Tables for the Grip Strength Test (Hand Dynamometer), 946
23. Locator and Data Tables fot the Grooved Pegboard Test (GPT), 957
23m. Meta-Analysis Tables forth~ Grooved Pegboard Test (GPT), 969
24. Locator and Data Tables for the Category Test (CT), 976
24m. Meta-Analysis Tables for the Category Test (CT), 985
25. Locator and Data Tables fot the Wisconsin Card Sorting Test, 989
Copyright Acknowledgments, 1013
Index, 1015
BACKGROUND
1
Introduction
Clinical neuropsychology is an applied science After administration of the te:.t battery, the
concerned with the behavioral expression of neuropsychologist is faced with making sense
brain dysfunction (Lezak et al., 2004). Neuro- of a plethora of numerical and qualitative data.
psychologists are neurobehavior specialists To make optimal use of the test data, the
who administer tests and test batteries that are neuropsychologist must have an understand-
typically tailored to answer specific referral ing of what constitutes "normal" performance
questions. Ideally, a neuropsychological test on the tests before an opinion regarding the
battery consists of well-validated, reliable, strengths and/or weaknesses of various neu-
standardized, and normed measures that help robehavioral capacities can be offered.
to elucidate and quantify behavioral changes To be meaningful, test scores must have an
that may have resulted from brain injury empirical frame of reference. Normative data
or other central nervous system disturbances. provide this empirical context and represent
A neuropsychological examination provides a the range of performance on a particular test
comprehensive evaluation of cognitive do- of a group of medically/neurologically healthy
mains putatively associated with various brain individuals with relatively homogeneous de-
substrates. The cognitive domains typically as- mographic characteristics. These normative
sessed include language, attention/concentra- reference groups are considered the "gold
tion, visuospatial perception and constructional standard" against which an individual's test
abilities, frontal systems/executive functions, performance is compared and contrasted.
and verbal and nonverbal learning and mem- However, while normative data are a critical
ory. Sensory and motor functions as well as starting point for the interpretive process, they
general intellectual functioning are also rou- do not provide the sole basis for the interpre-
tinely assessed. Findings from a neuropsycho- tation of test scores. Test data alone do not pro-
logical examination can help to highlight areas vide an adequate basis for making sound
of functional strengths and weaknesses that clinical judgments regarding cognitive function-
may have focal or lateralizing significance. A ing. As Lezak et al. (2004) stress, interpretation
neuropsychological evaluation is therefore of the data obtained from a neuropsychological
considered an essential component in the examination must take into account qualitative
diagnosis, treatment planning, and care of observations and the patient's history, back-
patients with suspected congenital or acquired ground, present circumstances, motivation,
brain dysfunction. attitudes, and expectations regarding self and
3
4 BACKGROUND
the examination. A formal evaluation of the nervous system, central nervous system, and
patient's emotional functioning and personal- medical and/or emotional dysfunction to the
ity characteristics is also an intrinsi' part of clinical picture.
neuropsychological evaluation. All tl}is infor- It is also important to observe the qualitative
mation taken together provides a .ework process of performance leading to a specific
for an accurate understanding of a patient's score on a test. Reporting a score without re-
cognitive strengths and limitations. : vealing how it was obtained can sometimes be
To illustrate the relationship betwe$1 differ- misleading. For illustrative purposes, assume
ent sources of information, consider~.: 1.1. that the criterion for passing a particular
As can be seen, no single section of the yramid component of a driving test is that the car is
in Figure 1.1 should be used alone to form an parked in the garage. We look and, yes, the car
opinion about neuropsychological 'oning. is in the garage-criterion met. The driver
All interpretive elements (raw data/n , s, ob- passed the test. Or did he? We interview an
servations of test-taking behaviors, an medical observer of the event and sadly learn that the
history/presenting symptoms) must person drove through the garage door to get
play in building evidence that forms there! Obviously not a stellar performance.
for a professional opinion. Further s Similarly, with neuropsychological testing, not-
interpretive process, of course, is th neuro- ing how an individual obtained a score can be
psychologist's clinical judgment, which. is influ- quite illuminating. Consider two 75-year-old
enced by his or her education, professional architects (patient A and patient B) who each
experience, and research knowledge b.e. obtain a score of 36/36 on the Rey-Osterrieth
The history (including medical, ~c. ed- test. On a nonnative basis alone, performance
ucational, vocational, and avocational)*pre- scores for both patients would be considered
senting symptoms are important in und ding within normal limits, but how did these two
test data. A detailed medical!psychia ·c his- architects obtain the score? Patient A quickly
tory is an especially important source of in- recognized the overall gestalt of the drawing,
formation given that neuropsychol~al test drew a box, and filled in the details. Patient B
performance can be greatly influe~d by failed to appreciate the drawing's overall shape
medical or psychiatric conditions. Doqument- and built up the figure by accretion, taking 8
ing risk factors known to affect neuropsycho- minutes to do so, with numerous erasures. Al-
logical test performance is essential slnce an though patient B produced the same score as
important task of neuropsychologists is to prop- patient A, the dramatically different approach
erly attribute the contribution of peiipheral that patient B followed to complete the design
OBSERVATIONS
lntarpratation --+ REPORT
RAW DATA,
NORMS
Fisure 1.1. Graphic representation of the relatio~hip between different sources of information contributing
to the decision-making process in neuropsycholoir.
INTRODUCTION 5
suggests significant loss in spatial/construc- In neuropsychological practice, one exam-

tional ability. ines the pattern of test performance within a
In observing a patient's test-taking style, it is functional domain to assure greater probability
important to assess attitude, effort, and moti- of accurate interpretation. No single test score
vation. In all evaluations, one must ascertain is sufficient to render a judgment regarding
whether papents are offering their best per- brain dysfunction. For instance, a single score
formance. However, low test scores can be a on a verbal memory test has little meaning
product of lack of effort, incorrectly following unless interpreted in the context of the indi-
test instructions, or deliberate failure. Fortu- vidual's pattern of performance on other tests
nately, several free-standing effort tests as well which also assess verbal memory functioning.
as cut-offs for standard cognitive tests sensi- The reason for this is quite simple. In most
tive to the presence of suspect effort are now neuropsychological examinations, one can ex-
available to detect noncredible performance. pect to 6nd an instance of unexplained per-
When suspect scores are obtained on such formance deviation. This deviation may be
measures, it may be important to confirm that due to an attentionallapse or to measurement
the patient correctly followed the instructional error. Obviously, an occasional poor test find-
demands of the test. In our clinical experience, ing, more or less in isolation, has less meaning
we have had cases (although rare) of patients than a pattern of poor test findings occurring
with well-documented significant brain within a specific cognitive domain. Also, it is
injuries, their diagnosis supported with struc- important to keep in mind that no test is a
tural and functional neuroimaging data (CAT, perfect measure of what it purports to assess.
MRI, fMRI, SPECT, PET), who nonetheless Therefore, it is often important to adminis-
scored below chance level on some of the most ter more than one test to assess a particular
popular forced-choice and probability theory- functional domain, especially if a lower than
based tests specifically designed to assess for expected score is obtained on an initial ex-
symptom validity and motivation. Therefore, amination.
when questionable scores are obtained on mea- For instance, consider a neuropsychologist
sures assessing for symptom validity, motiva- who administers the Grooved Pegboard Test
tion, and effort, we recommend that the to a right hand-dominant patient as the only
examiner, if possible, obtain from patients a measure of motor functioning. The medical
description of what they thought they were history establishes that the patient has never
asked to do and how they solved the test. The previously sustained any significant bone or
examiner may be very surprised by the an- soft tissue injuries to his fingers, hands, arms,
swers. As an example, because of the forced- shoulders, or neck. Yet, when administered
choice nature of its design, we sometimes use the Grooved Pegboard Test, the patient ob-
the Warrington Facial Recognition Test as one tains a lower score for the dominant hand than
of several measures to assess for symptom va- for the nondominant hand, and the neuro-
lidity, motivation, and effort. However, it is psychologist regards this as evidence of left
possible to score below chance level if the pa- hemisphere dysfunction. The problem with
tient decides that he or she is going to select all this interpretation, of course, is that poor
the "unpleasant" faces, determined by whether performance on the Grooved Pegboard Test
the face is looking away from the viewer, not might be due to factors other than a motor
smiling, or otherwise not staring the viewer in deficit, such as a "practice effect," which ben-
the eyes. efits the second hand to be tested. Poor
Generally, the overall profile of scores and dominant hand performance might also be
the consistency of successes and failures across explained by a patient's idiosyncratic approach
different measures addressing similar cogni- and slow start on the task. Therefore, in this
tive domains greatly aid in determining the situation, it would be important to administer
validity of test results. This issue is especially other motor tests, such as the Finger Tapping
important in medical-legal evaluations due to Test or Grip Strength (aka, Hand Dyna-
the possibility of identifiable secondary gain. mometer) Test, to determine whether the
6 BACKGROUND
original findings can be corroborated. Thus, conditions. Therefore, any deviation from a
more than one test should be administered standardized test environment should be well
when assessing performance within a specific documented in the report because such situa-
functional domain so that the internal consis- tions almost always adversely affect the reliabil-
tency of performance findings can be judged ity and validity of the test data and, therefore,
before offering an opinion about function. the reliability and validity of the professional
In summary, failure to consider all aspects opinions derived from the data. Observers,
of the interpretive process will increase the court reporters, and video/DVD and/or audio
probability of faulty inferences being drawn recording equipment should never be allowed
from the neuropsychological data obtained. to document a neuropsychological testing ses-
sion. Unfortunately, requests for observers
and recording are not uncommon from attor-
neys seeking to insure and document that
TEST-TAKING ENVIRONMENT
the neuropsychological tests administered
Neuropsychologists evaluate behavior with to their client were fairly and correctly
neuropsychological test procedures. The task conducted. However, patients may behave
of the neuropsychologist is to obtain from differently when observed or recorded; Con-
the patient the best possible performance. To stantinou et al. (2002) observed that memory
achieve this, the neuropsychologist must de- scores declined under audiotaping conditions,
velop a rapport with the patient, gain coop- while Kehrer and colleagues (2000) found that
eration, and conduct the evaluation in an the presence of an observer was associated
environment that is as free as possible from with lower scores on measures of attention,
distracting influences. This is referred to as information-processing speed, and verbal flu-
the "ideal" test environment. As Lezak et al. ency. Further, the presence of observers and/
(2004) commented, or recording equipment alters the standard-
ized examination environment. Because a non-
It is not difficult to get a brain damaged patient to do
standardized test environment is created
poorly on a psychological examination, for the
under these circumstances, the data obtained
quality of the performance can be exceedingly vul-
from the evaluation could be considered in-
nerable to external inHuences or changes in internal
valid. Further, the normative data used to in-
states. All an examiner need do is make these pa-
terpret test scores were collected under
tients tired or anxious, or subject them to any one of
standardized testing conditions and not in the
a number of distractions most people ordinarily do
presence of third-party observers or electronic
not even notice, and their test scores will plummet.
recording devices.
In neuropsychological assessment, the difficult task
An added problem is that the use of
is enabling [emphasis ours] the patient to perform as
recording equipment in a testing session may
well as possible. (p. 130)
place neuropsychologists in potential conflict
Indeed, discovering and documenting what a with state laws regulating the practice of psy-
patient is capable of doing, or not doing, un- chology, as well as federal copyright law. In
der the best of circumstances is of tremendous addition to legal consequences, ethical con-
clinical value from diagnostic and prognostic cerns are raised since Ethical Standard 9.11 of
viewpoints. The neuropsychologist is ethically the American Psychological Association Code
bound to insure that the testing session is of Conduct requires that "psychologists make
conducted in an atmosphere that promotes reasonable efforts to maintain the integrity and
cooperation, accuracy and honesty and that security of test materials and other assessment
minimizes any chance of collecting less than techniques . . . ." The key problem is that
optimal data from the patient because of neuropsychologists usually have no control of
negative influences of the examiner and/or the recordings once they leave the office. The
other influences, demands, or distractions. CD/DVD or tapes can be disseminated with-
Neuropsychological test norms have been out regard to maintaining test security and
standardized under ideal test environment assessment techniques. As was noted in the
INTRODUCTION 7
official statement regarding test security by the are employed in "publish-or-perish" environ-
National Academy of Neuropsychology, ments, such research takes a low priority.
These obstacles have remained in place despite
The potential . . . likely and foreseeable conse-
the awareness that normative research is sorely
quence of uncontrolled test release is widespread
needed and, indeed, essential. So, although
circulation, leading to the opportunity to detennine
neuropsychological assessment procedures are
answers in advance, and manipulation of test per-
widely available, there remains a relative scar-
formance. This is analogous to the situation in
city of normative data for most tests.
which a student gains access to test items and the
For a few of the most popular tests (e.g.,
answer key for a final examination prior to taking
Trailmaking Test A and B, Rey Auditory-
the test. (Axelrod et al., 2000b)
Verbal Learning Test [AVLT]), however, nu-
Should a test become invalidated due to ex- merous reports are available which provide
posure to the public domain, redevelopment normative comparison data for performance
and replacement would be very costly in terms across the greater part of the life span. The
of time and money. In the interim, the problem for clinicians and researchers then is
professional community would be deprived of which set of normative data to use? Large
an effective assessment instrument. differences in reported scores among studies
examining performance on the same test in
groups of individuals that have nearly identical
demographic characteristics, as well as vague
TEST NORMS
descriptions of how the data were collected,
Despite the critical importance of having compound the problem of identifying appro-
access to normative data to facilitate clinical priate norms for comparison purposes. A fre-
interpretation of tests findings, there still are quent difficulty is that use of one set of norms
relatively few large-scale normative reports may suggest that the patient is performing in
in the literature. This is especially evident the impaired range while use of another set
when one considers the large number of may suggest that performance is within nor-
validation studies on the utility of neuro- mal limits. Unfortunately, use of the wrong set
psychological tests in discriminating groups of of comparison data may pave the way for faulty
patients with lateralized or focal lesions. To inferences being drawn, perhaps also resulting
some extent, the relatively small body of lit- in either unnecessary treatment or therapeutic
erature regarding normative research is re- neglect. To demonstrate this point, consider
lated to the formidable logistic problems and the hypothetical medical-legal case of a phys-
expense associated with the execution of such ically healthy 80-year-old male, retired factory
studies. However, the larger problem is that worker, born and educated in the United
researchers generally have not been supported States (1 year of college), who presents with
either financially or professionally in conduct- complaints of memory difficulty. The patient
ing normative research. Although test norms would like to continue to handle his financial
are essential to proper interpretation in clinical affairs, but his family is concerned about his
and research settings, major funding agencies competence to do so. In the course of evalu-
have not favored such studies. The key prob- ating the patient, you administer several
lem is that these studies are inevitably de- memory tests including the Rey AVLT. In the
scriptive, and descriptive studies are generally unlikely event that you were only interested in
not considered "scientific" since no hypothe- his performance score for trialS, there would
ses are being tested. Until very recently, be eight normative reports listed in this vol-
journal editors have generally been loath to ume from which to choose, which are pre-
publish normative reports, opting instead for sented in Table 1.1.
more scientific works. As a result, a substantial Let us assume that the patient's trialS score
number of normative data sets are imbedded was 6/1S correct. Reliance on the Query and
in publications of clinical studies, making Megran (1983) normative data would suggest
them difficult to locate. Since most researchers that test performance is in the average range
8 BACKGROUND
Table 1.1. Published Rey Auditory-Verbal Learning Test Nonnative Reports

Age Group Education TrialS Delayed
Investigator {years) n {years) {#of words) Recall {# of words) Test lDcation
Rey, 1964 70-90 15 No info. 9.5 Switzerland

{2.2)
Query& 70-81 23 11.4 5.86 3.45 N. Dakota {all Veterans
Megran, 1983 {2.04) {2.92) Administration inpatients
with physical complaints)
Cohen et al., 75-89 4 13.8 9.25 8.25 Peoria, IL
pers. comm. {1.89) {2.63)
Bleeker et al., 1988 80-89 11 13-18 9.2 Muyland
{2.1)
Geffen et al., 1990 70-86 10 11.2 8.2 5.6 S. Australia (all male
{2.5) {2.6) subjects)
Ivnik et al., 1990 80-84 49 2:12 9.0 5.5 Minnesota
{2.5) {3.3)
Mitrushina et al., 76-85 26 13.3 9.7 s. California
1991 {3.6) {2.8)
Mitrushina & 76-85 16 14.0 10.3 s. Californla
Satz, 1991a {3.6) {2.4)
(53rd percentile). However, close reading of et al., Rey, and Mitrushina and Satz. As noted
that normative report would reveal that the in Spreen and Strauss (1998), use of Rey's
data were collected on a sample of Veterans norms (reported in 1964 but collected in
Administration patients hospitalized for a va- 1944) should also be avoided because of test
riety of physical complaints. Thus, overall per- content and administration differences. These
formance scores of the comparison sample data were collected over 50 years ago in
were probably artiflcially lowered because of Switzerland, raising serious concerns about
hospitalization effects, chronic pain effects, cohort and cultural effects. Similarly, data
and dysphoria. Therefore, applying the Query from the Geffen et al. (1990) report should be
and Megran data would lead the examiner to avoided due to cultural differences in com-
conclude that the patient's performance was paring North American vs. Australian samples
better than it probably was. Depending on and the fact that the educational level of the
which other remaining normative reports were samples was low. We had previously elimi-
used, the patient's score would fall in the low nated the Query and Megran (1983) study for
average (Geffen et al., 1990, 19th percentile; the reasons mentioned earlier. Of the two
Ivniketal., 1990, 12th percentile) or borderline normative reports remaining (Ivnik et al.,
(Rey, 1964, 6th percentile; Cohen et al., per- 1990; Mitrushina et al., 1991), that by Ivnik
sonal communication, 4th percentile; Bleecker et al. would be selected because of the larger
et al., 1988, 6th percentile; Mitrushina et al., sample size. The subject and procedural
1991, 9th percentile; Mitrushina & Satz, 1991a, characteristics of these two studies are other-
4th percentile) range. wise nearly identical. Finally, and most im-
Unfortunately, all the studies reporting trial portantly, the demographic characteristics of
5 data for this age group suffer from small the patient being evaluated matched well with
sample size (n <50). In terms of selecting the the demographic characteristics of the par-
"best" study for comparison purposes, those ticipants in this normative study.
with the smallest sample size should probably If one were interested in examining the
be first rejected. This would eliminate the delayed recall performance of this patient, four
studies of Cohen et al., Bleecker et al., Geffen studies would be available for normative
INTRODUCTION 9
comparison purposes (Query & Megran, 1983; similar to other patient groups would be a
Cohen et al., personal communication; Geffen welcome addition to the field. The availability
et al., 1990; Ivnik et al., 1990). Query and of clinical comparison data sets would permit
Megran's report would be avoided because of such analysis.
the sampling problems noted earlier. The
Cohen et al. report would not be considered
primarily due to the small sample size.
In narrowing the choices down to the two re- TESTS
maining studies, the Geffen et al. report would
Standard and Experimental
be rejected in favor of the Ivnik et al. report for
reasons of sample size and educational and Neuropsychologists use published, standard-
cultural issues and the fact that the length of ized measures that are well normed and gen-
delay for the delayed recall condition was erally accepted as standard tools of assessment
30 minutes. However, if a better normative in the field. There are several reasons why
data match is not available, it should be neuropsychologists must utilize standardized
considered that the greatest rate of forgetting measures. Patients presenting for neuropsy-
occurs during the first 20 minutes. Conse- chological evaluation are frequently reeval-
quently, the effect of variability in the length uated over time, often by different examiners.
of delay between different norms from 20 to As Figure 1.1 illustrates, the test data collected
30 minutes (even 45 minutes to 1 hour) on rates during the examination are an essential ingre-
of forgetting is minimal. Because no normative dient in forming a professional opinion re-
study ordinarily allows a perfect fit to the de- garding neuropsychological functioning. Since
mographic characteristics of the patient, the the written report of findings is a professional
examiner should be aware of the specific lim- communication from one clinician to another,
itations of any data set used for inferential for the report to be meaningful and therefore
purposes. useful to other professionals, the data must
Because patients are often evaluated more have been obtained from tests that are familiar
than once on the same test (often by different to, or can be easily referenced by, any clinician.
examiners), clinicians are urged to document The use of standard tests and administration I
the source of comparison data used to arrive scoring procedures is especially critical in ini-
at conclusions within the body of any report. tial examinations of patients, to establish a
This recommendation is especially relevant meaningful baseline. Baseline data are essen-
for tests with multiple and/or overlapping sets tial for subsequent comparison with retest
of normative data (D'Elia et al., 1989). data, to document whether there has been
In addition to normative data, there is an- improvement or decline in functioning. If tests
other set of data that can be used for com- are employed in the initial exam which are
parison purposes when it is available. Clinical unique to the neuropsychologist administering
comparison data (aka, abnorms) represent the them, then subsequent examiners will be un-
range of test performances of distinct groups able to make comparisons between baseline
of medically, psychiatrically, and/or neurolog- and subsequent performances.
ically compromised individuals with relatively In a forensic context, the use of nonstandard
homogeneous demographic characteristics. tests and procedures impedes the fact-finding
Because of the general lack of clinical process. Unfortunately, it is not uncommon
comparison data, however, neuropsycholo- to review reports of medical-legal examina-
gists have largely relied on normative com- tions that have relied heavily or exclusively
parison data for interpretive purposes. As a upon findings from experimental measures
result, interpretive comments have been lim- as the basis for an opinion of neuropsycholo-
ited to reporting how the patient under study gical impairment. Not surprisingly. the exper-
differs from a healthy sample. Having the imental measures employed almost always are
ability to discuss how a patient under study is purported to be tests of memory, attention/
10 BACKGROUND
concentration, and/or frontal system func- results . . . psychologists take into account
tioning, three cognitive domains that are es- the . . . characteristics of the person being
pecially sensitive to any brain insult, regardless assessed . . . that might affect . . . judg-
of etiology. The use of experimental tests and ments or reduce accuracy of their interpre-
the claims made to courts that the measures tations." Ethical Standard 9.09(c) states "Psy-
are at the "cutting edge" and provide objective chologists retain responsibility for the
evidence of impaired brain function often re- appropriate application, interpretation, and use
sult in, at least for a while, increased business of assessment instruments, whether they score
for the provider of this service. Rather than and interpret such tests themselves or use au-
providing neuropsychological service, how- tomated or other services."
ever, neuropsychologists conducting them- Of course, experimental measures are oc-
selves in this way are practicing a form of casionally used during the course of an evalu-
modem-day alchemy: they have found a way to ation. This is tolerated by our field because
tum their time into gold. Such practice not experimental test development is a natural
only harms the patient through mislabeling but evolution in a discipline that advocates and
also undermines the legitimacy of neu- uses research to advance knowledge about
ropsychological practice in the forensic arena. clinical assessment and diagnosis. However,
Increasingly, neuropsychologists are being findings from experimental tests should be
called upon to provide expert-witness court- used only to supplement and support findings
room testimony as to whether brain-behavior from standardized procedures that were also
functioning is "intact," "impaired," or otherwise administered. Findings from experimental
"compromised" after a brain insult (Hom, 2003; tests should never be used as the primary basis
Leckliter & Matarazzo, 1989; Matarazzo, 1990; for forming opinions about impaired neu-
Satz, 1988). This development seems reasonable ropsychological functioning. Serious profes-
since neuropsychology is an applied science sional and ethical concerns are raised when
concerned with the behavioral expression of neuropsychologists substantially deviate from
brain function and dysfunction. However, the standard test administration and scoring pro-
days of being able to state in court "It is my cedures and/ or rely heavily on experimental
professional experience that . . ."without also tests for clinical judgments.
providing the objective basis to support an
opinion (i.e., normative data, clinical comparison
data, history, medical records, symptoms, ob- When Is a Test Considered
servations) are over. Neuropsychologists are Experimentalt
now regularly asked by knowledgeable attorneys
There are at least four levels of experimental
to produce the objective basis-particularly the
tests (presented in order from most experi-
normative data-for their opinions. This devel-
mental to least):
opment stems, in part, from the Faust et al.
(1991) two-volume tome entitled Brain Damage
Levell tests have never been peer-reviewed
Claims: Coping with Neuropsychological Evi-
or published and are typically uniquely uti-
dence, which provided scathing reviews of the
lized by the neuropsychologist who developed
field of neuropsychology for attorneys.
them. Often, normative data are sparse or
Neuropsychologists must be prepared to
nonexistent.
state why they used a particular set of normative
data for comparison purposes. Although this Level2 tests although never published, may
line of inquiry by attorneys is a relatively new have been reviewed by peers and used by a
development in the courtroom, the practice is specific group of neuropsychologists conduct-
sound and should be welcomed. Why? Let us ing a multisite research study. Again, norma-
look at what our professional rules of practice tive data may be sparse or lacking.
tell us. Ethical Standard 9.06 of our code of Level 3 tests often hpve been widely distrib-
ethics (American Psychological Association, uted to interested parties for experimental use;
2002) states "When interpreting assessment however, there have been no published studies
INTRODUCTION 11
using these tests. Preliminaty normative, reli- adequate normative data are often sparse
ability, and validity data are usually available. or lacking. Fortunately, several research
Level 4 tests have been carefully described groups are continuing to address the
in peer-reviewed journals as having been need for norms within the upper age
included in a study where several other stan- ranges and for various cultural subsets.
dardized tests were administered. Preliminaty 2. The test stimulus and materials should be
normative data are available, and there is some standardized. A manual describing test
information about the test's reliability and administration and scoring procedures
validity. These tests generally have not yet and providing information on reliability
been formally published by a recognized test and validity should be available. These
publisher/distributor but are made available requirements usually imply that the
by the author(s) to interested parties for test has been formally published by a
research/experimental use. respected test publisher/distribution com-
pany. However, several tests that are
Whereas findings based on level 1, 2, and considered standard have never been
3 tests should be given little clinical weight and formally published. For instance, the Rey
viewed with extreme caution, a level4 test has AVLT has never been formally published
at least undergone formal peer review before nor have the administration procedures
publication of the findings in a recognized pro- been standardized; but most of the nec-
fessional journal. Therefore, although caution essruy information on norms, reliability,
is warranted when discussing findings from a and validity for this test can be found in
level 4 experimental test, the results may be the neuropsychological literature.
used if they are buttressed by similar findings 3. Research using the test must have been
from more formal, standardized tests. peer-reviewed and published in recog-
nized professional journals.
4. The test has been reviewed in the Mental
What Determines Whether a Test
Measurements Yearbook arullor in more
Is Considered 11StandardJ"
than one neuropsychological text by au-
At a minimum, three of the following four thors not connected with its development.
criteria should be met before a test is re-
garded as being in standard use: In conclusion, neuropsychologists are re-
sponsible for choosing the particular tests to
1. The test must be readily available to the answer referral or research questions and the
professional community and adequately best possible clinical comparison data (both
normed. Using this definition, however, norms and "abnorms"). Neuropsychologists
it is possible for a test considered "stan- are also responsible for insuring that the test
dard" to slide into the "experimental" measures are administered properly and in a
range when not adequately normed for comfortable, nonthreatening, and distraction-
the age or sociocultural group under free environment that will enable the patient
study. For instance, the majority of to perform to the best of his or her ability.
our "standard" tests must be consid- Being more accountable for what is done and
ered "quasi-experimental" when asses- being able to elucidate why it was done will
sing adults over age 80, members of not only raise the credibility of neuropsychol-
minority groups, or any individual who ogy as a distinct specialty but also enhance the
does not speak English or uses it as a use of neuropsychological services within the
second language. In each such case, medical and legal communities.
2
Use of Methodological Concepts
in Neuropsychology Practice
INTERFACE OF NEUROPSYCHOLOGY commands, calculations, abstract reasoning,

WITH OTHER CLINICAL DISCIPLINES fund of general information, and judgment
Assessment of cognitive and affective com-
The notion of mental status has three components of mental status also constitutes pJUt
ponents: of a neurological evaluation that addresses
higher cortical and limbic system functions. In
1. Mood and affect addition, the neurological evaluation focuses
2. Perception and content/process of on the integrity of the lower levels of the ner-
thought vous system through assessing functions of the
3. Cognitive status cranial neiVes, motor systems, senso.ry systems,
reflexes, coordination, station, and gait.
Factors such as appearance, motor activity, Psychiatric and neurological approaches are
insight, and motivation can be incorporated redefined in the context of neuropsychiatry, a
into these three components of mental status. recently rediscovered medical discipline that
Clinicians specializing in different disciplines has its roots in the notion of "psychosomatics".
approach the mental status evaluation from Neuropsychiatry is concerned with both neu- .
different perspectives. rological and psychiatric symptoms of brain-
For psychiatrists, the presence of the follow- related disorders and is supported by advances
ing symptomatology is of primary concern: af- in neuroimaging, psychopharmacology, genet-
fective symptoms (e.g., depression, mania, rapid ics, and molecular biology.
cycling), perceptual disturbances (e.g., halluci- In the context of psychiatric, neurological,
nations), disturbance in the content of thought and neuropsychiatric evaluations, cognitive
(e.g., delusions), and disturbed process of status is assessed by unstructured questioning
thought (e.g., tangentiality, loosening of associ- or through administration of structured screen-
ations, flight of ideas). Assessment of cognitive ing instruments, allowing quantification of a
status represents one of the aspects of psychi- patient's cognitive status (e.g., Mini-Mental
atric evaluation and includes brief appraisal of State Examination: Folstein et al., 1975). This
the level of consciousness, orientation, attention, assessment is brief (limited to 10-20 minutes)
memo.ry, language, ability to follow verbal and, therefore, yields only a gross estimate of
12
METHODOLOGICAL CONCEPTS IN NEUROPSYCHOLOGY 13
cognitive abilities. If indications of cognitive plementation of rehabilitative efforts to opti-

decline are evident on the cognitive screening, mize functional level of an individual.
the patient is referred for a neuropsychological Recent surveys and consensus statements
evaluation. document the utility of neuropsychological
Assessment of the cognitive component of evaluation in different clinical settings: for
mental status is the primary focus of neu- example, in the differential diagnosis of de-
ropsychological evaluation, though affective mentia (McKhann et al., 1984; Roman et al.,
state and content/process of thought are also 1993), in screening for cognitive impairment in
attended to. The standardized neuropsycho- the psychiatric emergency service (Copersino
logical instruments that are used in a com- et al., 2003), in assessment of neurobehavioral
prehensive assessment of different cognitive outcomes after cardiac surgery (Murkin et al.,
domains are more sensitive to subtle func- 1995), etc. The role of neuropsychological as-
tional deficits than the gross screening tools sessment in evaluating patients undergoing
used in psychiatric, neurological, and neuro- epilepsy surgery and in management planning
psychiatric examinations. for patients with suspected dementia, multiple
sclerosis, Parkinson's disease, traumatic brain
injury (TBI), stroke, or HIV encephalopathy is
addressed in the report issued by the Ameri-
APPLICATIONS OF
can Academy of Neurology (1996). The report
NEUROPSYCHOLOGICAL EVALUATION
describes the contribution of neuropsycholo-
The utility of psychological testing is in the gical assessment to our understanding of neu-
midst of a heated controversy. Over the years, rological disorders in adults and documents its
the psychometric approach has been criticized strengths and limitations. Although the report
for being too mechanistic and biased against offers endorsement of neuropsychological as-
underrepresented groups. Neuropsychologi- sessment as "appropriate" and rates it as "es-
cal evaluation is not limited to administration tablished," it stirred up a controversy in the
and interpretation of the tests; instead, it pro- field of clinical neuropsychology (Bieliauskas
vides a comprehensive picture of the patient, et al., 1997a; Bigler & Dodrill, 1997; Hartlage,
placing test performance in the context of 2001; Reitan & Wolfson, 2001).
expectations specific for a particular individ- Clinical neuropsychology, equipped with
ual. This approach overcomes the flaws inher- recent theoretical and clinical advancements,
ent in testing. alliance with neuroimaging and interdisci-
Clinical neuropsychology is a relatively plinary affiliations, has a wide range of
young discipline that continues to shape itself applications:
in response to the pressing needs of clinical
practice. With changing environmental de- I. Neuropsychological evaluation is used in
mands, clinical neuropsychology redefines its the differential diagnosis of conditions
objectives, its applications, and its relationship involving cognitive dysfunction (e.g.,
with other disciplines and revises its arma- dementia of Alzheimer's type vs. vascu-
mentarium. Whereas in the 1940s its main role lar dementia vs. depression). It is espe-
was in the localization of lesioned brain areas, cially useful in identifYing patterns of
this need has faded with the rapid advance- impairment in patients with subtle defi-
ment of neuroimaging. The focus of clinical cits (as confirmed in the report by the
neuropsychology has shifted toward develop- American Academy of Neurology, 1996)
ing methods for diagnosing developmental and and sensitive to abnormalities in brain
acquired disorders of brain function, which function that are not detectable by neu-
constitute a part of the multidisciplinary neuroimaging.
rodiagnostic work-up, and toward under- One of the applications of neuro-
standing the cognitive status of an individual psychological evaluation is to determine
and his or her functional efficiency in dealing the effect of medical disorders, specifi-
with daily tasks, as well as planning and im- cally involving the central nervous system,
14 BACKGROUND
TBis, infections, metabolic abnormali- possible changes in lifestyle, and prompts

ties, oxygen deprivation, exposure to reassignment of responsibilities within
neurotoxins, or developmental and psy- the family.
chiatric disorders on cognition. Similarly, the rate and pattern of
In the past, differential diagnostic cognitive deterioration associated with
questions addressed to a neuropsychol- dementia, as identified by follow-up
ogist were framed in terms of organic vs. probes, provide useful diagnostic and
functional etiologies of disturbance. prognostic information.
With the growing evidence of neuro- Repeated assessment is also used to
pathological and chemical correlates of identify changes in cognitive status due
functional disorders, the organic vs. to intervention such as cognitive reme-
functional dichotomy becomes obsolete. diation, to determine the efficacy/toxicity
2. Neuropsychological evaluation is useful of a medication, and to identify the ef-
in defining baseline levels of cognitive fect of radiation, chemotherapy, or other
functioning for longitudinal comparisons treatment modalities on cognitive status.
with follow-up data. For example, even a Pre- and postsurgery assessments allow
subtle age-related decline in attention, identification of cognitive deficits result-
decision-making, judgment, and visuo- ing from neurosurgery and follow-up on
spatial abilities in pilots might lead to the progress in cognitive remediation.
failure to respond adequately in a critical 3. Neuropsychological data are used in eval-
situation. In view of the tremendous re- uating patients' employability and con-
sponsibility for human lives imposed on stitute a basis for determining whether
pilots, early signs of age-related decline they meet the criteria for disability. Fur-
need to be identified through longitudi- thermore, these data are relied upon in
nal comparisons of performance. determining patients' competence in
In another example, a baseline profile handling their legal and financial affairs,
obtained on a child with a suspected capacity to participate in medical and
learning disability before entering pri- legal decision making, and ability to func-
mary school can be compared with the tion independently in an everyday envi-
results obtained 1 year later. This defines ronment. For example, a severe deficit in
the rate of acquisition of a particular skill executive functions related to amnesia
and facilitates decision making regarding associated with Korsakotrs syndrome
the necessity of remedial intervention renders a patient unable to maintain ba-
and special education/resource class- sic self-care in spite of retaining a high
room placement. level ofpsychometric intelligence. In such
Longitudinal follow-up is also used cases, the results of an evaluation identify
to identify the rate of improvement or the need for close supervision of the pa-
deterioration in cognitive status and re- tient's everyday activities to protect the
sponse to treatment. For example, follow- patient from inadvertent self-neglect and
up evaluations of a head injury patient to help him or her structure and organize
help to identify the rate of improvement activities and, thus, improve the quality
with or without treatment in comparison of life.
to an initial evaluation taken a few weeks Information about the pattern of cog-
after the accident. This allows prediction nitive abilities is also used in decision
of the extent of future recovery, the na- making regarding enrollment/return to
ture and severity of residual deficits, and school to obtain advanced degree, need
the highest functional level to be achieved for special accommodations at work, edu-
upon recovery. Availability of this infor- cational and exam-taking settings, choice
mation affords realistic expectations as to of psychosocial interventions, determi-
the limits of recovery, facilitates adjust- nation of factors primarily responsible for
ment of patients and their families to patient's complaints (e.g., postconcussion
syndrome vs. somatization tendency, with the consistency in performance of

factitious disorder, symptom magnifica- high-ability subjects being higher than
tion, or malingering), and prediction of corresponding values in average- and
efficiency in vocational and daily func- low-ability subjects.
tioning. In the context of forensic crimi- 2. Test Administration: In spite of a me-
nal evaluations, this information is used ticulous description of procedures aimed
to assess the integrity of cognitive pro- at uniformity in test administration, an
cessing in claims of insanity defense and examiner's individual style and personal
in determining competence to stand trial. tempo as well as idiosyncratic features of
The neuropsychological performance a subject's test-taking style introduce
pattern can also shed light on the dif- variability in administration.
ferential contribution of more recent, 3. Scoring: Test authors define scoring cri-
transient factors vs. longstanding factors teria for each test item. At times, how-
in a patient's cognitive profile. This dis- ever, patients' responses are ambiguous
tinction is especially relevant in the fo- and present a scoring dilemma. As a re-
rensic context. sult, interrater reliability for the majority
4. Neuropsychological evaluation identifies of tests is less than perfect.
cognitive strengths and weaknesses in 4. Norms: In using norms, error can be
patient's functional status. This knowl- introduced by comparing individual
edge facilitates the selection of remedial performance to an inappropriate target
techniques and rehabilitative strategies. population, outdated norms, or norms
It guides clinicians in their choice of re- based on a low sample size.
habilitation strategy, focusing on reme- 5. Interpretation: Sources of error in test
diation of weaknesses or focusing on interpretation are outlined in the next
compensation for cognitive losses using section. In addition, caution should be
intact abilities. taken when performance on an individual
test is interpreted as being representative
of a circumscribed cognitive ability. Cog-
nitive abilities are highly interrelated, and
DIFFERENT LEVELS OF DATA
performance on any test is dependent on
INTEGRATION IN NEUROPSYCHOLOGY
the integrity of many different abilities
PRACTICE
and on the overall level of alertness.
6. Test-Taker Characteristics: Test perfor-
As suggested in Chapter I, sole reliance on mance is influenced by motivational fac-
test scores and their normative references tors, emotional status, familiarity with
would frequently result in misinterpretation of test strategy, and the cultural and lin-
a patient's cognitive profile. The high wlner- guistic background of the test-taker.
ability of such interpretation to an erroneous
outcome hinges on multiple sources of error Incorporating behavioral observations and
in different components of the testing process. qualitative performance indices in the inter-
These sources can be identified as follows: pretation and decision-making process con-
siderably improves the accuracy of attributions
I. Test Construction: Error is inherent in relating low performance scores to faulty cog-
test construction due to the fact that the nitive mechanisms. Luria's approach to test
psychometric properties of the test are interpretation heavily emphasizes the qualita-
specific to the ability level of the subjects tive aspect of performance. This direction
on which they were obtained; the values in neuropsychological practice was further
of item difficulty and item discrimination promoted by the efforts of Edith Kaplan.
are sample-specific and sensitive to cul- Her introduction of the Wechsler Adult
tural factors, and the variance of errors of Intelligence Scale-Revised (WAIS-R) as a
measurement is unequal across subjects, Neuropsychological Instrument (WAIS-R-NI)
16 BACKGROUND
(Kaplan et al., 1991) attests to the importance According to this model, the following two
of the qualitative performance indices even in levels of data integration should be considered
the context of structured batte:ry assessment. in neuropsychological practice:
This movement toward attending to perfor-
mance quality without compromising stan- I. Testing refers to the psychometric aspects
dardization of test administration procedures and addresses the quantitative appraisal
(or with minimal modification of the proce- of a patient's performance on different
dures) is also reflected in Lezak et al.~s (2004) measures. It yields a score or a set of
distinction between "optimal" and "sftzndard" scores that allow comparison with nor-
testing conditions. mative data or with a patient's own scores
Based on a comprehensive review qf recent across different tests and over time.
developments in this area, Caplan and Shech- 2. Assessment incorporates qualitative as-
ter (1995) formulated the distinction between pects of test interpretation in addition to
testing and evaluation as follows: psychometric determination of a pa-
tient's relative standing in reference to
We view the fanner as a largely mechahical en-
the normative data. It is reliant on be-
terprise that, because of its rigidity, lends tself well
havioral observations to allow better un-
to group or computer-based applicationsi Evalua-
derstanding of the nature of difficulties in
tion is, by contrast, an art applied on an ijdividual
test performance and of dysfunctional
basis that involves not only testing skills,' but also
cognitive mechanisms contributing to
professional creativity, observational ~rtise,
low test scores. The clinician integrates
flexibility, and ingenuity in the service of ~lop
various sources of information to place
ing a multidimensional understanding of Pttients-
the interpretation of a patient's psycho-
their abilities and deficits, their emotiorfd state,
metric profile in the context of his or her
self-regulatory functions, the impact of ~nviron
history and current condition. Informa-
mental variables on test perfonnance, and JIO forth.
tion is based on behavioral observations
(pp. 359-360)
and an interview with the patient, in ad-
The authors passionately advocate flexibility dition to the patient's test performance.
in testing procedures-specifically in the re- Additional information can be obtained
habilitation setting-to allow patients to from medical and school records, inter-
maximally express their potential in t'st per- views with significant others, school-
formance. teachers, nursing staff, etc.
A similar appeal to "see beyond the test
data" in offering an opinion on psych.,logical The following issues constitute the essence of
functioning in the litigation setting wa9, voiced a neuropsychological assessment: the psycho-
by Matarazzo (1990). In forensic evaloations, metric aspect of a patient's performance across
it is especially important to address numerous cognitive domains; qualitative interpretation of
sources of bias in the test data (see vail Gorp dysfunctional mechanisms; the patient's be-
& McMullen, 1997), which affect the accuracy havior and interaction with the clinician; effort/
of interpretations. Matarazzo propose4 a dis- motivation to perform on the tasks; other aspects
tinction between psychological testi~ and of mental status, including affective state; per-
psychological assessment, where the lllJter in- sonality characteristics impacting information
corporates historical information, rfledical processing; demographic information, includ-
histo:ry, and other relevant informa.on in ing educational and occupational histo:ry; med-
clinical decision making. Meyer et al. ;(2001) ical and psychiatric history; family history;
refined the distinction with an emphisis on current symptomatology, progression of symp-
usage of multiple test methods in the; latter, toms, and treatment; sources of sociaVfinancial
incorporated in the context of historic:al in- support and living conditions; motivation to
formation and behavioral observations, and improve and future plans.
addressed applications of the obtaine& infor- Neuropsychological evaluations based on
mation. sound assessment techniques, with proper
consideration of a patient's background and which were neurological exams alone or with
sources of bias, increase the accuracy of the the addition of neuroimaging results. As neu-
clinician's judgment and decision making. ropsychological evaluations are more sensitive
to mild brain dysfunction than the criterion
measures used in these studies (e.g., mild ab-
normalities identified by neuropsychological
JUDGMENT AND DECISION MAKING evaluation would not be detectable by cr or
IN CLINICAL NEUROPSYCHOLOGY
MRI scans), the moderate rating of the validity
In various applications of neuropsychological of neuropsychologists' judgments is likely to be
evaluation, the clinician makes judgments as an underestimation. To remediate this situation
to differential diagnosis, effect of remedial in future research, the authors propose con-
intervention, employability, capacity to func- struct validation strategies that would be more
tion independently in daily life, etc. How valid appropriate in the assessment of judgment
are these judgments? Different opinions have accuracy, such as a multitrait-multimethod
been expressed in the literature as to the va- matrix approach. They also advocate stronger
lidity and basis for judgments in neuropsy- emphasis on behavioral prediction to improve
chology. clinical judgment.
Limitations of judgments based solely on Overall, the utility of clinical neuropsychol-
clinical data are addressed in several publica- ogy in the clinical and forensic settings has
tions (Dawes et al., 1993, 2002; Faust, 1991; been widely recognized. However, researchers
Wedding & Faust, 1989). The authors present and clinicians have been developing stronger
a rationale for the superiority of actuarial methodological bases in both the behavioral
methods over clinical methods (superiority of and actuarial domains, to improve the accuracy
"formula over head;" Dawes, 1993) and argue of judgments and decision making. Among the
for increased utilization of actuarial methods methodological advances, we will focus on
and decision aids. Among factors impeding the those addressing test selection, norming, and
accuracy of judgments and decision making in interpretation of test scores as they are most
neuropsychology, they discuss imperfect reli- relevant to the topic of this book.
ability and validity of tests, uncertainty inher-
ent in normative information and in effects of
modifying variables on test scores, failure to
STRATEGIES IN TEST SELECTION
recognize statistical artifacts, reliance on small
samples for normative information, overreli- A variety of instruments have been developed
ance on scatter analysis, and failure to account over the years to measure cognitive abilities.
for probability estimates for a given diagnosis Tests vary in their reliance on theoretical
in making diagnostic decisions. constructs, complexity of cognitive functions
However, there is an abundance of literature assessed, length and ease of administration,
to support the validity of judgments based on psychometric properties, relevant populations,
neuropsychological evaluations in both clinical and availability of norms.
and forensic contexts. Based on data from Decisions regarding tests to use depend on
more than 125 meta-analyses on test validity, the preferred approach. The fixed battery
Meyer et al. (2001) concluded that psycholog- approach advocates administration of a com-
ical assessment is as valid as medical tests in prehensive battery of tests to all patients in
detecting neurological conditions such as de- invariant order. Any additional information is
mentia. Similarly, a review and meta-analysis of gathered after the battery is administered and
11 studies addressing reliability and validity of results are analyzed, which prevents any bias
judgments (Garb & Schramke, 1996) indicates in interpretation of the results. Clinical inter-
that judgments made by neuropsychologists pretation of the obtained data is compared to
are reliable and moderately valid. The authors information available from other sources for
questioned the appropriateness of the external consistency. An example of such a data-driven
criterion measures in the studies reviewed, approach is the Halstead-Reitan Battery. The
18 BACKGROUND
Luna-Nebraska and Benton batteries also ex- a child). Based on the pattern of weak-
emplify the fixed battery approach. nesses identified by this battery and a priori-
In contrast, the flexible approach is based generated hypotheses, additional tests might
on a patient-centered model. The choice of be administered to specifically address the
tests is guided by the hypotheses formulated extent and nature of the deficits. The flexible
by the clinician after reviewing all available battery approach is enjoying growing popu-
information about the patient. The battery is larity. According to Sweet et al.'s (2000a)
individually tailored for each patient to in- survey of the practices and beliefs of clinical
clude measures used to test a priori hypoth- neuropsychologists, endorsement of the flexi-
eses regarding possible patterns of cognitive ble battery approach increased from 54% of a
dysfunction. Luria's (1980) view of neuro- surveyed sample in 1989 to 70% in 1999.
psychological evaluation is most descriptive of In spite of the preferential use of flexible
the hypothesis-driven flexible approach. It batteries in clinical practice (Groth-Marnat,
has been further expanded in the process- 2000; Retzlaff et al., 1992; Sweet et al., 1996),
oriented assessment strategy exemplified by the relative merits of flexible vs. fixed batteries
Christensen's (1974) standardization of Luria's remain controversial (Williams, 2001). Gold-
techniques and by Kaplan's (1988) Boston stein (1997) suggested that the choice of ap-
Process Approach. proach should depend on the setting, nature of
The advantage of the fixed battery approach the disorder, theoretical approach, and spe-
is in systematic acquisition of data on a wide cific questions to be addressed by the evalua-
range of measures, allowing comparisons across tion. For further discussion of this issue, the
patients and across diagnostic groups and reader is referred to Bornstein (1990), Gold-
building extensive databases for research pur- stein (1997), and Tupper (1999).
poses. Use of this approach overcomes the
effect of base rates on test selection, thus min-
imizing the probability of error in the early
NORMATIVE REFERENCES AND
stages of clinical decision making. However,
INTERPRETATION OF CLINICAL DATA
administration of an extensive battery to all
patients, irrespective of their individual needs, Selection of the most appropriate tests alone
leads to excessive testing and uneconomical does not assure accuracy in understanding the
expenditures of resources. In addition, the patient's cognitive profile. After the tests are
accuracy of the assessment is compromised if administered and scored, the test scores need
the fixed battery does not include tests sensi- to be interpreted in reference to an appro-
tive to the deficits in specific functional do- priate set of norms.
mains suspected in a given patient. When you happen to stumble across a ref-
The flexible approach overcomes the short- erence to a normative report in the discussion
comings of the fixed approach. However, it is section of someone else's paper, you often en-
vulnerable to the effect of base rates and does counter words to the effect that the data are
not lend itself to across-patient or across- based on biased samples (because the average
group comparisons, given extensive missing FSIQ of the group is 120), with the conclusion
data due to differences in tests administered. that the data are "therefore of limited use.''
In a compromise dictated by the realities of There is no such thing as the best nom&ative
clinical practice and the economic environ- data for any test since only the clinician can
ment, many clinicians use a "flexible battery" determine what report is best applied given a
approach, where a screening battery is spe- specific patient and situation. All normative
cifically tailored for the respective diagnostic data are of limited use. The data are limited to
group or differential diagnosis in question use with patients whose demographic charac-
(e.g., differential diagnosis of dementia in an teristics are similar to those of the normative
elderly patient, assessment of cognitive deficit data sample (e.g., Heaton et al., 1986; Kalech-
pattern and rate of recovery in head injury, stein et al., 1998; Ross & Lichtenberg, 1998;
or determination of learning disabilities in Van Gorp & McMullen, 1997) and match the
administration I scoring procedures of the test the impaired range on the Controlled Oral
utilized. The inHuence of the demographic Word Association Test, when their perfor-
factors is even more apparent for neurologically mance was scored in reference to the updated
normal individuals than for those who have normative data compared to the original nor-
cerebral dysfunction (Heaton et al., 1986). mative sample.
Another consideration in choosing an ap- To choose the best set of norms for com-
propriate normative data set is its sample size. parison purposes, it is therefore essential to
The sampling distribution of the variance in a know the subject characteristics and test ad-
small sample is positively skewed, which un- ministration procedures for the normative
dermines the accuracy of estimation of stan- sample. Subject characteristics are specific
dard scores (see the next section). Although identifiers regarding the subjects under study,
there is a controversy in the field as to what such as their age, education, IQ, and gender.
constitutes a sufficient sample size to ensure Procedural variables represent details of test
precision of estimates of test psychometric administration such as whether a 30-minute
properties (see Charter, 1999; Cicchetti, 1999), vs. a 1-hour delay was followed. In general, it
a sample size of 50 is typically viewed as ade- is advised that selection of the normative data
quate (Crawford & Howell, 1998a). be based on careful review of the subject and
Normative sample, by definition, implies procedural variables employed by the nor-
that participants are not affected by illnesses mative study since the population to which the
that might lead to cognitive compromise. As- reported findings apply may be either re-
sessment of health status in many normative stricted or ambiguous. Important subject and
studies is based on a self-report, whereas few procedural variables to consider can be found
studies use medical evaluation or neuroima- in Table 2.1.
ging results to rule out neuropathology or In neuropsychological practice, once the
other conditions. Stanczak et al. (2000) showed appropriate normative data set has been
that self-report of a negative history of neuro- identified, each raw test score is compared to
pathology and psychopathology is sufficient for the distribution of performance scores on the
the pwposes of inclusion into a nonnative same test obtained by a normative sample
sample. with similar demographic characteristics.
In addition, more recently collected data With this comparison, one can determine
should be preferred to older sets of data, given whether the test score is below average, above
that the samples are comparable in other re-
spects. A notable increase in normative ex-
pectations over time has been documented in Table 2.1. Subject and Procedural Variables to
the literature. Flynn (1984, 1998) showed an Consider When Trying to Locate an Appropriate
increase of about 0.3 IQ points per year, which Normative Comparison Data Set
necessitates periodic renorming of tests mea-
Subject Varitlblea:
suring intelligence. The reasons for this rise in
IQ scores are unclear. One possibility is a Sample Composition Description
greater exposure and availability of informa- 1. Age
2. Sample size
tion over time, which results in an increase in 3. Education/IQ
the fund of knowledge possessed by an average 4. Gender
individual. According to Kanaya et al. (2003), 5. Handedness (if appropriate)
"the norms are most valid at the time the norms Procedurdl Varitlblea:
are released." This phenomenon, termed the
1. Method of administration and scoring of the test (e.g.,
Flynn effect, has not been systematically ex- if a memory test, it should be reported whether
amined in application to neuropsychological delayed recall and recognition conditions were
tests. However, evidence confirming this administered)
phenomenon is provided by Iverson et al. 2. Mean, standard deviation, range (base rate information
is preferred as well)
(1999), who showed a substantial increase in
3. Testing history (including order of testing)
the number of TBI patients scoring within
20 BACKGROUND
average, or just average relative to the nor- and the Finger Tapping Test score is ex-
mative group data (cf. Anastasi, 1988). pressed in terms of the number of taps made
For instance, knowing that a fledically in 10 seconds. The ability to convert each of
healthy 76-year-old male obtained 18 out of these various scores to a standard score equiv-
36 possible points on 3-minute delayed recall alent, regardless of the previously expressed
of the Rey-Osterrieth Complex Figure h~ lit- units of measurement (seconds, number of
tle meaning by itself because the ~w s~re holes punched, kilograms, etc.) allows deter-
conveys no information regarding the exptcted mination of a subject's relative standing in one
performance score. We have no idea;wh ther distribution and permits its comparison with
this is a good, bad, or average scole. ven relative standing in another.
knowing that 50% of the figure was reailled has The underlying assumption when using z or
little meaning because there is no w4y to dis- T scores is that the distribution of scores ob-
cern what percent recall would be $peeled. tained by the normative sample follows what is
In this example, the subject (i.e., ~edically known as the "standard normal distribution,"
healthy 76-year-old male) and procedfral (i.e., which approximates the bell-shaped normal
direct copy of drawing followed by ~-minute curve (see Chapter 3). Therefore, there is a
delayed recall without warning, witll scoring fixed relationship between the standardized
following Taylor's method) variables~ known test scores, z scores, and percentile ranks.
and used to locate an appropriate ntrmative Table 2.2 illustrates the interrelationship be-
sample. When the raw performance lscore is tween z scores, percentile ranks, and corre-
contrasted with the range of scores obtiuned by sponding WAIS-III IQ equivalents. A positive
the normative sample, one can deternline that z score will translate to a percentile rank of
a recall score of 18/36 is in the high''average 50 or greater (refer to left side of the percentile
range (80th percentile; in referenc to the rank column) and to a WAIS-III IQ of 100 or
norms reported by Boone et al., 1992). n other greater (left side of the WAIS-IJI IQ column).
words, we now know the subject's !relative A negative z score will translate to a percentile
standing compared to the normativ' group rank below 50 (use right side of column) and a
(namely, that performance is better ~ 80% WAIS-III IQ below 100 (right side of column).
of all normals who took the test). Consider the following example. You have
To more precisely judge the nature of per- just assessed Mr. Smith's right (dominant)
formance on a test relative to the r~erence hand performance on the Grooved Pegboard
normative (e.g., standard) group, the raw score Test, and you note that it took him 68 seconds
is converted to a standard score (typica1Jy a z or to complete. Mr. Smith is 35 years old and has
T score, which is expressed in terms ofstandard finished 11 years of formal schooling. He has
deviation units from the mean, see Ch~ter 3). lived almost his entire life in a large western
Such conversion permits not only de~rmina Canadian city and only recently moved to the
tion of the subject's relative standing coptpared city where you evaluated his performance.
with the normative group but also dir~t com- After surveying the available normative data
parison of scores across different tests.\ for possible comparison purposes (see Chapter
The development of "standard" measure- 23), you decide that use of Bornstein's (1985)
ment scales is especially important to~ neuro- normative data for the Grooved Pegboard per-
psychologists since test scores collecteJI while formance would be optimal. Examining the
assessing the same functional domain !are of- normative table, you note that males in his age
ten expressed in different units of ~asure and education group performed the test with
ment. For instance, when assessing motor their dominant hand in 65.3 (8.5) seconds.
functioning, the Grooved Pegboard Te$t score
is based on the number of seconds tt> com- 68-65.3
plete placing metal pegs in all the grooved (8.5) = 0.32
slots on the pegboard, the PIN Test 1core is
based on the number of holes punc~d, the Considering that higher scores on this test
Dynamometer score is expressed in kildgrams, reflect poorer performance (since it took
Table 2.2. Percentile Ranks and WAIS-III IQ Equivalent Scores for Corresponding z Scores
Percentile WAIS-III Percentile WAIS-III
Rank IQ Equiv. Rank IQ Equiv.
SD or SDor
z Score +SD -SD +SD -SD z Score +SD -SD +SD -SD
2.17-3.00 99 1 ~133 '5,67 0.63-0.65 74 26

1.96-2.16 98 2 130-132 68-70 0.60-0.62 73 27 109 91
1.82-1.95 97 3 127-129 71-72 0.57--0.59 72 28
1.70-1.81 96 4 126 73-74 0.54--0.56 71 29
1.60-1.69 95 5 124-125 75-76 0.51--0.53 70 30 108 92
1.52-1.59 94 6 123 77 0.49-0.50 69 31
1.44-1.51 93 7 122 78 0.46-0.48 68 32 107 93
1.38-1.43 92 8 121 79 0.43--0.45 67 33
1.32-1.37 91 9 120 80 OA0-0.42 66 34
1.26-1.31 90 10 119 81 0.38-0.39 65 35 106 94
1.21-1.25 89 11 0.35-0.37 64 36
1.16-1.20 88 12 118 82 0.32-0.34 63 37 105 95
1.11-1.15 87 13 117 83 0.30-0.31 62 38
1.06-1.10 86 14 116 84 0.27--0.29 61 39 104 96
1.02-1.05 85 15 0.25-0.26 60 40
0.98-1.01 84 16 115 85 0.22-0.24 59 41
0.94--0.97 83 17 0.19-0.21 58 42 103 97
0.90-0.93 82 18 114 86 0.17--0.18 57 43
0.86--0.89 81 19 113 87 0.14--0.16 56 44
0.83--0.85 80 20 0.12-0.13 55 45 102 98
0.79-0.82 79 21 112 88 0.09-0.11 54 46
0.76--0.78 78 22 0.07--0.08 53 47 101 99
0.73--0.75 77 23 111 89 0.04--0.06 52 48
0.70-0.72 76 24 0.02-0.03 51 49
0.66-0.69 75 25 110 90 0.00-0.01 50 100 100
SD, standard deviation.
longer to complete placing all those pegs in Performance across different tests, ex-
the slots), you know that Mr. Smith has per- pressed in standard scores, is frequently com-
formed below the mean for the group pared in neuropsychology practice, to identify
(z = - 0.32). Obviously, this z score will be strengths and weaknesses across cognitive do-
converted to something less than the 50th mains. This approach, traditionally viewed as
percentile. Indeed, when you locate a z score the "pattern analysis," should be used with
of - 0.32 in Table 2.2, the corresponding va- caution as the probability of obtaining abnor-
lue (using the right side of the percentile rank mal scores in an intact individual increases
column) is the 37th percentile.
Percentile ranks permit one to indicate
whether the performance is very superior,
superior, high average, average, low average, Table 2.3. Converting Percentiles to Performance
Levels
borderline, or impaired. By convention, neu-
ropsychologists use the percentile cutoffs Percentile Level
presented in Table 2.3 in describing perfor-
~98 Very superior
mance levels. However, it should be noted that 91-97 Superior
these cutoffs may vary depending on psycho- 75-90 High average
metric properties of a given test. Therefore, 25-74 Average
whenever possible, the clinician is advised to 9-24 Low average
use the cutoffs for performance levels that are 2-8 Borderline
<2 Impaired
provided by test authors.
22 BACKGROUND
with the number of tests administered. Chris- in the diagnostic norms approach, the
tensen et al. (1999) and Schretlen et al. (2003) relative probability of impairment for a
reported large intraindividual variability in given test score increases as age in-
neurologically intact elderly samples, with 27% creases, whereas in the comparative
of the latter sample producing discrepancy norms approach it decreases with in-
values exceeding 3 standard deviations, which creasing age.
might explain the increase in probability of An important contribution of base
abnormally large deviations from the baseline rate information to the accuracy of dif-
performance level for an individual patient. ferential diagnosis and detection of ma-
lingering in test performance is further
underscored by Elwood (1993), Gouvier
(1999, 2001), Gouvier et al. (1998, 2002),
ALTERNATIVE METHODS FOR
King et al. (1998), Labarge et al. (2003),
INTERPRETATION OF CLINICAL DATA
Rosenfeld et al. (2000), Slick et al.
1. Sliwinski et al. (1997, 2003) refer to the (2001), and Woods et al. (2003).
conventional approach, described above, 2. Another limitation of the commonly used
as "comparative" and point out that cor- age-corrected norms is that data are
rection for age has an undesirable con- grouped into age intervals. No change in
sequence: the same proportion of the test performance is assumed within the
population falls below any impairment interval, with an abrupt shift in scores
cutoff score for all ages. Using probable between intervals. For example, in a nor-
Alzheimer's disease (pAD) as an example, mative data set partitioned by age decade,
they argue that this approach takes into performance of 61- and 69-year-old indi-
account only test performance of intact viduals would be referenced to the same
persons and the relationship between test normative expectations, whereas 69- and
performance and age, disregarding the 70-year-olds would be referenced to dif-
strong associations between age and the ferent normative expectations. As an in-
prevalence of pAD and between age and dividual passes from one age band to
the probability of having pAD. another, a notable shift in the interpreta-
In contrast, the authors advocate tions of the same test scores occurs.
norms that rely on statistical methods To remediate this problem, several
that model the probability of having pAD regression-based techniques have been
as a function of test performance and developed. Gorsuch (1983) developed a
age. They propose a "diagnostic norms" continuous norming method for gener-
approach, which takes into consideration ating continuously adjusted age norms,
the performance of the clinical group and using an analytic smoothing procedure.
prevalence rates (base rates) of the dis- Calculation of such continuous norms is
ease for the relevant setting/demographics based on a few simple equations. Russell
of the patient in addition to the infor- used this technique in renorming the
mation used by the comparative norms. Wechsler Memory Scale and in devel-
Diagnostic norms are derived using val- oping norms for the Halstead-Russell
ues of sensitivity and specificity for all Neuropsychological Evaluation System
possible cutoff scores (which can be (HRNES and HRNES-R) (Russell, 1987,
displayed on the receiver operating 1988; Russell and Starkey, 1993, 2001).
characteristic [ROC] plot). Using this Zachary and Gorsuch (1985) and Taylor
information, the probability of impair- (1998b) applied the continuous norming
ment for a certain test score can be method to calculate WAIS-R indices.
computed for different prevalence rates. A different regression-based norming
Since an increase in the prevalence of system was used by Heaton et al. (1991,
the disease is taken into consideration 2004) in their Comprehensive Norms for
an Expanded Holstead-Reitan Neuro- Ivnik et al. (2000) demonstrated the

psychological Battery and its revised usefulness of cutoff scores in interpreting
edition. data from the Mayo Cognitive Factor
Crawford and Howell (1998b) advocate Scales. The utility of cutoff scores was
use of the regression approach over con- further emphasized by Soukup et al.
ventional normative data with a single (1998), who recommended reporting
predictor (e.g., age) as well as with mul- cutoff scores that represent borderline
tiple predictors and describe two methods (15th percentile) and defective (<5th
for determining differences between the percentile) performances in addition to
test score predicted by a regression equa- the means and SDs, to offset problems
tion and an individual's obtained score. A associated with the skew in test scores.
large discrepancy between predicted and Recent literature suggests growing inter-
obtained scores suggests a cognitive defi- est among neuropsychologists in examin-
cit. They also address application of the ing the outcomes of neuropsychological
regression method to the assessment of services using alternative methods to tra-
change over time, where a regression ditional hypothesis testing, such as odds
equation can be generated to predict level ratios and relative risk analysis, as well as
of performance at retest based on the in- measures of diagnostic usefulness such as
dividual's score at initial testing. The au- predictive values and likelihood ratios (see
thors point out that such an approach Chapter 3), which are based on a diag-
simultaneously factors in the effects of nostic dichotomy that places the perfor-
practice and regression to the mean. mance of an individual patient above or
Conversely, Fastenau (1998) and below the cutoff scores (Bieliauskas et al.,
Fastenau and Adams (1996) criticize 1997; Cerhan et al., 2002; Cicerone &:
regression-based norms, specifically in Azulay, 2002; Haddock et al., 1998; Ivnik
reference to Heaton et al.'s (1991) com- et al., 2000, 2001).
prehensive norms. They question the Several studies address the utility of
validity of the regression-based norms cutoffs in determining criteria for ab-
when sample sizes in each demographic normality in test performance across
cell are low, especially when scores are multiple measures (Cicerone &: Azulay,
not normally distributed, variances are 2002; Drebing et al., 1994; Ingraham &:
not uniform across the range of each Aiken, 1996). For example, Drebing et al.
demographic variable, and demographic (1994) compared classification accuracy
relationships are not linear across the of the following criteria: one, two, or
sample used to derive the norms. three scores in a battery falling below 1,
3. Cutoff scores have long been used in 1.5, and 2 SDs.
testing practice. In neuropsychology, this In the context of signal detection
method has been used most widely in theory, selection of cutoff scores is based
defining cutoffs for impairment in the on the tradeoff between sensitivity and
context of Halstead-Reitan Battery. Cut- specificity (more precisely, between
offs for classification into normal and true-positive and false-positive rates of
clinical groups, as well as for differenti- classification), as reflected in the ROC
ation between normal and impaired plot. This method has been used for
functional range, have been used in a determining the optimal cutoff scores for
number of tests and interpretive ap- diagnostic classification (i.e., contrasting
proaches. Some tests use multiple cutoffs normal vs. clinical range) in neuropsy-
to further refine the differentiation be- chology. Typically, the score with the
tween performance levels (e.g., normal, highest combination of sensitivity and
mildly impaired, moderately impaired, specificity values is selected as a cutoff. A
severely impaired). more elaborate method of determining
24 BACKGROUND
a cutoff involves computation of the area overlapping interval strategies described

under the ROC curve, which ~presents by Pauker (1988) to maximize the sample
the most useful index of diagQostic ac- size of the normative distribution at each
curacy (Swets, 1996). The score associ- midpoint age interval (Ivnik et al., 1996;
ated with the largest area under the Smith&: Ivnik, 2003). Ten-year age bands
curve is the most sensitive cu~g score. were staggered at successive 3-year mid-
In spite of their popularity, t}l.e use of point intervals. For example, all partici-
cutoff scores has been criticized for im- pants between 62 and 71 years of age
posing an artificial dichotomy cin a con- contributed to the 67 year midpoint in-
tinuous distribution and for s~bjective terval, whereas part of this sample also
judgment involved in the sel~on of contributed to the 65-74 year age band
cutoff points (Dwyer, 1996). It poes not with the midpoint of 70 years and to the
allow consideration of a spectnfn of re- 68-77 year age band with the midpoint of
lated disorders among diagnostc possi- 73 years.
bilities. According to a nu.ber of The data were co-normed on the same
studies, use of a single cutoff fc;r a spe- normative cohort for a large battery of
cific test, without conside~ other tests. The raw score distribution for each
clinical information and dem.,graphic test at each midpoint age was normalized
factors, results in a large nu~ber of by assigning standard scores with a mean
false-positive misclassifications. Asimilar of 10 and SD of 3, based on actual per-
effect on classification accuracy is ren- centile ranks. Formulas based on linear
dered by the failure to account for the regressions were generated for each test
base rates (see Chapter 3) of ~e crite- in the battery to be applied to the nor-
rion condition in the normative sample. malized, age-corrected MOANS Sc~ed
In addition, an ROC curve ytelds re- Scores to adjust for education.
liable cutoff scores for diagnosl classi- Duff et al. (2003) used the overlapp-
fications only when an external riterion ing midpoint age interval technique with
measure provides a reliable bas' for di- 5-year midpoint intervals to report nor-
agnosis in the clinical group. Conditions mative data for the Repeatable Battery
resulting in subtle cognitive dysfunction for the Assessment of Neuropsychologi-
frequently do not have a reliable external cal Status (RBANS; Randolph, 1998).
diagnostic criterion, which undermines Age-corrected scaled scores were further
the accuracy of classification. converted into education-corrected scaled
Some investigators suggest that many scores using the same method across four
of the current cutoffs are too conserva- education levels.
tive (Fromm-Auch &: Yeudall.~ 1983), In spite of the complexity of the
thereby generating too many false nega- procedures used in these studies for der-
tives. However, their work has be,n done ivation of the normative data, these tech-
primarily with highly educated, '-igh-IQ niques hold great promise as they allow
samples; and of course, cutoffs based on maximization of the sample size for each
average performers would generate a age interval, "smoothening" of the tran-
high false-negative rate. Unfortunately, a sition in normative expectations as the
large number of studies docum.nt un- patient passes from one age group to
acceptably high false-positive ~classi another, and direct comparisons between
fication rates, placing normal Sllbjects various tests as they are co-normed on
into impaired ranges across dtfferent the same sample.
tests which are interpreted using i cutoff 5. Crawford and colleagues proposed a
criterion (see chapters on Halst~d-Re single-case approach, where an individ-
itan Battery tests in this book). ual is treated as a sample of n = 1.
4. The authors of the Mayo Older ~erican Crawford and Howell (1998a) described
Normative Studies (MOANS) used the a modified t-test method for interpreting
an individual's test score when the nor- respective confidence limits using the
mative sample is small (n <50). Accord- single-case approach.
ing to this method, the t score represents 6. As discussed earlier in this chapter, both
an estimate of the rarity or abnormality of fixed and flexible battery approaches
the individual's test score. The authors have their merits and shortcomings. One
argue that the t-test procedure is more of the weaknesses of the flexible battery
appropriate than the standard z scores approach is an increase in the number of
for any comparison of an individual scores falling below the expected range
against a normative sample because the as the number of tests included in the
norms are derived from samples rather battery increases. The fixed battery ap-
than from populations. Therefore, nor- proach allows one to minimize the im-
mative data are treated by the authors as pact of this problem by introducing
sample statistics, rather than population global indices, such as the Halstead-
parameters. Whereas with large or Reitan Average Impairment Rating
modest sample sizes, this distinction has (Russell et al., 1970), Halstead Impair-
minimal effect on sampling distribution ment Index (Reitan & Wolfson, 1985),
and the difference between the values of Global Neuropsychological Deficit Scale
t and z is trivial, in small samples t rep- (Reitan & Wolfson, 1988), as well as
resents a more accurate estimate of the summary indices derived through factor-
rarity of an observation as the sampling analytic or correlational studies.
distribution of the variance in a small Rohling's Interpretive Method (RIM)
sample is positively skewed (leading to (Miller & Rohling, 2001; Rohling et al.,
underestimation of variance) and, there- 2003a,b) is designed to generate global
fore, z scores are likely to be over- indices (summary statistics) in the con-
estimated. text of the flexible battery approach. This
This approach was expanded by the method produces summary results anal-
authors to provide point estimates of ogous to those generated in fixed battery
the abnormality of differences between approaches, yet it offers the advantage of
scores achieved by an individual on two allowing the clinician to choose varied
tests (Crawford et al., 1998a) or between test batteries depending on the needs of
a patient's mean score on several tests a particular case.
and the score on a specific test con- The authors describe 24 steps out-
tributing to the mean (Crawford & lining the sequence of the procedures
Garthwaite, 2002). underlying RIM, including steps com-
Introduction of an intraindividual monly used in neuropsychology practice
measure of association (liMA) as an es- (e.g., test administration and conversion
timate of abnormality (Crawford et al., of test scores to a common metric), as
2003) represents a further advancement well as additional statistical procedures
in the single-case approach. The liMA is leading to generation and graphical
expressed as a parametric or nonpara- representation of 14 summary statistics.
metric correlation coefficient or the Additional steps used to integrate and
slope of a regression line. The estimate interpret the obtained statistics are based
of abnormality of an individual score is on psychometric procedures yet require
based on comparison of the magnitude the subjective judgment of the clinician.
of the index of association for a specific The results can be viewed at the global
patient with corresponding statistics for level, the domain level, or the test mea-
normative or control samples. sure level.
The authors underscore that only The authors point out that clinical
summary statistics (SDs or measures of judgment involved in decision making in
association between tests) are needed the context of RIM is facilitated by sta-
to derive estimates of abnormality and tistical techniques (e.g., those assessing
26 BACKGROUND
impact of heterogeneity, sa~ple size, content of knowledge available to that

or unequal effect sizes). They view the individual. This property of the IRT
RIM as a statistical interpretfle model overcomes the above-mentioned weak-
and point to several weaknesses of this ness of the norm-referenced approach by
method (e.g., different tests art normed relating a test score to what the person
on different populations). Ne~rtheless, can do.
they assert that the method i$ "a solid As IRT represents model-based mea-
step toward increasing the statif;tical and surement, two or more scales measuring
methodological strength of interpretive different abilities can be psychometri-
arguments" (Miller & Roj1g, 2001, cally matched to assure equivalent levels
p. 168). of reliability throughout the ability con-
7. Another approach which alloWs one to tinuum. This is achieved by selecting
tailor the psychometric pro~rties of items which fit a specified model for
measures to specific measure~nt goals scale construction.
is represented by the item [response Another important distinction be-
theory (IRT). It challenges me~~ement tween the two theories is related to the
principles underlying classi~rst the- values of the standard error of measure-
ory, which is most commonly, used in ment (SEM). According to classical test
neuropsychology. theory, the SEM is constant across scores
In classical test theory, interpretation in a particular population but differs be-
of test performance is based on) a norm- tween populations. Therefore, the SEM
referenced standard, where ~e raw value reported in the manual for a par-
score obtained on a test is : linearly ticular test is applied across the entire
transformed into a standard ~core to range of test scores. However, in the
determine a relative position 1of that context of a test battery, different tests
score in comparison to the n~rmative are associated with different SEMs as
distribution. Embretson (1996); pointed estimates of reliability and variability
out an objection to the norm-referenced (which underlie calculations of the SEM)
approach in that scores have no meaning differ across tests/populations. Since in-
for what the person actually cari do. terpretations of test performance profiles
In IRT, the meaning of a scote is ref- are based on an assumption of equal
erenced to item difficulty. A rell4:ionship SEMs across different tests, inequality of
between ability level, which is ~Hected SEMs undermines the accuracy of in-
in a test score, and probability of a cor- terpretation.
rect response to a given item is summa- In the context of the IRT, the SEM
rized in the item characteristk curve increases with a departure from the
(ICC). The ICC is determined tby item mean of the score distribution but gen-
difficulty and discrimination. It~ diffi- eralizes across populations. This increase
culty is the ability level at w~ch the in the SEM is a function of the distri-
probability of failing that item is ~qual to bution of item difficulty as the number of
the probability of passing. (If the indi- items that appropriately measure a given
vidual's ability level exceeds th. item's ability level decreases toward the ex-
difficulty level, the individual wpuld be tremes of the score distribution. Conse-
likely to pass that item.) Item ~crimi quently, confidence intervals increase
nation is manifested in the slope of the and precision of measurement decreases
ICC at the level of item difficttlty and toward extremes of distributions, whereas
reHects the precision of measuref1ent. If the pattern of these changes is invariant
the items are structured by co*ent in across populations. In application to
order of increasing complexity, ~f,sing a neuropsychological test performance,
certain item by an individual c~ _he in- the magnitude of the SEM can be re-
terpreted from the perspective . of the presented as a composite value, which is
the mean of individual SEM values distorted profile of strengths and weaknesses.
across ability levels. Several factors influencing test performance
The IRT methods are applicable at the need to he considered beyond applying sta-
scale, test, or item level. Some of the IRT tistical methods to data interpretation. Some
models are applied by neuropsychology are discussed below.
researchers and clinicians, such as the
Rasch model, which is used in the anal-
Effort and Motivation
ysis of dichotomous items, rating scales,
partial credit scoring, and multidimen- A unique, rapidly evolving role for neu-
sional models (see Embretson, 1996, for ropsychological assessment is the identification
review). and documentation of noncredible test per-
Mungas et al. (2003) used IRT meth- formance in the context of malingering, somato-
ods to derive psychometrically matched form disorder, or other conditions in which the
measures of global cognition, memory, patient is motivated, either consciously or
and executive function from commonly nonconsciously, to present him- or herself as
used neuropsychological tests based on more cognitively impaired than is actually the
a sample of 400 elderly participants, case. If we cannot verify that patients are per-
who ranged in cognitive functioning forming with adequate effort, our evaluations,
from normal to demented. The authors no matter how lengthy or elegantly written, are
demonstrated utility of this method worthless. Clinical neuropsychology has the
for "creating psychometrically matched objective methods to determine whether cog-
measures of clinically relevant domains nitive complaints secondary to claimed brain
for use across a broad range of ability" damage and/or psychiatric dysfunction are
(p. 386). accurate or fabrications. Numerous well-
validated cognitive effort tests are now avail-
The majority of studies reviewed in this able, as are indices derived from standard
book report common descriptive statistics, neuropsychological tests found to be highly
such as means, SDs, or percentiles. Cutoff sensitive to feigned performance, which pos-
scores or other criteria for abnormality were sess sensitivity and specificity values in most
reported in a few studies. The regression ap- cases superior to those of standard cognitive
proach was used by us to derive predicted tests used to discriminate clinical disorders
scores for selected tests based on meta- from normal function. It is beyond the scope
analyses of available data sets (see Chapter 3). and mission of this book to include these tests
Although the emphasis in our literature re- and indices, but we will stress that, as with
views is placed on conventional descriptive measurement of traditional cognitive domains,
statistics, we urge clinicians and researchers to assessment of effort requires administration of
consider the merits of more advanced tech- more than one test or index, preferably inter-
niques. In spite of the psychometric sophisti- spersed throughout the test battery, tapping
cation of many of the above measures, their use the veracity of performance on differing types
in clinical practice can be facilitated by com- of cognitive skills, such as memory, motor
puter software available from their authors. function, processing speed, math, problem-
solving/abstraction, etc. While most neuropsy-
chologists realize that effort must be assessed in
forensic settings, it has become apparent that
FACTORS INFLUENCING
effort should be regularly assessed in clinical
PERFORMANCE ON
evaluations; we have found that when effort
NEUROPSYCHOLOGICAL TESTS
measures are routinely administered, evidence
When interpreting performance on neuro- of poor effort can emerge in unexpected situ-
psychological tests, sole reliance on the actu- ations (e.g., an "Alzheimer's patient" in a clin-
arial data might lead to misinterpretation of ical drug trial who was later determined to have
the patient's level of functioning and to a a factitious disorder rather than dementia).
28 BACKGROUND
Assessment of effort performs the impor- difficult in other ethnic groups, such as Asians
tant gatekeeper function of insuring that pa- (Wong, 2000) and Middle Easterners (Es-
tients receive compensation only for actual candell, 2002), where there is large diversity
disabilities and that neuropsychological as- in the languages or dialects and cultures.
sessments are not misused by individuals Another methodological problem in adapting
perpetrating fraud. In the case of somatofonnl English-standardized tests into other languages
conversion disorder patients, identification of involves the validity of such measures when
noncredible cognitive complaints can steer used with other cultural groups (Ponton &
treatment strategies away from reinforcing the Ardila, 1999, Puente & Ardila, 2000; Nell,
medical patient role to addressing the under- 2000). Simple language translations may fail to
lying issues/concerns driving the symptom take into account the impact of familiarity and
creation. relevance of the test items in different ethnic
groups (Escandell, 2002; Puente & Ardila,
2000; Wong, 2000), possibly compromising the
Issues in Cross-Cultural and Multicultural
validity of the results. Similarly, once a neu-
Neuropsychological Assessment
ropsychological test has been translated into
Issues of ethnic diversity pose difficult and another language, it may no longer measure the
serious challenges for the field of neuropsy- same cognitive functions it was once thought to
chology, particularly as clinicians are more measure in its standard form. For example,
frequently being requested to assess func- Escandell (2002) points out that according to a
tioning in patients from varied ethnic, cultural, study by Loewenstein et al. (1995), a pattern of
socioeconomic, and linguistic backgrounds correlations for a measure of daily functioning,
(see Ferraro et al., 2002; Fletcher-Janzen et al., the Direct Assessment of Functional Status
2000; Nell, 2000; for review). Practitioners are (OAFS) test, was different when it was trans-
faced with a number of problems when con- lated into Spanish relative to its original English
ducting such assessments. One problem may version. Similarly, Puente and Ardila suggest
be finding standardized neuropsychological that tests such as Digit Span in the WAIS or
instruments in languages other than English. WISC may require other cognitive processes
Alternatively, if the patient speaks English, the when administered in Spanish than in English
problem may be finding ethnicity-specific since naming the digits requires a different
normative data that take into account issues number of syllables.
such as culture and bilingualism. Some of the While developing cross-cultural tests can be
critical issues in using these approaches in challenging, when specific procedures and
cross-cultural and multicultural neuropsycho- guidelines are used, adequate outcomes can
logical assessment will briefly be discussed be achieved. Along with other approaches,
below. Puente and Ardila (2000) recommend using
Due to the lack of availability of neuro- Brislin's (1983) three-step procedure when
psychological instruments in other languages, translating tests, particularly for the Hispanic
clinicians often have to translate tests or find population. These steps include the initial
translated versions of standardized tests in the translation, back translation, and resolving dif-
literature. However, Puente and Ardila (2000) ferences between the original version and the
describe a number of methodological prob- resulting translated version (for a more detailed
lems with such an approach. These authors discussion of this approach, see Puente &
point out that "translation and adaptation re- Ardila, 2000). Additionally, translation and test
quire much time and expertise." For example, development for specific ethnic groups re-
translation of neuropsychological tests from quire a thorough understanding of the group's
English to Spanish is quite complex given that culture and a familiarity with the language.
there are a number of Hispanic subgroups Another issue is that normative data need to
that use varied idioms and expressions of the take into account various cultural factors in
language. The issue of test translation is just as addition to the usual demographic factors.
METHODOlOGICAl CONCEPTS IN NEUROPSYCHOlOGY 29
Even when standardized tests can be admin- administered word-list learning test (i.e., Ca-
istered in English, we lack understanding of lifornia Verbal Learning Test). Use of verbal
the effects of biculturalism and bilingualism learning and memory strategies, such as se-
on test performance in various ethnic groups mantic clustering, were found to be related to
(Ardila et al., 2002). Manly and colleagues level of language proficiency.
found that in a sample of African Americans Culture-specific and cultural/language ad-
the level of acculturation and the use of "black aptations of specific neuropsychological tests
English" accounted for significant proportions have been developed and, when available,
of variability in various neuropsychological test have been discussed in the individual chapters
scores, particularly those that relied on verbal of this book. Additionally, Artiola i Fortuny
skills (Manly et al., 1998), and that reading et al. (1999) developed the Bateria Neu-
level, not educational attainment, attenuated ropsychologica en Espanol, a standardized and
the difference in most neuropsychological test validated battery of neuropsychological tests
scores of older African Americans and Cau- culturally adapted for Spanish-speaking indi-
casians (Manly et al., 2002). viduals. Normative data based on 390 partici-
Several large-scale normative projects for pants, which were collected in Spain, Mexico,
African-American individuals have been con- and the United States, are stratified by geo-
ducted in response to the pressing need to graphical area x age x education. The battery
generate normative data specifically for this includes adaptations of commonly used neu-
ethnic group: the Consortium to Establish a ropsychological tests, with a visual memory
Registry for Alzheimer's Disease (Fillenbaum task, 16-word list-learning task, story learning
et al., 1997; Unverzagt et al., 1996; Welsh et al., task, letter Huency (P, M, R), attentional tasks
1995), the Washington-Heights-Inwood- (e.g., digit repetition), Stroop test, and Wis-
Columbia Aging Project (Manly et al., 1998), consin Card Sorting Test.
and the San Diego African American Norms Ostrosky-Solis et al. (1997) developed the
Project (Diehr et al., 1998; Gladsjo et al., 1999). NEUROPSI, a short neuropsychological test
In their most recently updated normative battery for Spanish-speaking adults. The bat-
manual, Heaton et al. (2004) present T-score tery includes tests of orientation (time, place,
conversions separately for groups of African person), attention (e.g., digits backwards),
Americans and Caucasians adjusted for age verbal and nonverbal recall and memory, lan-
and education level. These are among the most guage (e.g., naming, repetition, comprehen-
comprehensive norms presented for African sion, verbal Huency), concept formation (e.g.,
Americans to date. similarities), and motor skills (e.g., alternating
For Hispanics, Gasquoine (2001) suggests movements). Expanded normative data, strat-
that bilingualism and acculturation factors ified by four age groups and three educa-
should be used to determine whether English- tion levels, for this battery are presented in
or Spanish-language tests are to be adminis- Ostrosky-Solis et al. (1999).
tered. Harris et al. (1995) examined the It is clear that culturally sensitive tests that
cognitive performance of three groups: those are valid in use with individuals from various
proficient in both English and Spanish, those cultural and linguistic backgrounds need to
more proficient in Spanish, and non-Hispanic continue to be developed. Additionally, nor-
Caucasian monolingual English-speaking mative data for existing standardized tests
adults. They found a small difference between need to take factors of ethnicity, acculturation,
Spanish-dominant and monolingual English and bilingualism into account if valid inter-
speakers on a nonverbal reasoning task but pretations are to be made for different ethnic
noted that this difference was probably not groups. It is also clear that adequate train-
clinically significant. However, they did find ing in cross-cultural and multicultural neuro-
that degree of bilingualism represented a psychological issues is needed. Fastenau et al.
significant variable in the learning and reten- (2002) propose one such model of multi-
tion of verbal information on a commonly cultural training that may be used at the
30 BACKGROUND
graduate, internship, and postdoctoral training simplistic scenario because decisions of this
levels. magnitude should be based on multiple data
points. However, the example does illustrate a
clinical problem which, while raised by Ax-
elrod and Goldman in 1996, has apparently
FINAL CAVEATS
received no further attention or discussion by
A major focus of this book is to assist the the field of clinical neuropsychology, and as a
clinician to identify and select the most result, no consensus has emerged as to how to
demographically appropriate normative data address the dilemma.
for specific patients, but the issue of demo- One partial remedy is that all neu-
graphic adjustment is not as simple as it might ropsychological reports should indicate for a
appear and some caveats need to be raised. given score which demographic corrections
While neuropsychologists have come to ap- were applied, as in the following example:
preciate the impact of demographic variables
Trails B performance was borderline impaired (4th
on test performance, there has been little or no
percentile for age and education; Seines et al.,
dialog as to how best to use and communicate
1991).
this information for the best interest of specific
patients. By way of example, let us examine the It might also be preferable at times to com-
case of a 58-year-old male physician who ob- pare patient test scores against the general
tains a score of 80" on Trails B. This score population and demographic peers:
would be in the average range for his age (43rd
Dr. X scored within the average range on Trails B,
percentile for men aged 40--59 with 12 or more
when compared to the general population of
years of education; Bomstein, 1985), but to
58-year-olds (43rd percentile for age; Bornstein,
report this performance as average would
1985), but within the borderline impaired range as
overlook the probable very real decline in
compared to individuals of the same educational
function for this individual with 20 years of
level (4th percentile for age and education; Seines
education. To illustrate the decline, the clini-
et al., 1991).
cian would need to use norms which adjust for
the high level of education, with the result that Assuming that other executive skills showed a
the Trails B score drops to the borderline im- similar pattern of performance, the report
paired range (4th percentile for men aged summary might conclude:
45-60 with more than college education;
While executive/problem-solving skills were intact
Seines et al., 1991). However, a treating pro-
for the patient's age, they were significantly low-
fessional who receives the neuropsychological
ered (i.e., borderline impaired) as compared to
report and sees a description of "borderline
individuals of the same educational level.
impaired" performance might be tempted to
limit the patient's control over his finances, his The consumer of the report would then
access to driving, etc.-an unwarranted inter- be able to use the appropriate interpretation
vention and one that would do the patient a when making decisions regarding patient
grave disservice, given that he is in fact average management. Specifically, if the issue is fi-
relative to the "typical" 58-year-old. Con- nancial decision making, in which the general
versely, to omit the demographic correction population of 58-year-olds can successfully
and report the patient's performance as aver- engage, then the health-care professional
age would be of concern if a decision needed would probably be correct in allowing the pa-
to be made regarding whether the patient was tient to make financial decisions. However, if
capable of returning to work as a surgeon; if the issue is return to work as a surgeon, which
the patient is borderline impaired relative to requires a finely developed skill executed only
other 58-year-olds with 20 years of education, by a highly educated subset of 58-year-olds,
continued surgical practice might in fact place then the treating professional would be wise to
future patients in danger. Admittedly this is a be concerned and should take steps to insure
that, at the least, the patient's professional dependently). As a field, it will be critical
activities are closely monitored. for clinical neuropsychology to develop algo-
A second issue in the clinical application rithms, perhaps based on quantification of
of demographic corrections has to do with school records and psychosocial data, which
whether the relationship between demo- allow for determination as to whether a
graphic variables and test scores is unidirec- patient's low educational level is primarily
tional or bidirectional. In the case of age and due to low cognitive capability vs. environ-
gender, both can impact test scores and, by mental factors, with these data then used to
extrapolation, cognitive function, but test determine whether education corrections to
scores I cognitive function do not affect age test data should be applied.
and gender; thus, these variables have a Corrections for ethnicity may be similarly
straightforward unidirectional relationship distorting. Ethnic differences in test perfor-
with cognitive scores, and it is appropriate to mance appear to be primarily driven by cul-
correct test scores for these factors when they tural inftuences, such as level of acculturation
account for test score variance. However, ed- to the predominant Anglo-American culture,
ucation leads to a more complicated situation. English Huency, quality of education, and
Educational level can impact test scores and, socioeconomic factors that affect cognitive
by inference, cognitive function, but cognitive development, and not race or ethnicity per
function can also impact educational level; i.e., se. Unfortunately, our lack of understanding
individuals of low cognitive ability do not of such factors combined with the focus on
typically complete much schooling. There are ethnic/racial differences in test performance
two groups of individuals with low education: has often led to faulty assertions regarding the
those who did not have the opportunity to cognitive capabilities of some ethnic minority
complete much schooling but who would have groups. Ultimately (and hopefully in future
benefited from it and those who discontinued editions of this book), adjustments to test
schooling because they could not comprehend scores will be made for factors such as accul-
and learn the material and would not have turation, not ethnicity.
profited from additional education. It is ap-
propriate to correct for educational level in
Data Inclusion in Neuropsychological
the former group but not the latter group.
Reports
This issue has very important real-world
ramifications in that individuals of mentally The writing of psychological reports, includ-
retarded intelligence and low education might ing neuropsychological reports, has gradually
be "corrected" out of their lowered cognitive evolved to include more actual scores and data.
scores when lowered educational level is fac- Those of us who attended graduate school in
tored into test interpretation, thereby poten- the 1970s and 1980s were frequently instructed
tially losing benefits to which they are entitled to summarize psychological results in a narra-
(Regional Center resources, disability com- tive format and specifically admonished not to
pensation) or, even more seriously, losing life include actual test scores because these would
if the restrictions against capital punishment be misused/misinterpreted by the reader and
for mentally retarded individuals no longer patients would be harmed. However, one sus-
apply to them. On the other hand, to fail to pects that a secondary motive of at least some
correct for education in individuals with nor- report writers for not including scores was that
mal premorbid cognition who did not have the many interpretations were not in fact empiri-
opportunity to complete at least high school cally based; if no raw scores were reported, the
would portray them as more cognitively im- interpretations could not be questioned.
paired than is accurate, and as a result, they More recent pressures for research-
might lose privileges to which they are enti- validated treatment approaches and account-
tled and for which they are competent (e.g., ability has led to a greater rigorousness
managing financial affairs, driving, living in- within the subspecialty of psychological testing.
32 BACKGROUND
Practitioners who are confident in their neu- It is of interest that although the most
ropsychological skills and knowledge are not commonly stated reason for omitting actual
reticent to provide raw test data and to reveal test scores from reports is protection of
how interpretations were derived. Matarazzo the patient from misinterpretation of results
(1995) asserts that, in fact, the inclusion of test by nonpsychologists, no empirical data have
scores serves to minimize and clarify any in- emerged in the decades of psychological
terpretation biases or idiosyncrasies on the testing showing any harmful effects caused by
part of the writer. The inclusion of raw test inclusion of scores in test reports. In fact,
scores in reports is also critical for comparing Freides (1995) and Matarazzo (1995) assert
the results of initial and subsequent neu- that there is a greater potential for harm from
ropsychological evaluations. Further, access to interpretations of scores without the scores
some services (e.g., Regional Center resources than from scores without interpretations. The
for individuals with extremely low intelli- interested reader is directed to more thorough
gence) and some criminal sentencing deci- discussions of this topic by Freides (1993,
sions (e.g., ineligibility for the death penalty in 1995), Pieniadz and Kelland (2001), Donders
mentally retarded individuals) require the (2001), Matarazzo (1995), and Naugle and
reporting of actual scores. McSweeny (1995, 1996).
Pieniadz and Kelland (2001), summarizing We recommend that neuropsychological
the results of a survey completed in the 1990s reports contain (1) all raw scores and per-
by 81 directors of neuropsychology training centiles, standard scores, and/or T scores; (2)
programs, indicated that only 35% of respon- the normative studies used to derive percen-
dents routinely appended test scores to re- tiles if other than the published test manual;
ports. The reasons most commonly given for and (3) which demographic factors were ad-
including numerical data were "thoroughness" justed for in each test. Reporting of raw
(100%) and "facilitation of comparison" of test scores is important for various reasons. For
records (96%). Of those who did not append example, a superior normative data set may
actual test results, 80% indicated that their emerge after an initial neuropsychological
decision was based on a desire to avoid mis- assessment; on retesting, the examiner might
interpretation by unqualified persons. In con- want to score both sets of test scores ac-
trast, Donders (2001), summarizing the results cording to the more recent norms, which
of a survey completed by 414 U.S. members of would not be possible if the initial raw scores
American Psychological Association Division were not available. Further, inclusion of the
40, revealed that 88% of respondents included raw scores enables the reader to check that
numeric data in their reports, although raw the scores were in fact converted and inter-
scores were provided less frequently. preted properly.
3
Statistical and Psychometric Issues
The administration, scoring, and interpreta- rehabilitation, evaluating a patient's ability to

tion of neuropsychological tests are major function independently in an everyday envi-
sources of information used in the clinical ronment, choosing an academic field and
practice of a neuropsychologist to make de- professional career, making diagnostic differen-
cisions about patients' cognitive status, diag- tiation between disorders that affect cognitive
nosis, prognosis, and treatment. However, functioning, assessing the rate of improve-
accurate decisions based on test results cannot ment or deterioration in functional capacities,
be made without a clear understanding of the and making prognostic predictions.
issues related to the measurement of psycho- The concept of measurement implies nu-
logical phenomena and the statistical proper- merical representation of certain properties.
ties of the tests. This chapter reviews basic Such physical properties as dimensions, inten-
statistical concepts of importance to neuro- sity, speed, and gravity, which represent a core
psychologists. No intent was made to provide of scientific exploration, lend themselves to ac-
a comprehensive review of statistics. The goal curate and reliable measurement. In contrast
of this chapter is to help a novice understand to direct measurement of physical phenom-
and interpret psychometric data. ena, psychological attributes such as cognitive
abilities, personality traits, and emotional sta-
tus cannot be measured directly. To assess
these psychological constructs, we need to ob-
MEASUREMENT AND INTERPRETATION
tain a sample of behavior that can be quanti-
OF NUMERICAL VALUES fied and represented in numerical scores.
Measuring abilities and traits is an inherent Well-validated psychological tests are designed
part of clinical work. It facilitates decision to elicit behaviors that are representative of
making by relating the performance of a given the underlying psychological constructs.
individual to an appropriate reference group Numerical values derived from an individ-
or by uncovering a change in an individual's ual's performance on a test are identified as
behavior over time. The nature of decision raw scores and may represent the number of
making is specific to a given situation and incorrect responses, time required for comple-
cludes a wide range of decisions, such as iden- tion of the test, number of errors, rating of the
tifying cognitive strengths and weaknesses, quality of a drawing, or different combinations
choosing an appropriate course of cognitive of the above criteria.
33
34 BACKGROUND
In contrast to physical measuring scales that normative or standardization sample is

have an absolute 0 point, scaling ofpsychological assumed to be representative of the population
measures does not start at the point of "no from which it is drawn and is used as an ex-
ability at all"-i.e., a raw score of 0 on an ternal standard of perfonnance for interpre-
arithmetic test does not indicate that the pa- tation of individual scores. There are several
tient has no ability to solve arithmetic prob- methods of relating individual performance to
lems. If the test included basic operations of the norms:
addition or subtraction of single digits, the a. Raw scores obtained on a test can be
patient would be likely to succeed on these converted to age or grade equivalents,
items. As a result, we cannot infer that a patient which allow interpretation of a partic-
who received a raw score of 50 on a test is ular score in the context of expected
twice as good at arithmetic as someone with a performance for a specific age or grade
score of 25. Due to the lack of the absolute 0 level. This method is highly useful in
point in psychological measurements, ratios of assessing the developmental standing
scores are meaningless and most psychological of children in comparison to their
tests are scored on an interval scale. Despite peers, provided that a considerable
some disadvantages of the interval scale in and well-identified increment in abil-
comparison to the ratio scale, both of these ity with age and grade advancement
scales provide measurements that lend them- is expected. However, this method
selves to advanced statistical analyses. loses its effectiveness when the rate
The raw score obtained on a test says little of development becomes uneven and
about an individual's level of ability or maste:ry the relationship between levels of
of the subject. To interpret the raw score, it ability and developmental markers
should be related to the content of the test weakens.
or compared to the performance of a group b. Measures of relative standing of indi-
of individuals on the same test. Various vidual scores within a distribution
interpretive strategies are outlined in Chap- provide an alternative method of eval-
ter 2; however, the statistical and psychomet- uating individual performance. Percen-
ric issues related to the most frequently tile rank (PR) reftects the percentage
used interpretation strategies are discussed of the standardization sample that
below. scored lower than the individual score
Test performance interpretation can be (plus one-half of that portion of the
based on different reference criteria: standardization sample who achieved
the same score as the individual being
1. Domain- or content-referenced interpre- assessed). The PR is useful for pro-
tation indicates how proficient an individual is viding the relative standing of a score;
in the domain tapped by the task presented however, it indicates only the ordinal
on the test. Content maste:ry is usually reposition of the individual score within
ported as a percentage of correct responses the distribution. It does not show dis-
on the test. persion of the remainder of the dis-
2. Criterion-referenced interpretation re- tribution below that score and does
lates an individual's performance on a test to not indicate the absolute amount of
an external criterion measure, such as a difference between scores. For exam-
practical situation which requires skills as- ple, percentile transformations mag-
sessed by the test. For example, expectancy nify differences between individuals
tables tie different levels of test performance close to the center of the distribution
to expected practical outcomes. and compress the differences at the
3. Norm-referenced interpretation com- extremes.
pares scores achieved by one individual to the Let us consider a distribution of
performance of a respective group of individ- scores (A) representing the number of
uals who have similar characteristics. This words recalled on the fifth trial of the
STATISTICAL AND PSYCHOMETRIC ISSUES 35
Rey Auditory-Verbal Learning Test STANDARDIZATION OF RAW SCORES

(RAVLT):
Compare the distribution of RAVLT scores
(A) used in the above example with an-
A= (5, 7, 8, 8, 9, 9, 9, 9, 10, 10, 10, 10,
other distribution of scores on this test (B),
10, 11, 11, 11, 11, 12, 12, 13, 15)
both of which range from 5 to 15 with a mean
of 10:
There are 21 observations in this dis-
tribution. An individual who obtained
A= (5, 7, 8, 8, 9, 9, 9, 9, 10, 10, 10, 10,
a score of 9 performed better than
10, 11, 11, 11, 11, 12, 12, 13, 15)
four individuals who received scores
B = (5, 5, 5, 5, 6, 6, 6, 7, 7, 8, 10, 12, 13,
lower than 9, plus two out of four in-
13, 14, 14, 14, 15, 15, 15, 15)
dividuals (0.5) who achieved a score of
9. Relating this proportion of six in-
Visual examination of these distributions sug-
dividuals to the total number of 21
gests that the variability around the mean is
observations yields a PR of 29, as fol-
much greater for distribution B. Therefore, a
lows from the formula:
score of 7 would indicate very poor perfor-
mance relative to distribution A and a much
PR(9) = 4 + (0.5 ) (4) x 100 = 29 better performance relative to distribution B.
21
To account for the degree of variability in the
Similar calculations of PR for scores 8 normative distribution, individual measure-
and 7 in the above distribution yield ments are converted into z scores.
14 and 5, respectively: In another example, assessing recall of an
individual on the fifth trial of the RAVLT, we
use a reference sample with a mean of 10 (see
PR(s) = 2 + (0.5 )(2) x 100 = 14 graphs below). If the individual recalled seven
21
words, comparison of the raw score with the
and mean for the reference sample (X- M) sug-
gests that this individual's recall was three
1
PRm = - x 100 = 5 words below the expected score of 10 for his
21
or her age. However, this does not tell us how
low his or her performance was relative to the
As follows from the above calculations,
distribution of the normative sample.
the difference in PR between scores 9
If a high degree of variability is expected
and 8 (29 - 14 = 15) is greater than the
and the standard deviation (SD) for the ref-
difference between scores 8 and 7
erence sample is 6 (graph A), then a score of
(14- 5 = 9). This example illustrates
7 falls halfway between a mean of 10 for the
the main disadvantage ofPRs: whereas
reference sample and a score of 4 represent-
they reflect the position of an individ-
ing 1 SD below the mean, which results in a
ual score relative to the standardiza-
z score of - 0.5.
tion sample, they do not indicate the
absolute differences between scores.
-1SD X M
c. To accommodate the absolute differ-
ences between scores, interpretation of A) ~--+--+--+--+--+--+--+--+--+--+--+--+--+-1
a raw score should be based on the rel- 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
ative standing of the score with respect Number of words
to the mean for the distribution and the
variability of the scores within the dis- In reference to another sample with the
tribution. This can be accomplished same mean but a much lower degree of vari-
through converting a raw score into a ability reflected in an SD of 1 (graph B), a score
standard score. The most frequently of 7 lies 3 SDs below the mean (z = -3).
used standard scores are z and T scores. Therefore, the recall of seven words indicates
36 BACKGROUND
much poorer performance in relation to dis- In spite of the obvious advantages of using z
tribution B than in relation to distribution A. scores over raw scores, some of the properties
of z scores are viewed as undesirable: (1) z
X-1SDM scores have fractional values, which are car-
B) ~--+--+--+--+--+--+--+--+--+--+-+--+--+--~ ried to at least one decimal place; (2) half of
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 the z scores in the standardized distribution
Number of words are negative and half are positive, which leads
to the zero-sum problem (i.e., corresponding
values on both sides of the distribution cancel
Thus, to account for the variability within
each other when totaled).
the nonnative distribution, raw scores are
Parameter values of the standard distribu-
standardized, i.e., converted into z scores that
tion are arbitrarily designated. Therefore, they
relate the difference between an individual
can be easily changed through simple arith-
score and the group mean (X- M) to the SD
metic transfonnations of z scores. T-score
for the reference group:
transfonnations overcome these disadvantages
through multiplying z scores by 10 (thus elim-
z=--
X-M
inating fractional values) and adding a constant
SD
of 50 (which eliminates negative values and
A negative z score indicates that the raw places all the scores on a scale of0--100 with a
score lies below the mean for the reference mean of 50 and SD of 10):
group, a positive z score represents higher
performance than the mean for the group, T= 10z+50
and a z score of 0 indicates that the raw
score is equal to the mean of the reference For example, a z score of - 1.6 can be ex-
group. 1 pressed in T scores as follows:
The z score (SD units) shows not only how
much an individual perfonnance deviates T=(-16) +50=34
from the mean of the sample but also how
likely it is that other individuals in the sample An example of a test which uses T-score
would achieve scores as high or as low as the conversion is the Minnesota Multiphasic
person being tested. Personality Inventory (MMPI) and its recent
Standardization of raw scores, e.g., their revision. Clinically significant elevations on
conversion into z scores, allows comparison of the scales are judged relative to a mean of
the relative standing of individuals across dif- 50 and an SD of 10, which equates the scale of
ferent tests in spite of the differences in the measurement across all validity and clinical
measurement scales or the means and SDs for scales on this test.
these tests. A standardized distribution of
z scores has a mean of 0 and an SD of 1 be-
cause the mean is subtracted from each score STANDARD SCORES AND NORMAL
and the result is divided by the SD. It pre- DISTRIBUTION
serves the same shape as the distribution of
the raw scores from which it was derived. Many biological measures and human charac-
Therefore, differences in standard scores are teristics are distributed so that the highest fre-
proportional to the differences in the corre- quency of scores is observed around the
sponding raw scores. distribution mean, with a gradual decrease in
the frequency further away from the mean,
1For those tests that measure performance in terms of time which eventually tails off on both sides. Score
or number of errors, where the higher scores reHect lower distributions of many psychological tests ap-
performance, z scores represent an inverse of the obtained
score. Mathematically, in these cases the numerator proximate this model, which in its ideal hypo-
should be multiplied by -1, i.e.-(X- M). thetical fonn represents a normal distribution. It
is convenient to treat test score distributions as if distribution of scores in the population from
they were normally distributed because the which the sample was drawn.
properties of this model are known: This assumption of normality of the test score
distribution allows it to be converted into a
1. The distribution of hypothetical score distribution of z scores with a mean of 0 and an
frequencies arranged from the lowest to SD of 1, which represents a standard normal
the highest values is bell-shaped and distribution. Use of this conversion facilitates
symmetrical; i.e., the left and right sides interpretation of the test scores because it al-
are mirror images of each other. lows comparison of a variety of otherwise not
2. The frequency is highest in the middle comparable distributions through equating
of the distribution; therefore, the mean, their means and SDs. The proportion of cases
median, and mode have the same values comprising a certain area under the curve be-
and divide the distribution into two equal tween two points along the z axis is known,
parts. which permits conversion of z-score units
3. The normal distribution stretches from into percentiles. For example, it is known
minus to plus infinity; thus, the "tails" of that 34.13% of all scores lie between z=O
the distribution get closer and closer to and z = + 1. Since the mean of the distribution
the x axis as they get farther away from (z = 0) divides the distribution in half, we know
the mean but never touch the x axis. that 50% of all scores lie below the mean. Thus,
4. The normal distribution is described by adding 34.13% of scores above the mean to
a specific mathematical formula. 50% of scores below the mean suggests that the
84.13th percentile corresponds to z = + 1.
Although the test score distributions do not Figure 3.1 illustrates the corresponding
perfectly match this model, if the number of conversion values for selected z scores. The
cases were increased and smaller class inter- proportion of scores (i.e., the area under the
vals were used, the shape of the sample standard normal curve) for each value along
distribution would become relatively smooth the z axis can be easily determined using ta-
and symmetrical, thus approximating the bles provided in any basic statistics textbook.
Standard -3 -2 -I Mean +I +2 +3
deviations
Pen:cntile ranks s 20 so 80 9S 99
zscores -3 -2 -I 0 +I +2 +3
Tscores 20 30 40 so 60 70 80
figure 3.1. Illustration of the relationship between proportion of scores {represented by the area under the
normal curve), percentile ranks, z scores, and T scores.
38 BACKGROUND
INTERPRETATION OF INFREQUENT g. Emotional factors and level of endur-

(OUTLYING) SCORES ance (fatigue)
h. Response style and response bias
As follows from Figure 3.1, 68.26% of all
i. Motivational factors
scores fall within 1 SD from the mean in both
j. Previous exposure to similar tests
directions, 95.44% fall within 2 SDs, and al-
(practice effect)
most all scores except for 0.0026% are in-
3. Outlying scores may result from an ex-
cluded between - 3 and + 3 SDs from the
ecution error. In this case, an individual score
mean.
is foreign to the distribution used for com-
This correspondence between the propor-
parison. For example, an elderly individual of
tion of cases and z-score values is important
low average ability might appear impaired
for interpretation of individual test perfor-
when compared to a normative sample of
mance since such interpretation is based on
highly functioning, independently living, rel-
the relative frequency of the score obtained
atively healthy elderly individuals. This bias in
by the individual being assessed with respect
the normative sample, which is not repre-
to the distribution of scores. For example, a
sentative of all functional and economic levels
test score falling outside the range of 2 SDs
of the population for the respective age
above or below the distribution mean is highly
group, would result in an inflated mean and in
infrequent; only 4.56% of the scores deviate
upward "slippage" of the entire distribution.
that far from the mean in both directions.
To avoid execution errors, a clinician should
Therefore, individuals obtaining these infre-
be highly sensitive to the appropriateness of
quent scores may be viewed as outliers. The
the norms used for each individual being
decision criteria for defining scores as outlying
evaluated.
might vary from more conservative to more
liberal in different clinical situations, de-
pending on the cost-benefit ratio of making
false-positive vs. false-negative errors. INTERPRETATION OF SCORES
Outlying scores can have different origins: THAT ARE NOT
NORMALLY DISTRIBUTED
1. They might be due to the inherent Interpretation of individual test scores with
variability in the population. Indeed, in any respect to the normative distribution is based
population, innate levels of a trait or ability on an assumption of normality of this distri-
range from very low to very high, which is bution. To avoid interpretive errors, the basis
modeled by a normal distribution. Therefore, for test score interpretation should be differ-
a certain proportion of extreme scores is a ent if distribution is asymmetrical. Standard-
natural feature of the population. ized distribution has the same shape as the
2. There might be purely deterministic original distribution of test scores, which is
reasons accounting for too low or too high highly dependent on the characteristics of the
scores, such as the following: individual items comprising the test.
a. Inadequate reliability of the measur- If a test is designed in such a way that the
ing instrument majority of individuals can succeed on most of
b. Variations in test administration and the items, test scores are compressed into a
scoring strategy few discrete values at the upper extreme of the
c. Errors in data recording or in calcu- score range, with only a few observations at
lating appropriate statistics the lower part of the score range. In this case,
d. Demographic factors and physical the distribution is negatively skewed and the
handicaps affecting performance variability of scores falling within or above the
e. Situational factors (e.g., external normal range is highly limited. The test has its
noise} highest discriminative power at the lower
f. Sensitization and anxiety associated ability levels; i.e., it is most useful in identify-
with the testing situation ing impaired individuals. For example, the
distribution of scores on the Boston Nam- cific ability. However, a certain degree of
ing Test and the Rey-Osterrieth Complex variability is inherent in test performance. It is
Figure (copy condition) in a sample of highly due to transient factors associated with the
functioning individuals would acquire this testing situation and the patient's state at the
shape. time of testing.
When test items present difficulty for the Thus, the score on a test reflects the con-
majority of subjects, the score distribution is tribution of the following two factors:
positively skewed. Variability within the lower
range of scores is highly limited, whereas the X=T+e
highest sensitivity is obtained in the upper
part of the distribution. Such a test would be where X is the score on a test, T is a "true"
most appropriate for the selection of a few score representing the actual level of ability
outstanding individuals from a large group. measured by a test, and e is an error of mea-
The distribution of scores for Raven's Ad- surement reflecting random variability.
vanced Progressive Matrices can be used as an With an increase in test reliability, a con-
example. siderable proportion of the variability in test
In both of the above cases resulting in scores is due to differences between subjects
skewed score distributions, use of z-score con- in the "true" scores. In other words, reliability
versions is inappropriate since such conver- can be expressed as the proportion of variance
sions are based on the assumption of normality in test scores which is accounted for by the
(particularly symmeby) of the distribution. "true" differences between subjects on the
ability being measured. Therefore, a reliability
coefficient provides a measure of test reli-
ability representing the ratio of "true" score
PSYCHOMETRIC PROPERTIES OF TESTS
variance to the total variance of the test
In view of the proliferation of psychological scores:
tests with a high overlap in terms of abilities
assessed, we are frequently faced with a di-
lemma: which test to select in a particular
situation. All published tests have to meet the
requirements outlined by the Standards for
Educational and Psychological Testing (Amer- Methods of Estimating Test Reliability
ican Psychological Association, 1999). Yet, the Several procedures have been developed to
choice of a test should be made in the context determine the reliability of a test by measur-
of a certain clinical situation. In choosing an ing the proportion of "true" variance vs. the
appropriate test, one has to keep in mind proportion of "error" variance. Different
several criteria for evaluating its psychometric methods define measurement error with re-
properties. spect to different sources of error. The four
most common methods are described below:
Reliability
1. The test-retest method assesses the
When measuring a certain aspect of func- consistency of test scores from one test
tioning, our main assumption is that the scores administration to the next. It is measured
on a particular test would be consistent over as the correlation between the scores
repeated administrations. If an individual on the first test and the retest and re-
sometimes receives high scores and some- flects the stability of scores over time.
times low scores on the same test, no infer- 2. The alternate forms method assesses the
ences can be made regarding the level of correlation between scores obtained by
ability being measured. In other words, we the same subject on alternate forms of a
have to be assured that the test is a reliable test. This method is the closest approxi-
measure of a stable construct such as a spe- mation to the parallel tests model.
40 BACKGROUND
3. The split-half method involves splitting Typical levels of reliability attained by

the test into two equivalent halves after a neuropsychological tests range from 0.95 to
single administration. There are different 0.80, which represents a high to moderate
ways of splitting a test. The hi~est com- range of reliability. For a test with a reliability
parability of the two halves is aQhieved by estimate of 0.80, 20% of the variability in
an odd~en split in which one form scores is due to measurement error. Thus,
contains all odd-numbered itef$s and the tests with reliability below 0.80 introduce a
other form, all even-numbered_etems. considerable proportion of "noise" in scores,
4. The internal consistency metJfWd esti- which compromises their interpretability. For
mates the reliability of a test bafd on the screening tests, reliability between 0.80 and
number of items and the aver~ed inter- 0.60 would be acceptable, whereas reliability
correlations among them. This ptethod is estimates below 0.60 are usually judged as
mathematically related to the! split-half unacceptably low.
method. Coefficient oc is the moJt general
form of this method and repr+ents the Standard Error of Measurement
mean reliability coefficient obt$ed from The reliability estimate provides a relative
all possible split-half comparjsons. In measure of the accuracy of test scores. As any
essence, internal consistency ~stimates correlation, it is influenced by the variability
compare each item on a test' to every of scores. In a sample with a heterogeneous
other item. · score distribution, reliability will be higher
. than in a more homogeneous sample.
There is no universally agreed best method The reliability estimate does not indicate
to evaluate test reliability. Each me~d has its how much variability should be expected due
advantages and disadvantages. The ;split-half to measurement error and how accurate the
reliability method overcomes theore!tical and individual test scores are. Therefore, in addi-
practical problems associated with the test- tion to reporting reliability coefficients, test
retest and alternate forms methodsl such as developers report the size of the standard
difficulty in developing two equivalent forms error of measurement (SEM), which is useful
of a test, carry-over effects, reactivity effects, in interpreting the observed scores of each
and the effect of random variability on two individual patient. The SEM is determined by
test probes. However, the reliability estimate the reliability of the test (rxx) and the vari-
obtained by the split-half method varies de- ability of test scores (ax):
pending on the arbitrarily chosen method of
splitting. In addition, the split-half reliability SEM=a,Jl-rxx
coefficient underestimates the reliability of
the full test and requires the use of • correc- Since no test provides a perfect measure of
tion formula. · ability, a certain degree of variability in the
The level of reliability varies for different scores obtained by the same subject is ex-
tests. Ideally, a highly reliable test would be pected. The SEM indicates how much an in-
preferred to a test with low reliability. How- dividual's score might vary if he or she is
ever, many practical considerations might in- retested repeatedly with the same test (as-
fluence a clinician's test selection. The cost of suming that there is no practice effect or fa-
error in a decision-making situation iS, another tigue effect). According to measurement
factor which needs to be considered in se- theory, the scores obtained by one subject
lecting an appropriate test for the gi•en situ- across an infinite number of retests with the
ation. Test reliability should be high when a same test would result in a normal distribution,
patient's test performance is considered as with the mean equal to this subject's "true"
one of the factors in making a final diagnos- score and the SD equal to the SEM.
tic determination. Tests with lower reliability Since in most clinical situations we obtain
might be acceptable in preliminary screening only one score on a test, we may treat it as an
situations. estimate of the theoretical "true" score. Using
STATISTICAl AND PSYCHOMETRIC ISSUES 41
the SEM, we can form a confidence interval are a representative sample of the abil-
(CI) around this score, which provides a range ity being measured. It is not measured
in which a subject's "true" score is likely to fall. statistically but determined by agree-
For example, a 70-year-old patient obtained ment among expert judges with respect
a WAIS-111 Full Scale IQ (FSIQ) of 110. to a detailed description of the content
According to the test manual, the size of the domain that is measured by each test
SEM for FSIQ in this age group is 2.19 item.
(Wechsler, 1997). Since we know that 95% of 2. Construct validity determines how well
all scores in a normal distribution fall within observable behaviors measured by the
1.96 SD of the mean, the 95% CI for this test represent underlying theoretical
score will fall between 1.96 SEM below and construct. This relationship can be es-
1.96 SEM above the obtained IQ score tablished through high correlation of
(110 ± 1.96 SEM). Calculation of a CI by mul- the test with other tests measuring the
tiplying the SEM by 1.96 (2.19 x 1.96=4.29) same construct (convergent validity) or
suggests that we would expect 95% of the IQ through low correlation with tests mea-
scores obtained by this patient to fall in the suring different constructs (discriminant
range of 110 ± 4.29, or between 105.71 and validity). If more than one method is
114.29. used to measure several constructs, the
If we want to increase the level of certainty correlations among them can be re-
in constructing a range in which a patient's presented in a multitrait-multimethod
true score is likely to fall, we can use the 99% matrix, which establishes whether the
Cl. In other cases, we might use lax Cis that results of a certain test are determined
provide accuracy below the 95% level. by the construct being measured or by
The drawback in using SEMs to determine the method of measurement. Construct
the accuracy of test scores is the fact that they validity can be further assessed by cor-
do not have the same size for all scores: they are relating one test with many other tests
smaller for extreme scores and larger for using factor analysis. In this case, con-
moderate scores. Another limitation in using struct validity is established through the
SEMs is that test scores are generally further high loading of a particular test on the
away from the mean than "true" scores be- factors that represent those constructs
cause of the tendency for regression toward the presumed to be measured by the test.
mean. To overcome this distortion, the CI can Thus, content validity and construct
be formed around the estimated "true" score, validity represent two different strate-
which is obtained from a regression equation. gies in determining that the test mea-
sures what it is supposed to measure:
"Content validity is established if a test
Validity looks like a valid measure; construct va-
When a test is used to assess a certain aspect of lidity is established if a test acts like a
functioning, it is assumed that the test mea- valid measure" (Murphy & Davidshofer,
sures what it is supposed to measure and that 1991).
it is useful in making accurate decisions. Dif- 3. The usefulness of a test in decision-
ferent validation strategies are used to under- making situations represents another as-
stand the meaning and implications of the pect of test validity. Criterion-related
scores achieved on the test. Content validity validity reflects the relationship between
and construct validity indicate whether a test is test scores and measures of decision
a valid measure of a specific ability. Criterion- outcome, i.e., criteria. Any measurable
related validity refers to the accuracy of deci- behavior can be used as a criterion. For
sions that are based on the test scores. example, the choice of a rehabilitation
strategy can be evaluated using a mea-
1. Content validity reflects the extent to sure of symptom reduction as a criterion,
which the behaviors sampled by the test or the accuracy of a screening test can be
42 BACKGROUND
assessed using a patient's psychiatric di- Decision Theory

agnosis as a criterion. The correlation In clinical practice, a clinician has to make
between test scores and the criterion decisions which range from assigning a certain
measure, which is derived without using diagnosis to applying a specific course of
the test, reflects the accuracy of predic- treatment. Since predictions based on the in-
tions or decisions made on the basis of formation available to the clinician are never
the test scores. perfect, each decision may have several pos-
Criterion measures can be obtained sible outcomes. In the context of decision
after decisions are made based on the theory, the predictor and criterion values are
test scores in a random sample of a reduced to only two categories, in spite of the
population about which decisions are continuous nature of these values. Compari-
made (predictive validity) or at the same son of predictions with criterion values sug-
time that decisions are made in a pre- gests that there are four possible outcomes
selected sample (concurrent validity). of decisions: correct decisions include true-
Whereas predictive validity is superior to positive (TP) and true-negative (TN) out-
concurrent validity in that it is a direct comes, whereas incorrect decisions include
measure of the relationship between test false-positive (FP) and false-negative (FN)
scores and a criterion measure for the outcomes. Tests are used to maximize the
general population, it has a number of number of correct decisions and to minimize
practical and ethical drawbacks. For this the number of errors. The contribution of the
reason, the most practical and commonly criterion-related validity of a test to improve-
used measure of criterion-related valid- ment in the accuracy of decisions depends on
ity is concurrent validity, despite the fact the base rate and selection ratio.
that its coefficient underestimates the
predictive validity.
Base rates
Theoretically, an estimate of the
The base rate reflects the proportion of an
correlation between test scores and a
unselected population who meet the criterion
criterion measure obtained in a criteri-
standard. Clinically, this term is used inter-
on-related validity study can range be-
changeably with incidence or prevalence of a
tween -1 and +I. Validity coefficients
disorder. In a hypothetical example, assume
for most of the tests are relatively low,
that among 500 normal elderly, 9% would
ranging between 0.2 and 0.5. This is due
have scores below the cutoff for dementia of
to the imperfect reliability of the test and
24 on the Mini-Mental State Exam (MMSE).
the criterion measure: whereas a crite-
Assume that 80% of patients with the diag-
rion is assumed to represent the "true
nosis of dementia of Alzheimer's type (DAT)
state" of a patient, it is frequently based
score below a cutoff of 24. If 100 DAT pa-
on subjective clinical judgment, which is
tients were added to 500 intact elderly (total
inherently unreliable.
number of subjects 100 + 500 = 600), the base
If the correlation coefficient between
rate would be 100/600, or 17%. In this situa-
a test and a criterion is 0.3, the pro-
tion, the outcomes would be as follows:
portion of the variance in the criterion
that is accounted for by the test (?', or
100 x 80%=80 TP
coefficient of determination) is 0.09.
This means that only 9% of the vari- 500 X 9%=45 FP
ability in the criterion can be accounted
500-45=455 TN
for by the test scores. Although these
numbers look discouraging, they should 100-80=20 FN
be interpreted in the context of other
measures that contribute to the accu- Thus, follow-up of subjects who score be-
racy of decisions. low the cutoff (80 TP + 45 FP = 125) will yield
a hit rate of 80/125 = 64%. In other words, the Selection ratio

diagnosis of DAT will be confirmed in 64% of Another factor affecting the accuracy of de-
those subjects who scored below the cutoff of cisions is the selection ratio, which is defined
24 on this test. as the ratio of TP + FP outcomes to the total
In contrast to the above hypothetical ex- number of subjects. Assume that a psychiatric
ample, in the general population, the base rate ward has 30 beds for severely depressed pa-
for DAT is considerably lower than 17%. For tients. If only 32 patients are referred for
example, if the base rate for DAT is 5%, then hospitalization at any one time, the selection
out of 600 subjects, 30 would be suffering ratio would be high (30/32 = 0.94). The hos-
from DAT and 570 would be intact with re- pital cannot be very selective in this situation,
spect to this diagnosis. Assuming that 80% of and most of the referred patients could be
DAT patients and 9% of the intact elderly hospitalized. In another scenario, 100 patients
score below the cutoff on the MMSE, as was could be referred for hospitalization at any
the case in the above example, the table of one time due to severe depression. Since the
outcomes would look different because of the selection ratio is low (30/100 = 0.30), a certain
lower base rate for DAT: strategy needs to be used to identify those
who are acutely suicidal for immediate hos-
30 X 80%=24 TP pitalization. When a selection ratio is low, a
test with even modest validity can make a
570 x 9%=51 FP considerable contribution to the accuracy of
570-51=519 TN decisions.
30-24=6 FN Incremental validity

The utility of a test has to be assessed in terms
The ratio of TP scores to the total number of an increase in the accuracy of decisions
of subjects scoring below the cutoff (24 obtained using the test, which extends beyond
TP +51 FP = 75) will yield a hit rate of 241 the base rate or beyond information obtained
75 = 32%, which is considerably lower than in from other sources. In other words, incre-
the example with a higher base rate. mental validity reHects the unique contribu-
With a decrease in base rates, most of the tion of a test to understanding the patient.
population are negatives; positives become Incremental validity is affected by the base
more rare, and therefore, an attempt to iden- rate, selection ratio, and criterion-related va-
tify this group will lead to an increase in the lidity of the test. When decisions are made at
number of FP decisions. Low base rates also random, the frequency of different outcomes
lead to a large number of TN decisions since a can be computed directly from the base rate
majority of the population do not suffer from and the selection ratio. The incremental va-
DAT. Following the same logic, in the case of lidity of a test indicates the degree of im-
high base rates, as the number ofTP decisions provement in the accuracy of decisions, i.e.,
increases, the frequency of FN errors also in- frequency of TP and TN outcomes, beyond
creases. An optimal base rate of about 50% the random level, which are made using the
minimizes decision errors and maximizes ac- test.
curate decisions, providing that the test used to The incremental validity is highest when the
assist in decision making has sufficient validity. base rate is moderate, selection ratio is low,
In the general population, the base rates for and criterion-related validity is high. Values of
certain disorders are usually low and most of incremental validity for different combinations
the "red Hags" represent false alarms. Base of base rates, selection ratios, and criterion-
rates among individuals referred for evalua- related validity coefficients are provided in
tion due to progressive symptomatology are Taylor-Russell (1939) tables.
higher, and therefore, the expected number of Thus, the validity coefficient alone does not
false alarms would be lower. determine the usefulness of a test in each
44 BACKGROUND
clinical situation. Test usefulness depends Thus, manipulation of the cutoff affects the
largely on the context in which the test is used. balance between sensitivity and specificity
and results in different cost-benefit ratios.
Cutoffi and diagnostic accuracy of a test Based on the empirical evidence, the cutoff is
or interpretive strategy usually set at a value that ensures a reason-
As pointed out above, in the framework of a able balance between sensitivity and speci-
decision theory approach, both the predictor ficity so that only "borderline" patients will
(test) and criterion values are reduced to only likely be misidentified. Setting the optimal
two outcomes. Thus, the continuous nature of cutoff yields the highest Hit Rate, i.e., ability
test scores is reduced to categories of pass I of the test to correctly identify the presence
fail, impaired/unimpaired, etc. Selection of a and absence of impairment (expressed as the
cutoff point dividing a sequence of test scores ratio of [TP +TN] to all individuals in the
into these two categories is another factor that sample [TP + FP + FN +TN]).
affects the accuracy of decisions. Through In making a diagnostic decision, the clinician
manipulating the cutoff, the frequency of a is concerned with the utility of a test in cor-
certain type of correct decision can be maxi- rectly identifying impairment in an individual
mized at the expense of increasing the fre- patient, i.e., in the test's predictive value, ra-
quency of another type of error. ther than in its accuracy in discriminating
For example, test sensitivity, or the ability between groups. Positive Predictive Value
to correctly identify impaired individuals (ex- represents the probability that the patient is
pressed as the ratio of TP to all impaired in- indeed impaired, given an impaired test score
dividuals [TP + FN]), can be increased by (expressed as the ratio of TP to all individuals
fixing the cutoff at a small number of incorrect identified by the test as impaired [TP + FP]).
responses. This will reduce the frequency of Negative Predictive Value represents the prob-
FN errors but increase the proportion of FP ability that the patient is intact given a non-
errors. In other words, this will assure correct impaired test score (expressed as the ratio of
identification of the majority of individuals TN to all individuals identified by the test as
with even mild impairment and very few non-impaired [FN +TN]).
misidentifications of impaired individuals as The probability of the condition based on
being intact. At the same time, this will yield a the test result (predictive value) is referred to
large number of intact individuals who will be as the posttest probability. However, the use-
misidentified as impaired. The costs of such fulness of a test in aiding diagnostic decisions is
misidentification include inappropriate treat- also determined by the base rates (prevalence)
ment, psychological distress, and adverse social! of the condition in a given setting (see above),
economic consequences. which represents the pretest probability. These
On the other hand, test specificity, or the probabilities can be converted into odds of
ability to correctly identify the absence of having the condition, which are expressed as
impairment (expressed as the ratio of TN to all the ratio of the probability of having the con-
intact individuals [TN+ FP]), can be in- dition to (!-probability of having the condi-
creased by setting the cutoff at a large number tion). Posttest odds (which represent the
of incorrect responses. This will reduce the likelihood that the individual who obtained a
proportion of FP errors but result in a large score X on the test has the condition) take into
number of FN errors. In other words, only account the pretest odds and likelihood ratio:
those patients who have pronounced impair-
ment will be identified as impaired, and very Pretest odds x Likelihood ratio = Posttest odds
few intact individuals will be misidentified.
However, many individuals with mild symp- where the likelihood ratio represents the odds
tomatology will be missed. This will preclude of a specific test result occurring in an indi-
timely therapeutic intervention which other- vidual who has a condition over the odds of
wise would allow stabilization or reversal of that test result occurring in an individual who
these patients' symptomatology. does not have the condition. In other words, it
represents the likelihood that, given a score X Withering's summary review on the use of
on the test, an individual is impaired vs. un- digitalis for treatment of heart disease, pub-
impaired. Likelihood ratios relate the speci- lished in 1785, is one of the first examples of a
ficity and sensitivity of a test to a given setting. systematic review.
They can be defined over the range of possi- Statistical methods to combine data from
ble values, identifying a degree of abnormal- different studies in medicine were first in-
ity rather than representing the presence/ troduced in 1904 by British mathematician
absence dichotomy, and therefore allow con- Karl Pearson. These early efforts spearheaded
sideration of the greater predictive power of the development of statistical methods to
scores at the extremes of a distribution. Like- synthesize research findings in social sciences,
lihood ratios are particularly useful in deter- particularly in psychology and educational
mining the probability of having a condition research. In 1976, the American psychologist
based on the results of a test battery, rather Gene Glass coined the term "meta-analysis"
than an individual test score. to describe research synthesis based on sta-
An alternative to tests of statistical signifi- tistical techniques. In the 1980s, meta-analysis
cance between groups is the Odds Ratio sta- became popular in medicine, particularly
tistic, which reflects the association between for summarizing results of clinical trials ad-
the incidence of a condition given specific dressing the effectiveness of treatment tech-
situational factors versus incidence of that niques and of observational (epidemiological)
condition in the absence of those factors and studies examining the accuracy of diagnostic
approximates relative risk estimates when the methods.
incidence of the condition is low. Conversely, In response to a pressing concern regarding
it measures the strength of dissociation be- the lack of summary reviews for those who
tween individuals with different test results need to use evidence from unmanageable
and reflects the probability that the condition amounts of information to make informed
is present in an individual with an abnormal decisions in medicine, British physician and
test result. In other words, it shows that an epidemiologist Archie Cochrane founded
individual who obtained an abnormal score the Cochrane Collaboration in the 1990s. This
(falling below the cutoff for impairment) on a international network of health-care profes-
specific test is X times more likely to have the sionals promotes accessibility of systematic
condition than an individual who scores in the reviews through maintaining registers of con-
nonimpaired range (above the cutoff). trolled trials and preparing/updating system-
For further discussion of the utility of di- atic reviews, which are published in the
agnostic tests, see Fletcher et al. (1996) and Cochrane Library available on the internet
Sackett et al. (2000). (Antes & Oxman, 2001; Egger et al., 2002).
To achieve a consensus across disciplines on
how to report the results of systematic reviews,
the conference on the Quality of Reporting of
SYNTHESIS OF RESULTS OF DIFFERENT Meta-Analyses (QUOROM) was held in 1999,
STUDIES IN A META-ANALYSIS bringing together clinical epidemiologists, cli-
nicians, statisticians, and researchers from the
Historical Overview and the Rationale for
United Kingdom and North America. As a
Using Meta-Analysis in This Book
result of this conference, the QUOROM
Historically, researchers and clinicians strug- statement was published, which includes a
gled with the amount and diversity of infor- checklist and a How diagram that identify the
mation available in the literature on any topic. type and format of information to be included
Efforts to summarize results of studies in in systematic reviews (Moher et al., 1999). The
medicine can be traced back to the 18th cen- goal of this "gold standard" is to help readers
tury, marked with the conception of two to evaluate the quality of reports and to ap-
journals published in Germany that provided praise the likelihood of systematic error, i.e.,
critical appraisals of new publications. William bias in data reporting (Shea et al., 2001).
46 BACKGROUND
In the past decade, neuropfjychology be insufficient to detect or rule out a

researchers have turned to comparalji.ve (case- modest but important property of a spe-
control) observational meta-analytic tech- cific parameter.
niques in an effort to examine the cj.iagnostic 2. Meta-analysis identifies areas where
accuracy of test batteries by comp~fing per- consistent evidence is absent. It high-
formance profiles of clinical and matched lights the need for further research if
control groups. The outcome measures in conclusions drawn from individual stud-
such studies are sensitivity, specificey, likeli- ies are contradictory or high-quality
hood ratio, effect size, and/or a sunimary re- studies in the area of interest are not
ceiver operating characteristic (ROC) curve. available.
A few studies, primarily in medicGte, have 3. Analysis of heterogeneity in study results
used noncomparative (descriptive) observational helps to identify subgroups that differ in
meta-analyses to identify mean value$ and the an estimated parameter and draws at-
expected variability of a certain pt:u-ameter tention to factors mediating the out-
(e.g., blood pressure) in non-disea.1ed indi- come.
viduals. The studies described in tftis book
were subjected to such non-comparat:fve meta- Sources of Bias
analyses, the results of which are ~erted in 1. Publication bias has been of considerable
the relevant test chapters. The out e mea- concern and was described by Rosenthal
sure in our analyses is an expected · 'bution (1979) as a "file drawer problem." It
of scores in a nonclinical population, ediated points to the fact that studies yielding
by demographic variables, when apptopriate. statistically significant findings are more
' likely to be published, published without
delay, and published in English than
Application of Meta-Analysis
studies with "negative" findings, which
in Clinical Practice
tend to be filed away. This causes a Type
The advantages and limitations of meta- I publication bias error and results in a
analytic techniques are addressed i~ several spurious effect of the parameter un-
publications (Cooper & Hedges, 1994; Egger der investigation. To remediate this bias,
et al., 2001, 2002; Green & Hall, 1984; the influence of unpublished studies and
Harris & Rosenthal, 1985; Hasselblad & those published in languages other than
Hedges, 1995; Hedges, 1982; Hedges & Olk- English on the outcome should be taken
in, 1985; Hunter et al., 1982; Kuli~, 1983; into consideration.
Light & Pillemer, 1984; Rosenthal 1983, 2. Methodological and design quality dif-
1984; Sterne et al., 2001; Sutton et aJ., 2000; ferences between studies in terms of
Wolf, 1986) and are briefly sumhtarized degree of experimenter blindness, ran-
below. domization, sample size, controls for re-
cording errors, and type of dependent
Advantages variable (e.g., self-report vs. objective)
1. Karl Pearson was the first mathe$latician represent another source of bias. Some
to point out that individual stu~es are researchers suggest that studies of higher
too small to allow definitive con~usions, quality (with larger samples sizes and
in view of the size of the prob.ble er- well-controlled sampling) tend to have
ror. To solve this problem, he proposed lower variance and effect sizes.
combining individual studies. ~uch a 3. The issue of combinability of studies in a
synthesis provides a solid basis for evi- single meta-analysis is rooted in homo-
dence-based clinical decisions in Ptodern geneity of parameter estimates. Homo-
clinical research and practice, thereas geneity across studies is assumed, based
conclusions drawn from several tndivid- on the expectations that all studies
ual studies might be contradictory or the are testing the same hypothesis and es-
sample size of an individual study might timating the same population parameter
and that variations in study estimates are meta-analysis refers to reproducibility of

random. results (i.e., the likelihood that inde-
However, heterogeneity of parameter pendent meta-analysts replicating the
estimates is a common problem. There analysis will locate and include the same
are several distinct points of view on how studies and measures) and agreement
to deal with heterogeneity: among raters in the coding of study
characteristics (Rosenthal, 1984; Wolf,
a. Heterogeneity is analogous to indi- 1986; Zakzanis, 1998). External validity
vidual differences among subjects and internal validity are directly related
within single studies and represents to the choice of studies to be aggregated
variations within the same parameter. (coding strategies, examining mediating
b. The studies should be grouped into effects, testing homogeneity of results)
homogeneous subsets and combined and methodological quality of the stud-
in separate meta-analytic syntheses. ies, respectively.
c. Outliers contributing to heterogeneity Guidelines for conducting meta-
should be subjected to close exami- analyses to evaluate diagnostic tests and
nation, to test for mediating effects writing meta-analytic reviews have been
that may contribute to the heteroge- published by Hasselblad and Hedges
neity and to better understand the (1995), Irwig et al. (1994), and Rosenthal
properties of the parameter of inter- (1995).
est and suggest new hypotheses.
In addition to statistical methods di-

rected at reduction of heterogeneity, a SELECTION OF STUDIES AND
practical approach to this problem rests PROCEDURES FOR META-ANALYSES
on treating meta-analyses differently PRESENTED IN THIS BOOK
from systematic reviews. It is expected
Literature Search and Selection of Studies
that all available data will be systemati-
cally reviewed. However, it might be The initial pool of studies considered for in-
inappropriate to pool data from all het- clusion in the meta-analysis was generated
erogeneous studies in a meta-analysis. through a computer-based search of the PSY-
This is especially true for case-control CHINFO and MEDLINE databases. Names
epidemiological studies, where combin- of the tests of interest and the neuropsycho-
ing a set of confounded studies may re- logical functions measured by these tests were
sult in spuriously precise but biased entered as key words in separate runs, with the
estimates of association. Thus, careful search limited to English-language publica-
examination of the data to determine tions dating from 1998 until the present. The
sources of heterogeneity is advocated in intent of this search was to add the most recent
the literature. articles to a presumably comprehensive set of
4. Another source of bias stems from in- articles containing normative data included in
cluding multiple tests of a hypothesis the first edition of this book. References in the
from a single study in a single meta- articles generated as a result of this search
analysis, which inflates sample size were reviewed to identify earlier relevant
beyond the number of independent publications that might have been missed in
studies. A practical solution to this "ap- prior searches. In addition, unpublished sets of
ples and oranges" criticism is to perform normative data that have been sent to the au-
separate meta-analyses for each type of thors after publication of the first edition of the
outcome variable. book were evaluated for inclusion.
Careful attention to the sources of bias The meta-analytic tables are presented in
in meta-analytic synthesis enhances its this book only for those neuropsychological
reliability and validity. Reliability of the tests that have a sufficient number of
48 BACKGROUND
homogeneous studies that are based on the sample size). Data points that have a larger
same version of the test or the same admin- sample size and a smaller variance contribute
istration format. For those chapters .that con- more to the analysis. This helps to control for
tain the meta-analytic tables, not all studies study quality since higher-quality studies tend
available in the literature were necessarily in- to have larger weights. Stata's analytic weights
cluded in the database for the analyses. Those were used, which represent the number of
data sets that are based on clinical groups not elements that gave rise to the statistic re-
well identified in terms of methodol•gy or on presenting the data point.
administration of the tests by medical staff, Fixed effect with a cluster option was used
rather than by trained examiners, were not for all regressions. A cluster option was used
reviewed. Among studies that were Jleviewed, to identify data points that were derived from
those that do not contain test means ;and SDs the same study, to account for a lack of in-
(or data that can be converted into these sta- dependence of data points within each study.
tistics), do not have demographic de*riptions Ordinary least square regressions ("regress"
of the sample, or are based on idi~cratic command) were used, as opposed to the meta-
samples (e.g., data collected in China) or analysis regression ("metareg" command),
nonstandard administration procedUfes were because "metareg" does not allow for the
not included in the meta-analyses. ..4.n effort cluster option. We opted for the fixed effect
was made to identify multiple puijications based on an assumption that all data came
based on the same study and to inclu4e a data from the same population. Preliminary tests
set from each study only once, to avqid over- with the "Meta" command for all data sets
lapping data sets. Similarly, when ~ta are revealed that pooled estimates of the fixed and
presented in overlapping age groups, only random effects were comparable (e.g., 42.42
nonoverlapping data points were us~. Data and 42.22, respectively, for the FAS).
sets based on medical patients and on ~atients Tables of predicted values are based on the
referred for neuropsychological e¥uation parameters identified in the above regressions
which yielded no neurological findirts were and include 95% Cis (expected to include 95
not included. The resulting data sets;include out of 100 estimated values if the trials were
data collected primarily (but not exclusively) replicated 100 times), calculated according to
in the United States and Canada, and the vast the following formula: 95% CI =value ± 1.96
majority of the participants across the studies v'var(vaiue).
are Caucasian.
Data Editing
Procedures Used in the Analyses
After the relevant literature was selected,
Data were analyzed using Stata, which is a mean test scores with their respective SDs,
general-purpose, command line-driven statis- demographic variables, and study character-
tical package for data management and anal- istics were recorded in the Stata database
ysis. It reads data into storage memory and is partitioned by age and/or education group or
programmable, allowing the user to add new for the entire sample as reported by study
commands. This package was used for our authors. When data allowed gender compari-
purposes because it contains a comprehensive sons in addition to the overall scores, they
set of user-written commands for meta-anal- were also recorded in a separate file to avoid
ysis, in addition to commonly used ordinary double sampling. Every entry in the database
regression analysis tools. is viewed as a data point. For example, a study
Data in all analyses were inversely weighted that provides test performance data stratified
on standard errors for the means since such into 4 age x 2 education groups would gen-
weighting allows one to account for both erate eight data points.
sample size and the dispersion around the Data were examined for consistency and for
mean for each data point (calculated as a outlying scores. To aid us in this examination,
square root of the ratio of squared SD to the we used the "meta" test, which tests data for
heterogeneity and assesses the influence of a reviewed. Although variability across studies
single study in meta-analysis. It has an option and age/education groups is expected, we
of generating a graph depicting weighted strived to identify data included in error.
means for all data points with their respective Outlying data points that can be explained by
standard errors, centered around a vertical clerical errors in published sources, deviations
line demarcating the combined estimate of from standard administration procedures, or
the mean (e.g., see, Fig. 3.2). Inspection of the idiosyncratic samples were excluded from
resulting forest plot allows one to visualize the further analyses. The "meta" test was rerun on
overall distribution of scores and to identify the remaining data to assure a decrease in the
outlying data points. The degree of deviation Q value and in an estimate of between-study
from the estimated distribution parameters variance in comparison to the initial analysis.
was further examined with a box plot and with Information related to the analyses for outlying
an "iqr" test, which classifies outliers into mild scores is not included in the tables in the ap-
vs. severe categories based on the analysis of pendices, to avoid "information overload."
an interquartile range. The presence of outli- Data reported in the tables describe hetero-
ers typically resulted in a high Q value and a geneity only in the final data set, after data
high estimate of between-study variance, re- editing. It should be noted that the final Q
presenting heterogeneity among studies in- values for all data sets are significant at the
cluded in the data. 0.000 level, indicating heterogeneity for all
Outlying data points identified by visual in- data. This outcome likely stems from the fact
spection and through the above analyses were that the data come from different studies and
Yeucllll
Noma
Yeucllll
·:1 - - + - -
:--.-
-
Gonion
Gonion •
.....
Yeucllll
IJIIal*l
Holf
I
---1 -
a..o
-
Yeucllll
Frleclmln
.....
Tam.ugh
.....
Tam.ugh
Slmldna
Allllrod
Troyer
a-.
Bola
Allllrod
F-
c:n..y I
I
Tam.ugh
Tontleugh ---at-
1
Monlc:h
P.tcln
Non1a
Beelly
Allllrod
Cantined
15 20 25 30 35 40 45 50 55 60
fasmean
STaTa-
Figure 3.2. Example of a forest plot, which was used to assess the influence of a single study in the
meta-analysis (data for the Verbal Fluency-FAS test were used). Vertical line demarcates the combined
estimate of the mean. Data points are depicted as weighted means with their respective standard errors.
50 BACKGROUND
is dealt with using the cluster optioo (which diet test scores. The strength of the rela-
identifies the study from which the data were tionship between age and the test means is
derived) in the regression analyses. reflected in the dispersion of data points
Furthermore, for those tests that are sensi- around the regression line. A scatterplot il-
tive to an education effect, data were txamined lustrating the dispersion is included in each
for consistency of represented educl¢ion ran- relevant chapter, with the size of the bubbles
ges. If a large gap was detected, data falling reflecting the weight of the data point (larger
beyond the empirically supported distribution bubbles indicate larger standard error [SE]
of education ranges were excluded from the and smaller weight).
analyses, to avoid extrapolating a prediction Information for each term of the regression
rule over ranges that are not supported by model is provided in the tables. The coeffi-
existing data. This process is describtd in the cient for a predictor variable indicates the
respective chapters. extent of gain or loss in the test performance
given a one-unit change in the value of that
predictor variable (given that all other vari-
Regression
ables in the model are held constant). For
It has been widely documented in tle litera- example, for the FAS version of the Verbal
ture that expected test performance ~es as a Fluency test, the coefficient for the Educa-
function of demographic characteris~cs of an tion term is 0.498 (see Table llm.l in Ap-
individual. Age has been shown to cqntribute pendix Urn, under "Effect of demographic
most to this variability. To identify 1the rule variables"). This means that for each 1-year
describing a relationship between age and increment in education we expect a 0.498-unit
test performance, as reflected in the corre- increment in word production. In other
sponding study means, data were subjected to words, with every additional year of schooling,
regression analyses. Ordinary least square rean individual is expected to increase verbal
gressions with fixed effect and a clust~r option fluency by almost 0.5 of a word, irrespective of
were used. The shape of the distribution of age. The 95% CI for the coefficient shows how
means was visually inspected to ass~t in the high and how low the actual population value
decision on whether linear or quadratic re- of the coefficient might be.
gression was appropriate for a specf£ic data Dividing the coefficient by the SE for that
set. In addition, both linear and quadratic parameter yields the t value, which is used in
solutions were subjected to the test fdr model testing the null hypothesis that the coefficient
fit. An increase in the R2 for the <fladratic for a given term is 0. In our example, a t value
model and comparison of the Baye$ian In- of 2.47 with a two-tailed p = 0.025 indicates
formation Criterion generated for eacl,t model that the coefficient for Education of 0.498 in a
were used to guide the decision-maldng pro- model based on 29 observations is significantly
cess. This information is presente~ in the different from 0; thus, we can infer that edu-
relevant tables. ' cation significantly contributes to test perfor-
The results of the regression analYJiiS yield mance. This information was used in the
rich information on the relationship l>etween tables to provide a correction factor for pre-
age as a predictor variable and the te. means dicted test scores for different levels of edu-
as an outcome variable. R2 indicates ~e pro- cation.
portion of variance in the test scores ac- It should be noted that significance tests for
counted for by the model. It should be noted the term age in the quadratic equations do not
that we used R2 rather than adjUfted R2 accurately reflect the linear effect of age on
(which corrects for chance variation) $nee we test performance due to collinearity with the
had only one predictor for a relativqly large quadratic effect, i.e., with the age2 term in the
number of observations in each mocJel and, equation. To address the linear effect of age,
therefore, both values are very close; The F avoiding the collinearity, we present signifi-
value and a corresponding probabilfy level cance tests for age centered (by subtracting
indicate how reliably predictor varia~es pre- the mean age for the aggregate sample from
the mean age for individual samples) in the which underlie calculations of the predicted
footnotes to the relevant summaries of the scores. The equations are based on the coef-
regression models. ficients for all predictor variables used by the
program. The equation for a linear model is as
follows:
Prediction
The model that was estimated using regression Predicted test score =constant+ (Page) x age
command was used to make out-of-sample
predictions on another data set, which included That for a quadratic model is as follows:
values for age distributed in 5-year increments
(with smaller intervals at the extremes of the Predicted test score= constant+ (Page) x age
age distribution in some cases), representing + (pag_.) x age2
mathematical centers of respective age cate-
gories. For example, for the F AS, the data set where ~age is the coefficient for age and ~agei
includes values 19.0, 22.5, 27.5, 32.5, 37.5, is the coefficient for age2 , respectively. For
42.5, 47.5, 52.5, 57.5, 62.5, 67.5, and 72.5. example, a quadratic equation derived from
These numbers represent the age categories the model estimation for the F AS is 34.298 +
18-19,20-24,25-29,30-34,35-39,40-44,45- 0.554 x age- 0.007 x age2 (see Table llm.1
49, 50-54, 55-59, 60-64, 65-69, and 70-74. in Appendix 11m). The equations are pro-
Care was taken to avoid out-of-range estimates. vided below the prediction tables; coefficients
For example, when the available data extended used in these equations are listed among the
only to the age of 82, the 80-84 category was results of the analyses provided in the ta-
not used in the prediction table because an bles, specifically under the subtitle "Ordinary
assumption that the same rule applies to ages least square regression."
83 and 84 would not have an empirical basis. In As reflected in the shape of the regression
some cases, when a partial age group was well line in the scatterplot in Appendix lim, FAS
represented, a predicted value for this age performance is expected to increase some-
group is listed. It should be noted that distri- what up to approximately age 40, with a sub-
butions of the data used for model estimation sequent decline. The value for age when
were examined for continuity, to avoid gaps performance reaches its maximum can be de-
within the distribution. When such gaps were rived from the regression equation using the
detected, the extreme data points were ex- following formula:
cluded from analyses. Tables of predicted val-
ues with corresponding Cis for the relevant
neuropsychological tests are presented in the
meta-analytic tables in the appendices along Using coefficients from the regression equation
with supporting statistics. Critical reviews of for the F AS this value is - [.55412 x (- .007)] =
strength and limitations of predicted values 39.57. The obtained value represents the age at
are included in the text of the respective which the curve turns over to the declining
chapters. direction.
In clinical practice, situations might arise
where an estimated score is needed for an age
Standard Deviations
that falls beyond the range of age categories
included in a prediction table. We strongly To test for a possible relationship between the
recommend that the clinician seek the needed variability in scores at different ages, regres-
data in individual studies included in this book sions of SDs for test scores on age were run.
(using locator tables to facilitate the search) or When age accounts for a significant amount of
tum to the data accumulated in that specific variability in the SD, the predicted values for
clinic. However, if everything else fails, the SDs and Cis (calculated using the same
needed score can be calculated using the re- approach as above} are reported along with
gression equations included in the tables, the predicted test scores. The results of
52 BACKGROUND
significance tests for regressions on~ SDs are is based on the assumption that the distribu-
reported. Tests for model fit for the lsolutions tion is normal. Thus, high values of p indicate
on SDs were performed using the same ap- that we cannot reject the hypothesis that the
proach as for the performance scqres. The variable is normally distributed. The normality
results of these tests were used for decision- of residuals was also assessed using the
making purposes, but they are not P,resented "kdensity" plot (Kernel Density Estimate),
in the meta-analytic tables in the apjendices, which approximates the probability density of
to avoid information overload. a variable, and through visual inspection of
When the results suggested tha~e does residuals regressed on age. Close approxima-
not account for any notable amoun of vari- tion of the estimated curve to the normal
ability in SD, as reflected in a ve low R2, density overlaid on the plot and no pattern in
mean SDs derived from the original ~ta are the dispersion of residuals support the results
listed in the tables as they are appliciahle for of the Shapiro-Wilk test. Kernel Density Es-
all age groups. : timate and plot of the residuals regressed on
age for the F AS are reproduced in Figures 3.3
and 3.4 for illustration purposes (the size of
Testing Model Fit and Parameter
the bubbles in Fig. 3.4 reflects the size of the
Specifications I SEs of the data points, reciprocal to their
Postestimation tests of parameter s 1 weights). However, they are not included in
tions were performed to ensure cy of the meta-analytic tables in the appendices.
the prediction. Though violation of e nor- Hmnoscedo.sticity, or homogeneity of vari-
mality of the residuals would not affi esti- ances of the residuals, is one of the main as-
mates of regression coefficients and p ·cted sumptions of the regression analysis. We used
values, it would affect the validity of ypoth- White's general test for heteroscedasticity,
esis testing; in other words, significan devia- which regresses the squared residuals on all
tion from normality would affect the ·dity distinct regressors, cross-products, and squares
of p values for the t-test and F -te . The of regressors. It tests the null hypothesis that
Shapiro-Wilk W test was used to as ss the the variance of the residuals is homogenous.
normality of residuals for the variables bsed in Low values of the derived Lagrange multiplier
the regressions. The p value for the W $tatistic statistic and high values of p indicate that we
.2
.15
I .1
.05
figure 3.3. Kernel Density Estimate, which compares the estimated curve to the normal density (data for
the Verbal Fluency-FAStest were used).
5
0
0 0
0
0
0
oo 0 0
0
I 0 0
0
0 o o0
0
0
0
00 of6
0
0
0
-5 0
20 30 40 50 60 70 60
mean age
~
Figure 3.4. Plot of residuals regressed on age (data for the Verbal Fluency-FAStest were used). The size of
the bubbles reflects the size of the standard errors of the data points, reciprocal to their weights.
cannot reject the hypothesis of homogeneity of not reproduced in the meta-analytic tables in
variance in the residuals. A dispersion of re- the appendices.
siduals plotted vs. fitted values on the residual- Independence assumption refers to the ex-
vs.-fitted plot (rvf plot) was visually inspected pectation that errors associated with one ob-
for each regression. In a model with a good fit servation are not correlated with errors
and homogenous residual variances, this dis- associated with any other observation. Our
tribution should have no pattern. An rvf plot data clearly do not meet this assumption be-
for the F AS is included (see Fig. 3.5) to illus- cause the data points derived from the same
trate this technique. However, these plots are study (e.g., when the scores are stratified by
4.41668 0
0 0
0
0
0
0 0
..
1
0 0
0 0 0
0
'0 0 0
1a: 0
0 0 0
0 0
0
0
0
0
0
0
-5.174 0
36.6217 45.1 7
Fitted values
STaTa-
Figure 3.5. Residual-vs.-fltted plot (data for the Verbal Fluency-FAStest were used).
54 BACKGROUND
I
age group) are likely to be related and B allow such adjustments by adding or sub-
subjected to the same source of errqr. To ac- tracting the appropriate correction factor to/
count for the lack of independ$lce, we from the predicted scores provided in the
used the cluster option for model e~mation, prediction tables. The SD to be used with the
which specifies that observations ;-e inde- education-corrected score is that for the per-
pendent across studies (clusters) ;but not son's actual age group (which is relevant to the
within studies. TMT tables but not to the F AS as the same
SD is used for all age ranges for the latter). It
should be noted that the range of years of
Effect of Demographic Variables
education for the correction tables is limited.
The effect of education was exploted with This limitation is due to a lack of empirical
the "metareg" command, which yiel~ an es- data for individuals with lower levels of edu-
timated between-study variance taq2 , mea- cation in the studies reviewed. We do not know
suring residual heterogeneity adjtqted for whether the pattern of education/test perfor-
covariates. The value of the tau2 estirpate was mance relationship linearly continues into the
compared for regressions of test ~ans on lower educational levels. Therefore, extrapo-
additive components of variance ~th and lation of the suggested correction pattern onto
without education. If the tau2 valu1·for the educational levels falling below the empiri-
regression with education was mu lower, cally supported range might undermine the
indicating that education explains a nsider- accuracy of the prediction.
able amount of heterogeneity in tesq perfor- The effect of gender on test performance
mance, education was entered as a ltedictor was assessed by adding a variable accounting
variable into the equation used for th~ model for a percent of males in the sample as a pre-
estimation. If R2 considerably improted as a dictor variable into a regression of test means
result of addition of the education tdrm and on age. In addition, a t-test was run on the
the t value for education was high wi~ a low data that are reported for males and females
p value, the coefficient for education, !derived separately. Male/female differences in mean
from the latter regression, was used ~ a cor- test scores are reported in the tables, where
rection factor in the tables for relevapt tests. appropriate. If a sufficient number of studies
Where education accounts for a ~e pro- for a specific neuropsychological test report
portion of variance in test performal)ce, the the data stratified by gender and a significant
predicted scores listed in the age-s~tified relationship between gender and test scores is
tables are accurate for individuals wijh edu- highlighted in the literature, age-predicted
cation at the mean level for the origijal data scores are presented for males and females
set. With every year of education a~ove or separately (e.g., GPT). For a number of neu-
below the mean, expected gains or lcfs~es in ropsychological tests, the differences between
test performance are equal to the coctfticient genders were not large enough to warrant
for the education term. For exampleJ values separate predicted tables or addition of a
listed in the prediction table for the PAS are correction factor for gender.
accurate for individuals with appro'4mately Although it is widely known that intelli-
14 years of education since the meanfeduca- gence level makes a considerable contribution
tion across all samples for this test i$ 14.31 to performance on certain tests, we could not
(see Table llm.1 in Appendix llm,i under provide corrections for IQ level because of
"Description of the aggregate sample"). Thus, the paucity of reported data on IQ in the
an expected score for a 37-year-old individual samples aggregated for the analyses in this
with 12 years of education (2 years bebw the book.
mean of 14 cited above) is 45.17- 2(b.50) = Similarly, the volume of information on
44.17. : ethnicity or other demographic variables gath-
Correction tables provided for the FAS and ered from the studies reviewed was not suffi-
the Trailmaking Test (TMT) parts A and cient to conduct statistical analyses.
I
Comments on the Applicability of the speed, education for tests emphasizing verbal
Meta-Analyses Presented in This Book achievement). Regression-based predictions are
best considered an aid in selecting an appro-
As discussed earlier, an advantage of any meta- priate table when results from different studies
analysis is in increased power to direct in- yield contradictory values.
formed clinical decisions based on synthesis of Regression-based norms have been criticized
empirical data. Data derived from individual (Fastenau, 1998; Fastenau & Adams, 1996;
studies underlying our meta-analyses might Heaton et al., 1996; Morgan & Caccappolo-van
be biased by imperfect sampling procedures, Vliet, 2001; Moses et al., 1999). Major criti-
random individual differences due to small cisms refer to the concerns of violation of
sample sizes in each demographic cell, and assumptions undermining the accuracy of pre-
deviations from standard administration pro- diction and extrapolation of the rule summa-
cedures. In addition, they are setting-specific rizing the relationship between predictor and
and contain data for limited age ranges or outcome variables to the ranges of the pre-
demographic categories. Thus, choosing a dictor variable that are not supported by
normative data set as a reference for a specific available data. As it follows from the above
patient might become a time-consuming un- description of the procedures for heterogene-
dertaking. Meta-analytically derived regres- ity and parameter specification testing applied
sion estimates are based on large aggregate in our analyses, the issue of violation of as-
samples and represent the mathematical cen- sumptions was closely attended to. In addition,
ter of all studies across demographic groups. all predicted values fall strictly within the
As such, regression-based tables of normative empirically supported ranges of the predictor
data are relatively free of chance factors affect- variables.
ing individual studies. However, regression- In spite of these efforts, the scope and
based norms should not be used as a substitute quality of our analyses are limited by the scope
for empirically derived tables presented in the of the data available in the literature. The ac-
context of study reviews. curacy of regression solutions presented in this
Any averaging results in a loss of specific book is undermined by several factors:
qualities. We intended to present corrections
for variables that are in theory expected to 1. Age groupings provided in the literature
affect test performance. However, we were vary greatly between studies. Whereas
limited by the data available in the literature, for one study the mean age of 48 years
which in many cases seem to be at odds with might represent a range of 45--50, in
the theory. Individual data sets based on a another study the mean age of 48 rep-
sample of participants who are similar in terms resents a range of 20--86. The perfor-
of setting, demographic characteristics, and/or mance score reported in the latter study
functional level to the patient for whom nor- is much less meaningful in terms of age-
mative comparisons are sought would provide referenced prediction than in the for-
more accurate estimates of expected perfor- mer. This situation was mitigated by
mance than regression-based tables. It has weighting data points on SEs for the
been emphasized by a number of investigators means as this weighting takes into ac-
(e.g., Heaton et al., 1986; Kalechstein et al., count the dispersion around the mean.
1998; Ross & Lichtenberg, 1998; Van Gorp & 2. Evaluation of the effects of demographic
McMullen, 1997) that a selection of the nor- variables on estimates of test perfor-
mative data set should be guided by the com- mance was limited by scarcity of demo-
parability of the patient's demographic graphic data provided in the literature.
characteristics to those of the normative data For example, an important variable
and, more specifically, by the moderating such as IQ, which is expected to con-
variable that is most likely to affect perfor- tribute significantly to variability in sev-
mance (e.g., age for tests tapping psychomotor eral neuropsychological tests, had very
56 BACKGROUND
limited variance across the data sets. for each relevant neuropsychological
Only few studies reported IQ. test, available data seem to cluster at the
3. Levels of education and IQ for the ma- young and advanced ages, with more
jority of data sets are high. Therefore, scarce data points in-between. Further
the predicted values overestimate ex- investigations are needed to assure con-
pected performance for individuals with sistency in the relationship between pre-
a high school education or below and dictor and outcome variables across all
with average or lower than average range ages. However, large gaps in the ranges
of intelligence. of predictor or moderator variables were
4. We cannot describe our aggregate sam- avoided by eliminating extreme scores
ple in terms of ethnic distribution be- from the analyses. As a consequence of
cause of scarcity of information on such adherence to empirically supported
participants' ethnicity in the individual data, ranges of demographic catego-
articles. We believe that the underlying ries covered in prediction tables are re-
samples are not representative of the stricted; e.g., age groups are limited from
mixture of ethnic groups according to both ends, and lower levels of education
U.S. Census figures since many samples are not represented.
were dominated by Caucasian partici- 6. The suggested predictions for age (and
pants. Those data that were collected education in a few cases) are based on the
exclusively on representatives of specific data for largely intact samples. It is un-
ethnic groups (e.g., Chinese, African known if the same relationship between
American, or Hispanic) were not in- demographic variables and test perfor-
cluded in the meta-analyses as they in- mance holds for individuals with brain
crease the heterogeneity of the data. pathology. Ultimately, normative data-
Ideally, separate analyses on data for dif- bases should be expanded to include
ferent ethnic groups should be con- meta-analyses based on various clinical
ducted in the future, providing that a samples across test batteries, to acquire
sufficient number of studies reporting information on expected performance
normative data specifically for different proffies for different diagnostic categories.
ethnic groups will be generated.
5. Increments in the values of predictor or In spite of the weaknesses addressed above,
moderator variables extracted from the we hope that the predictions presented in this
literature are uneven. As reflected in book will facilitate the process of clinical de-
scatterplots depicting the distribution of cision making, which encompasses historical,
data points around the regression line clinical, and psychometric information.
II
TESTS OF ATTENTION AND
CONCENTRATION: VISUAL
AND AUDITORY
4
Trailmaking Test
BRIEF HISTORY OF THE TEST alternate forms of the test in which the order
of the progression was reversed but the loca-
The Trailmaking Test (TMT) is included in the tions of the circles were not altered. The re-
Halstead-Reitan Battery (HRB) and was origi- sulting coefficients were 0.89 and 0.92 in the
nally part of the Anny Individual Test Battery normal sample with over 300 participants and
(1944). 0.95 and 0.94 in the mixed sample for Trails A
Part A is an 8" x 11" page on which the and B, respectively. The standard errors of
numbers 1-25 are scattered within circles. The measurement were 8.05 and 21.7 for Trails A
patient is instructed to draw lines connecting and B. Dikmen et al. (1999) reported test-
the numbers in order as quickly as possible. retest reliabilities of 0. 79 for Trails A and 0.89
Part B is a page with the numbers 1-13 and for Trails B over a 9-month interval in a mixed
letters A-L within circles. The patient is in- sample. Data on repeated administration are
structed to draw lines connecting the numbers also presented by McCaffrey et al. (2000).
and letters in order, alternating between Reliability and validity of the TMT are further
numbers and letters (e.g., 1-A-2-B, etc.). addressed in Franzen (2000), Lezak et al.
Specific administration procedures are pro- (2004), and Spreen & Strauss (1998). A chil-
vided in Reitan's (1979) Manual for Admin- dren's version of the tasks (age 9-14) is avail-
istration of Neuropsychological Test Batteries able, which incorporates fewer items.
for Adults and Children. Two scores are ob- The TMT enjoys considerable popularity
tained, reflecting the total time in seconds to due to its high sensitivity to the presence of
complete each task. In the Reitan (1979) ad- cognitive imp~ent. In addition, a number of
ministration format, errors are not scored, but studies document the usefulness of the TMT as
when they occur, the patient is alerted to the a predictor of instrumental activities of daily
mistake and instructed to correct it, thus living in the elderly (Cahn-Weiner et al., 2002)
slowing overall performance time. The patient and of functional outcome following acquired
is presented with short sample items prior to brain injury (Acker & Davis, 1989; Millis et al.,
the administration of each task. Detailed ad- 1994; Ross et al., 1997; Schmidt et al., 1996).
ministration instructions are provided in Lezak According to surveys of test usage in neuropsy-
et al. (2004) and Spreen and Strauss (1998). chology practice (Butler et al., 1991; Camara
Charter et al. (1987) reported reliability et al., 2000; Lees-Haley et al., 1996; Sellers and
coefficients expressed as correlations with the Nadler, 1992; Sullivan & Bowden, 1997), the
59
60 TESTS OF ATTENTION AND CONCENTRATION
TMT is one of the most frequently used tests. motor speed, impaired working memory, poor
The TMT is a standard component of screen- visual scanning, or a combination of several
ing batteries designed to detect cognitive im- cognitive deficits.
pairment in different neuropsychological Factor analytic studies indicated that both
conditions. For example, in 1990, the TMT was parts A and B load on a visual perceptual
adopted as a measure of cognitive impairment factor (Groff & Hubble, 1981), a spatial factor
by the Drug Abuse Treatment Outcome Study (Moehle et al., 1990), a visuomotor scanning
(DATOS), sponsored by the National Institute factor (Shum et al., 1990), a visuomotor speed
on Drug Abuse of the National Institutes of and coordination factor (Swiercinsky, 1979), a
Health. The DATOS was a naturalistic, pro- motor problem-solving factor (Goldstein &
spective cohort study of adults enrolled in drug Shelly, 1975), and a sustained attention and
abuse treatment programs, which collected mental tracking factor (Lamar et al., 2002).
data on 10,010 adults in 96 programs across Because of the complexity of mechanisms
11 cities in the United States between 1991 contributing to TMT performance, poor per-
and 1993 (Horton & Roberts, 2003). The TMT formance on this test is a nonspecific finding,
data for a subsample of 8,521 adults were an- which can be attributable to visual perceptual,
alyzed and presented by Horton and Roberts in motor, executive, motivational, or other fac-
a series of 19 articles published by the Inter- tors (Anderson et al., 1995; Crowe, 1998b;
national Joumtd of Neuroscience between 2001 Heilbronner et al., 1991; Iverson et al., 2002;
and 2003. The findings reflected in these Lezak et al., 2004; Lorig et al., 1986; Reitan &
publications point to significant effects of age, Wolfson, 1995b).
education, and ethnicity on many indices of To tease out a contribution of executive
TMT performance across various groups of functioning to TMT performance, investigators
drug users. However, the authors emphasized turned to part B as a more complex measure
that these demographic effects are weak. requiring sequence alternation. According
to the literature, several factors contribute to
greater difficulty of part B in comparison to
Contributions of Cognitive Mechanisms
part A, which include cognitive demands and
and Physical Layout Differences to
physical layout. Part B was found to place ad-
Performance on Parts A and B
ditional demands on the ability to alternate
The TMT is described as a measure of visual (Crowe, 1998b; Gaudino et al., 1995; Salt-
conceptual and visuomotor tracking (Lezak house et al., 2000) and to flexibly modify a
et al., 2004); complex visual scanning with a course of action (Arbuthnott & Frank, 2000;
motor component (Shum et al., 1990) with a Kortte et al., 2002; Lamar et al., 2002; Lamb-
contribution of motor speed and agility erty et al., 1994; Pontius & Yudowitz, 1980)
(Schear & Sato, 1989); simple motor-spatial with a task-set inhibition component (Ar-
skills and basic sequencing abilities (Lamberty buthnott & Frank, 2000). Conversely, several
et al., 1994); visual tracking, mental flexibility, investigators have identified additional de-
and attention (Crowe, 1998b); visual percep- mands on the ability to maintain two response
tual abilities (Groff & Hubble, 1981); motor sets simultaneously as the cognitive mecha-
speed and visual attention (Gaudino et al., nism contributing to the greater difficulty of
1995); attention, simple motor and spatial part B (Eson et al., 1978; Lezak et al., 2004;
skills, and sequencing abilities (Martin et al., Reitan, 1971).
2003); and executive function (Burgess, 2003). Recent studies suggest that differences in
Based on the results of a neuroimaging study physical layout further contribute to the
exploring cognitive correlates of brain aging, greater difficulty of part B. Rossini and Karl
Coffey et al. (2001) concluded that the neural (1994) reported that part B is 32% longer than
substrates for the functions measured with the part A. According to Gaudino et al. (1995),
TMT part B involve multiple systems distrib- mean distances for parts A and B are 7.8 (3.2)
uted throughout the brain. They attributed and 10.2 (4.5) em, respectively, which in-
age-related slowing on part B to reduced creases trail length for part B by 56 em in
TRAILMAKING TEST 61
comparison to part A. In addition, analysis of (O'Donnell et al., 1994) and by findings of

visual interference indicated that part A has clinical studies indicating that samples of pa-
on average less than one visually interfering tients with frontal lobe damage or traumatic
stimulus between each target, whereas part B brain injury demonstrate lower performance
averages more than one. Vickers et al. (1996) on part B than normal samples or clinical
stated that the two parts of the TMT differ samples with intact frontal lobes (Cicerone &
with respect to length and angular variability. Azulay, 2002; Corrigan & Hinkeldey, 1987;
However, there is no difference in structural Pontius & Yudowitz, 1980; Reitan, 1971; Stuss
complexity. Fossum et al. (1992) alternated et al., 2001). In contrast, Cicerone (1997)
the configura! arrangement of test stimuli by demonstrated low sensitivity of Part B to mild
placing stimuli of part A in the spatial con- traumatic brain injury. Similarly, Anderson
figuration of part B and vice versa. The au- et al. (1995) and Reitan and Wolfson (1995a)
thors concluded that differences in symbolic did not find the TMT useful in detecting
complexity and spatial arrangement, as well as frontal lobe damage. These contradictory find-
interactions between these factors, contribute ings might be explained by differences in the
to the greater difficulty of part B. severity of pathology in the study samples or
Arnett and Labowitz (1995) developed a by differences in the anatomy of the affected
modified version, which used a standard lay- frontal regions (dorsolateral convexities vs. me-
out of part B with numbers substituted for dial or basal-orbital regions).
letters, thus eliminating the alternation com- Several investigators have examined the re-
ponent. The authors found that it takes about lationship between performance on the two
1.4 times as long to complete part B relative to parts of the TMT, using the B-A difference
part A because of the more complex layout of and B:A ratio, in an attempt to identify an in-
part B. Based on analysis of differences in crement in time associated with the additional
time to completion between parts A, B, and processing demands imposed by part B. Golden
the new version, the authors related longer (1981) examined the properties of the B:A
completion time for part B to three factors: a ratio and found that it has a curvilinear rela-
cognitive processing factor that is unique to tionship with impairment. Both high and low
part B, the more complex layout of part B, and ratios may indicate neuropsychological im-
a psychomotor-attentional factor that is com- pairment, with a ratio score lower than 2 being
mon to both parts A and B. indicative of deficient performance on part A
and a ratio score greater than 3 reflecting de-
ficient performance on part B. Heaton et al.
Utility of the Derived Measures, which
(1985) recommended use of the B-A differ-
Are Based on Differences in Performance
ence as a measure of cognitive efficiency.
Times for Parts A and B
Corrigan and Hinkeldey (1987) found the dif-
The above review suggests that TMT part B ference and the ratio measures to be sensitive
differs from part A in cognitive demands, to the increased cognitive demands of part B.
length of trail, and perceptual complexity. They recommended use of the ratio as an in-
However, studies that removed confounds of trasubject comparative index, allowing one to
physical layout or visual complexity still doc- control for individual variability.
umented a significant increase in time to Lamberty et al. (1994) demonstrated the
completion with addition of an alternating usefulness of the ratio measure as an index of
component to the trailmaking condition (Crowe, cognitive flexibility controlling for intrasubject
1998a; Gaudino et al., 1995). This suggests variability as it is relatively free of age and
that sequence alternation places additional education confounds. The authors concluded
demands on executive function beyond the that the ratio measure is most useful in a
confounds of physical layout, which accounts screening evaluation when strong diagnostic
for the increase in time to completion on part information is not available, in the context of
B. This assumption is further supported by the age and education confounds. They also de-
report that part B loads on an attention factor scribed its usefulness in the forensic context as
"fakers" are expected to have a smaller ratio patients into subgroups, based on the number
than brain-damaged individuals. The authors of errors, indicated that damage in dorso-
suggested that ratios of 2.0-2.5 represent lateral frontal areas was associated with the
normative performance, with 3.0 being iden- greatest degree of impairment, whereas dam-
tified as a cutoff for the presence of neu- age to the inferior medial aspects of the frontal
ropsychological impairment. The usefulness of lobes did not significantly affect performance.
this cutoff is supported by Arbuthnott and Steffens et al. (2001) reported greater fre-
Frank (2000), who found an especially large quency of subsequent errors after controlling
cost for alternating switches in participants for the overall initial error rate in a sample of
with a B:A ratio greater than 3.0. However, geriatric depressed patients in comparison to
Drane et al. (2002) concluded that rates of a control elderly sample. The authors inter-
false-positive misclassifications for the 3.0 cut- preted this finding as a performance feedback
off were unacceptably high in their sample of deficit in geriatric depression which is linked
normal adults, especially for older age groups. to a dysfunction of the orbital frontal cortex.
In addition, Martin et al. (2003) did not find Ruffolo et al. (2000) investigated the diag-
the B:A ratio to be sensitive to the severity of nostic utility of TMT errors by comparing
traumatic brain inju:ry. It also failed to identify error rates in two head-injured groups (vary-
examinees who were dissimilating, according ing in degree of injury severity), patients with
to independent psychometric indicators. The suspect effort, controls, and experimental ma-
authors concluded that the ratio measure does lingerers. No differences in error rates be-
not enhance the clinical utility of the TMT in tween both head-injured and control groups
individuals with traumatic brain injury. were found. However, error rates on part B
Axelrod et al. (2000a) examined the sensi- were significantly higher in both malingering
tivity and specificity of different cutoffs for the groups in comparison to the head-injured and
ratio measure based on the Hebrew version of control groups. The authors concluded that
the TMT. They recommended use of a more performance errors lack diagnostic utility for
conservative cutoff when performance is con- persons with head injuries but that they may
sidered to be pathological. be helpful in the assessment of malingering if
used in conjunction with time to completion.
To explore the specific cognitive mecha-
Utility of the Error Analysis
nisms contributing to poor performance on
The diagnostic utility of the rate of perfor- the TMT in clinical samples, several investi-
mance errors on the TMT was examined in gators have examined the frequency of dif-
several studies. ferent categories of errors.
Rasmusson et al. (1998) reported an in- Klusman et al. (1989) categorized errors on
crease in error rate on part B, but not on part part B into shifting (from number to letter and
A, with each decade of life in their sample of from letter to number) and sequencing
nondemented participants over the age of 60. (number and letter) errors. The authors found
However, there was no longitudinal change in that neither error type nor frequency nor
the error rate on a 2-year follow-up. In the percentage of individuals making errors dif-
demented sample, dementia status was signif- fered significantly between head-injured and
icantly associated with the proportion of par- control groups.
ticipants making errors on both parts A and B, McCaffrey et al. (1989) identified two types
independent of age. of errors: sequential errors, which involve
Stuss et al. (2001) found error analysis to be omission of the consecutive element in the
more useful than time to completion in dis- series, and perseverative errors, indicating a
tinguishing between patients with frontal lobe failure to alternate between categories, with
injuries and those with damage to nonfrontal the latter type being applicable only to part B.
areas or normal controls. All patients who Stability of both types of error between two
made more than one error on part B had administrations over a 7-10 day period was
frontal lesions. Further division of the frontal evaluated by the authors on a sample of
TRAILMAKING TEST 63
polysubstance users. The error analyses re- the worst. Reitan (1958) initially recom-
vealed a significant improvement in perfor- mended use of this scaling method and sug-
mance from the test to retest. The authors gested cutoffs for impaired performance
interpreted this improvement as a practice ef- based on the scaled scores. Given the signifi-
fect and pointed to the questionable utility of cant association between TMT performance
the error rates as indicators of stable central and age, IQ, education, and possibly gender,
nervous system (CNS) dysfunction in poly- use of single cutoff scores does not appear to
substance users. be appropriate, as has been confirmed by
Amieva et al. (1998) used similar error cate- several studies. Bornstein et al. (1987b) found
gories in an investigation of cognitive failures that, using a cutoff of ~40 seconds on part A
contributing to low TMT performance in De- and ~92 seconds on part B, 33% and 39% of a
mentia of Alzheimer's Type (DAT). Every tran- healthy elderly sample were misclassified as
sition between two items in their demented and brain-damaged. Ernst (1987) found that use of
elderly control samples was examined, and cutoffs ~39 and ~92 seconds resulted in a
errors were further broken down into sub- misclassification rate of 48% for both parts A
categories. The sequential errors (SE) category, and B. Bak and Greene (1980) reported that
identified by the authors as a failure to efficiently at least 40% of their elderly normal partici-
pursue the letter or the number series by omit- pants were misclassified on part B. Dodrill
ting an item, was further subdivided into prox:- (1987) documented misclassification rates of
imitySE (characterized by spatial proximity), SE 11.7% and 13.3% when using cutoffs of ~39
to rectify (an attempt to move back and rectify seconds on part A and ~89 seconds on part B,
an initial error), displacement SE (displacement respectively, in a young control sample.
of the subsequent sequence), and unexplained Bornstein and colleagues (1987b) noted
SE (not falling under any of the above descrip- that 96% and 98% correct classification rates
tions). The perseverative errors (PE) cate- for parts A and B, respectively, were obtained
gorywas conceptualized as a failure to alternate when cutoffs of ~55 seconds and 2:137 sec-
between the series of numbers and letters. The onds were used (but 46% and 40% of brain-
results yielded a significantly greater frequency damaged participants were then misclassified
of proximity SEs and PEs in the patient group, with these cutoffs). The authors emphasized
which was interpreted as an inhibitory deficit that cutoff scores may be useful but only if
largely accounting for poor TMT performance considered in the context of other neu-
in demented patients. ropsychological information obtained in a test
Thus, review of the literature suggests that battery and if age, education, and other ap-
error analysis is not sensitive to cognitive propriate adjustments are made.
deficits associated with head injury or poly- Cahn et al. (1995) used cutoffs of ~66 and
substance use. However, analysis of frequency ~172 seconds for parts A and B, respectively,
and type of errors might be diagnostically in a study comparing DAT patients with a
useful in dementia and might contribute to large sample of neurologically normal indi-
the detection of localized frontal lobe dys- viduals. They report sensitivity and specificity
function, suboptimal effort, and age-related indices of 69% and 90% for part A and 87%
decline in performance accuracy when used in and 88% for part B. The authors underscored
conjunction with time to completion. Further the diagnostic effectiveness of part B, which
research is needed to replicate these findings was one of the few measures contributing to
and to investigate the usefulness of these in- optimal differentiation between DAT and
dices in other clinical settings. control participants. The part B cutoff of 172
seconds was used by Rasmusson et al. (1998)
in distinguishing between nondemented el-
Utility of the Cutoffs for Impairment
derly and those participants who met criteria
The manual for the Army Individual Test for DAT. Obtained sensitivity and specificity
(1944) provides a 10-point scale for converting indices of 77% and 89.4% support the use-
raw scores, with 10 being the best score and 1 fulness of this cutoff.
The utility of the cutoff scores was further a mean age of 59.1 years (standard deviation
emphasized by Soukup et al. (1998), who [SD] =9.3). At the fourth testing probe 6
recommended reporting cutoff scores that months after the initial assessment, practice
represent borderline (15th percentile) and effect gains were partially lost. Practice ef-
defective (< 5th percentile) perforinance in fects, specifically on part B, were reported by
addition to the means and standard deviations DesRosiers and Kavanagh (1987). Frank et al.
in future studies, to offset problems associated (1996) also reported a significant practice ef-
with the positive skew in the disbibution of fect for part B on the 2-year retest for a sam-
TMT scores. i ple of 380 elderly over the age of 65.
To minimize the practice effect over re-
peated administrations, several alternate ver-
Effect of the Order of Presentation and
sions of the TMT were developed. Lewis and
Practice Time, Practice Effect, an~
Rennick (1979) developed alternate forms for
Alternate Versions of the TMT .
part B, which were included in the Repeatable
The effect of the order of presentation on Cognitive-Perceptual-Motor Battery. Further
performance on parts A and B was examined discussion of the comparability of these forms
by Taylor (1998a) in a sample of pa*nts with to part B can be found in Kelland and Lewis
neurological disorders and by Miner and (1994) and Lezak et al. (2004).
Ferraro (1998) in a sample of undergraduate DesRosiers and Kavanagh (1987) devel-
students. Both studies revealed a ;gnificant oped Trail C (TMC) and Trail D (TMD)
time x order interaction, with time~ to com- versions, which retained the same relative
pletion being lower for part A and lii.gher for position of each circle but inverted the labeled
part B, for the reverse order of pretentation. sequences respective to their equivalents,
Taylor (1998a) explained this trend i~ terms of TMA and TMB. Administration of both sets to
a slight effect of practice in visual1scanning 16 normal adults in the pilot study yielded
and noted that part B can be used if4 isolation high correlations between alternate forms
as omission of part A will not lead ., serious (r=0.73 and 0.80 for TMA/TMC and TMB/
distortion of part B performance. TMD comparisons, respectively). The equiv-
Thompson et al. (1999) examined the utility alence of the alternate forms was further in-
of practice times in predicting success or fail- vestigated in an orthopedic control group and
ure on the full version of the test. ~ authors in a sample of closed head injury patients.
presented tables of classification ac<llracy for Alternate forms for both conditions were sta-
various practice times. They found: that 20- ble and consistent in both groups. The equiv-
and 30-second cutoffs on practice times for alence of these alternate versions was further
parts A and B, respectively, optimize4 the pre- tested by McCracken and Franzen (1992) and
diction of successful completion o~ the full Franzen et al. (1996). Based on the data from
version (within< 180 seconds for pl¥f A and clinical samples and patients referred for
< 300 seconds for part B). The au..ors un- neuropsychological evaluation, the correlation
derscored the usefulness of practice! times in analyses as well as a comparison of solutions
decision making regarding discontinuation of for two runs of the principal component anal-
the full version. ysis provide support for the equivalence of the
Significant practice effects over repeated standard and alternate test versions. On a
administrations of the test were reported by sample of healthy adults, LoSasso et al. (1998)
Craddick and Stem (1963), Dye (1$79), and found that the TMT-D version is somewhat
Mitrushina and Satz (1991a), althougil Dodrill more difficult than TMT-B. Therefore, the
and Troupin's (1975) data did not i~dicate a authors concluded that TMT-D can serve as
practice effect in their sample. ~cCaffrey an excellent alternate form to the TMT-B, if it
et al. (1992, 1993) reported signific~t prac- is administered on the retest after the TMT-B.
tice effects for part B within 7-10 Clays and In the same study, the authors reported that
then 3 months following initial asseSJlment in there is no clinically meaningful difference in
their group of chronic cigarette smo1ers with scores with respect to whether the test is
TRAILMAKING TEST 65
performed with the preferred or nonpreferred 74 years of age, are reported by Lannoo and
hand. Vingerhoets (1997). The data are stratified
Charter et al. (1987) developed alternate into 3 age x 2 education groups.
versions for parts A and B using similar Siegert and Cavana (1997) presented nor-
methodology. The authors report high corre- mative data for a sample of 127 New Zealan-
lations (r > 0.9) between the new and stan- ders over 60 years of age, divided into five age
dard test versions in a large sample of patients groups.
and normal controls. The Color Trails Test (CTI) and TMT were
Kilander et al. (2000) used two new ver- administered in Hong Kong to 84 Chinese-
sions, letters A-Z and following arrows indi- English bilingual and English monolingual
cating the order, in addition to the standard A participants in a study designed to examine
and B parts. A different approach to gener- the effect of different language backgrounds
ating TMT alternate forms was proposed by on trailmaking performance (Lee et al., 2000).
Vickers and Lee (1998). The authors stated Data for participants between 20 and 50 years
that there is no established procedure for of age with at least a university education are
generating equivalent but stochastically dif- reported for both groups. No significant dif-
ferent test forms and proposed a neural net- ferences in test performance between groups
work method as a practical solution. were found.
In a follow-up article, Lee and Chan (2000a)
provided comparative data for the CTI and
Culture-Specific Sets of Normative Data
TMT on a sample of 108 Chinese adults in
and Cultural Adaptations for the TMT
Hong Kong. The data are reported in 2 age x
Stewart et al. (2001) administered the TMT 2 education groups.
among other tests to 285 African-Caribbean In another study, Lee et al. (2002) provided
participants between 55 and 75 years of age, data for the TMT, among other neuropsycho-
residents of south London. Normative data logical tests, collected in Hong Kong on a
are presented for part A, stratified by two age sample of 475 Cantonese-speaking Chinese
groups, gender, three education groups, and between 13 and 46 years of age. Time to
three occupational classes. completion and the number of errors are re-
Preliminary normative data collected on 190 ported in 3 education x 2 achievement x
Greek adults between 18 and 89 years of age 2 gender groups for adolescents and in 3 ed-
are reported by Vlahou and Kosmidis (2002). ucation x 2 gender groups for adults.
Elwan et al. (1996, 1997) reported data for Lu and Bigler (2000) administered the
211 normal Egyptians between 20 and 72 TMT to 60 American and Chinese students
years of age. It is unclear, however, how the (age 21-32) from an American university.
performance was measured. The Chinese group was also given an equiv-
Giovagnoli (1996) provided data for 287 alent of the TMT part B for native Chinese
healthy Italian adults between 15 and 79 years speakers, where English letters were re-
of age. Time to completion for parts A and B placed by numbers in Chinese characters.
as well as the B-A difference were reported. The data for the two groups are presented as
Nielsen et al. (1989, 1995) administered the T scores using the Heaton et al. (1991)
TMT among other neuropsychological tests to norms. The results indicate significantly lon-
a sample of Danish adults. In their 1989 ar- ger performance time for part B than for the
ticle, the authors reported data for 101 vol- modified Chinese version of part B for the
unteers and patients who had undergone Chinese group.
minor surgery between ages 20 and 54 years, In a follow-up article, Lu and Bigler (2002)
stratified into three age groups. In their 1995 presented normative data for part A and the
article, they presented data for elderly par- Chinese version of part B (C-Trails B), col-
ticipants between ages 64 and 83. lected on a sample of 110 adults between 21
Flemish normative data for the TMT, col- and 75 years of age, who were born in main-
lected on 200 healthy adults between 18 and land China, Taiwan, or Hong Kong but lived
in the United States and spoke Chinese as the the test are described by the authors in the
first and primary language. D-KEFS manual. Normative data for ages 8-
The psychometric properties of the Arabic 89 years are based on a large, nationally rep-
version of the Expanded TMT were examined resentative sample. Wecker et al. (2000) used
by Stanczak et al. (2001). In this version, a preliminary version of this test to examine
English symbols are replaced with Arabic the effect of age on component mechanisms
numbers and letters. The words begin and end contributing to test performance.
are also translated into Arabic. The authors Another version of the TMT, the Compre-
compared test performance by Sudanese hensive Trail-Making Test (CTMT), which
normal and brain-damaged participants and includes five conditions, was developed by
equivalent groups from the United States, Reynolds (2002) and published by Psycho-
who completed the standard English version logical Assessment Resources. Normative data
of the test. The data are presented as mean for ages 11-74 are based on a large nationwide
logarithmic scores for the four groups. sample matching the U.S. Census data.
The Hebrew version of the TMT was used In an effort to remedy the shortcomings of
by Axelrod et al. (2000a) in a study comparing the standard TMT and investigate effects of
performance of normal control participants different component mechanisms on age-
and post-head injury outpatients. In the He- related slowing, Salthouse et al. (2000) de-
brew version of the test, English letters are veloped the Connections Test, which is based
replaced with Hebrew letters on part B. Time on the Zahlen Verbindsungs Test introduced
to completion and the B:A ratio are reported by Oswald and Roth (1978) and described by
for both groups. The sensitivity and specificity Vernon (1993). It consists of four conditions:
of different cutoffs for the B:A ratio are numbers, letters, alternating numbers and let-
presented. ters, and alternating letters and numbers. The
instructions are to connect successive targets
presented in 7 x 7 matrix of circles. The place-
Modified Versions of the TMT
ment of successive targets is fixed to one of
Given the multidimensional nature of cogni- eight positions relative to the prior target. The
tive mechanisms contributing to TMT per- authors examined the equivalence of several
formance, several versions of the trailmaking alternate forms of the test. In the context of
tests were developed in an effort to tease out the elementary process model, the authors
the effects of different component mecha- pointed to target identification, search-
nisms on test performance. comparison, response, and sequence update
A variant of the TMT is included in the processes as being involved in the nonalter-
Delis-Kaplan Executive Function System nating conditions, with the addition of a se-
(D-KEFS) published by the Psychological quence switching process in the alternating
Corporation (Delis et al., 2001). This version conditions. Based on the performance of a
includes five sections measuring component sample of 207 adults between 18 and 80 years
mechanisms contributing to trailmaking per- of age on this test, the authors concluded that
formance: the number sequencing task is a all age-related effects are mediated by per-
variation of part A; the letter sequencing ceptual speed.
task is a letter counterpart of the number se- Stanczak et al. (1998, 2001) developed an
quencing task; the visual scanning task taps the Expanded TMT (ETMT), which is aimed at
ability to identify targets of a specific shape; systematically varying the stimuli within the
the motor speed task taps the speed of tracing TMT format in order to isolate the cognitive
over a dotted line; and the number-letter components. The authors designed experimen-
switching task taps the flexibility of thinking on tal forms X, Y, and Z, which are administered
a visual-motor sequencing task and is similar in addition to parts A and B. Form X consists of
to part B of the standard TMT. Performance is a series of 25 clock faces with the hour and
evaluated based on time to completion and minute hands pointing to the time arranged in
error analysis. The psychometric properties of 15-minute increments. Form Y consists of
TRAILMAKING TEST 67
a sequence of 24 solid black dots of increasing The oral version of the TMT (OTMT) was
diameter. The two forms are arranged in the designed to eliminate visual and graphomotor
vertically inverted mirror-image format of components in trail-making performance (Abra-
parts A and B, respectively. The examinee is to ham et al., 1996; Ricker and Axelrod, 1994;
connect the clock faces in order of advancing Ricker et al., 1996). Two conditions, per-
time and the dots in order of increasing di- formed orally, are presumed to be comparable
ameter. Form Z requires alternation between to parts A and B of the standard version of the
clock faces and dots (apparently, form Z was test. The instructions are to count from 1 to 25
used only in the Arabic version of the test). The as quickly as possible in the first condition and
psychometric properties of the adult version of to orally switch between numbers and letters
the test were examined by Stanczak et al. in the second condition. The second condition
(1998) on a sample of 164 brain-damaged in- concludes with number 13. Although the clin-
dividuals and 252 normal controls. The authors ical validity of the test was not established in
found adequate concurrent and criterion va- the initial study, the authors reported that
lidity of the new forms and highlighted their more recent studies, based on clinical sam-
utility in explaining between-group variance ples, supported the validity of the OTMT as a
and in cross-cultural assessment. measure of executive functioning. Ruchinskas
The midrange adaptation of the ETMT was (2003) cautioned against use of unadjusted
initially validated by Davis et al. (1989), who performance indices for the OTMT with older
found that forms B, X, and Y discriminated medical patients or individuals with low edu-
well between normal and learning-disabled cational levels, due to moderate correlations
children. A modification of this version (Mid- of the test scores with the Mini-Mental State
Range Expanded TMT, METMT), which in- Exam (MMSE) and level of education.
cluded five new forms (X, Y, Z, G, and B4) in Two other versions of the OTMT have been
addition to the standard parts A and B, was described in the literature: the Alphanumeric
cross-validated by Stanczak and Triplett (2003). Sequencing procedure is incorporated in the
The authors report adequate psychometric Behavioral Dyscontrol Scale (Grigsby et al.,
properties of this version and provide nor- 1994; Grigsby &: Kaye, 1995); the Mental Al-
mative data. ternation Test (Jones et al., 1993) was devel-
Bamcord and Wanlass (2001) pointed out oped as a bedside screening test of cognition
that the TMT versions designed to reduce the for patients with HIV infection. The psycho-
cultural confound associated with use of the metric properties of these tests need to be
Latin alphabet have not addressed the con- addressed in future studies.
found associated with the Arabic counting
system. To fill in this void, the authors devel-
oped an alternative to the TMT, the Symbol
TMT (STMT), which uses symbols that are
RELATIONSHIP BETWEEN TMT
PERFORMANCE AND DEMOGRAPHIC
not language- or number-based. The test con-
FACTORS
sists of three trials. Two of them require
matching of the symbols to the sequential key The literature to date indicates that TMT
presented at the top of the page, with an performance is associated with age: increased
added alternating background coloration re- age is related to poorer test scores. The asso-
quirement in the second trial. The third trial ciation between age and TMT scores is pres-
measures incidental memory for symbols ent in both normals and patients but appears
presented in trials 1 and 2. The authors report to be of smaller magnitude in brain-damaged
modestly similar psychometric properties for samples (Corrigan&: Hinkeldey, 1987). Anum-
the STMT compared to the TMT, established ber of studies have reported either significant
on a sample of 210 normal participants. correlations with age or significant differences
Another culture-fair version of the TMT, the across various age groups (Alekoumbides et al.,
CIT, which was developed for the World 1987; Anthonyet al.,1980; Bak &: Greene, 1980;
Health Organization, is described in Chapter 5. Bomstein,1985; Davies,1968; Elias et al.,1993;
Ganguli et al., 1991, 1996; Giovagnoli, 1996; performing poorly). In spite of the reported
Gordon, 1972; Goul & Brown, 1970; Heaton nonsignificant correlations, the percent of
et al., 1986, 1991, 1999, 2004; Horton & participants correctly classified as normal for
Roberts, 2001, 2002, 2003, see Brief History the oldest age group (60-64) did fall precipi-
of the Test for the description of the study; tously compared to other age groups, sug-
lvnik et al., 1996; Kennedy, 1981; Lamberty gesting that there was a decline with age in
et al., 1994; Lannoo & Vingerhoets, 1997; Lee test performance at least in this age group.
and Chan, 2000a; Libon et al., 1994; Lu and Gonzalez et al. (2001) did not find a rela-
Bigler, 2002; Lyness et al., 1994; Matthews tionship between age and TMT part B per-
et al., 1999; Parsons et al., 1964; Rasmusson formance in a sample of homeless individuals
et al., 1998; Reed & Reitan, 1962; Salthouse & who were receiving medical care. The variance
Fristoe, 1995; Saxton et al., 2000; Siegert & for the test scores was very large (probably
Cavana, 1997; Small et al., 2000; Soukup et al., due to varied degree of impairment in mental
1998; Stanton et al., 1984; Stuss et al., 1987; status), which possibly obscured the relation-
Vlahou & Kosmidis, 2002; Wahlin et al., 1996; ship between age and performance time.
Wiederholt et al., 1993). In two other studies, Many studies have documented a signifi-
which do not formally acknowledge the asso- cant relationship between education and TMT
ciation between age and TMT performance, scores in normal individuals, with higher ed-
TMT data are presented by age groupings, ucation levels being tied to better test per-
and the mean scores of the groups obviously formance (Alekoumbides et al., 1987; Anthony
increase with age (Fromm-Auch & Yeudall, et al., 1980; Bornstein, 1985; Bornstein &
1983; Harley et al., 1980). Suga, 1988; Drane et al., 2002; Ernst, 1987;
Coffey et al. (2001) report a significant re- Finlayson et al., 1977; Ganguli et al., 1991;
lationship between TMT part B performance Giovagnoli, 1996; Gonzalez et al., 2001; Gor-
and age-related brain changes documented on don, 1972; Heaton et al., 1986, 1991, 1999,
quantitative MRI in a sample of 320 elderly 2004; Horton & Roberts, 2001, 2002, 2003a,
nonclinical volunteers, with poorer perfor- 2003b; Kennedy, 1981; Lamberty et al., 1994;
mance being related to cerebral atrophy (re- Lannoo & Vmgerhoets, 1997; Lee & Chan,
flected in decreased cerebral hemisphere 2000a; Lu & Bigler, 2002; Matthews et al.,
volume and increased peripheral cerebrospi- 1999; Parsons et al., 1964; Portin et al., 1995;
nal fluid volume) and ventricular enlargement. Saxton et al., 2000; Stanton et al., 1984; Stuss
The authors comment on negative findings in et al., 1987; Vlahou & Kosmidis, 2002; Wie-
similar neuroimaging studies and suggest that derholt et al., 1993); however, a few studies
differences in subject characteristics and sam- did not find a significant correlation between
ple sizes, brain imaging methods, measure- education and TMT scores (Fastenau, 1998;
ment techniques, and approach to statistical Ivnik et al., 1996; Wahlin et al., 1996; Yeudall
analysis might account for this discrepancy. et al., 1987). Heaton and colleagues (1986),
Yeudall et al. (1987) and Boll and Reitan assessing the combined effect of age and ed-
(1973) detected no association between age ucation on TMT part B in normal participants,
and part A or B. Yeudall et al.'s (1987) negative documented a significant interaction, suggest-
findings may be due to the restricted age ing that for individuals less than 60 years old
range of their sample (15-40); examination of lower levels of education are associated with
other data sets suggests that declines with age greater amounts of age-associated impairment
appear to occur after age 40 (Goul & Brown, and for those more than 60 years old level of
1970; Stuss et al., 1987) or age 50 (Kennedy, education has less of an effect than for youn-
1981) for part A. The reason for Boll and ger individuals.
Reitan's (1973) failure to document age ef- Another aspect of the age/education inter-
fects is less obvious, but it could involve small action in reference to part B performance was
sample sizes in the very young and very old presented by Richardson and Marottoli (1996).
groups and problems in the linearity of the The mean performance for community-
data (e.g., very young and very old participants residing elderly participants with less than
TRAILMAKING TEST 69
12 years of education was stable across youn- Auch & Yeudall, 1983; Heaton et al., 1986,
ger-old (76-80) and older-old (81-91) age 1991; Ivnik et al., 1996; Stuss et al., 1987; Wah-
groups; it was considerably lower than for the lin et al., 1996; Waldmann et al., 1992; Yeudall
better-educated participants in this study and et al., 1987). The studies that found gender
well below expectation in comparison to the effects differed as to which gender performed
Heaton et al. (1991) norms. However, for par- better, and the differences in performance
ticipants with 12 or more years of education, between the genders were small: Davies
performance for the younger-old age group (1968) and Wiederholt et al. (1993) found that
was superior to that of the older-old group and men scored higher than women on part B;
comparable to the Heaton et al. (1991) norms. Portin et al. (1995) reported superiority of
The relationship between TMT scores and men on part A, whereas Arbuthnott and Frank
education in brain-damaged samples is more (2000) reported superiority of women on
equivocal, with some investigators document- part A; Bornstein (1985) and Saxton et al.
ing an associ,ation (Anthony et al., 1980) and (2000) reported that women outperformed
others failing to detect a relationship (Finlay- men on both parts A and B; and Gaudino et al.
son et al., 1977) or reporting a statistically (1995) found significant superiority of women
significant but clinically negligible association on experimental versions of the test.
(Corrigan & Hinkeldey, 1987). Multiple regression analyses of the effects
General intelligence (in the majority of of age, IQ, gender, and education on TMT
studies expressed as Full Scale IQ) has con- scores in normal individuals by Greene
sistently been found to be related to TMT and Farr (1985) suggested that age accounts
outcome in both patients and normal partici- for most of the variance, followed by FSIQ,
pants, with higher intellectual levels being for both parts A and B. Gender was a con-
associated with superior test scores (Anthony tributor to part B performance only, while
et al., 1980; Boll & Reitan, 1973; Corrigan & education had a negligible association with
Hinkeldey, 1987; Dodrill, 1987; Giovagnoli, both parts. In contrast, Heaton et al. (1991)
1996; Goul & Brown, 1970; Kennedy, 1981; found that while age accounted for the most
Lamberty et al., 1994; Parsons et al., 1964; test score variance on parts A and B, 16%-
Siegert & Cavana, 1997; Tremont et al., 1998; 27% of the unique test score variance was
Waldmann et al., 1992; Warner et al., 1987; attributable to educational level; gender did
Wiens & Matarazzo, 1977). However, infor- not account for any appreciable test score
mation regarding whether TMT data are more variance. The relationship between FSIQ and
related to Performance IQ (PIQ) vs. Verbal TMT scores was not assessed in this study.
IQ (VIQ) has been contradictory. For exam- The Greene and Farr (1985) and Heaton et al.
ple, Yeudall et al. (1987) found a significant (1991) data appear to be contradictory regard-
correlation between PIQ and both parts A and ing the effect of education on TMT perfor-
B but no relationship between TMT scores mance. However, the findings can be reconciled
and VIQ. Conversely, Wiens and Matarazzo if the contribution of education to TMT scores
(1977) found a significant correlation between is viewed as occurring through its association
part B and VIQ but not PIQ. However, their with IQ.
data should be viewed with caution due to Hays ( 1995) reported a considerable effect
unexpected and unexplainable findings; spe- of intelligence level and age on TMT perfor-
ciflcally, a significant positive correlation was mance, demonstrated by multiple regression
found between PIQ and part A (the higher the analysis of data collected on a sample of
PIQ, the longer it took participants to com- 661 psychiatric inpatients. The authors pro-
plete the task), and significant correlations vided normal standard score conversions from
were found for one control group but not the raw scores and information derived from the
second one. regression analysis, allowing correction ofTMT
The literature has generally indicated that scores for IQ and age.
there are no gender differences in TMT scores The effect of culture and acculturation on
in normal participants (Dodrill, 1979; Fromm- TMT performance was examined in several
studies. Arnold et al. (1994) reported a sig- Subject Variables

nificant effect of acculturation on TMT, and
Manly et al. (1998) found that black English Sample Size
use was associated with poor performance on Fifty cases are considered a desirable sample
Trails B. size. Although this criterion is somewhat ar-
Soukup et al. (1998) examined the effects bitrary, a large number of studies suggest that
of demographic and other variable; on TMT data based on small sample sizes are highly
performance and compiled norm4tive data influenced by individual differences and do
for adolescents, adults aged 20-54 years, and not provide a reliable estimate of the popu-
older adults aged 55-85 years, based on a lation mean.
review of relevant studies. The authors ad-
vocated the use of sample-spe~fic nor- Sample Composition Description
mative comparisons and emph~zed the Information regarding medical and psychiatric
importance of considering the in~ence of exclusion criteria is important. It is unclear if
demographic and other variables; on test gender, geographic recruitment region, socio-
performance. ! economic status, occupation, ethnicity, hand-
Medical condition and physical s'tus have edness, or recruitment procedures are relevant.
been shown to affect TMT performance. The Until this is determined, it is best that this in-
data for a large sample used in a mflticenter formation be provided.
prospective study of risk factors fori osteopo-
rotic fractures suggest that TMT Part B per- Age Group Intervals
formance is associated with bone; mineral This criterion refers to grouping of the data
density (Yaffe et al., 1999b), diabet~ (Gregg into limited age intervals. This requirement is
et al., 2000), depression (Yaffe et al~ 1999a), especially relevant for this test since a strong
exposure to toxic chemicals (Mo~ et al., effect of age on TMT performance has been
2001), and apolipoprotein E phenotype (Yaffe demonstrated in the literature.
et al., 1997) in females, as well as )with sex
hormone levels in both males and\ females Reporting of Education Levels
(Yaffe et al., 1998, 2002). Binder et ~· (1999) Given the association between education and
found that TMT part B performapce was TMT scores, information regarding educa-
significantly associated with objectiv~ evalua- tional level should be reported for each sub-
tion of physical function in a sampl~ of 125 group, and preferably normative data should
elderly participants. In addition, Kilander et al. be presented by educational levels.
(2000) reported a strong relationship l>etween
diastolic blood pressure at age 50 and per- Reporting of Intellectual Levels
formance on the TMT part B 20 years later in Given the relationship between TMT perfor-
a population-based study conducted fin Swe- mance and IQ, information regarding intel-
den on a sample of 502 men. lectual level should be reported for each
subgroup, and preferably normative data should
be presented by IQ levels.
METHOD FOR EVALUATING
THE NORMATIVE REPORTS
Procedural Variables
To adequately evaluate the TMT normative
reports, six key criterion variables were deemed Data Reporting
critical. The first five of these relate tq subject Means and standard deviations, and prefera-
variables, and the remaining dimensi~ refers bly ranges, for total time in seconds for each
to a procedural issue. Minimal requifements part of the TMT should be reported. Given
for meeting the criterion variables ivere as the demonstrated utility of the B-A differ-
follows. · ence, B:A ratio, and error analysis with some
TRAILMAKING TEST 71
clinical groups, reporting of means and SDs medical/psychiatric patients, V.A. inpatients
for these indices would facilitate interpreta- and outpatients, and homosexual/bisexual males.
tion of the results. Similar concerns regarding variability between
studies were raised by Soukup et al. (1998). In
addition, some investigators set the maximum
time for both parts A and B, which varies from
SUMMARY OF THE STATUS
180 to 300 seconds.
OF THE NORMS
The majority of studies report mean age,
Our review of the literature located TMT education, and gender distribution for the sam-
normative reports for adults, as well as three ple and/or for the age groups. Some studies
interpretive guides for the HRB (Gilandas report WAIS-R IQs or estimated intelligence
et al., 1984; Golden et al., 1981b; Reitan & level, handedness, occupational level, and eth-
Wolfson, 1985). Hundreds of other studies nic composition.
have also reported control subject data, and Many studies present data divided into age
we have included a discussion of those inves- groups. Few studies classify participants into
tigations based on well-defined samples that education or IQ groups or present data for
involved some unique features, such as large males and females separately; few studies re-
sample size, retest data, elderly population, port data for males only or present data in age
cutoff score analysis, reporting of derived by education by gender cells. Geographical
measures, error analysis, etc. origin of the data also varies widely: British,
It should be noted that Russell and Starkey Australian, and Canadian data sets are pre-
(1993) developed the Halstead-Russell Neu- sented in this chapter. Data for other cultural
ropsychological Evaluation System (HRNES), groups are also available in the literature (see
which includes the TMT among 22 tests. In above). The data are most commonly reported
the context of this system, individual perfor- as time to completion for parts A and B.
mance is compared to that of 576 brain- Some studies present raw data converted to T
damaged participants and 200 participants who scores, error rate, percentile ranks, median
were initially suspected of having brain damage time, total time for parts A and B, B-A dif-
but had negative neurological findings. Data ference, and B:A ratio. One study provides
were partitioned into seven age groups and regression equations to correct raw data for
three educationaVIQ levels. The authors pub- age and education. Few studies present clas-
lished an appendix to the manual (HRNES-R; sification rates for different cutoff criteria.
Russell and Starkey, 2001), which contains ta- Test-retest data are reported in some stud-
bles of scale scores based on the original ies with intertrial intervals ranging from 1 week
HRNES norms, demographic corrections, and to 24 weeks (in some studies, which are not
regression-based predicted scores. These data reviewed for the purposes of data collection,
will not be reviewed in this chapter because the up to 2 years). Issues of reliability and/or prac-
"normal" group consisted of the Veterans Ad- tice effect are discussed in these studies.
ministration patients who presented with symp- Given that use of the TMT has typically
toms requiring neuropsychological evaluation. been within the context of the HRB, the Re-
For further discussion of the HRNES system, itan data and interpretation recommendations
see Lezak et al. (2004, pp. 676-677). will be reported first, followed by a summary
There is a great deal of variability in the of the other interpretation formats and then
methodological aspects of studies summarized by the normative publications and control data
in this chapter. Sample sizes vary from 19 to from clinical studies, presented in ascending
over 700. Age represented in the studies var- chronological order.
ies from 15 to over 90 years. Sample compo- The text of study descriptions contains
sitions have been diverse and have included references to the corresponding tables iden-
neurologically normal individuals (according tified by number in Appendix 4. Table A4.1,
to stringent exclusion criteria), job applicants, the locator table, summarizes information
provided in the studies described in this Golden, Osmon, Moses, and Berg, 1981b
chapter. 1 (pp. 22-23)
The authors provided recommendations re-
garding the detection of laterality of brain
SUMMARIES OF THE STUDIES damage:
Reitan and Wolfson, 1985 part A is generally considered more a measure of right
hemisphere integrity (i.e., visual scanning skills),
The authors provided general guidelines for
where part B is more indicative of left hemisphere
TMT score interpretation in the form of test
intactness (i.e., language symbol manipulation and
completion times (in seconds), which corre-
direction of behavior according to a complex plan).
spond to "severity ranges" for part B only:
Therefore when one part indicates impairment rela-
0-60 sec: perfectly normal (or better than tive to the other part, a lateralized injwy may be
average) present. . . . Part A is considered to indicate greater
61-72 sec: normal impairment if the score on part B is less than twice the
73-105 sec: mildly impaired score on part A. Part B indicates greater impairment if
~106 sec: seriously impaired its score is more than three times the score on part A.
Tests in which the part B score lies between two times
No other information was provided, such as
and three times the part A score suggest that perfor-
score means, SDs, or any data regarding the
mances on the two parts are essentially equal.
normative sample on which these guidelines
were developed. These cutoffs represent a However, lateralizing properties of perfor-
substantial departure from cutoffs published mance time ratios for two conditions have
earlier; the definition of normal performance been repeatedly refuted in the literature
here is approximately 20 seconds less than in (Hom & Reitan, 1990; Salthouse et al., 1996).
the 1958 and 1979 guidelines.
Considerations regarding use of the study [TMT.l] Davies, 1968 (Table A4.2)
The authors argued that these norms were The author published TMT data on 540 British
meant as "general guidelines" and that "exact participants as a part of her investigation of
percentile ranks corresponding with each pos- the influence of age on TMT performance.
sible score are hardly necessary because the Test scores were obtained on 50 men and 40
other methods of inference are used to sup- women in each of six decade age groups. The
plement normative data in clinical interpreta- reference Davies cited as containing a further
tion of results of individual participants" (p. 97). description of her subject sample could not be
However, we maintain that more precise scores located.
as well as separate normative data for different Mean times in seconds corresponding to
age, IQ, and educational levels are necessary to lOth, 25th, 50th, 75th, and 90th percentile
avoid false-positive errors in diagnosis. ranks for parts A and B are provided for each
age decade, with the exception that the data on
Gilandas, Touyz, Beumont, and Greenberg, the participants in their 20s and 30s were
1984(p.l02) collapsed. Davies also reports optimal cutoff
points for young vs. middle-aged individuals.
The authors provided the percentile ranks
No significant gender differences were ob-
associated with Davies' (1968) TMT norma- served within any specific decade, although in
tive data and concluded that a percentile rank the group as a whole men performed slightly
of 25 is "mildly suggestive of brain damage" but significantly more quickly on part B.
and scores at the lOth percentile and lower
are "moderately suggestive of brain damage."
Study strengths
'Nonns for children are available in Baron (2004) and 1. Presentation of the data in 10- or 20-year
Spreen and Strauss (1998). age intervals.
TRAILMAKING TEST 73
2. Very large sample size, large Ns within 4. Small sample sizes in the upper age
each age subgroup, and fairly equal ranges.
representation of males and females. 5. Data were collected in Canada, raising
questions regarding their usefulness for
Considerations regarding use of the study clinical interpretation in the United
1. Lack of IQ and education data or States.
description of exclusion criteria.
2. Lack of test score SDs. [TMT.l] Wiens and Matarazzo, 1977
3. Tested in England, which may limit (Table A4.4)
generalizability for clinical interpretive The authors collected TMT data on 48 male
purposes in the United States. applicants to a patrolman program in Port-
land, Oregon, as a part of an investigation of
[TMT.2] Goul and Brown, 1970 (Table A4.3) the WAIS and Minnesota Multiphasic Per-
The authors tested 103 (or 106) Canadian sonality Inventory (MMPI) correlates of the
workers' compensation board non-brain- Halstead-Reitan battery. All participants pas-
injured patients who had been hospitalized for sed a medical exam and were judged to be
at least 3 months. These data were collected neurologically nonnal. Participants were di-
as a part of the authors' analysis of the effects vided into two equal groups, which were
of age and intelligence on TMT perfonnance. comparable in age (23.6 vs. 24.8 years), edu-
Participants had negative neurological histo- cation (13.7 vs. 14.0 years), and WAIS FSIQ
ries and included amputees, burn victims, and (117.5 vs. 118.3). TMT mean time in seconds
patients with lumbosacral fusions. Educational and SDs are provided for each group. A ran-
levels ranged from 6 to 13 years of fonnal dom subsample of 29 of the applicants was
schooling; no means are reported. Participants readministered the TMT 14-24 weeks fol-
were classified into five age groups: 20-29, lowing the original administration. Means and
30-39, 40-49, 50-59, and 60-72. Individual SDs for TMT times in seconds for both the
group sizes ranged from 15 to 26. Mean (SD) original testing and retest are reported. None
WAIS FSIQs for the five groups were 103.8 of the 29 original participants obtained scores
(12.1), 110.1 (8.9), 105.3 (7.9), 112.7 (8.6), and lower than Reitan's suggested cutoff for part
104.2 (12.2), respectively. B; however, one subject fell below the re-
TMT parts A and B data are presented in commended cutoff for part A on the second
tenns of mean time in seconds, SDs, ranges, administration. Correlations between test per-
medians, and recommended cutoff scores for fonnance and IQ scores were not meaningful.
the five age groups. Perfonnance declined In the first group, significant negative corre-
significantly with age. Contrary to expecta- lations were obtained between part B perfor-
tions, IQ was significantly positively correlated mance and FSIQ and VIQ, but no significant
with TMT scores. correlations were obtained between the sec-
ond control group and IQ measures. In group
Study strengths 1, a significant negative correlation between
1. Presentation of the data by age groups. part A and FSIQ and a significant positive
2. Infonnation on mean IQ and SD for correlation between part A and PIQ were
each age group. documented; again, no significant correlations
3. Infonnation on educational level and were obtained between part A scores and IQ
geographic area provided. measures in the second control group.
4. Means and SDs are reported.
Study strengths
Considerations regarding use of the study 1. Demographic characteristics of the sam-
1. Participants were medical patients with ple are presented in tenns of gender,
extensive hospitalizations. age, education, IQ, recruitment proce-
2. Lack of data regarding education means. dures, and geographic area.
3. Some variability in IQs across age groups. 2. Adequate medical exclusion criteria.
3. Means and SDs are reported: (n =20). Mean educational level and percent
4. The data are provided in a restricted age included in each of the diagnostic classifica-
range. tions are reported for each age group. The
5. Information on test-retest performance authors also provide test data on a subgroup of
is provided. 160 participants equated for percent diag-
nosed with alcoholism across the five age
Considerations regarding use of the study groups. The "alcohol-equated sample" was
1. High IQ level. developed "to minimize the influence that
2. Relatively small sample size. cognitive or motor/sensory differences un-
3. All-male sample. iquely attributable to alcohol abuse might
have upon group test performance levels"
[TMT.4] Eson, Yen, and Bourke, Personal (p. 2). This subsample remained heteroge-
Communication (Table A4.5) neous regarding representation of the other
The authors collected normative data for the diagnostic categories. Mean time in seconds,
TMT on a sample of 63 older patticipants. SD, and ranges are reported for parts A and B
Mean time in seconds and SDs are;provided for each age interval for the whole sample and
for four age groups, with mean agdi of 63.2, for the alcohol-equated sample.
67.0, 72.0, and 78.3 years. Samplel sizes for
each age group range between 15 ar!d 16. No Study strengths
other information is provided, such ias exclu- 1. Large sample size, with some individual
sion criteria or demographic data. : cells of approximately n =50.
2. Reporting of IQ data, geographic area,
Study strengths age, and education.
1. Data on an elderly sample are provided 3. Data presented in age groupings.
and stratified by age group. 4. Means and SDs are reported.
Considerations regarding use of the study
Considerations regarding use of the $rudy 1. The presence of substantial neurological
1. No reported exclusion criteria ,or other (chronic brain syndrome), substance
demographic, IQ, or geographic data. abuse, and major psychiatric disorders in
2. Age range and SDs for each group are the sample.
not reported. 2. Low educational level, though IQ levels
3. Relatively low sample sizes. are average.
3. No information regarding gender, but
[TMT.S] Harley, Leuthold, Matthews, given the V.A. setting, it is likely that
and Bergs, 1980 (Table A4.6) most or all of the sample was male.
The authors collected TMT data on 193 V.A.-
hospitalized patients in Wisconsin, ranging in Other comments
age from 55 to 79. Exclusion criteria if'tcluded The scores for the two oldest age groups are
FSIQ less than 80, active psychosis, qnequiv- identical in the whole sample and the alcohol-
ocal neurological disease or brain damage, and equated group because these two groups did
serious visual or auditory acuity p~blems. not have overrepresentation of alcoholics and,
Patients with chronic brain syndronle were thus, did not need to be adjusted.
included. Patient diagnoses were as follows:
chronic brain syndrome unrelated to ~cobol
ism (28%), psychosis (55%), alcoholis~ (37%), [TMT.6] Anthony, Heaton, and Lehman, 1980
neurosis (9%}, and personality disorddr (4%). (Table A4.7)
Mean educational level was 8.8 ye~. The The purpose of the study was to cross-validate
sample was divided into five age groups: two computer programs designed to deter-
55-59 (n =56), 60--64 (n = 45), J 65-69 mine the presence, location, and process of
(n=35), 70-74 (n=37), and 75-7~ years brain lesions using scores from the HRB and
TRAILMAKING TEST 75
the WAIS. Patients with structural brain documented between the two groups on both
lesions and normal controls were compared. parts of the test.
The control group consisted of 100 volunteers
with no medical or psychiatric problems and Study strengths
no history of head trauma, brain disease, or 1. The study provides data on a very elderly
substance abuse. The study was conducted in cohort not found in other published
Colorado. normative data.
TMT data are presented in terms of mean 2. Adequate exclusion criteria.
times in seconds and SDs for part B only. 3. Sample composition is well described in
terms of age, gender, education, 8uency
Study strengths in English, handedness, and geographic
1. Information regarding education, IQ, area.
age, and geographic area is provided. 4. Means and SDs are reported.
2. Large sample size.
3. Adequate exclusion criteria. Considerations regarding use of the study
4. Means and SDs are reported. 1. Sample sizes are small.
2. High IQ and educational level for the
Considerations regarding use of the study older age grouping.
1. Undifferentiated age grouping. 3. The older age grouping spans nearly two
2. The IQ range is high average. decades and may be too broad for opti-
3. No information is available regarding the mal clinical interpretive use.
gender ratio.
4. Data are provided for part B only. [TMT.B] Kennedy, 1981 (Table A4.9)
The author collected TMT data on 150 Ca-
[TMT.7] Bak and Greene, 1980 (Table A4.8) nadian participants as a part of his analysis of
The authors gathered TMT data on 30 right- the effects of age on TMT performance. Par-
handed Texan participants as a part of an ticipants were employees of a mental health
investigation of the effect of age on perfor- center "who represented diverse work roles"
mance on the HRB and the Wechsler Mem- randomly selected from five age groups: 20-
ory Scale. Participants were equally divided 29,30-39,40-49,50-59,60-69.
into two age groupings: 50-62 and 67-86. Participants were excluded who reported
Participants were 8uent in English and denied histories of "central nervous system disorders,
a history of CNS disorders, uncorrected sen- illnesses, or incapacities which would bias test
sory deficits, or illnesses or "incapacities" results;" exclusion criteria were not further
which might affect test results; participants in specified. Mean education was 13. 73, 13.53,
poor health were excluded. The mean (SD) 13.11, 11.59, and 12.50 years, respectively;
ages of the two groups were 55.6 (4.44) and those 50-59 years old were significantly less
74.9 (6.04), respectively. Participants in the educated than those 20-29 or 30-39 years old.
first group were born between 1916 and 1929, The Ammons Quick Test was used as an es-
and participants in the second group were timate of intelligence level; average estimates
born between 1892 and 1912. Nine individu- for the five groups were 123.43, 127.10,
als in the first group were female, and 127.40, 123.30, and 128.54, respectively.
10 participants in the second group were fe- Males and females were equally represented
male. Four WAIS subtests were administered in each group.
(Information, Arithmetic, Block Design, Digit The mean time in seconds and SDs for
Symbol); the mean scores on these measures parts A, B, and A+ B for each group are
suggested that IQ levels were within the high provided. Performance decreased significantly
average range or higher. with age, and significant negative correlations
Mean times in seconds and SDs for parts A between TMT test scores and education and
and B are presented for the two age groups. IQ suggest that lower education and IQ are
Significant differences in performance were adversely related to test performance.
Study strengths Study strengths

1. Large sample size, although the indi- 1. The large overall sample size.
vidual cells had only 30 participants per 2. Data are partitioned into five age groups.
cell. 3. Sample composition is described in
2. Presentation of the data in terms of age terms of IQ, educational level, age,
groupings. gender, handedness, recruitment proce-
3. Reporting of education, IQ estimates, dures, and geographic area.
gender, and geographic area. 4. Some psychiatric and neurological
4. Means and SDs are reported. exclusion criteria are used.
1. Very high mean intelligence scores. Considerations regarding use of the study
2. Some variability in educational level 1. High intellectual and educational levels
across groups, which may have led to of the sample.
some unusual findings; inexplicably, 2. Sample size for some age groups is very
those 60--69 years old performed either small.
as well as or slightly better than those 3. Data were collected in Canada, which
50-59 years old. may limit their usefulness for clinical
3. Vague exclusion criteria. interpretation in the United States.
4. Lack of reference to ethnicity/language 4. Essentially no differences in perfor-
issues and the fact that data were ob- mance were noted between those 18-
tained on Canadians, possibly reducing 23 years old and those 24-32 years old,
its generalizability for clinical interpre- suggesting that use of a single age
tation in the United States. grouping for 18-32 would have been
appropriate.
[TMT.9] Fromm-Auch and Yeudall, 1983
(Table A4.10) [TMT.10] Bornstein, 1985 (Table A4.11)
The authors obtained TMT data on 193 The author collected data on 365 Canadian
Canadian participants (111 male, 82 female) individuals (178 males and 187 females)
recruited through posted advertisements and recruited through posted notices on col-
personal contacts. Participants are described lege campuses and unemployment offices,
as "nonpsychiatric" and "nonneurological." newspaper ads, and senior-citizen groups.
Eighty-three percent of the sample were Participants were paid for their participation.
right-handed. Mean (SD) age was 25.4 (8.2) Participants ranged in age from 18 to 69 years,
years (range= 15--64). Mean (SD) education with a mean of 43.3 (17.1) years, and had
was 14.8 (3.0) years (range= 8-26) and completed 5-20 years of education, with a
included technical and university training. mean of 12.3 (2. 7) years. Ninety-one and a half
Mean (SD) WAIS FSIQ, VIQ, and PIQ were percent of the sample were right-handed. No
119.1 (8.8, range= 98-142), 119.8 (9.9, other demographic data or exclusion criteria
range=95-143), and 115.6 (9.8, range=89- are reported.
146), respectively. Of note, no subject ob- Mean time in seconds and SDs for parts A
tained an FSIQ which was lower than the and B are reported for three age groupings
average range. (20-39, 40-59, and 60--69 years), two educa-
Mean time in seconds, SDs, and ranges for tional levels (less than high school, greater
parts A and B are reported for five age than or equal to high school), and gender,
groupings: 15-17, 18-23, 24-32, 33-40, and resulting in a total of 12 separate groups.
41-64 years. Sample sizes range from 10 to 75. Individual group sample sizes ranged from 13
The two oldest age groupings had sample sizes to 86. Significant correlations were obtained
less than 20. No gender differences were between TMT scores and age and education,
documented, and male and female data were suggesting that better performance was asso-
collapsed. ciated with younger age and more years of
TRAILMAKING TEST 77
education. Females generally outperformed Testing was conducted by trained techni-

males on both parts A and B. cians, and all participants were judged to have
expended their best effort on the task. The
Study strengths TMT mean time in seconds for part B is
1. Very large overall sample size. reported for the six subgroups, as well as per-
2. Data are stratified by age, gender, and cent classified as normal using Russell et al.'s
educational level. (1970) criteria. Approximately 30% of the test
3. This data set is unique in that it reports score variance was accounted for by age and
data for participants with less than a high approximately 20% was associated with edu-
school education. cation level. Significant group differences in
4. Information on handedness, recruitment TMT scores were found across the three age
procedures, and geographic area is pro- groups and across the three education groups,
vided. and a significant age-by-education interaction
5. Means and SDs are reported. was documented. No significant differences in
performance were found between males and
Considerations regarding use of the study females.
1. Individual sample sizes of some cells are
small. Study strengths
2. Lack of any reported exclusion criteria. 1. Large size of overall sample and indi-
3. Data were collected on Canadian citi- vidual cells.
zens, which may limit generalizability for 2. Information regarding age, education,
their use in the United States. gender, handedness, and geographic
4. Lack of IQ data. The concern over the area is provided.
lack of IQ data is somewhat mitigated by 3. Adequate exclusion criteria.
the fact that the mean education level 4. Data are grouped by age and educational
was not unduly elevated (12.3 years), level.
which might suggest that mean intellec-
tual levels were within the average Considerations regarding use of the study
range. 1. No reporting of data for part A.
2. SDs are not provided.
[TMT.11 1 Heaton, Grant, and Matthews, 1986 3. Mean scores are reported for individual
(Table A4.12) WAIS subtest scaled scores but not for
The authors obtained TMT data on 553 nor- overall IQ scores.
mal controls in Colorado, California, and 4. Age groupings are quite large in terms of
Wisconsin as a part of an investigation into the ranges.
effects of age, education, and gender on HRB
performance. Nearly two-thirds of the sample [TMT.12] Alekoumbides, Charter, Adkins, and
were male (males= 356, females= 197). Seacat, 1987 (Table A4.13)
Exclusion criteria were history of neurological The authors report data on 118 medical and
illness, significant head trauma, and substance psychiatric inpatients and outpatients without
abuse. Participants ranged in age from 15 to cerebral lesions or histories of alcoholism or
81 years, with a mean of39.3 (17.5) years, and cerebral contusion from V.A. hospitals in
mean education was 13.3 (3.4) years, with a southern California as a part of their devel-
range of 0-20 years. The sample was divided opment of standardized scores corrected for
into three age categories (less than 40, 40-59, age and education for the HRB. Among the
and greater than or equal to 60 years) with 41 psychiatric patients, nine were diagnosed
319, 134, and 100 participants respectively, as psychotic and 32 as neurotic. In addition to
and into three education categories (less than psychiatry services, patients were drawn from
12 years, 12-15 years, and greater than or medicine (n =57), neurology (n = 22), spinal
equal to 16 years) with 132, 249, and 172 cord injury (n = 9), and surgery (n = 6) units.
participants, respectively. Mean age was 46.85 (17.17) years, ranging
from 19 to 82 years, and mean education was history of neurological or psychiatric illness.
11.43 (3.20) years, ranging from 1 to 20 years. Mean Verbal IQ was 105.8 (10.8), ranging
Frequency distributions for age and years of from 88 to 128, and mean Performance IQ
education are provided. Mean WAIS FSIQ, was 105.0 (10.5), ranging from 85 to 121.
VIQ, and PIQ were within the average range: Participants were administered the HRB in
105.89 (13.47), 107.03 (14.38), and 103.31 standard order both on initial testing and
(13.02), respectively. Means and SDs for in- again 3 weeks later. Means, SDs, and ranges
dividual age-corrected subtest scores are also for time in seconds to complete parts A and B
reported. All participants except one were for both testing sessions are provided, as well
male; the majority were Caucasian (93%), as raw score change and SD, median raw
with 7% African-American. The mean score score change, and mean percent of change.
on a measure of occupational attainment For part A, no significant correlations
was 11.29. between mean change and age or education or
No differences were found in test perfor- between mean percent of change and age or
mance between the two psychiatric groups education were documented. For part B, no
and the nonpsychiatric group, and the data significant correlations between mean change
were collapsed. Mean times to complete parts and age or education or between mean percent
A and B in seconds and SDs are reported. In change and education were found; however, a
addition, regression equation information to significant correlation did emerge between
allow correction of raw scores for age and mean percent of change and age.
education is included.
Study strengths
Study strengths 1. Information on short-term (3-week) re-
1. Large sample size. test data is provided.
2. Information regarding IQ, age, educa- 2. Sample composition is described in
tion, ethnicity, gender, occupational terms of age, VIQ, PIQ, and gender.
attainment, and geographic area is 3. Minimally adequate exclusion criteria.
provided. 4. Means, SDs, and ranges are reported.
3. Regression equation for computation of
age- and education-corrected scores is Considerations regarding use of the study
provided. 1. Undifferentiated age range.
4. Means and SDs are reported. 2. Small sample size.
3. No data on educational level.
1. The sample was heterogeneous in terms [TMT.14] Dodrill, 1987 (Table A4.15)
of medical diagnoses; psychiatric The author collected TMT data on 120 par-
patients were included in this sample, ticipants in Washington during the years 1975-
which was supposedly representative of 1976 (n = 81) and 1986-1987 (n = 39). Half of
"normal" participants. the sample was female, and 10% were minor-
2. Undifferentiated age range (mitigated by ities (six black, three Native American, two
the regression equation information). Asian American, one unknown). Eighteen
3. Nearly all-male sample. were left-handed, and occupational status
included 45 students, 37 employed, 26 unem-
[TMT.13] Bornstein, Baker, and Douglass, 1987a ployed, 11 homemakers, and one retiree.
(Table A4.14) Participants were recruited from various
The authors collected TMT test-retest data on sources, including schools, churches, employ-
23 volunteers (14 women, nine men) who ment agencies, and community service agen-
ranged in age from 17 to 52, with a mean age cies, and either paid for their participation or
of 32.3 (10.3), as part of an examination of the offered an interpretation of their abilities.
short-term retest reliability of the HRB. Exclusion criteria were history of "neurologi-
Exclusion criteria consisted of a positive cally relevant disease (such as meningitis or
TRAilMAKING TEST 79
encephalitis);" alcoholism; birth complications differ in years of education. Participants were

"of likely neurological significance;" oxygen recruited primarily through random selec-
deprivation; peripheral nervous system injury; tion from the Queensland State electoral roll
psychotic or psychosis-like disorders; or head (n = 97), with the remainder (n = 13) solicited
injury associated with unconsciousness, skull through senior-citizen centers. Exclusion cri-
fracture, persisting neurological signs, or di- teria were history of significant head trauma
agnosis of concussion or contusion. Of note, or neurological disease. Nearly one-half of the
one-third of potential participants failed to sample was diagnosed with at least one
meet the above medical and psychiatric crite- chronic disease (hypertension = 33, heart dis-
ria, resulting in a final sample of 120. Mean age ease= 9, thyroid dysfunction = 7, asthma= 5,
was 27.73 (11.04) years, and mean education emphysema= 2, diabetes= 1) for which they
was 12.28 (2.18) years. The participants tested were receiving treatment described as "well-
in the 1970s were administered the WAIS, controlled." Sixty-six of the participants were
whereas the participants assessed in the receiving medications, primarily for the dis-
1980s were administered the WAIS-R; WAIS eases listed above.
scores were converted to WAIS-R equivalents The test was administered according to
by subtracting 7 points from the VIQ, PIQ, Reitan's instructions. All participants were
and FSIQ. Mean FSIQ, VIQ, and PIQ administered the TMT first, followed by ei-
scores were 100.00 (14.35),100.92 (14.73), and ther the Tactual Performance Test or Booklet
98.25 (13.39), respectively. The IQ scores Category Test. Using the standard cutoffs of
ranged from 60 to 138 and reflected a normal 39 seconds and 92 seconds, 48% and 48%
distribution. of all participants were misclassified as im-
Mean time in seconds and SDs for parts A paired for parts A and B, respectively. Gender
and B are reported as well as IQ-equivalent differences were not significant; however,
scores for various levels of intelligence. education was significantly related to perfor-
Between 10% and 15% of the sample were mance on part B. There were no significant
misclassified as brain-damaged using cutoffs effects of chronic disease or medication
of 39 seconds for part A and 89 seconds for intake.
part B.
Study strengths
Study strengths 1. Large sample size in a restricted age
1. Large sample size. range.
2. Comprehensive exclusion criteria. 2. Presentation of the data by gender.
3. Sample composition is described in 3. Sample composition is described in terms
terms of education, IQ, occupation, of age, education, geographic recruit-
gender ratio, age, handedness, ethnicity, ment area, recruitment procedures, and
recruitment procedures, and geographic ethnicity.
area. 4. Information regarding test administra-
4. IQ-equivalent scores are provided. tion order effects is provided.
5. Data for different IQ levels are provided. 5. Means, SDs, and error rates are re-
6. Means and SDs are reported. ported.
Considerations regarding use of the study Considerations regarding use of the study
1. Undifferentiated age range. 1. Approximately half of the participants
had at least one chronic illness, and over
[TMT.15] Ernst, 1987 (Table A4.16) half were taking prescribed medications.
The author obtained TMT data on 110 pri- 2. No information regarding IQ.
marily Caucasian (99%) residents of Brisbane, 3. Low mean educational level.
Australia, aged 6~75. Fifty-nine were female 4. Data were collected in Australia and may
and 51 were male, with a mean educational be unsuitable for clinical use in the
level of 10.3 years; men and women did not United States.
[TMT.16] Stuss, Stethem, and Poirier, 1987 4. Adequate exclusion criteria.

(Tables A4.17 and A4.18) 5. Information regarding practice effect.
The authors collected normative data on 60 6. Means and SDs are reported.
Canadian English- or French-speaking par-
ticipants, who were recruited through per- Considerations regarding use of the study
sonal contacts or employment agencies and 1. Small sample sizes within each age
paid for their participation. Tests were ad- group.
ministered in each subject's native language. 2. Variability in mean educational levels
Participants were tested twice at 1-week across age groups; of importance, those
intervals. Exclusion criteria were abnormal 50-59 years old had the lowest mean
vision (even after correction); history of sub- educational level, the lowest mean test
stance abuse; presence of medical, neurolog- scores, and the largest SDs relative to
ical, and/ or psychiatric disorders; and current the other age groups.
use of psychotropic medication (Stuss, per- 3. Lack of IQ data.
sonal communication). Ten participants were 4. Unknown influence of language differ-
assigned to each of six age ranges: 16-19, 20- ences.
29, 30-39, 40-49, 50-59, and 60-69. Fifty-five 5. Data were obtained in Canada and may
percent of the sample were male, and 18% be of limited usefulness for clinical
were left-handed. Mean education was 14.3 interpretation in the United States.
(2.62) years. Data are provided regarding
[TMT.17] Yeudall, Reddon, Gill, and Stefanyk,
handedness, gender distribution, and educa-
1987 (Table A4.19)
tion.
Mean time in seconds and SDs for the two The authors obtained TMT data on 225
parts of the TMT for the first, second, and Canadian participants recruited from posted
combined testing sessions are reported for advertisements in workplaces and personal
each age interval. Mean time and SDs are also solicitations. The participants included meat
provided for males, females, those with less packers, postal workers, transit employees,
than or equal to 12 years of education, and hospital lab technicians, secretaries, ward
those with greater than 12 years of education, aides, student interns, student nurses, and
collapsed across age groupings. summer students. In addition, high school
Older participants and those with a high teachers identified for participation average
school education or less performed signifi- students in grades 10-12. The participants
cantly poorer than younger participants or (127 males and 98 females) did not report any
those with some college or university educa- history of forensic involvement, head injury,
tion. Educational level was somewhat irregu- neurological insult, prenatal or birth compli-
larly distributed across age groups, and the cations, psychiatric problems, or substance
authors suggest that the normative data be abuse.
used with caution. A practice effect was pres- Data were gathered by experienced testing
ent, but the authors question the clinical technicians who "motivated the participants
relevance of the improvement. No signifi- to achieve maximum performance" partially
cant gender differences in performance were through the promise of detailed explanations
present. of their test performance.
Means and SDs for time in seconds to
Study strengths complete parts A and B are presented for four
1. Presentation of the data by age group- age groupings (15-20, 21-25, 26-30, and
ings, education groupings, and gender. 31-40) for males and females combined and
2. Extensive information on educational separately. Information regarding percent
level. right-banders, mean years of education, and
3. Sample composition is described in terms mean WAIS/WAIS-R FSIQ, VIQ, and PIQ is
of age, gender, handedness, geographic reported for each age grouping and age-
location, and recruitment procedures. by-gender grouping. For the sample as
TRAILMAKING TEST 81
a whole, 88% were right-handed and had (n = 85). The average (SD) age for the sample
completed an average (SD) of 14.55 (2.78) was 62.7 (4.3), and the mean ages of the three
years of schooling. The mean FSIQ, VIQ, and educational groups were comparable: 62.3,
PIQ were 112.25 (9.83), 114.77 (10.34), and 62.9, and 63.0 years, respectively. Exclusion
108.50 (10.34), respectively. criteria were history of neurological or psy-
chiatric disorder.
Study strengths Significant group differences in perfor-
1. Large sample size. mance on both TMT A and B were obtained
2. Grouping of data by age. across the three education groups, which were
3. Data availability for a 1~20 year age due to the group with ~10 years of education
group. performing significantly worse than both of
4. Adequate medical and psychiatric the other education groups (which did not
exclusion criteria. differ from each other). Mean time in seconds
5. Information regarding age, handedness, and SDs for parts A and B are reported for the
education, IQ, gender, occupation, three education groups.
recruitment procedures, and geographic
area is provided. Study strengths
6. Means and SDs are reported. 1. Large overall sample size and individual
cell sizes are adequate.
Considerations regarding use of the study 2. Data are partitioned into three educa-
1. High educational level of the sample. tion groups; the study is unique in terms
2. Data were obtained on Canadian par- of representation of participants with
ticipants, which may limit their useful- less than 12 years of education.
ness for clinical interpretation in the 3. Information regarding gender, age, and
United States due to possible subtle geographic area is provided.
cultural differences. 4. Means and SDs are reported.
5. Minimally adequate exclusion criteria.
Other Comments 6. Reasonably restricted age grouping.
No significant correlations were found be-
tween age or education and part A; a signifi- Considerations regarding use of the study
cant correlation emerged for age and part B 1. No information regarding IQ.
(r = 0.27), but no significant relationship was 2. Greater than 12 years of education is too
documented between education and part B. large a category.
Significant correlations emerged between 3. Data collected in Canada, which may
parts A and B and PIQ but not VIQ. No sig- limit generalizability for use in the Uni-
nificant gender differences were observed for ted States.
part A or B. The authors recommend use of
the combined age group norms for part A and [TMT.19] Stuss, Stethem, and Pelchat, 1988
the separate age-grouped norms for part B. (Table A4.21)
In this publication, Stuss and colleagues
expanded the data presented in Stuss et al.
[TMT.18] Bomstein and Suga, 1988 (1987). The size of the sample is increased,
(Table A4.20) and the participants are collapsed into three
As part of their evaluation of the effect of age groupings of 30 participants each: 16-29,
educational level on neuropsychological test 30-49, and 50-69. Gender distribution was
performance in the elderly, the authors report essentially equal across groups. Mean years of
TMT data on 134 healthy elderly Canadian education for the youngest to oldest groups
paid volunteers aged 5~70 according to were 14.1 (1.34), with a range of 11-18; 14.9
three educational levels: ~10 (n =46), 11-12 (3.95), with a range of ~20; and 13.2 (2.38),
(n = 44), and greater than 12 (n = 44) years. with a range of 8-18, respectively (compare to
Nearly two-thirds of the sample were female TMT.16).
Mean time in seconds and SDs for the two Study strengths
parts on the initial test and retest 1 week later 1. Large overall sample size.
are reported for each age interval. The,authors 2. Small age range within each grouping.
call attention to the skewness and i lack of 3. Adequate exclusion criteria.
normal distribution of the test datli, which 4. Information on age, IQ, education, gen-
they suggest have implications for t~t score der, and geographic recruitment Mea is
interpretation. pro~ded.
5. Means and SDs Me reported.
Study strengths
1. Increased sample size per age iqterval.
2. Adequate exclusion criteria. I
1. High intellectual level of the sample.
3. Information regarding age, e~cation,
2. Relatively high educational level.
gender, and handedness is provif!ed.
3. Inexplicably, part B performance was
4. Data regarding retest at 1-week ibtervals lower in those 66-70 years old relative to
Mepro~ded. :
those 71-75 yeMs old, and there
5. Means and SDs Me reported.
appeMed to be considerably more varia-
tion in performance in the 66-70 age
Considerations regarding use of the stpdy
grouping. Given that increasing age is
a;
1. Considerations remain the same for the
associated with a worsening of perfor-
initial report except for the imprqvement mance, the data for the 66-70 yeM group
in sample size. ! Me problematic.
[TMT.20] Van Gorp, Satz, and Mitrushh..,
1990 (Table A4.22) I [TMT.21] Heaton, Grant, and Matthews, 1991
The authors present TMT data for 156 pealthy The authors pro~ded normative data on the
elderly participants ranging in age fro~ 57 to TMT from 486 (378 in the base sample and
85, recruited from an independent-li~ng 108 in the validation sample) urban and rural
retirement community in California. llle data participants recruited in several U.S. states
were collected as a part of their investigation (California, Washington, Colorado, Texas,
of cognitive changes in normal aging1 Infor- Oklahoma, Wisconsin, Illinois, Michigan, New
mation regMding general medical status was York, Virginia, and Massachusetts) and
collected. Participants with a history of neu- Canada. Data were collected over a 15-year
rological or psychiatric disorder or substance period through multicenter collaborative
abuse were excluded. Sixty-one percent: of the efforts.
sample were females. Mean educati~n was Sixty-five percent of the sample were males.
14.14 (2.86) years, and mean FSIQ (\VAIS-R Mean age for the total sample was 42.06 (16.8)
Satz-Mogel format) was 117.21 (12.59, years. yeMs, and mean educational level was 13.6
Mean time in seconds and SDs to cdmplete (3.5) yeMs. The majority of participants were
parts A and B were listed for the sample as a administered the WAIS; mean FSIQ, VIQ,
whole and for four age groups: 57-65,:66-70, and PIQ were 113.8 (12.3), 113.9 (13.8), and
71-75, and 76-85. Sample sizes for e~h age 111.9 (11.6), respectively. Exclusion criteria
group ranged from 26 to 57. Mean VJQ and were history of learning disability, neurologi-
PIQ and SDs for each age range M~ listed. cal disease, illness affecting brain function,
Mean VIQs were consistently within ~e high significant head trauma, significant psychiatric
average range, except for those 71-75 years old, disturbance (e.g., schizophrenia), and alcohol
who fell within the superior range. Me~ PIQs or other substance abuse.
were within the high average range for those The TMT was administered according
57-65 years old and 76-85 yeMs old~ Older to procedures outlined by Reitan and Wolfson
participants (70 or older) did not diffec signif- (1985), with the exception that attempts
icantly from younger participants (less ~an 70) to complete part B were limited to 10 min-
in VIQ, PIQ, or years of education. ' utes. In those situations when part B was
TRAILMAKING TEST 83
discontinued at 10 minutes, the time score 5. The normative data are presented
was prorated by dividing 300 seconds by the in comprehensive tables in T-score
number of items completed and then multi- equivalents for males and females sepa-
plying the resulting figure by 25. Participants rately in 10 age groupings by six educa-
were generally paid for their participation and tional groupings.
judged to have provided their best efforts on
the tasks. Considerations regarding use of the study
The normative data, which are not re- 1. Above average mean intellectual level
produced here, are presented in compre- (which is probably less of an issue given
hensive tables in T-score equivalents for that these are WAIS rather than WAIS-R
test scaled scores for males and females sep- IQ data).
arately in 10 age groupings (20-34, 35-39,
40-44, 45-49, 50-54, 55-59, 60-64, 65-69, [TMT.22] Seines, Jacobson, Machado, Becker,
70-74, 75-80 years) by six educational group- Wesch, Miller, Visscher and McArthur, 1991
ings (6--8, 9-11, 12, 13-15, 16-17, ~18 (Table A4.23)
years). The investigation used participants from the
For part A, 30% of the score variance was Multi-Center AIDS Cohort Study (MACS).
accounted for by age, while 16% was attrib- The article presents data for seronegative
utable to educational level; gender accounted homosexual and bisexual males collected in
for a negligible amount of unique variance in Los Angeles for the purpose of establishing
performance (1%). A total of 35% of the test normative data for neuropsychological test
score variance was accounted for by demo- performance based on a large sample. Partic-
graphic variables. For part B, 34% of the score ipants with a history of head injury with loss of
variance was accounted for by age, while 27% consciousness greater than 1 hour and who
was attributable to educational level; again, reported drinking 21 or more drinks per week
gender accounted for a negligible amount of in the previous 6 months were excluded.
unique variance (1%). A total of 45% of the The majority of the sample consisted of
test score variance was accounted for by de- Caucasian participants. African-American par-
mographic variables. ticipants ranged from 3.4% to 4.1% for dif-
For the sample as a whole, mean time in ferent age groups. Left-banders ranged from
seconds for part A was 29.0 (12.5) and that for 11.3% to 14.9%.
part B was 75.2 (42.8).
The interested reader is referred to the Study strengths
Fastenau and Adams (1996) critique of Hea- 1. The overall sample size and individual
ton et al. (1991) norms and Heaton et al.'s cell sizes are large.
(1996a) response to this critique. 2. Normative data are stratified by age and
In 2004, the authors published the revised education.
norms, which are based on a sample of over 3. The demographic composition of the
1,000 normal adults. In addition to age, sample is described in terms of age, gen-
education, and gender stratification, the data der, sexual orientation, handedness, eth-
are partitioned by race/ethnicity (African- nicity, and geographic area; demographic
American and Caucasian). composition is described for each age and
education cell separately.
Study strengths 4. Means, SDs, as well as scores for the 5th
1. Large sample size. and lOth percentiles are presented.
2. Comprehensive exclusion criteria. 5. Minimally adequate exclusion criteria.
3. Detailed description of the demographic
characteristics of the sample in terms of Considerations regarding use of the study
age, education, IQ, geographic area, and 1. All-male sample.
gender. 2. No information on IQ is reported.
4. Administration procedures are outlined. 3. Very high educational level of the sample.
[TMT.23] Elias, Robbins, Walter, and Schultz, of the Alzheimer's Type (DAT) in a community-
1993 (Table A4.24) dwelling elderly sample. The participants are
The authors explored the influence of gender stable, upper middle-class, retired older adults
and age on performance on tests included who entered the Rancho Bernardo Study,
in the HRB. The sample consisted of 427 surveying for heart disease risk factors, be-
community-dwelling volunteers. As per med- tween 1972 and 1974. The initial sample in-
ical interview and self-report on the Cornell cluded 5,052 adults between 30 and 79 years of
Medical Index, none of the participants had a age, who have been followed until the present.
history of treatment for neurological disorder, Participants over the age of 65 who returned
senility, alcoholism, brain trauma, mental ill- for a reexamination in 1988 and later and
ness, cerebral vascular or catastrophic disease, screened positive for cognitive impairment
or a diagnosis of senile dementia. To achieve were seen in clinic for diagnostic pmposes
equivalence between age groups in terms (n = 199). A matched control sample of 203
of education, the lower and upper limits for normal elderly participants who screened neg-
education were set at 12 and 19 years, respec- ative for cognitive impairment was randomly
tively. All participants had normal or corrected- selected for the comprehensive evaluation,
to-normal vision. Occupations ranged from which included neurological examination, neu-
blue-collar to professional. Non-age-corrected ropsychological assessment, standard medical
WAIS Vocabulary scaled scores ranged from history and examination, and, in some cases,
13.9 to 14.7, and Information scores ranged CT scans of the brain.
from 13.2 to 13. 7. On the basis of the diagnostic evaluation, the
Mean time in seconds and SDs to complete group composition was re-assessed. The final
parts A and B were reported for six age groups sample of normal elderly included 238 partic-
(15-24, 25-34,35-44,45-54,55-64, and ;:::65) ipants (97 males, 141 females), with a mean age
for males and females separately. of 78.4 (6.8), education of 13.8 (2.6), and De-
The authors found significant linear trends mentia Rating Scale (DRS) score of136.8 (5.4).
across age cohorts for parts A and B. The TMT was administered as part of a
larger battery by a trained psychometrist who
was blind to the participants' group assign-
Study strengths
ment. Time to completion was reported for
1. Large overall sample and adequate
the entire sample. In addition, the authors
sample size for individual cells.
2. The sample composition is well described provided optimal cutoff scores and sensitivity/
in terms of age, education, gender, and specificity of the TMT for the diagnosis of
DAT: 69%/90% for part A at the cutoff of
WAIS Vocabulary and Information scaled
66 seconds and 87%/88% for part B at the
scores.
cutoff of 172 seconds.
3. Rigorous exclusion criteria.
4. Means and SDs for the test scores are
Study strengths
reported.
2. The sample composition is well described
Considerations regarding use of the study in terms of age, education, gender, DRS
1. Education and estimated intelligence score, geographic area, history of the pro-
level for the sample are high. ject, and recruitment procedures.
2. Age range for the oldest group is not 3. Rigorous exclusion criteria.
reported. 4. Test administration procedures are
specified.
[TMT.24] Cahn, Salmon, Butters, Wiederholt, 5. Means and SDs for the test scores are
Corey-Bloom, Edelstein, and Barrett-Connor, reported.
1995 (Table A4.25) 6. Sensitivity and specificity for optimal
The study examined the accuracy of neuro- cutoff scores for the two parts of the test
psychological measures at detecting Dementia are reported.
TRAILMAKING TEST 85
Considerations regarding use of the study the tables, to adjust the patient's score for ed-
1. The data are not partitioned by age ucation. Age- and education-corrected scores
group. for the TMT (A&E-MSS) can be calculated as
2. No information on IQ is reported. follows:
[TMT.25] lvnik, Malec, Smith, Tangalos, A&E-MSSn1r=K+(W 1 * A-MSSn.n)

and Petersen, 1996 (Table A4.26) - (W2 * Education)
The study provides age-specific norms for the
TMT obtained in Mayo's Older Americans where the following indices are specified for
Normative Studies (MOANS) projects, which the two parts of the TMT:
aim at obtaining normative data for elderly in-
Part A Part B
dividuals on different neuropsychological tests.
The total sample consisted of 746 cognitively
1.99 3.38
normal volunteers over age 55; however, only
1.10 1.06
359 volunteers participated in TMT testing.
0.21 0.29
Mean MAYO FSIQ (which differs somewhat
from standard WAIS-R FSIQ) for the whole
sample was 106.2 (14.0) and mean Mayo Gen- Education should enter the formula in years
eral Memory Index on the WMS-R was 106.2 of formal schooling.
(14.2). For a description of their samples, the The tables of scaled scores per age group
authors refer to their earlier publications. Par- provided by the authors should be used in the
ticipants were independently functioning, context of the detailed procedures for their
community-dwelling persons who were recently application, which are explained in Ivnik et al.
examined by a physician and had no active (1996). Therefore, they are not reproduced in
neurological or psychiatric disorder with the this book. Interested readers are referred to
potential to impact cognition. the original article. Table A4.26 in Appendix 4
Age categorization used the midpoint summarizes sample sizes for different demo-
interval technique. The raw score distribution graphic groups.
for each test at each midpoint age was "nor-
malized" by assigning standard scores with a
Study strengths
mean of 10 and SD of 3, based on actual
1. Information regarding age, education,
percentiJe ranks. The authors provided tables
gender, ethnicity, occupation, recruit-
of age-corrected norms for each age group.
ment procedures, and geographic area is
The procedure for clinical application of these
reported.
data are described in the original article (Ivnik
2. The data were stratified by age group
et al., 1996) as follows:
based on midpoint interval technique.
first select the table that corresponds to that per- 3. The innovative scoring system was well
son's age. Enter the table with the test's raw score; described. The authors developed new
do not use corrected or final scores for tests indices of performance.
that might present their own age- or education- 4. The sample sizes for most groups are
adjustments. Select the appropriate column in the large.
table for that test. The corresponding row in the 5. Restricted age range in each cell.
left-most column in each table provides the
MOANS Age-Corrected Scaled Score . . . for
your subject's raw score; the corresponding row in
1. The measures proposed by the authors
the right-most column indicates the percentile
are quite complicated and might be dif-
range for that same score.
ficult to use in clinical practice.
Further, linear regressions should be 2. Participants with prior history of neuro-
applied to the normalized, age-corrected logical, psychiatric, or chronic medical
MOANS scaled scores (A-MSS) derived from illnesses were included.
Other comments 2. Information on age, education, gender,

1. The theoretical assumptions underlying and geographic location is reported.
this normative project have been pre- 3. Exclusion criteria are described.
sented in lvnik et al. (1992a,b). 4. The data are classified into two age
2. The authors cautioned that the validity groups by two education groups.
of the MAYO indices depends heavily on 5. Means and SDs are reported.
the match of demographic features of
the individual to the normative sample Considerations regarding use of the study
presented in this article. 1. Only part B part of the test was admin-
3. Correlations of parts A and B with age istered.
were 0.30 and 0.53, respectively, whereas 2. No information on IQ is reported.
correlations with education and gender 3. Sample sizes for each age-by-education
were negligible. cell are relatively small.
[TMT.26] Richardson and Marottoli, 1996 [TMT.27] Hoff, Riordan, Morris, Cestaro,
(Tables A4.27 and A4.28) Wieneke, Alpert, Wang, and Volkow,
The authors report data for 101 autonomously 1996 (Table A4.29)
living elderly participants who comprise a The authors used the TMT in a study
subsample of a cohort of participants in Pro- exploring the relationship of cocaine use to
ject Safety, a study on driving performance performance on neuropsychological tests tap-
conducted in New Haven, Connecticut. Indi- ping functions of frontal and temporal brain
viduals with a history of neurological disease regions. The performance of crack cocaine
or excessive use of alcohol or those who were users was compared to that of a control
at risk for dementia based on MMSE scores group consisting of 54 paid male volunteers
were excluded. with a mean age of 32.1 (9.7) years and mean
The sample includes 53 males and 48 fe- education 15.4 (2.4) years. The sample in-
males, with a mean age of 81.47 (3.30) years cluded 48 white, four black, and two His-
and mean education of 11.02 (3.68) years. panic participants. Exclusion criteria were a
Part B was administered and scored according history of medical, neurological, or psychiat-
to the standard instructions provided in the ric problems; more than moderate use of al-
test manual. cohol (12 oz./week); history of intravenous
The data were divided into two age groups of drug use; and self-reported history of learn-
younger-old (76--80 years) and older-old (81- ing disability (with enrollment in special ed-
91 years) and two education groups. The results ucation classes).
indicated that the mean performance for par-
ticipants with less than 12 years of education Study strengths
was stable across the younger-old and older-old 1. Relatively large sample size
age groups; however, it was considerably lower 2. The sample composition is described in
than for participants with 2:12 years of educa- terms of age, education, and ethnicity.
tion and well below expectation in comparison 3. Rigorous exclusion criteria.
to the Heaton et al. (1991) norms. For the 4. Means and SDs for the test scores are
participants with 2: 12 years of education, per- reported.
formance for the younger-old age group was
superior to that of the older-old and comparable
to the norms published by Heaton et al. (1991).
1. Wide age and education range. No in-
formation on IQ or gender distribution.
Study strengths 2. Recruitment procedures were not re-
1. Data for a relatively large sample of very ported.
elderly participants are presented. 3. Education level for the sample is high.
TRAILMAKING TEST 87
[TMT.28] Salthouse, Toth, Hancock, of Aging (BLSA). The sample has been
and Woodard, 1997 (Table A4.30) recruited continuously since 1958, and par-
The authors examined controlled and auto- ticipants were asked to return for testing every
matic processes underlying memory and atten- other year. The majority of the sample are
tion using the process-dissociation procedure, white (37% female); working or retired from
as well as the uniqueness of age-related influ- scientific, professional, or managerial posi-
ences on these processes. Participants were tions; graduated from college (71% ); and
115 healthy adults (47% male, 53% female) married. All participants aged 70 years and
between the ages of 18 and 78 years, who were older and some younger participants who met
recruited from appeals to groups and specific criteria were seen by a neurologist for
acquaintances. They were included in the a clinical evaluation, who classified partici-
study if they were in "reasonably good health," pants in three categories: cognitively normal,
were not currently students, and had at least suspect for early dementia, or dementia.
11 years of education. No other exclusion cri- The 667 nondemented participants who
teria are reported. Participants were adminis- were included in the TMT portion of the
tered a battery of neuropsychological tests in study were 60 years of age or older at the last
their homes. The data were stratified into visit at which the TMT was administered. The
three age groupings: 18-39 (mean age= 29.0, mean age of the sample was 74.4 (8.2) years,
SD = 4.8; mean education= 15.5, SD = 1.7), mean education 16.0 (2.9) years, mean MMSE
40-59 (mean age=49.1, SD=5.1; mean ed- score 28.6 (1.6), and mean number of errors on
ucation= 15.2, SD=2.5), and 60-78 (mean the Blessed Mental Status Exam 1.3 (1.8).
age= 69.2, SD = 5.1; mean education= 15.3, The TMT was administered according to
SD = 2.6) years. the standard procedures. A maximum of
The TMT was administered according to 300 seconds was allowed for each part. The
the standard instructions. authors provided a detailed description of the
administration procedures.
Study strengths The authors found a significant effect of age
1. Sample size is large. on completion times for both parts A and B.
2. Sample composition is well described in Incidence of errors increased with age only for
terms of age, education, gender, and part B. Dementia status was significantly
various health indices. associated with the proportion of participants
3. Recruitment procedures are specified. making errors on both parts A and B, inde-
4. Data are partitioned into three age pendent of age. The error rates did not
groups. increase over a 2-year longitudinal comparison
5. Test administration procedures are made on a subset of the nondemented sample.
specified. The authors described the sensitivity and
6. Means and SDs for the test scores are specificity of various cutoff scores in distin-
reported. guishing between nondemented participants
and those with cognitive dysfunction, based
on receiver operating characteristic (ROC)
Consklerations regarding use of the study
analyses. They evaluated on their sample
1. Exclusion criteria are not well identified.
sensitivity and specificity of the previously
2. High educational level for each age
reported optimal dementia cutoff score on
group.
part B of 172 seconds reported by Cahn et al.
(1995). All significant effects were replicated.
[TMT.29] Rasmusson, Zonderman, Kawas,
and Resnick, 1998 (Table A4.31) Study strengths
The authors explored the effect of age and 1. Large sample size.
dementia status on TMT performance within 2. The sample composition is well
the scope of the Baltimore Longitudinal Study described in terms of age, education,
gender, geographic area, and recruit- [TMT.31] Crowe, 1998b (Table A4.33)
ment procedures. The TMT and a series of measures derived
3. Mental status was assessed with MMSE from it were administered to 98 undergradu-
and Blessed Mental Status Exam. ate students from La Trobe University in
4. Adequate exclusion criteria. Melbourne, Australia, in order to examine
5. Performance for very old group (90- cognitive mechanisms contributing to perfor-
96 years) is reported. mance on both parts. Participants were
6. Test administration procedures are thor- screened for a history of loss of consciousness
oughly described. or other neuropathology. The mean age for
7. Means and SDs for the test scores and the sample was 23.4 (3.1) years, mean edu-
the percentage of participants who made cation 14.0 (2.3) years, and mean Wide Range
errors on parts A and B are reported. Achievement Test (WRAT) Reading score
8. Data are partitioned by four age groups. 101.0 (9.0).
The authors developed modified proce-
Considerations regarding use of the study dures in an effort to separate cognitive
1. Education levelfor the sample is very high. mechanisms contributing to TMT perfor-
2. No information on IQ is reported. mance. They concluded that visual search
and motor speed contributed to performance
[TMT.30] Miner and Ferraro, 1998
on part A, whereas visual search and cognitive
(Table A4.32)
alternation contributed to performance on
The study examined the role of different part B. The latter was further influenced by
information-processing factors and presenta- reading level, ability to mentalJy maintain two
tion order in TMT performance. The sample simultaneous sequences, attention, and work-
consisted of 110 undergraduate students (88 ing memory.
females and 22 males) from the University of Time to compJetion for both TMT parts is
North Dakota, with a mean age of 21.7 (5.24) provided.
years, who received a course credit for their
participation. Their health was assessed with a Study strengths
background information questionnaire and 1. Large sample size.
with the Geriatric Depression Scale. 2. The sample composition is described in
The TMT was administered in a counter- terms of age, education, gender, WRAT
balanced order as part of a larger battery. Those Reading score, and geographic area.
participants who received the test in the part 3. Minimally adequate exclusion criteria.
B-part A order demonstrated considerably 4. Test administration procedures are
slower performance on part Bin comparison to specified.
the group tested in the standard order. 5. Means and SDs for the test scores are
reported.
Study strengths
1. Relatively large sample.
2. The sample composition is described
1. No information on IQ is reported.
in terms of age, education, gender, and
2. High educational level of the sample.
incentive for participation.
3. The data were obtained on Australian
participants, which may limit their use-
4. Test administration procedures are
fulness for clinical interpretation in the
specified.
United States.
reported.
[TMT.32] Tremont, Hoffman, Scott,
Considerations regarding use of the study and Adams, 1998 (Table A4.34)
1. Exclusion criteria are not described. The authors challenged Dodrill's (1997)
2. No information on IQ is reported. findings of no relationship between level of
TRAILMAKING TEST 89
intelligence and neuropsychological test per- later. The composition of the latter sample was
formance by presenting data collected from 48 Caucasian, one African-American, and one
archival files of the University of Oklahoma Hispanic, with a mean age of 32.5 (9.27) years,
Neuropsychological Laboratory, stratified by a mean education of 14.98 (1.93) years, and a
intelligence level. The data included files mean FSIQ of 109.30 (12.29) at baseline. At
for 157 patients (71 males and 86 females) each probe, participants were screened for
between 16 and 74 years of age, with a mean neurological disease, head injury, learning
age of 39.38 (15.80) and a mean education of disabilities, or other medical illnesses based
13.12 (3.26); 143 were Caucasian, nine Afri- on an informal interview. They were also
can-American, and the rest other races, and or screened for psychiatric disorders through a
unknown. All patients were evaluated for structured clinical interview. None was ex-
suspected neurological disease, which yielded cluded based on these screens.
no biomedical evidence for brain impairment. The TMT was administered according to
The TMT was administered as part of the standard procedures by thoroughly trained
HRB. The results are stratified by three in- and supervised technicians. The authors
telligence levels, based on patients' WAIS-R compared TMT performance at baseline and
FSIQ. on the retest using reliable change indices and
The authors concluded that performance on concluded that TMT scores did not change on
both parts of the test was affected by intelli- the retest.
gence level, with the greatest impact on part B. Performance on the TMT for the two
probes is reported for the entire sample.
Study strengths
1. Relatively large sample. Study strengths
2. The sample composition is well described 1. Adequate sample size.
in terms of age, education, gender, VIQ, 2. The sample composition is described in
PIQ, FSIQ, geographic area, and clinical terms of age, education, gender, ethni-
setting. city, FSIQ, and recruitment procedures.
3. It is presumed that standard adminis- 3. Adequate exclusion criteria.
tration procedures were used since 4. Test administration procedures are thor-
the TMT was administered as part of the oughly described.
HRB. 5. Means and SDs for the test scores are
4. Means and SDs for the test scores are reported.
reported.
5. Data are stratified by intelligence level. Considerations regarding use of the study
1. The data are not partitioned by age
Considerations regarding use of the study group.
1. Wide age range. 2. Education level for the sample is high.
2. Data were collected from patients' files.
Though biomedical evidence for brain [TMT.34] Crews, Harrison, & Rhodes, 1999
impairment was negative, this is not a (Table A4.36)
normal sample. A control sample of 30 nondepressed women
was used in a study on the effect of depression
[TMT.33] Basso, Bornstein, and Lang, 1999 on executive functions in young women.
(Table A4.35) Control participants were recruited via flyers/
The study examined the practice effect on re- sign-up sheets from town and university set-
peated administration of several tests over a tings. They did not meet diagnostic criteria
12-month interval. The baseline sample con- according to the ADIS-R and scored within
sisted of 82 men recruited through newspaper the nondepressed range on the Beck Depres-
advertisements, who were not paid for their sion Inventory (BDI). The exclusion criteria
participation. Fifty men out of this sample were past or present history of neurological
returned for the repeated testing 12 months problems or psychiatric disorders, alcoholism
or drug abuse, learning disabilities, concurrent brain injury. The rest of the participants
medication/drug usage, eating disorders, or in these samples denied any history of condi-
current medical illness. tions that might be expected to affect brain
The TMT was administered according to function. The third group, mixed normal con-
the standard procedures. Test performance is trols, consisted of 125 participants who had no
reported for the entire sample. history of trauma or disease involving the
brain. They were enrolled in longitudinal re-
Study strengths search projects at multiple sites under the
1. The sample composition is described in supervision of the neuropsychology laborato-
terms of age, education, gender, scores ries at the University of Colorado and the
on selected WAIS-R tests, and recruit- University of California at San Diego. Their
ment procedures. mean age was 43.6 (19.6) years and mean
2. Adequate exclusion criteria. education was 12.0 (3.3) years; 68% of the
3. Test administration procedures are sample were males, and the test-retest inter-
specified. val was 5.4 (2.5) months. The data are re-
4. Means and SDs for the test scores are ported for all groups combined. Demographic
reported. information for all groups combined is also
provided. The mean WAIS FSIQ (Wechsler,
Considerations regarding use of the study 1955) on the initial testing for the three
1. The sample is small. groups combined was 108.8 (12.3).
2. Education level for the sample is high. Trails A and B were administered according
to the procedures specified by Reitan and
[TMT.35] Dikmen, Heaton, Grant, Wolfson (1993). Time limits were imposed of
and Temkin, 1999 (Table A4.37) 100 seconds on Trails A and 300 seconds on
The TMT was used in a study on the psy- Trails B.
chometric properties of a broad range of The authors provide raw scores for perfor-
neuropsychological measures, based on a mance at two time probes, as well as various
sample of 384 normal or neurologically stable measures of test-retest reliability and magni-
adults who were tested twice as part of several tude of practice effect. The test-retest
longitudinal studies. A group of "friend con- reliability over an 11-month interval for Trails
trols" consisted of 138 individuals who had no A was r = 0. 79 and that for Trails B was
history of recent trauma and were friends of r=0.89.
head-injured patients. Their mean age was
28.5 (12.2) years and mean education was 12.2 Study strengths
(1.9) years; 60% of the sample were males, 1. Large sample sizes for the three groups.
and the test-retest interval was 11.1 (0.6) 2. The sample composition is well described
months. A group of "trauma controls" con- in terms of age, education, gender, IQ,
sisted of 121 individuals who had a recent geographic area, and setting.
traumatic injury that did not involve the head. 3. Test administration procedures are spec-
They were tested at baseline 1 month after ified.
trauma and then 11 months later. Their mean 4. Means and SDs for the test scores are
age was 31.2 (13.6) years and mean education reported.
was 12.0 (2.6) years; 70% of the sample were 5. Information on test-retest reliability is
males, and the test-retest interval was 10.7 provided.
(0.6) months. Both of these groups were tes-
ted at the University of Washington under the Considerations regarding use of the study
direction of one of the authors. Twenty per- 1. Exclusion criteria are not clear]y de-
cent of friend controls and 46% of trauma scribed. As the authors pointed out, 20%
controls had preexisting conditions that might of friend controls and 46% of trauma
affect test performance, the most significant controls had preexisting conditions that
being alcohol abuse or a significant traumatic might affect test performance, the most
TRAILMAKING TEST 91
significant being alcohol abuse and a Study strengths

significant traumatic brain injury. 1. Large sample size.
2. The data are not partitioned by age 2. The sample composition is well described
group. in terms of age, education, gender, eth-
3. Time limits were imposed on test per- nicity, indices of physical health, Blessed
formance that deviated from test score, Geriatric Depression Scale score,
administration procedures. However, geographic area, and research setting.
these limits should not have had a no- 3. Adequate exclusion criteria.
ticeable effect on the results. 4. Test administration procedures are
specified.
[TMT.36] Binder, Storandt, and Birge, 1999 5. Means and SDs for the test scores are
(Table A4.38) reported.
The authors examined the relationship
between performance on psychometric tests Considerations regarding use of the study
and a modified Physical Performance Test 1. The data are not partitioned by age
(modified PPT) in a sample of 125 adults aged group.
75 years and older, who participated in trials 2. No information on IQ is reported.
of exercise or hormone replacement therapy.
The study was approved by the Washington [TMT.37] Ruffolo, Guilmette, and Willis, 2000
University School of Medicine, St. Louis. The (Table A4.39)
mean age for the sample was 82.3 (4.4) years, Time to completion and number of errors in
mean education was 13.5 (3.0) years, 25% TMT performance are compared for four
were male, and 87% were Caucasian. Indices clinical I experimental groups and a control
of physical health, Blessed score, and Geriat- group. The latter sample included 49 intro-
ric Depression Scale score are reported. Pre- ductory psychology students, graduate stu-
liminary screening included a medical history; dents, and employees of a local social services
physical examination; the Short Blessed Test agency, who were screened for any prior head
of memory, concentration, and orientation; injuries. The TMT was administered accord-
blood and urine chemistries; a chest X-ray; ing to standard instructions.
and a cross-validated self-report regarding
health problems in the previous 12 months. Study strengths
Exclusion criteria were inability to walk 50 feet 1. Adequate sample size.
independently, active medical problems that 2. The sample composition is described in
would contraindicate performance of a graded terms of age, education, and setting.
exercise stress test, inability to complete the 3. Minimally adequate exclusion criteria.
graded exercise stress test or the modified 4. Test administration procedures are
PPT, a score greater than 8 on the Short specified.
Blessed Test, inability to provide informed 5. Means and SDs for the test scores and
consent due to cognitive impairment, and in- error rates are reported.
ability to follow the directions for the psy-
chometric tests due to visual or auditory Considerations regarding use of the study
impairments. 1. The data are not partitioned by age
The standard administration procedure was group.
used except that the maximal allowed time 2. Education level for the sample is high.
for both parts A and B was 180 seconds. Time 3. No information on gender and IQ.
to completion and the number of lines cor-
rectly drawn within the allotted time were [TMT.38] Saxton, Ratcliff, Newman, Belle,
recorded. Fried, Vee, and Kuller, 2000 (Table A4.40)
The authors found that part B performance The TMT was administered as part of
was significantly associated with total modified the Memory and Aging Study (MAS) con-
PPT score. ducted as an ancillary project to the CHS,
a multicenter observational study of heart protocol included a standardized general

disease and stroke in Washington County, medical history and physical examination;
Maryland, and Pittsburgh, Pennsylvania. No a detailed neurological and mental status ex-
selection criteria were used. Data were ana- amination; hematological, metabolic, and se-
lyzed for a sample of 989 participants (444 rological tests; and neuroimaging when
males and 545 females), who completed all of appropriate. Relevant medical records were
the cognitive tests included in the battery. The abstracted. The sample included 302 females
mean age for the sample was 73.63 (4.45) and 181 males, with a mean age of 74.9 (4.4)
years, and mean education was 13.23 (2.85) years; 31.9% of participants had less than a
years; 93.9% of the sample were white. This high school education.
sample was divided into two clinical groups Times to completion for the two parts of the
and a "no disease" group, based on cardio- TMT were reported for the entire sample.
vascular status. Results of the ROC analysis suggested that
Times to completion for the TMT for the TMT part B was one of the tests that had the
"no disease" sample of 357 participants are highest accuracy in discriminating between
reproduced in Table A4.38. Demographic nondemented participants and those who
characteristics for this sample are not reported were in the preclinical stages of DAT (area
by the authors. However, we assume that they under the curve=0.773).
are similar to the demographics for the entire
sample described above. Study strengths
Study strengths 2. The sample composition is well described
1. Large sample size. in terms of age, education, gender, his-
2. The sample composition is described in tory of the project, and geographic area.
terms of age, education, gender, setting, 3. Rigorous exclusion criteria.
geographic area, and recruitment pro- 4. Means and SDs for the test scores are
cedures. reported.
3. Means and SDs for the test scores are 5. Information on the diagnostic accuracy
reported. of part B is provided.
1. No exclusion criteria. 1. The data are not partitioned by age
2. The data are not partitioned by age group.
group. 2. No information on IQ is reported.
3. No information on IQ is reported. 3. The number of participants with less
4. Demographic characteristics for the "no than a high school education is reported.
disease" group are not reported. However, mean education and SD is not
reported.
[TMT.39] Chen, Ratcliff, Belle, Cauley,
DeKosky, and Ganguli, 2000 (Table A4.41) [TMT.40] Small, Graves, McEvoy, Crawford,
A control sample of 483 elderly nondemented Mullan, and Mortimer, 2000 (Table A4.42)
individuals was derived from a community- The authors examined the relationship
based multiwave prospective study, the Mon- between APOE genotype and cognitive func-
ongahela Valley Independent Elders Survey tioning in normal aging based on a sample of
(MoVIES), in southwestern Pennsylvania. The 413 adults between 60 and 85 years of age,
purpose of the study was to identify cognitive with a mean age of 72.90 years, who were
measures that are most accurate in discrimi- randomly selected from a larger sample of
nating between individuals with presymptom- participants in the community-based, cross-
atic DAT and nondemented individuals. The sectional Charlotte County Healthy Aging
control participants remained nondemented Study conducted in south Florida. The sample
over a 10-year follow-up period. The study was stratified into two age groups, young-old
TRAILMAKING TEST 93
(00-73 years, n = 202) and old-old (74-85 scores and their logarithmic transformations.
years, n = 211), and further divided into two Only four control participants made one error
groups according to the presence of the on part B.
APOE-e4 allele. The sample was almost The resu1ts suggest that error analysis is a
exclusively white. Education, gender distribu- more useful method of categorizing perfor-
tion, and self-rated indices of health status are mance than time to completion. All patients
reported for each group. Intelligence levels who made more than one error on part B had
were estimated using the Spot the Word Test. frontal lesions.
The TMT was administered according to
standard procedures. Study strengths
1. The sample composition is described in
Study strengths terms of age, education, gender, esti-
1. Large sample sizes per group. mated IQ, and clinical setting.
2. The sample composition is well described 2. Adequate exclusion criteria.
in tenils of age, education, gender, geo- 3. Test administration procedures are
graphic area, and research setting. specified.
3. Test administration procedures are 4. Means and SDs for the test scores and
specified. derived measures are reported.
4. Data are stratified by two age groups.
5. Estimated intelligence levels are
reported.
1. The sample size is small.
2. Age range is wide.
reported.
3. The data were obtained on Canadian
fulness for clinical interpretation in the
1. Exclusion criteria are not clearly
United States.
described.
2. Education level for one of the groups is
high. [TMT.42] Bell, Hermann, Woodard, Jones,
Rutedd, Sheth, Dow, and Seidenberg,
[TMT.41] Stuss, Bisschop, Alexander, Levine, 2001 (Table A4.44)
Katz, and lzukawa, 2001 (Table A4.43) The TMT was administered as part of a larger
The study examined the relationship of the battery in a study examining the neurobeha-
TMT to focal frontal lobe lesions. Time to vioral status of patients with early-onset
completion and number of errors performed temporal lobe epUepsy. The control group
by the clinical groups with different lesion included 29 friends, relatives, and spouses of
localizations and the control group were patients (72% female), who were between
compared. The sample of 19 control partici- ages 16 and 60 years, with a mean age of 34.4
pants with a mean age of 53.4 (13.6) years and (12.5) years; FSIQ (as measured with the
a mean education of 13.7 (2.5) years was WAIS-111 7-subtest short form) between 69
drawn from a general popu1ation pool of vol- and 110, with a mean FSIQ o£97.7 (6.4); and
unteers. The participants were fluent English mean education of 13.0 (1.7) years. Exclusion
speakers with adequate ability to read and had criteria were current substance abuse, psy-
no prior history of any neurological or psychotropic medication use, medical or psychi-
chiatric disorders. atric condition that could affect cognitive
The TMT was administered according to functioning, an episode of loss of conscious-
standard procedures. All participants contin- ness longer than 5 minutes, developmental
ued working on the test until they completed learning disorder, and repetition of a grade in
the task. Time to completion for both TMT school.
parts, the B-A difference, and the propor- Time to completion for both TMT parts is
tional score (B-A)/A are reported in both raw provided.
Study strengths Considerations regarding use of the study

1. The sample composition is well described 1. The sample is small.
in terms of age, education, gender, FSIQ, 2. Wide age span.
and recruitment criteria. 3. No information on IQ is reported.
2. Adequate exclusion criteria. 4. All-female sample.
reported. [TMT.44] Drane, Yuspeh, Huthwaite,
and Klingler, 2002 (Table A4.46)
Considerations regarding use of the study The purpose of the study was to examine the
1. The sample is small and includes a wide relationship of TMT time to completion as
age range. well as derived indices, such as difference
2. The data are not partitioned by age group. scores and ratio scores, with demographic
variables. The sample consisted of 285 adults
[TMT.43] Stein, Kennedy, and Twamley, (205 males and 80 females) between 18 and 90
2002 (Table A4.45) years of age, who participated in a compre-
hensive neuropsychological normative project.
The authors compared cognitive functioning They were recruited through a variety of civic
in female victims of domestic violence and organizations. Participants did not have any
nonvictimized women. The control sample history of known psychiatric or neurological
included 22 participants who were recruited disorder, were living independently, had no
through posted advertisements and ongoing history of substance abuse, and were not
personal contacts. The study was conducted in treated with psychotropic medications at the
San Diego, California. The control group time of the examination, per clinical interview.
included 22 participants who had no lifetime All participants performed within the normal
exposure to a posttraumatic stress disorder range on the MMSE. Mean age for the sample
Diagnostic and Statistical Manual-IV criterion was 48.30 (19.68) years, mean education was
A stressor, spoke English fluently, and had at 12.98 (2.65) years, and mean MMSE score
least an 8th-grade reading ability. Exclusion was 28.63 (1.61).
criteria were use of any psychotropic medi- The TMT was administered according to
cations within 6 weeks before participation, standard procedures. Time to completion, B-
use of oral or intramuscular steroids within A difference, and B:A ratio are reported for
4 months before participation, history of eight age groups.
learning disability or attention-deficit disor- The authors evaluated the sensitivity of the
der, head injury with loss of consciousness B:A impairment cutoff score of 3.0 that was
greater than 10 minutes, seizure disorder, suggested by Lamberty et al. (1994) and
drug or alcohol use, and history of psychotic concluded that rates of false-positive misclas-
illness or neurological disorder. Mean age for sification are unacceptably high, especially for
the sample was 29.4 (10.7) years and mean older age groups.
education was 13.9 (1.5) years.
Time to completion for both TMT parts as Study strengths
well as the B-A difference are provided. 1. Large sample size.
Study strengths in terms of age, education, gender,
1. The sample composition is described MMSE scores, setting, and recruitment
in terms of age, education, gender, geo- procedures.
graphic area, setting, and recruitment 3. Adequate exclusion criteria.
procedures. 4. Test administration procedures are
2. Rigorous exclusion criteria. specified.
3. Means and SDs for the test scores and 5. The data are partitioned by eight age
the B-A difference are reported. groups.
TRAilMAKING TEST 95
6. Means and SDs for the test scores as 3. Test administration procedures are
well as derived indices are reported. specified.
Considerations regarding use of the study reported.
1. Demographic characteristics for age
groups are not reported. Considerations regarding use of the study
2. Overall sample is large, but some indi- 1. Participants had established coronary
vidual cells are small. disease. It is unclear if any neurological
3. No information on IQ is reported. exclusion criteria were used.
2. All-female sample.
[TMT.45] Grady, Yaffe, Kristof, Lin, Richards, 3. The data are not partitioned by age
and Barrett-Connor, 2002 (Table A4.47) group.
Data on TMT part B were collected for a 4. No information on IQ is reported.
subsample of 1,063 older women in a multi-
center study examining the effect of hormone [TMT.46] Miller, 2003, Personal
replacement therapy on cognitive functioning Communication (Table A4.48)
in postmenopausal women. This is a follow-up The investigation used participants from the
on the articles reporting normative data for MACS study. The data were collected from
different subgroups from the same study 949 seronegative homosexual and bisexual
(Barrett-Connor & Goodman-Gruen, 1999; males for the purpose of establishing nor-
Kritz-Silverstein & Barrett-Connor, 2002). mative data for neuropsychological test
The participants were younger than 80 years performance based on a large sample. These
old and had established coronary disease and data represent an update on the data pro-
an intact uterus. They were randomly assigned vided by Seines et al. (1991). Mean age for
to treatment vs. placebo groups in a double- the sample was 38.0 (7.5) years and mean
blind experiment. They were followed for 4.2 education was 16.3 (2.4) years; 91.5% were
(.04) years. At the end of the trial, cognitive Caucasian, 3.0% Hispanic, 4.5% black, and
functioning was measured in both groups. 1% other. All participants were native
The data are reported for 517 participants in English speakers.
the treatment group and 546 in the placebo The TMT was administered according to
group, separately. The mean age for the two standard instructions. The data are partitioned
groups at the time of testing was 66.3 (6.4) and by three age groups (25--34, 35-44, 45-59)
67.3 (6.3) years, respectively, and mean edu- times three education levels ~ 16, 16, > 16
cation was 12.7 (2. 7) years for both groups; years).
approximately 90% of the sample were white.
There are no notable differences between the Study strengths
groups on any demographic variables or 1. The overall sample size is large, and
physical indices. most of the individual cells have more
Trails B was administered according to than 50 participants.
standard procedures. 2. Normative data are stratified by age x
The authors concluded that there were no education.
differences between the treatment and pla- 3. Information on age, education, ethnicity,
cebo groups on any cognitive measures. and native language is reported.
Study strengths reported.
2. The sample composition is well described Considerations regarding use of the study
in terms of age, education, gender, phys- 1. All-male sample.
ical findings, clinical setting, and selection 2. No information on IQ is reported.
criteria. 3. No information on exclusion criteria.
[TMT.47] Tombaugh, 2004 (Table A4.49) 4. Test administration procedures are

The author provided normative data for 911 specified.
community-dwelling adults between 18 and 5. Means and SDs for the test scores are
89 years of age. The data for volunteers who reported.
participated in earlier studies were analyzed. 6. Data are stratified by age x education.
Out of this sample, 823 participants were
recruited through booths at shopping centers, Considerations regarding use of the stw:hj
social organizations, places of employment, 1. As the authors pointed out, the sample
psychology classes, and word of mouth. Ex- size of the oldest group is small.
clusion criteria were history of neurological 2. No information on the intellectual level
disease, psychiatric illness, head injury, or of the sample is reported.
stroke, per self-report; the remaining 88 par- 3. The data were obtained on Canadian
ticipants represent a subset of individuals who participants, which may limit their use-
had received a consensus diagnosis of "no fulness for clinical interpretation in the
cognitive impairment" made by physicians United States.
and clinical neuropsychologists, based on
history, clinical and neurological examination,
and an extensive battery of neuropsychological RESULTS OF THE META-ANALYSES
tests, over two successive evaluations sepa- OF THE TRAILMAKING TEST DATA
rated by approximately 5 years. The author (See Appendix 4m)
pointed out that all participants 18-24 years
old were university students. Data collected from the studies reviewed in
Mean age for the sample was 58.5 (21.7) this chapter were combined in regression
years, mean education was 12.6 (2.6) years, analyses in order to describe the relationship
and the male/female ratio was 4081503. All between age and test performance and to
participants scored above 23 on the MMSE, predict expected test scores for different age
with a mean of 28.6 (1.5), and below 14 on groups. Effects of other demographic vari-
the Geriatric Depression Scale, with a mean ables were explored in follow-up analyses. The
of 4.1 (3.4). Elderly participants were also general procedures for data selection and
excluded on the basis of a clinical evaluation analysis are described in Chapter 3. Detailed
of depression. results of the meta-analyses and predicted test
Trails A and B were administered as part of scores across adult age groups for parts A and
a larger battery according to the Spreen and B are provided in Appendix 4m.
Strauss (1998) guidelines. Educational range was unevenly repre-
The results indicated that test performance sented, with a large gap between 8.5 and 11.59
for both Trails A and B was affected by age. years at the lower extreme. Based on the pre-
Performance on Trails B was also related to liminary analyses, the data point with 8.5 years
education, particularly in individuals over 54 of education was retained in the main analyses
years of age. Therefore, tables of raw data and but dropped in the analyses generating an
percentiles are stratified into 11 age groups. education-correction factor (see below). After
For ages 55 and above, they are further par- data editing for consistency and for outlying
titioned into two education levels (~12 and scores, 28 studies for Trails A and 29 for Trails
12+years). B, which generated 89 data points for each
part based on totals of 6,317 and 6,360 par-
ticipants, respectively, were included into the
Study strengths analyses.
1. Large sample size. Quadratic regressions of the test scores on
2. The sample composition is well described age yielded R2 values of 0.905 for Trails A and
in terms of age, education, gender, set- 0.876 for Trails B, indicating that 91% and
ting, and recruitment procedures. 88% of the variance in test scores for the two
3. Rigorous exclusion criteria. parts, respectively, is accounted for by the
TRAILMAKING TEST 97
model. Based on these models, we estimated 13.87 years of education, rounded to 14 years
scores for both parts for age intervals between (which is the mean education for the original
16 and 89 years. If predicted scores are data set) in the education-correction tables.
needed for age ranges outside the reported With every year of education above or below
boundaries, with proper caution (see Chap- this level, we suggest correcting the obtained
ter 3), they can be calculated using the score by adding or subtracting 1.31 to or from
regression equations included in the tables, the predicted score given in the table for the
which underlie calculations of the predicted relevant age group (see Chapter 3 for an
scores. example). The coefficient for education for
It should be noted in the context of across- Trails B is -6.446, rounded to -6.45 in the
condition comparisons that mean age for education-correction table. Thus, we suggest
Trails B is somewhat higher than for Trails A correcting the obtained score by adding or
because data for one study based on the older subtracting 6.45 to or from the predicted
sample were reported for Trails B only. score given in the table for the relevant age
Quadratic regressions of SDs on age yielded group. The SDs for the person's actual age
R2 of 0.602 for Trails A and 0.676 for Trails B, group should be used with the education-
indicating an increase in variability with corrected scores.
advancing age, consistent with the literature. Correction factors for different education
Predicted SDs, based on these models, are levels for both Trails A and B are included in
reported. Appendix 4m. These corrections should be
Examination of the effects of demographic applied within the education range of 12-17
variables on the test scores revealed that years since this is the range available in the
education is a significant predictor of test original data set. Unfortunately, data for lower
performance for both parts A and B. Values of educational levels were not available in the
estimated between-study variance (tau 2 ) for literature. Any extrapolation of scores outside
regression of test means with education were the reported range should be made with
considerably lower than the corresponding caution.
values for regression without education. This IQ did not have a significant effect on the
suggests that education explains a consider- test scores in our data set. Given consistent
able amount of the heterogeneity in the out- evidence of the effect of intellectual level on
come variable. Inclusion of education into the test performance described in the literature,
regression of test means on age considerably our lack of association is likely due to insuffi-
improved the R2 (see Appendix 4m). In this cient data regarding IQ levels reported in the
analysis, regression with and without educa- studies reviewed (only seven studies reported
tion was rerun on a subset of studies that IQ levels, which generated 21 data points for
reported education for each data point. In each TMT part).
addition, the group with 8.5 years of education The difference in mean scores for the two
was dropped because of a large gap at the genders across 17 studies reporting scores for
lower extreme of the educational range, with males and 15 studies reporting scores for
the next lowest level available for analyses females separately was negligible: 0.704 in
being 11.59 years of education. Therefore, the favor of males for Trails A and 0.379 in favor
data set for Trails A was based on 25 studies of females for Trails B.
and that for Trails B, on 26 studies. The
t-value for education is -3.00 (p = 0.006) for Strengths of the analyses
Trails A and -2.56 (p = 0.017) for Trails B. 1. Total sample sizes of 6,317 for Trails A
The coefficient for education of -1.308, and 6,360 for Trails B.
rounded to -1.31, for Trails A indicates that 2. R2 of 0.905 for Trails A and 0.876 for
with a 1-year decrement in education we Trails B, indicating a good model fit.
expect a 1.31-second slowing in test perfor- 3. Postestimation tests for parameter spec-
mance. This suggests that the table of pre- ifications did not indicate problems with
dicted values is accurate for individuals with normality.
4. Effect of education was evident, which is concentrational and executive problems, as

consistent with the literature. Significant well as to psychomotor slowing. It would be
effect of education on both parts A and B misleading to view the TMT as the test for
called for corrections for education. organic brain pathology. For example, patients
with memory deficits associated with temporal
Limitations of the analyses lobe pathology may perform normally on this
1. Postestimation tests for parameter spec- test, and if the TMT is administered in isola-
ifications indicated lack of homoscedas- tion, the serious processing difficulties of this
ticity for both Trails A and B. Variability population might be overlooked.
in scores across age groups is greater Most commonly, clinical use of the TMT is
than expected by chance, with a con- based on a norm-referenced interpretation of
siderable increase in variability in the completion times for each condition. Use of
older age groups, as reflected in the size the TMT cutoff criteria for brain impairment
of the confidence intervals. Therefore, is now quite infrequent (Spreen & Strauss,
the predicted scores are less accurate for 1998). According to the literature, perfor-
the older age ranges than for the youn- mance on the TMT (especially part B) is
ger ranges. highly affected by age and education. This
2. Levels of education and IQ for the sam- finding is supported by the results of the
ples included in the review are high. meta-analyses discussed above. Thus, it is of
Although corrections for education are utmost importance to interpret individual
provided, mean IQ levels are 116.69 scores with reference to the relevant norma-
(7.48) for Trails A and 116.88 (7.80) for tive data.
Trails B. According to the literature, Among the derived measures, B:A ratio was
there is a strong relationship between test found to be diagnostically useful in several
performance and IQ level. Therefore, the studies, with modest support for use of the
predicted values are likely to underesti- B-A difference. There is little consensus on
mate expected time to completion for the utility of error analysis. Further research
individuals with average and lower than is needed to gain better insight into the
average intellectual levels. diagnostic utility of the derived measures
and error analysis with different clinical pop-
ulations.
The optimal format for data reporting in
CONCLUSIONS
future investigations is in age-by-education
and I or -by-intelligence level cells. Given the
The TMT has achieved high popularity as a demonstrated utility of the B-A difference,
screening tool for cognitive impairment. B:A ratio, and error analysis with some clinical
There is ample evidence supporting the sen- groups, reporting of statistics for these indices
sitivity of performance times for parts A and B would further facilitate interpretation of the
to cerebral dysfunction in mild traumatic results and contribute to diagnostic decision
brain injury, in the differential diagnosis of making.
dementia, in detecting attentional I concen- As to the use of the cutoffs, Soukup et al.'s
trational dysfunction in children and adults, (1998) recommendation to report cutoff scores
and in other conditions. Although poor perfor borderline (15th percentile) and defective
formance on the TMT is viewed as a non- (< 5th percentile) ranges in addition to the
specific finding due to the complexity of the descriptive statistics should be given careful
mechanisms contributing to test performance, consideration due to the positive skew in the
the TMT is most sensitive to attentional I distribution of TMT scores.
5
Color Trails Test
BRIEF HISTORY OF THE TEST et al., 1993). Because of the TMT's popularity
and availability in the public domain, it be-
Because of their ease of administration and came perhaps the most frequently photo-
sensitivity to brain damage, trail-making tasks copied neuropsychological test of the 20th
have long been among the most widely used century. Poor photocopy quality often blurred
measures in neuropsychological practice (Le- the target stimuli, and it was not uncommon to
zak et al., 2004). The original Trail Making Test discover TMT protocols in which the stimuli
(TMT) was developed in 1944 (see Chapter 4); closest to the edge of the page had been cut off
however, it relied upon the English alphabet as due to improper placement of the original on
part of the test stimuli, thereby limiting its use the photocopy machine. Successive genera-
in non-English-speaking countries. Further, its tions of photocopies yielded slightly smaller or
use in English-speaking countries was proble- slightly larger versions of the test, thereby
matic when assessing adults with language and changing the distance between stimuli. Be-
reading disorders, limited education, or Eng- cause the "time to complete" score obtained
lish as a second language. for the test reflects not only visual scanning
The Color Trails Test (CTT) was created in and psychomotor speed but also the distance
response to a request made in 1989 by the traveled between stimuli, the problem of not
World Health Organization (WHO) for a test having a standard version of the TMT would
that would be similar to the TMT (1944) in necessarily hamper the comparability of re-
terms of its sensitivity and specificity yet allow search and clinical findings. Therefore, it was
broader application in cross-cultural contexts. important to develop a format of the test that
The WHO wanted a test with standardized, would also discourage photocopying (D'Elia
equivalent, multiple forms for test-retest pur- et al., 1996).
poses. Additionally, although the TMT had The CTT is similar to the TMT in that it
been translated into other languages, its basic requires cognitive flexibility and visuomotor
linguistic and phonological properties con- skills to complete the task. Additionally, the
tinued to limit its application in special-needs CTT is similar to the TMT in that it is
contexts (e.g., language disorders, specific administered under timed conditions. How-
reading disorders, or illiteracy). ever, the CTT relies on the use of numbered,
The WHO also wanted standardized test colored circles and universal sign language
stimuli to insure the new test's reliability (Maj symbols to solve the task, rather than relying
99
on English (or any other) alphabet letters as (number), whereas the CIT2 requires the
part of the test stimuli. Instructions for the subject to simultaneously track both a specified
CTI may be administered verbally or non- number sequence and a separate color
verbally, using only visual cues. sequence. Therefore, an interference index
Both the TMT and CTI are paper-and- was developed to quantify and highlight the
pencil tests that are administered in two parts relative difference regarding the effects of
on an 8 ~ x 11" page. However, for the CTI1, visual attention and perceptual tracking re-
the numbers 1-25 are printed within colored quired on the CTI1 from the more demanding
circles. All even-numbered circles are printed sustained, divided attention and more complex
with a bright yellow background and all odd- perceptual tracking required by the CIT2.
numbered circles, with a vivid pink back-
ground. These background color differences Interference Index =
are perceptible even to color-blind individuals. (CTT2 time raw score- CITl time raw score)
The individual is instructed to quickly draw a C1Tl time raw score
continuous line that connects the numbers in
consecutive/sequential order. The incidental The interference index reflects the compar-
fact that color alternates with each succeeding ison of the subject's performance on the CTI1
number is not highlighted or discussed with relative to the CIT2. ThiS index is expressed
the subject since attention to color sequence is as a function of the level of performance on
not necessary for completion of the CTIL the CTil. Therefore, the index score is a
The CTI2 introduces a divided attentional relatively "pure" measure of the extent of
component, requiring attention to the alter- interference (if any) attributable to the more
nating and sequencing of the stimuli. For the complex divided attention and the alternating
CTI2, the number 1 circle is printed against a sequencing tasks required by the CIT2. For
vivid pink background; however, the numbers example, an interference index score of 0 in-
2-25 are presented twice: once with a vivid dicates that the subject's time to complete
pink background and once with a bright yellow the CTI1 was the same as that to complete
background. The subject has to again quickly the CTI2 (i.e., no interference). An inter- -
connect the numbers in sequence; however, ference index score of 1.0 indicates that the
the task requires alternation of colors as the subject required twice as long to complete the
sequence of numbers advances, so the subject CIT2 as the CTI1, whereas a score of 3.0
must ignore distracter circles that contain the indicates that it took the subject four times as
correct number but are printed in the wrong long to complete the CIT2 relative to the
color background (e.g., start with pink 1 and CTI1 (i.e., significant interference). As the
avoid pink 2, select yellow 2, avoid yellow 3, interference index score increases, the in-
select pink 3, avoid pink 4, select yellow 4, creasing score suggests the presence of greater
etc.). Therefore, there is always a distracter susceptibility to cognitive interference from
number that must be avoided because it is alternating and sequencing demands (i.e., de-
printed against a color background that is not creased cognitive flexibility).
appropriate to the sequence. Before the CTI1 The WHO's request for a test that would
and CTI2 are administered, nontimed prac- allow broader application in cross-cultural
tice trials are administered to insure that the contexts seems quite reasonable. Ideally, neu-
subject understands the task. When the CTI1 ropsychological procedures that assess the ef-
and CTI2 forms are administered, however, fects of conditions affecting neurological
the time required to complete each form is functioning, including brain injury, infectious
noted. Subjects must complete each form of diseases (e.g., HIV) and other pathologies,
the test in :5240 seconds, or that part of the should be as culture-free as possible; but is it
test is discontinued. possible to develop a totally culture-free neu-
The CTI1 is a less cognitively demanding ropsychological test? Perhaps not. If this is the
task because it requires the subject to per- case, then procedures should be developed
ceptually track only a single specified sequence that allow, at minimum, enhanced assessment
COLOR TRAILS TEST 101
in cross-cultural contexts. Although color for all forms. 'The alternate forms (i.e., forms
perception may not be a totally culture-free B, C, and D) are considered experimental and
phenomenon (Bomstein, 1973), color was used should be used only in research settings.
as the test stimulus for the categorical shifting The scoring of the CIT differs from that of
in the CIT because it typically transcends most the TMT, to allow quantification of the cognitive
cultural distinctions. Also, the decision to use slippage that often occurs following mild brain
numbers and colors was based on the fact that injury. For instance, following mild cerebral
both are universal symbols that place limited insults, patients commonly report subtle chan-
demands on language production or knowledge ges in sequencing, planning, and ability to
(D'Elia et al., 1996). In cross-cultural pilot tests inhibit specific responses. They frequently
of the CIT, it was found that individuals in poor complain that it takes extra effort to perform
Third World countries in Africa, Asia, and most tasks they formerly completed without
South America, with little or no formal educa- much thought or effort. Unfortunately, current
tion, know and recognize the Arabic numbers approaches to characterizing performance on
1-25, perhaps because they have to barter for most neuropsychological tests allow empirical
goods and services (Maj et al., 1991). quantification of only gross errors but not the
In developing the CIT, it was hypothesized more subtle forms of cognitive slippage fre-
that the alternating shift between number and quently described by these patients. The near-
color sequences would require more effortful miss score was developed to allow empirical
executive processing than the shift between quantification of this type of cognitive slippage.
numbers and letters of the alphabet. Specifi- This response occurs when a subject initiates an
cally, in the United States, the English alphabet incorrect response but self-corrects before
is learned at a very early age. Students are taught actual connection to a distracter circle. Report-
not only to recite the alphabet but to sing it as ing near-miss scores allows the examiner to
well. As such, the alphabet sequence is strongly comment on the degree to which a patient is
encoded. Indeed, it is not unusual to observe a susceptible to distracters. Other scoring criteria
premorbidly high-functioning individual pre- include quantification of prompts, number-
senting with a history of moderate brain injury sequence errors, and color-sequence errors.
who is able to call upon sufficient brain reserve In the course of preparatory work for the
capacity (Satz, 1993) to complete the TMT part WHO cross-cultural study on the neuro-
B within a nominally "normal" time limit. In- psychiatric aspects of HIV-1 infection, Maj
terestingly, these individuals have been occa- et al. (1993) evaluated the CIT in comparison
sionally observed to hum or sing the alphabet to translated versions of the TMT at four world
(although almost inaudibly) while solving part B. sites: Munich, Germany; Bangkok, Thailand;
Removal of reliance on the English alphabet to Naples, Italy; and Kinshasa, Zaire. Those pre-
solve the CIT2 was hypothesized to effectively liminary results suggested that the CIT was
eliminate this potential performance confound. not only sensitive to HIV-1-associated cogni-
Use of colors also permitted the develop- tive impairment but also more culturally fair
ment of identical, equivalent forms of the test than the TMT. 'The sensitivity of the test was
for repeat administration in longitudinal re- found to hold across the different cultures ex-
search. Currently, there are four versions of amined. However, whether it would hold in
the CIT (i.e., forms A, B, C, and D). Form A other cultures was unknown at that time, and
is the standard test form, on which normative more work still needs to be done.
data were collected. Therefore, form A is the
only one that should be used for clinical eva-
luation. The subsequent forms were created
RELATIONSHIP BETWEEN CTT
by printing a mirror-image version, a 90-
degree rotated version, and a 90-degree mirror-
FACTORS
image version of form A. This method of
creating alternate forms insured that the dis- 'There are currently four normative reports
tance traveled between stimuli was standard regarding the CIT. Analyses conducted on
the CTT data obtained from the U.S. stan- subject variables and the last to procedural
dardization manual revealed that increasing variables. Minimal requirements for meeting
age adversely affects performance on both the criterion variables were as follows.
CTT1 and CTT2. Increasing education was
found to enhance performance on CTT2 but
Subject Variables
not on CTTl. Gender and the interactions
between gender and age were not significantly Sample Size
related to CTT performance scores after the
Fifty cases are considered a desirable sample
effects of age were removed (D'Elia et al.,
size. Although this criterion is somewhat
1996). Ponton et al. (1996) and LaRue et al.
arbitrary, a large number of studies suggest
(1999), in examining their respective norma-
that data based on small sample sizes are
tive data from their Hispanic samples, also
highly inHuenced by individual differences
found a negative performance association
and do not provide a reliable estimate of the
between increasing age and CTT1 and 2 test
population mean.
scores. In addition, they found a positive
relationship between education and CTI1 Sample Composition Description
and 2 scores. No gender effects were found.
Information regarding medical and psychiatric
Similarly, Hsieh and Riley (1997), in examin-
exclusion criteria is important. It is unclear if
ing the normative data from their Chinese
geographic recruitment region, socioeconomic
sample, found a negative performance asso-
status, occupation, ethnicity, handedness, or
ciation between increasing age and CTI1 and
recruitment procedures are relevant. Until
2 test scores and a positive performance as-
this is determined, it is best that this in-
sociation between increasing education and
formation be provided.
CTI1 and CTT2 scores. No gender effects
were found in the Chinese sample. Age Group Intervals
In summary, research suggests that perfor- This criterion refers to grouping of the data
mance on the CTT is enhanced by education into limited age intervals. This requirement is
and negatively affected by increasing age. No especially relevant for this test since a strong
gender effects have been reported. The CTT effect of age on CTT performance has been
is available in adult and child formats from demonstrated in the literature.
Psychological Assessment Resources (see Ap-
pendix 1 for ordering information). Normative Reporting of IQ and/or Education level
data for the Children's CTT can be found in Given the association between educational
Uorente et al. (2003). level and CTT scores, information regarding
highest educational level completed should be
reported. Optimally, normative data should be
METHOD FOR EVALUATING THE categorically reported by age and education
NORMATIVE REPORTS level. It is unclear ifiQ is relevant, so until this
Our review of the literature located four is determined, it is best that information on IQ
normative reports: one for primarily Mexican- be provided.
American, Central and South American,
Spanish-speaking adults (Ponton et al., 1996); Procedural Variables
one for Mandarin-speaking mainland Chinese
(Hsieh & Riley, 1997); one for senior adult Data Reporting
bilingual Spanish/Mexican Americans (LaRue Means, standard deviations, and preferably
et al., 1999); and the U.S. standardization ranges for total time in seconds for each part of
manual (D'Elia et al., 1996). the CTT should be reported. Additional in-
To adequately evaluate the CIT normative formation regarding prompts, near-misses,
reports, five key criterion variables were errors, and interference index would facilitate
deemed critical. The first four of these relate to interpretation of test performance.
SUMMARY OF THE STATUS OF SUMMARIES OF THE STUDIES

THE NORMS
This section presents critiques of the nonna-
In tenns of subject variables, the standardi- tive studies for the CTT.
zation manual as well as the Ponton et al.
(1996) and LaRue et al. (1999) studies provide [CTT.1] D'Eiia, Satz, Uchiyama, and White,
perfonnance data grouped by ag~ and edu- 1996
cation categories. Hsieh and Riley (1997)
This is the original standardization of the CTT.
present data separately for age and for edu-
The manual reports nonnative data from a
cation.
sample of 1,528 healthy, nor.mal i.ndi~duals
Although the total sample for the U.S.
residing in a variety of settings m diverse
standardization study is 1,528, unfortunately
regions of the United States. Participants were
the manual does not indicate the sample size
excluded if there was a history of head trauma,
within each of the 30 age/education cate-
neurological disorder, or substance abuse. The
gories. Whereas in the LaRue et al. (1999)
data were collected during the course of sev-
study sample sizes for most of the age and
eral norming studies with distinct samples,
education categories are generally adequate,
including medically and psychiatrically normal
the sample size for each of the age and edu-
participants from a longitudinal cardiovasc~lar
cation categories reported by Ponton et al.
epidemiological study that has ~ee? ongomg
(1996) is small. Similarly, the sample size for
since 1960; medically and psychiatrically nor-
each of the age categories reported by Hsieh
mal pilots from four major U.S. commerci~ air-
and Riley (1997) is small.
line manufacturing corporations undergomg a
For the standardization study, Ponton et al.
yearly medical examination as part of a na~on
(1996), LaRue et al. (1999), the younger age
ally mandated Federal Aviation Administration/
group categories are generally narrowly de-
Equal Employment Opportunity Commission
fined and therefore adequate; however, the
study to obtain nonnative data on neuropsy-
older age group categories tend to be very
chological functioning of pilots across the age
broad. Hsieh and Riley (1997) report age data
span; medically and psychiatrically no~al re-
in 10-year increments as well as data accord-
sidents living in an independent retirement
ing to the age groupings found in the U.S.
community in southern California; medically
standardization manual. Regarding procedural
and psychiatrically healthy, HN-negative, bi-
variables, all studies report means and SDs for
sexual and homosexual men participating in a
time to completion for CTfl and 2. Only
multi-center epidemiological study; and medi-
the U.S. standardization manual reports data
cally and psychiatrically healthy ~rican
regarding errors, near-miss responses, and
American men living in Los Angeles With no
prompts. The standardization manual and the
history of drug!alcohol abuse, participating in a
Hsieh and Riley (1997) study provide data on
larger study of the neuropsychological, medi-
the interference index. The LaRue et al.
cal, and psychosocial consequences of poly-
(1999), Ponton et al. (1996), and D'Elia et al.
drug abuse and HN.
(1996) normative data were collected from
The data are stratified by age and educa-
participants residing in the United States. The
tion. There are five age categories: 18-29, 30-
Hsieh and Riley (1997) data were collected
44, 45-59, 60-74, and 75-89. For each age
from participants residing in the mainland
category, data are reported for perfonnance of
People's Republic of China.
those with education of :58, 9-11, 12, 13-15,
In this chapter, nonnative publications are
16, and 2::17 years.
reviewed in ascending chronological order.
The sample is primarily male; women com-
The text of study descriptions contains refer-
prise only 12% of the sample. The manual
ences to the corresponding tables identified
states:
by number in Appendix 5. Table A5.1, the
locator table, summarizes infonnation pro- Gender and the interactions between gender and
vided in the studies described in this chapter. age were not significantly related to CIT raw scores
after the effects of age were removed, explaining the initial report from an ongoing project.
between 0.4% to 2.4% of the variance. Therefore, The sample consists of 300 volunteers (180
the relatively small proportion of women in the female, 120 male) recruited from fliers and
normative sample does not constitute a threat to advertisements posted at community centers
either the validity or the utility of the CIT. (D'Elia and churches in Los Angeles County, Cali-
et al., 1996) fornia (Santa Ana, Pasadena, Pacoima, Mon-
tebello, and Van Nuys). The sample was
Spanish-language administration instruc-
primarily right-handed (95%). Regarding lan-
tions and preliminary normative data for
guage, 210 were monolingual Spanish and 90
Hispanics are provided in the manual. The
were rated by the examiner to be bilingual.
preliminary normative data are from a sample
The average (SD) duration of residence in the
of healthy, normal Hispanics living in south-
United States was 16.4 (14.4) years; however,
em California, participating in a large, on-
55% of the total sample had lived in the
going normative study. The Hispanic data are
United States less than 15 years, and half
reported separately since all participants in
of those participants had less than 6 years of
this subsample were educated outside the
residence in this country. Sixty-two percent of
United States and were primarily Spanish-
the sample were born in Mexico, 15% in
speaking or had Spanish as their first lan-
Central America, and 23% in other Latin
guage. Data for Hispanics are presented by
countries. Exclusion criteria included a his-
four age categories: 17-29, 30-39, 40-49, and
tory of neurological disease, psychiatric dis-
50-75 years.
order, alcohol or drug abuse, or head trauma.
The normative data contained in the stan-
Participants ranged in age from 16 to 75 years
dardization manual are not reproduced here,
(mean = 38.4 [13.5] years). Whereas the 30-
and the interested reader is referred directly
39 and 40-49 age groupings are adequately
to the publication for further information.
narrow, the 16-29 and 50-75 age groupings
are somewhat broad.
Study strengths
The data are reported by age and education
1. Sample composition is well described in
groupings. The tables separately present data
terms of exclusion criteria.
for males and females.
2. Performance is reported by age and
education intervals.
Study strengths
3. Data reporting includes means and SD
scores for each age/education interval.
terms of exclusion criteria.
4. Age group intervals are generally ade-
2. Educational levels are reported.
quate.
3. Mean and SD scores are reported.
4. Age group intervals are generally ade-
quate for younger samples (<50 years).
1. Sample size within each of the 30 age/
education categories is not indicated.
2. No information on the IQ of participants
1. Sample size is generally small per agel
is reported, although the data are pre-
education interval.
sented by age/education intervals.
2. The age group interval is too broad for
the older sample (50-75 years).
[CTT.2] Ponton, Satz, Herrera, Ortiz, Urrutia,
Young, D'Eiia, Furst and Namerow, 1996 Other comments
(Tables A5.2 and A5.3) 1. IQ scores are not reported; however,
This study presents normative data stratified scores are reported for Raven's Standard
by age and education for Spanish-speaking Progressive Matrices Test at each age/
adults' performance on the Neuropsycholo- education level. Raven's test is used
gical Screening Battery for Hispanics to provide an estimate of nonverbal
(NeSBHIS), which contains the C'IT. This is intelligence.
2. Although the data are reported by gen- in a brief battery of cognitive measures ad-
der, there does not appear to be a sig- ministered during the course of a community-
nificant gender effect. based epidemiological survey in Bernalillo
County, New Mexico. The authors report that
[CTT.3] Hsieh and Riley, 1997 the Hispanic population in that area reflects a
(Tables A5.4-A5.6) well-established ethnic and cultural group
This report provides normative data on tests since the majority of the residents evidently
of attention and concentration collected in trace their ancestry to the colonization by the
the People's Republic of China. The norma- Spanish in the 1600s. The majority of His-
tive sample consisted of 177 (93 male, 84 panics in the sample identified themselves as
female) urban, Mandarin-speaking participants Spanish Americans (83%). Most Hispanic
recruited across a broad range of educational participants were bilingual, with 94% report-
and occupational categories. The data are strat- ing that they spoke Spanish well and 83%
ified by age categories and by education cate- reporting that they spoke English well. Four
gories. The age categories are 30--39, 40--49, percent reported that they did not speak
5~9. 60-69, and 70--83 years. The education English. The authors reported that "Seventy-
categories follow the Chinese system of pri- nine percent of the Hispanic participants
mary school (1-6 years), middle school (7-9 completed the cognitive tests exclusively or
years), and high school (10-12 years). The vast primarily in English, 14% exclusively or pri-
majority of adults over the age of 60 had fewer marily in Spanish, and 7% in a combination of
than 6 years of education. English and Spanish." For the present report,
the authors excluded from analyses persons
whose age education-adjusted Mini-Mental
Study strengths
Status (Folstein et al., 1975) was 23 or lower.
1. Data include means and SDs for test
Adjustments for age and education were made
scores.
following a regression equation developed by
2. Age group intervals are adequate.
Mungas et al. (1996).
Because of time constraints, the authors
deviated from the standard administration
1. The sample composition description does
of the CIT in one important way: the de-
not sufficiently address study inclusion/
pendent measure for CIT1 and CTI2 is
exclusion criteria.
the number of digits correctly traced in 60
2. Sample sizes are generally small.
seconds.
The data are presented by ethnic group (non-
Other comments Hispanic white vs. Hispanic), by age and edu-
1. Years of education per age group are cation categories. The age categories are 65-74
presented in a table. and 75-97. The educational categories for His-
2. Performance is reported separately for panics are 0-6,7-9, 10-12, and >12years. The
age and for education categories. CIT educational categories for non-Hispanics are 0-
data would generally be more useful if 12 and > 12 years. The mean age of the non-
they were partitioned by an age/educa- Hispanic Caucasian men (n = 230) was 74.1
tion category. (5.8) years, and that of the women (n = 208)
was 73.8 (5.9) years. For the Hispanic group,
[CTT.4] LaRue, Romero, Ortiz, Chi Liang, and the mean age for men (n = 194) was 73.6 (6.5)
Lindeman, 1999 (Tables A5.7-A5.1 0) years, and that for women (n = 165) was 72.9
This report provides normative data for 797 (5.6) years. For the non-Hispanic group, the
community-dwelling, senior adult (age 65-97), mean educational level for men was 14.1 (3.5)
Spanish I Mexican American Hispanic and years, and that for women was 13.5 (2.5) years.
non-Hispanic Caucasian men and women, For the Hispanic group, the mean educational
reported by age and educational level. CIT level for men was 9.7 (4.4) years, and that for
data were collected as a result of its inclusion women was 9.3 (3.9) years.
Study strengths amounts of time to complete; however, the

1. Sample size of the various age/education CTI2 generally takes slightly longer to com-
groups is generally ~50 for both the plete than the TMT part B. The fact that
Hispanic and non-Hispanic groups (for the CTT2 takes longer to complete than the
exception, see "Considerations regard- TMT part B has been demonstrated in English-
ing use of the study," below). speaking (D'Elia et al., 1996), Turkish-
2. Sample composition is well des<$ibed in speaking (Dugbartey et al., 2000), and
terms of inclusion/exclusion crit,ria. Chinese (Mandarin)-speaking samples (Lee &
3. Performance is reported by 4ge and Chan, 2000a,b; Lee et al., 2000). In an intact
education intervals. individual, the reason for the added time to
4. Data reporting includes means ~d SD complete the CTI2 vs. the TMT part B most
scores for each age/education inferval. probably can be attributed to the design dif-
' ferences in the two test forms and, therefore,
the difference in test demands. The CTI2 has
Considerations regarding use of the sludy almost twice as many stimuli to scan and to
1. The age group interval for the upper age consider as the TMT part B (e.g., 25 circle
category is quite broad (75-97).; stimuli for TMT part B vs. 49 circle stimuli for
2. No information on IQ is rctported, CTI2), even though the distance the pencil
although the data are presented.by agel needs to travel to correctly complete the task
education intervals. is shorter (TMT part B = 243.6 em, CTT2 =
3. The standard test administration format 184.6 em).
was altered (for both forms of the test, Because the TMT part B uses the English
the data reflect the number oE circles alphabet as one of the alternating stimuli, one
containing digits that were cbrrectly must keep in mind not only the increasing
traced in 60 seconds). : sequence of numbers but also the increasing
4. Sample size for each of the education sequence of alphabet letters. In other words,
categories was <50 participants for the to correctly complete the TMT part B, one
75-97 year group of Hispanics. must keep in mind that G comes after F, then
H etc. The CTI2 similarly requires that the
test-taker keep in mind the correct numerical
sequence; however, even though there are
CONCLUSIONS
more distracter stimuli on the page, the al-
The CTI was developed to allow speeded/ ternating choice is between only two colors.
timed assessment and quantification pf cog- As such, for English-speaking individuals, it
nitive flexibility and sequencing skills :as well may be that the working-memory demands of
as assessment of mental processing s~d and the TMT part B are greater than for the
attention/concentration abilities. The~ skills CTT2. However, as discussed earlier, the
have long been associated with ex~utive/ sequence of the English alphabet seems to be
frontal lobe functioning (Lezak et al.,. 2004). an overlearned and multiply encoded string
The CTI was not designed to produce for individuals educated in the United States.
equivalent or even similar time to coitfletion Therefore, the actual difference in working-
scores when compared to the TMT. 1tather, memory demands may be quite minimal for
since the TMT has long been view~ as a intact individuals. This, of course, is an area
test of frontal lobe/executive functionipg, the for further research.
CTI was designed to tap similar frontaV Research to date suggests that the CTI
executive cognitive abilities and to allow broader holds promise for cross-cultural and long-
assessment applications in cross-cultu$1 con- itudinal research as well as clinical assessment
texts. Although the physical designs :of the of sequencing, visual scanning, and speed of
CTI and TMT are similar, they are J¥>t. nor mental processing abilities in non-English-
were they intended to be, identical. The CTI1 speaking adults and adults with limited edu-
and TMT part A generally take ~imilar cation, English as a second language, and
language and reading disorders. Further studies of different ethnic/cultural groups are
research is needed to compare CIT perfor- needed. Reporting the data by age/education
mance in cross-cultural settings. categories would allow performance compar-
Four equivalent forms have been developed ison across cultures.
for the CIT (A, B, C, and D). Currently, only In general, the age categories need to be
form A has been normed for clinical use. narrowed for reporting data on older adults.
Even though all four are physically equivalent, We recommend that future studies follow the
future research is needed to establish the WAIS-III age category groupings as an ex-
psychometric and normative equivalence of ample.
the alternate forms. Future research should Although some excellent work has been
also focus on establishing the reliability and done, further normative work still needs to be
equivalence of the alternate forms in samples done regarding the performance of Hispanic
of both normal participants and patients with individuals of Mexican descent on the CIT
specific neurobehavioral dysfunctions (e.g., above age 75 years. Fortunately, Ponton and
clinical comparison data; aka, abnorms). Fu- colleagues continue to collect normative in-
ture normative studies with any form of the formation on the NeSBHIS; therefore, a
test should also report base rate data regard- larger normative database will accumulate,
ing error and near-miss responses, data for allowing a sample size more appropriate for
prompts, as well as information regarding the inferential purposes with Hispanics. In their
interference index. For instance, no age- and initial report of Spanish-speaking individuals,
education-corrected normative data are avail- the sample size for the age- and education-
able for Hispanic samples regarding the corrected groups was quite small. Yet, com-
occurrence of near-miss and error responses, parison of these preliminary performance data
nor is there information regarding prompts with those found in the U.S. standardization
and the interference index. The one Chinese manual for the CIT at the same age and
normative study reports information regarding education levels does not suggest a significant
time to complete the CIT1, CIT2, and the difference. This finding coupled with the
interference index but no information regard- findings of Maj et al. (1991, 1993) further
ing prompts, near-misses, or errors. supports the notion that the CIT may allow
Normative data are needed for English- enhanced application in cross-cultural con-
speaking and non-English-speaking individuals texts. How many cultures this effect trans-
with low or no education. Further normative cends remains to be discovered. 1
'Meta-analyses were not performed on the CTr due to a

lack of sufficient data.
6
Stroop Test
BRIEF HISTORY OF THE TEST pathology (Perret, 1974), while subsequent

functional imaging studies have found the
'ne Stroop Test measures the relativf speed Stroop interference effect to be associated with
of reading names of colors, naming col+rs, and activation of anterior cingulate and/ or frontal
naming colors used to print an incoagruous cortex (Bench et al., 1993; Brown et al., 1999;
color name (e.g., the color red used to print Carter et al., 1995, 1997; George et al., 1994;
the word blue). The last task requires one to Pantelis et al., 1996; Pardo et al., 1990: Pe-
override a reading response. This conflict terson et al., 1996; Taylor et al., 1997).
interference situation has come to b~ called Poor performance on the Stroop Test has
the Stroop Effect. been associated with frontal system dysfunc-
The interference section of the Stroop Test tion secondary to closed head injury (Trenerry
has traditionally been viewed as a me¥ure of et al., 1989), discrete frontal lobe lesions
executive functioning involving cognitive inhi- (especially left frontal lobe; Perret, 1974; see
bition (Boone et al., 1990) and, specifically, the Regard, 1981, cited in Spreen &: Strauss,
ability to inhibit an overlearned response in fa- 1998), frontotemporal dementia (Pachana
vor of an unusual one (Spreen &: Straus,, 1998) et al., 1996), frontal lobe seizures (Boone et al.,
and "to maintain a course of action in th~ face of 1988), white-matter hyperintensities (Fukui
intrusion by other stimuli" (Comalli et al. 1962, et al., 1994; Ylikoski et al., 1993), Klinefelter's
p. 47). Factor analyses of sets of executiVe mea- syndrome (Boone et al., 2001), age-associated
sures suggest that the Stroop interfe~ trial memory impairment (Hanninen et al., 1997),
has more in common with timed eJeeutive transient global amnesia (Stillhard et al., 1990),
measures, such as verbal fluency (FA$), and depression (Boone et al., 1995; Trichard et al.,
measures of information-processing speed, such 1995), schizophrenia (Brebion et al., 1996;
as Digit SymboL than executive tests i~olving Buchanan et al., 1994; Schreiber et al., 1995),
set shifting (Wisconsin Card Sorting Test) or late-life psychosis (Miller et al., 1991), atten-
divided attention/working memory (.Ailditory tion-deficit hyperactivity disorder (ADHD;
Consonant Trigrams) (Boone et al., 1998). Seidman et al., 1997; Rapport et al., 2001), and
Initial lesion studies indicated that poor exposure to alcohol in utero (Connor et al.,
performance on the interference sectio* of the 2000); and Stroop scores have been observed
Stroop Test was associated with left fronfallobe to predict aggression (Foster et al., 1993).
108
STROOP TEST 109
In addition, Stroop scores are lowered in used the color-word interference test that now
cases of brain dysfunction not necessarily con- bears his name (Stroop, 1935).
fined to anterior brain areas, such as left and Stroop's original studies employed three
right cerebrovascular accident (Trenerry et al., cards, all with white backgrounds:
1989), Alzheimer's disease (Binetti et al., 1996;
Koss et al., 1984; Pachana et al., 1996), and 1. An achromatic color-word reading card,
myotonic dystrophy (Palmer et al., 1994). consisting of a series of 100 words for
Stroop performance is impaired in both left colors printed in black ink.
and right cerebral damage but may be partic- 2. A chromatic color-word naming card,
ularly pronounced with left-sided damage consisting of a series of 100 color names
(Perret, 1974; Trenerry et al., 1989), although printed in a color of ink incongruent
this may be an artifact of coexistent aphasia. with the word.
Specifically, Nehemkis and Lewinsohn (1972) 3. A pure color card, consisting of a series
found that patients with left cerebral damage of 100 squares printed in different solid
with aphasia performed particularly poorly on colors.
the Stroop, while patients with left cerebral
damage without aphasia actually performed For all cards, five colors and/ or color-words
better than patients with right hemisphere were used (red, blue, green, purple, and
dysfunction. brown). The words and the colors were gen-
Finally, there is evidence that Stroop per- erally arranged in a 10" x 10" grid of evenly
formance is unaffected by chronic caffeine use spaced rows and columns. As Stroop notes:
(Hameleers et al., 2000) but is influenced by "The colors were arranged so as to avoid any
endogenous cholesterol synthesis (Teunissens regularity of occurrence and so that each color
et al., 2003). would appear twice in each column and in
The Stroop Test paradigm is among the each row, and that no color would immediately
oldest in experimental psychology. Interest in succeed itself in either column or row. The
the relative speed of color naming and read- words were also arranged so that the name of
ing color-words has been active for over a each color would appear twice in each line."
century. In 1883, as a result of a suggestion For the chromatic color-word naming cards,
by Wilheim Wundt (who founded the first "no word was printed in the color it names but
psychological laboratory in Leipzig, Germany), an equal number of times in each of the other
America's first psychologist, James Cattell four colors: i.e., the word 'red' was presented
(then a student of Wundt), began conducting in blue, green, brown, and purple inks; the
what would later become the earliest pub- word 'blue' was printed in red, green, brown,
lished study (1886) examining the relative and purple inks, etc. No word immediately
speeds of color naming and color-word succeeded itself in either column or row"
reading. Over 40 years later, the first pub- (p. 648). An alternate form was also created
lished report of the conflict I interference sit- by printing all the cards in the reverse order.
uation (e.g., where one must name the color In three experiments, Stroop examined four
of the ink used to print the word when the different tasks using the above-mentioned
color and color name are incongruous) origi- three cards. Using cards 1 and 2, experiment 1
nated in the Marburg, Germany, laboratory of examined the differences in rates of reading
Erick Rudolf Jaenasch (Jensen & Rohwer, color-word names (task 1) when the word was
1966). printed in black ink vs. an incongruous ink
Some years later, John Ridley Stroop, then color (task 2). Using only cards 2 and 3, ex-
a graduate student working in the Jesup periment 2 examined the differences in rates
Psychological Laboratory at George Peabody of verbally identifying squares of color (task 3)
College for Teachers, began his doctoral vs. naming ink colors against the distraction of
research, examining interference in serial incongruous color-words (task 4). For experi-
verbal reactions in which he developed and ment 3, Stroop modified his test, shortening
the cards to 10 columns and five ~ws (so with a white background, others have
that there were only 50 responses tequired used cards with a black background or a
per card instead of 100) and using: colored color different from both the color ink of
swastikas on the pure color card instead of the word and the color name ("Super-
solid square color patches. For expeqment 3, Stroop;" Dyer, 1973).
Stroop administered each of the foor tasks 5. Number of stimuli cards: Various ver-
separately on different days. Stroop never sions of the test require the use of two,
administered all three cards in the same three, or four cards.
testing period; this procedure did ~ot be-
come standard until Thurstone's (1944) in- Administration Procedures
vestigations of perception using the Stroop 1. Scanning orientation: Some versions
paradigm. I require the examinee to scan across rows
As testimony to its popularity, the Stroop from left to right, whereas others require
has been translated into several laGguages, the examinee to read down columns.
including Spanish (Rosselli et al., : 2002b; 2. Stimuli sequence: Some versions present
Annengol, 2002), Chinese (e.g., <fhen &: word reading followed by color naming
Ho, 1986), Czechoslovakian (e.g., Sovclikova &: and vice versa.
Bronis, 1985), German (e.g., Perrel1974),
Hebrew (e.g., Ingraham et al., 1988), wedish Method of Scoring
(e.g., Hugdahl &: Franzon, 1985), ~anese Determination of the total score has ranged
(e.g., Fukui et al., 1994; Toshima et ., 1992, from the number of correct responses made in
1996; Yamazaki, 1985), Vietnamese ( oan &: 45 or 120 seconds to the total time to com-
Swerdlow, 1999), and Italian (B barotto plete each card to a difference score (color
et al., 1998). In addition, a version e Joying interference minus color naming or reading)
numbers rather than words for a "Iahguage- to the total number of errors made in
neutral" test has been examined in jvarious 45 seconds.
populations, including those of low sctctoeco-
nomic status and/ or education, lowf.'ading Current Administration Procedures
level, and Mandarin and Spanish eakers
(Sedo, 1998, personal communication . At present, there is no one recognized stan-
The major problem with the Stroo. litera- dard version of the Stroop Test. There are,
ture is the presence of numerous tersions however, three versions that are commer-
of the task. Following is a summary! of the ciallypublished: Charles Golden's (1978), Max
variations. i Treneny et al.'s (1989), and that contained in
the Delis-Kaplan Executive Function System
(Delis et al., 2001). The first two are available
Stimulus Cards from Psychological Assessment Resources,
1. Color and shape of items: Carcls have and the Delis-Kaplan Executive Function
contained three, four, or five; colors System can be purchased from Psychological
presented as either squares, rectangles, Corporation. Edith Kaplan's Stroop version
circles, dots, or swastikas. can be used as a Comalli et al. (1962) version
2. Number of items: Various stimul11S cards (reading words, naming colors, naming colors
have contained 17, 20, 22, 24, ~7. 50, with incongruous color names) or as the
100, 112, and 176 items. Comalli and Kaplan version (naming colors,
3. Size and presentation of stimuli cards: reading words, naming colors with incongru-
Stimulus cards have varied frorl! small ous color names). Carl Dodrill (1978a) and
Hash cards to wall charts, and some Otfried Spreen and Esther Strauss (1998) have
studies have used a tachistoscopi•• slide, also developed versions of the Stroop, which
or computer presentation. I can be obtained by writing to their respective
4. Stimuli background: Althou~ most laboratories (see Appendix 1 for ordering
investigators have used stimul~ cards information).
110
STROOP TEST 111
The Stroop versions reviewed in this chap- mistake, Edith Kaplan decided the error might
ter will be limited to five formats which are have been fortuitous. As a result, she decided
commercially published or readily available to permanently make the modification in card
from the authors. (The Delis-Kaplan Execu- order presentation in her laboratory for the
tive Function System version will not be following reasons: (1) administering the Color-
reviewed because the published normative Naming card first allows an immediate check
data set is large [n for adults= 875], and few on whether the subject is color-blind, a con-
studies have appeared using this version due dition which would invalidate the use of the
to the recency of its publication.) These five test, and (2), more importantly, administration
versions differ in format and will be briefly of the Word-Reading task immediately before
described below. the Interference task (in which the sub-
ject must now inhibit a reading response)
Comalli/Kaplan Stroop, 1962; may exert a priming effect on the degree of
Personal Communication interference, whereas presenting the Color-
The Comalli and Kaplan Stroops use the same Naming card before the Interference task (in
three cards originally developed by Comalli which the subject is again expected to identify
et al. (1962), although Comalli and Kaplan color) may serve to minimize the "Stroop
differ in the order of presentation of cards 1 effect" (personal communication).
and 2. All three cards are 9~" x 9~" with 100 Demick, Kaplan, and Wapner (personal
stimuli per card arranged in a 10 x 10 grid communication) have more recently proposed
against a white background. At the top of each a process-oriented scoring system for the
card is an additional row of 10 practice items. Stroop that utilizes the identification of spe-
The color-name reading card consists of color cific verbal errors (reflecting both deviant
words (red, blue, green) printed in black ink responses to items, e.g., inappropriate color
and presented in random order. The color nam- responses, and deviant responses to sequence,
ing card consists of rectangles (5/16" x 2/16"} e.g., inserted linguistic words or phrases);
of colors (red, blue, green) arranged in ran- nonverbal behaviors serving as cognitive
dom order. The third card presents color- devices (e.g., nodding, body rocking); and
words printed in a color of ink different from expanded temporal measures (e.g., time per
the color designated by the word. For each line, time between utterances). Based on
card, participants are instructed to proceed exploratory studies (Demick &: Wapner, 1985;
line by line down the page either reading Demick et al., 1986), they have documented
words or naming colors as quickly as they can. that while various developmental groups may
Each line is scanned from left to right, mir- not differ with respect to achievement mea-
roring English reading format. The time to sures (e.g., total time}, they are distinguishable
complete the 100 items on each card is on the basis of more process-oriented mea-
recorded, along with the number of errors sures (e.g., relative to young and middle-aged
made. Near-miss responses (i.e., self-corrected adults, older adults used significantly more
errors) are also recorded. The Comalli/Kaplan inefficient strategies, such as gazing across the
Stroop scoring protocol also allows for inde- cards to identify specific patterns and I or using
pendently tracking the response times for fingers as if counting in succession to modulate
the first half of each card separately from the responses, to meet task demands; on a down-
last half. ward extension of the Stroop for preschoolers,
In the Comalli et al. (1962) administration 4- and 5-year-olds employed a range of non-
format, the Word-Reading card is presented verbal strategies to maintain serial organiza-
first, followed by the Color-Naming card and tion, while 3-year-olds failed to do so).
then the Interference card.
The Kaplan alteration in the test adminis- Golden Stroop, 1978
tration format occurred by happenstance when This version of the Stroop uses three 8~" x 11"
a research assistant mistakenly reversed the pages. Each page has 100 items presented in
order of cards 1 and 2. In thinking about the five columns of 20 items. Page 1 consists of
the words red, green, and blue presented which the color-word is printed as quickly as
randomly and printed in black ink. Page 2 possible. Rows are scanned from left to right
contains blocks of Xs printed in either red, as the subject works down the page. Time
green, or blue ink. Page 3 is the Stroop effect to completion and the number of errors are
card and contains color-words printed in a recorded for each card.
noncongruent color (i.e., the word blue prin- Jensen and Rohwer (1966) provide a de-
ted in red ink, etc.). For each page, the tailed and fascinating review of the Stroop
examinee is required to scan the columns Test and its many reincarnations; and Dyer
vertically, starting on the left side and moving (1973), Golden (1978), and MacLeod (1991)
to the right. The score is the number of cor- review applications and research findings
rectly identified items per page within 45 sec- subsequent to Jensen and Rohwer's report.
onds. Errors are not counted. Lezak et al. (2004) also provide an overview of
the test for the interested reader.
Dodrill Stroop, 1978a
The Dodrill version of the Stroop consists of Trenerry et al. Stroop, 1989
two alternate administrations of one stimulus This version of the Stroop consists of two
card containing 176 color-words (red, green, cards: form C and form C-W. Form C con-
blue, and orange) randomly printed in 11 col- tains 112 color-words (red, blue, green, and
umns of 16 color-words. Each color-word is tan) randomly arranged in four columns of
printed in an incongruous color (e.g., the color- 28 color-words. Each color-word is printed in
word blue is printed in green ink, etc.). In an incongruous ink color (e.g., the word tan is
the first administration, participants read the printed in red, etc.). Form C-W follows the
color-words as they scan down the columns. In same format as form C; however, there is a
the second administration, participants name different random order of color-words.
the color of ink in which the words are printed. For form C, the examinee is requested to
Time to complete each card is noted. Two read the words as quickly as possible while
scores are generated: the total time to com- scanning down the columns. For form C-W,
plete part I (which is essentially an estimate of the examinee is instructed to name the color
the examinee's reading speed) and the total ink in which the color-word is printed as
time for part II minus that for part I, "which quickly as possible, again while scanning down
reflects an estimate of the degree of interfer- the columns. A maximum of 120 seconds is
ence induced by the test" (Dodrill, 1987, p. 6). allowed to complete each task. The score for
each task is the number of correct responses
Victoria Stroop, 1991 (Reported -by (or number of items completed) minus any
Spreen & Strauss, 1991, 1998) incorrect responses.
The Spreen and Strauss version of the Stroop Although the Dodrill version relies upon a
(also known as the Victoria version) uses three difference score between the reading and
21.5 x 14 em cards presented in the following interference cards, a discriminant analysis
order: part D, part W, and part C. Each card conducted by Trenerry and colleagues
has six rows of four items. Part D contains demonstrated that the data from form C-W
colored dots (red, green, blue, and yellow), alone provided the sharpest classification
and on this card the task is to name the colors accuracy; thus, the score from form C-W is
as quickly as possible. Part W has the words the only one used for interpretation purposes.
when, and, over, and hand printed in red,
green, blue, or yellow ink; and the examinee
must name the color of ink in which each
RELATIONSHIP BETWEEN STROOP
word is printed as quickly as possible. On part
TEST PERFORMANCE AND
C, the color-words red, green, blue, and yellow
DEMOGRAPHIC FACTORS
are printed in incongruous-colored ink (e.g.,
the word red is printed in green ink, etc.); and While some studies have found no significant
the examinee must name the color ink in age effect on the Stroop Test (Graf et al.,
STROOP TEST 113
1995; Lopez et al., 2003) or an age effect that Boone, 1999; Connor et al., 1988; Houx et al.,
was equaled by the effect of health status 1993; Ingraham et al., 1988; Jensen & Roh-
(Houx et al., 1993), others have found sig- wer, 1966; Swerdlow et al., 1995; Stroop,
nificant or nearly significant age-related de- 1935; Treneny et al., 1989), and others re-
crements in Stroop performance (Anstey et al., porting a female advantage confined to color
2000; Barbarotto et al., 1998; Boone et al., naming (Golden, 1978; Stroop, 1935; Jensen
1990; Cohn et al., 1984; Comalli et al., 1962; & Rohwer, 1966; Strickland et al., 1997;
Daigneault et al., 1992; Feinstein et al., 1994; Swerdlow et al., 1995) or word reading
Graf et al., 1995; Ivnik et al., 1996; Jensen & (Strickland et al., 1997). In addition, there is
Rohwer, 1966; Klein et al., 1997; Libon et al., some evidence that a female advantage may
1994; Moering et al., 2004; Panek et al., 1984; be limited to samples with ~ 12 years of
Rosselli et al., 2002b; Spreen & Strauss, 1998; education (Moering et al., 2004).
Sullivan et al., 2002; Swerdlow et al., 1995; Spreen and Strauss (1998) suggest that
Treneny et al., 1989; Uttl & Graf, 1997; there is a relationship between Stroop per-
Whelihan & Lesher, 1985), although Doan formance and intellectual level, although Tren-
and Swerdlow (1999) observed an age effect eny and colleagues (1989) indicate that Stroop
for English speakers but not Vietnamese scores are not strongly related to IQ in brain-
speakers. Of interest, younger and older damaged participants.
groups may show a different pattern of per- A recent examination of the relative con-
formance on color interference, with younger tribution of IQ and demographic factors to
participants performing the first half of the Comalli Stroop color-interference perfor-
task faster than the second half and older mance in a large sample of healthy, older par-
participants demonstrating the opposite patticipants found that age and FSIQ accounted
tern (Klein et al., 1997). for 15% and 13%, respectively, of unique test
At least part of the slowing observed score variance; education and gender did not
in Stroop performance with age may be due contribute to test performance (Boone, 1999).
to declines in visual function. Specifically, Similarly, Ivnik et al. (1996), using Golden's
reduced contrast sensitivity is associated with (1978) version of the Stroop, found that
more time needed in word reading, red/ green age strongly influenced Stroop performance,
color weakness is correlated with increased accounting for 13% of the test's raw score
time needed in color naming, and reduced variance on the Word-Reading card, 29% on
distance acuity negatively impacts speed on the Color-Naming card, and 27% on the In-
the interference trial (van Boxtel et al., 2001). terference card. Education accounted for 8%
van Boxtel and colleagues (2001) indicate that of the performance variance on the Word-
visual function accounts for half of the age- Reading card, 3% on the Color-Naming card,
related variance in Stroop performance. and only 2% on the Interference card. Gender
Several studies have reported that more accounted for <4% of the performance var-
highly educated participants perform better iance on any card, "suggesting that sex-
on the Stroop Test (Anstey et al., 2000; Bar- corrections were not needed" (p. 263).
barotto et al., 1998; Houx et al., 1993; Lopez The literature taken as whole suggests that
et al., 2003; Moering et al., 2004); however, age and IQ are substantial contributors to
other investigations have not been able to Stroop performance. Gender appears to have
document a significant relationship between a more minor relationship to Stroop scores.
education and Stroop performance (Treneny It is doubtful that educational level has an
et al., 1989). impact on test performance over and above
The literature on gender differences in that accounted for by IQ.
Stroop performance is equivocal, with some Regarding the effect of language on
studies observing differences (Barbarotto Stroop performance, Rosselli et al. (2002b),
et al., 1998; Martin & Franzen, 1989; Moering in studying monolingual Spanish speakers,
et al., 2004; Small et al., 2000), others finding monolingual English speakers, and bilingual
none (Anstey et al., 2000; Armengol, 2002; Spanish-English speakers, found that groups
did not significantly differ in Stroop scores Subject Variables

with the exception of slower perf~mance
in bilinguals relative to English speakers on Sample Size
color naming in English. Testing bilinguals Fifty cases are considered a desirable sample
in their second language was associated with a size. Although this criterion is somewhat
10%-15% increase in time for color naming arbitrary, a large number of studies suggest
and a So/o-10% increase in time f~ color that data based on small sample sizes are
interference. Bilinguals who were more facile highly influenced by individual differences
in Spanish were significantly slower jon the and do not provide a reliable estimate of the
English Stroop trials, while bilin~ more population mean.
fluent in English were slower on the Spanish
I
Stroop. 1 Sample Composition Description
As discussed in other chapters, information
regarding medical and psychiatric exclusion
METHOD FOR EVALUATING criteria is important; it is unclear if geographic
THE NORMATIVE REPORTS j region, ethnicity, occupation, handedness,
or recruitment procedures are relevant, so
Our review of the literature located two until this is determined, it is best that this
Stroop manuals (Golden, 1978; T*eneny information be provided. Fluency in English
et al .• 1989), as well as 10 normative reports appears to moderate Stoop performance; thus,
and 18 clinical studies reporting control data it is preferable that information on this vari-
on the Stroop Test. The manuals are for the able also be reported.
Golden and Treneny Stroop versions; while
the normative studies are for the Comalli Age Group Interval
(Demick &: Harkins, 1997), Kaplan (Schiltz, Given the consistent evidence of a signifi-
personal communication; Strickland ~et al., cant decline in performance with advancing
1997), Golden (Ingraham et al., 1~ Ivnik age, Stroop data should be presented in age
et al., 1996; Moering et al., 2004), T~neny groupings.
(Anstey et al., 2000), and Victoria (~egard,
1981, cited in Spreen &: Strauss, 1 1991; Reporting of Educational levels
Spreen&: Strauss, 1998) versions. Amqog the
Data on the relationship between education
control data studies, seven provide ~ta for
and Stroop performance is equivocal; thus,
the Comalli version (Boone et al., 1990,
until this issue is resolved, information re-
1991, 2001; Boone, 1999; Comalli et al.; 1962;
garding the educational level of the normative
Eson, personal communication; Stuss et al.,
samples should be included.
1985), one presents data for the Kaplan
version (D'Elia et al., unpublished da~). se-
Reporting of IQ
ven report data for the Golden ../ersion
(Cohen et al., 2003; Connor et al.,. 1988; Some data suggest that IQ may be nearly as
Daigneault et al., 1992; Doan &: Sw$'dlow, predictive of performance as age, although
1999; Fisher et al., 1990; Rapport et al., this finding needs to be replicated in addi-
2001; Rosselli et al., 2002b), and two present tional studies. For the time being, Stroop
data for the Dodrill version (Dodrill, ~978a; normative data probably do not need to be
Sacks et al., 1991). presented by IQ group intervals but infor-
To adequately evaluate the Stroop Test mation on the IQs of normative samples
normative reports, eight key criterion variables should be provided.
were deemed critical. The first six o£1 these
relate to subject variables, and the re~ing Reporting Gender Composition
two relate to procedural variables. Minimal Data on the relationship between gender
requirements for meeting the criteriOJl vari- and Stroop performance are also equivocal;
ables were as follows. ! thus, until this issue is resolved, information
STROOP TEST 115
regarding the gender composition of norma- Strauss, 1991; Swerdlow et al., 1995), four
tive samples should be provided. publications present data by educational levels
(Anstey et al., 2000; Lopez et al., 2003; Miller,
2003; Moering et al., 2004), and two provide
Procedural Variables corrections for education (Ivnik et al., 1996;
Description of Test Stimuli and Moering et al., 2004). Gender is indicated in
Administration Format Used all reports but five (Comalli et al., 1962; Eson,
personal communication; Golden, 1978;
'nlere are several different versions of the
Regard, 1981, cited in Spreen & Strauss,
Stroop Test, and each version differs regard-
1991; Spreen & Strauss, 1998), four reports
ing administration and scoring procedures.
present data for an all-male sample (Cohen
'nlerefore, normative data sets must indicate
et al., 2003; D'Elia, Satz, & Uchiyama,
the precise test stimuli and administration
unpublished data; Lopez et al., 2003; Miller,
format. (For the purposes of this chapter, only
2003), two reports provide data stratified by
those studies which met this criterion were
gender (Moering et al., 2004; Swerdlow et al.,
reviewed.)
1995), and one developed gender corrections
Data Reporting for raw scores (Moering et al., 2004).
Cell sample sizes are generally adequate in
Means and standard deviations, and prefer-
11 studies (Daigneault et al., 1992; D'Elia,
ably ranges, for time in seconds for each card
Satz, & Uchiyama, unpublished data; Dodrill,
are important. In addition, it is advantageous
1978a; Ivnik et al., 1996; Lopez et al., 2003;
for data to be provided on number of errors
Moering et al., 2004; Rosselli et al., 2002b;
(corrected and uncorrected).
Schiltz, personal communication; Miller, 2003;
Swerdlow et al., 1995; Treneny et al., 1989).
Medical and psychiatric exclusion criteria are
judged to be adequate in a majority of studies
(Boone et al., 1990, 1991, 2001; Boone, 1999;
OF THE NORMS
Daigneault et al., 1992; D'Elia, Satz, &
In terms of subject variables, all studies except Uchiyama, unpublished data; Dodrill, 1978a;
one provide information on age (Swerdlow Lopez et al., 2003; Miller, 2003; Rapport et al.,
et al., 1995) but 17 do not present data by age 2001; Rosselli et al., 2002b; Schiltz, personal
groupings (Boone et al., 1991, 2001; Connor communication; Strickland et al., 1997;
et al., 1988; Cohen et al., 2003; Doan & Swerdlow et al., 1995; Treneny et al., 1989).
Swerdlow, 1999; Dodrill, 1978a; Fisher et al., Geographic recruitment area is indicated in all
1990; Ingraham et al., 1988; Lopez-Carlos but seven reports (Eson, personal communi-
et al., 2003; Rapport et al., 2001; Regard, cation; Golden, 1978; Rosselli et al., 2002b;
1981, cited Spreen & Strauss, 1998; Rosselli et Spreen & Strauss, 1991; Swerdlow et al., 1995;
al., 2002b; Sacks et al., 1991; Schiltz et al., Treneny et al., 1989); most of the data were
personal communication; Strickland et al., collected in the United States, although some
1997; Stuss et al., 1985; Swerdlow et al., data were collected in Canada (Stuss et al.,
1995). Data on IQ are reported much less 1985; Daigneault et al., 1992; and it is as-
frequently, appearing in only 10 publications sumed that the Regard, 1981, cited in Spreen
(Boone et al., 1990, 1991, 2001; Boone, 1999; & Strauss, 1991, samples were also obtained
Cohen et al., 2003; Lopez et al., 2003; Rapport in Canada), Australia (Anstey et al., 2000;
et al., 2001; Regard, 1981, cited in Spreen & Sacks et al., 1991), Israel (Ingraham et al.,
Strauss, 1991; Sacks et al., 1991; Stuss et al., 1988), and Mexico (Lopez et al., 2003). Eth-
1985); data are stratified by IQ and age in one nicity is reported in 11 studies (Boone et al.,
report (Boone, 1999). Educational level is 1990, 1991; D'Elia, Satz, & Uchiyama, un-
indicated in all but five reports (Comalli et al., published data; Doan & Swerdlow, 1999;
1962; Eson, personal communication; Golden, Dodrill, 1978a; Ivnik et al., 1996; Lopez et al.,
1978; Regard, 1981, cited in Spreen & 2003; Moering et al., 2004; Rosselli et al.,
115
2002b; Miller, 2003; Strickland et al., 1997), by number in Appendix 6. Table A6.1, the
occupational status is indicated in ¥Y five locator table, summarizes information pro-
publications (Anstey et al., 2000; Dafgneault vided in the studies described in this chapter.
et al., 1992; D'Elia, Satz, & Uchiyaina, un-
published data; Dodrill, 1978a; Ivnik et al.,
1996), and handedness is described in only SUMMARIES OF THE STUDIES
four reports (Boone et al., 1991; Ooan &
Swerdlow, 1999; lvnik et al., 1996; Regard, Published manuals for the Stroop Test are re-
1981, cited in Spreen & Strauss, 1991). Lan- viewed first, followed by normative studies and
guage and/or fluency in English is rep,rted in control groups from clinical comparison stud-
10 studies (Boone et al., 1990, 199l, 2001; ies presented in ascending chronological order
Boone, 1999; Daigneault et al., 1992; Doan & for each version of the test separately. Studies
Swerdlow, 1999; Ingraham et al., 19~; Ros- using the Comalli version are presented first,
selli et al., 2000; Miller, 2003; Stuss et al., followed by those using the Kaplan, Golden,
1985), and recruitment procedures ar. speci- Dodrill, Victoria, and Treneny versions.
fied in 16 data sets (Anstey et al., 2000; Boone
et al., 1990, 1991; Boone, 1999; Conndr et al., Manuals
1988; Daigneault et al., 1992; D'Elia, Satz, &
Uchiyama, unpublished data; Dodrill,; 1978a; [STROOP.1] Golden, 1978 (Golden Version)
Fisher et al., 1990; Ingraham et al.~ 1988; The test stimuli and administration proce-
Ivnik et al., 1996; Lopez et al., 2003; Moering dures developed by Golden are well specified.
et al., 2004; Rosselli et al., 2000; Schilb, per- Primarily utilizing previously published nor-
sonal communication; Swerdlow et al.J 1995). mative reports, the norms presented in this
Regarding procedural variables, test ,stimuli manual have largely been empirically derived
and procedures are described in all ~ports. by calculating how many items the participants
Means are presented in all data sets, although in other studies would have obtained if the test
SDs are not reported in three studies 4Aostey were discontinued after 45 seconds. In addi-
et al., 2000; Comalli et al., 1962; Golden, tion to including data from his own studies
1978). Percentiles corresponding to ra~cores, (sample sizes unknown), Golden utilized nor-
stratified by age and education, are p.:ovided mative data provided by Stroop (1935), Jensen
by Anstey et al. (2000) and Moering et al. (1965), and Comalli et al. (1962) to generate
(2004). One study provides data only for the the norms. No information is provided re-
color-naming trial (Stuss et al., 1985), and six garding demographic, IQ, or other character-
studies report data only for the . color- istics of Golden's own normative samples.
interference trial (Boone et al., 1991; Boone, Using the tables in the manual, all raw
1999; Cohen et al., 2003; Daigneault! et al., scores can be converted to T scores. For
1992; Sacks et al., 1991; Schiltz, personal participants younger than 17 and older than
communication). Two studies provide· cutoff 45, age corrections need to be applied before
scores (Dodrill, 1978a; Treneny et al.,;1989), the T-score conversion can be made. The man-
and one study presents data for the fi~t half ual cautions that the age corrections for adults
and the second half of the color-inter&rence over age 65 and children under age 17 are
I
trial separately (Schiltz, personal communi- considered to be "experimental."
cation). Several studies report error s~res as The normative data contained in Golden's
well as time scores (Boone et al., 1990; P'Elia, manual are not reproduced here, and the
Satz, & Uchiyama, unpublished data; ~gard, interested reader is referred directly to this
1981, cited in Spreen & Strauss, ' 1991; publication for further information.
Spreen & Strauss, 1998), and one presents
information on alternate forms (Sacks et al., Study strengths
1991). 1. The Stroop cards developed by Golden
The text of study descriptions contains re- and test administration procedures are
ferences to the corresponding tables ideptified well described in the manual.
STROOP TEST 117
2. Mean scores are reported for number of Test administration procedures are care-
items completed. fully specified. Means and SDs for number
3. Some of the data are presented by age of items completed, number of incorrect
group intervals, although the age ranges responses, and color scores are provided for
within each grouping are broad (15-45, the Color task. Similarly, means and SDs for
46-64, 65--80) and of questionable clin- number of items completed, number of in-
ical utility. correct responses, and color-word scores are
reported for the Color-Word task. In addition,
Considerations regarding use of the study optimal cutoff scores are provided for each
1. A major problem with Golden's use of score for each age grouping, and percentile
other published normative data to gen- ranks for raw scores are also reported for each
erate his norms is that the Stroop for- age grouping. This is a proprietary test, and
mats and procedures used by the other the normative data are available in the test
investigators differed from Golden's manual. A significant age effect on performance
version (e.g., in other formats, partici- was detected, but no relationship between
pants scanned from left to right, while in education or gender and test performance was
his version participants scanned the observed.
stimuli vertically from top to bottom;
other versions required participants to Study strengths
complete the entire stimulus cards ra- 1. Cell sample sizes are adequate.
ther than stopping at 45 seconds, used 2. Minimally adequate exclusion criteria.
colored rectangles rather than colored 3. Information is provided for each age
Xs, used a wall chart presentation rather grouping on mean age, education, and
than a standard page presented up close gender distribution.
to the subject, etc.). 4. Test stimuli and administration proce-
2. No information is provided regarding dures are specified.
the number of participants in Golden's 5. Means and SDs are reported for each
sample. age grouping on a wide range of scores,
3. No information on exclusion criteria is as well as optimal cutoff scores and
available for Golden's sample. percentile ranks.
4. The gender, educational, and IQ levels
of Golden's sample are not reported. Considerations regarding use of the study
5. No SDs are provided. 1. The age group intervals are quite broad.
2. No information regarding IQ.
3. High educational level of the sample.
[STROOP.2] Trenerry, Crosson, DeBoe,
and Leber, 1989 (Trenerry Version)
This manual presents the standardization data Normative Studies and Control Groups
for this version of the Stroop Test. The sample from Clinical Comparison Studies
consisted of 156 adults ranging in age from Comalli Version
18 to 79, divided into two age groups: 18-49
[STROOP.3] Comalli, Wapner, and Wemer,
(n = 106) and 5~79 (n =50). Exclusion cri-
1962 (Comalli Version) (Table A6.2)
teria included history of neurological disorder,
major psychiatric illness, or physical handi- These authors provide data (on the Comalli
caps which might affect performance. The Stroop version) for 235 participants aged 7-80
younger group averaged 30.34 (8.57) years, apparently in Massachusetts as a part of their
with 14.68 (2.44) years of education, and study of the effects of aging on the Stroop
contained 43 males and 63 females. The older task. For the purposes of this review, the data
group averaged 62.68 (7.93) years, with 14.70 on the 63 adult participants will be reported.
(3.24) years of education, and contained 26 The adult participants were divided into four
males and 24 females. age groupings: 17-19 (n = 18), 25-34 (n = 14),
35-44 (n = 16), 65-80 (n = 15). Those 17-19 [STROOP.Sl Stuss, Ely, Hugenholtz, Richard,
years old were undergraduate stude'*s, those LaRochelle, Poirier, and Bell, 1985 (Comalli
25-34 or 35-44 years old were drawnifrom an Version) (Table A6.4)
evening college, and those 65-80 ~ars old These authors collected Stroop data on 20
were from a community old-age club; control participants (13 male, seven female) in
Mean time in seconds for each card is Canada as a part of their investigation of
charted in a graph for each age group~ no SDs the neuropsychological effects of closed head
are provided. : injury. Participants spoke either English or
French. Mean age was 29.2 (12.0), mean years
Study strengths 1 of education was 12.5 (2.0), and mean
l. Data are presented in narrt>w age WAIS IQ was 106.6 (13.4). Participants were
groupings. 1 paid $15 for their participation.
2. Good description of test sti~ and The Comalli test stimuli and administration
administration procedures. . procedures were employed. The mean and
3. Information on geographic rec¥tment SD for time in seconds to name colors were
area is provided. ' 64.0 (12.9) for the control group. (Data from
4. Mean time in seconds is reported. the two other trials are not provided.) Per-
formance on Color Naming was significantly
depressed in the head injury group relative to
Considerations regarding use of the study controls; groups did not differ in word reading
l. No exclusion criteria. . or color interference.
2. No information regarding gendet or IQ,
and cursory data on educational tvel are
provided only for those 17-19 years old. Study strengths
3. Small individual cell sizes. I l. Data provided on gender, age, educa-
4. No SDs reported. tion, IQ, language, and geographic area.
2. Test stimuli and administration proce-
[STROOP.4] Eson, Penonal Communica~n dures are speci6ed.
(Coma IIi Version) (Table A6.3) 3. Mean and SD for time to name colors
are reported.
Eson provides Stroop data on 63 older par-
ticipants in four age groupings that reftect the
following mean ages: 63.2 (n = 1~, 67.0 Considerations regarding use of the study
(n = 16), 72.0 (n = 16), and 78.3 (n = 16). l. No information on exclusion criteria.
The Comalli test stimuli and adminiktration 2. Data were collected in Canada with at
procedures were utilized. Means and ~Ds are least some participants French-speaking;
reported. cultural and linguistic factors may limit
usefulness for clinical interpretation in
Study strengths the United States.
l. Large overall sample size, an~ while 3. Data from the word-reading and color-
individual cell sizes are small, ttey are interference trials are not reported.
for very restricted age ranges. 1 4. Small sample size.
2. Test stimuli and administration 1 proce- 5. Undifferentiated age range.
dures are specified.
3. Mean time in seconds and Sps are [STROOP.6] Boone, Miller, Lesser,
reported. ' Hiii-Gutierrez, and D'Eiia, 1990 (Comalli
Version) (Table A6.5)
Considerations regarding use of the stt,dy Data were collected on 61 middle-aged and
l. No information on exclusion qriteria, older individuals ranging in age from 50 to 79
education, gender, IQ, or other char- recruited as controls in southern California
acteristics. through newspaper ads, flyers, and personal
STROOP TEST 119
contacts as part of a study of the effect of Considerations regarding use of the study
aging on executive abilities. Participants had 1. The sample size within each age group
no history of psychotic, major affective, or interval is small (see 1 above).
alcohol and other drug-dependence disorders 2. High educational and IQ levels of the
and spoke English ftuently (a handful of par- sample.
ticipants spoke English as a second language).
Participants were excluded if there was a [STROOP.7] Boone, Ananth, Philpott, Kaur,
history of neurologic disease, such as stroke, and Djenderedjian, 1991 (Comalli Version)
Parkinson's disease, or seizure disorder. Also (Table A6.6)
excluded were individuals with laboratory Stroop data were collected on 16 controls as
findings showing serious metabolic abnormal- part of a study on the neuropsychological
ities (e.g., low sodium level, elevated glucose characteristics of obsessive-compulsive disor-
level, or thyroid or liver function abnormal- der (OCD). Participants were recruited in
ities). Eighteen percent of the original sample southern California from newspaper adver-
of 74 were eventually excluded due to the tisements and from siblings (n =9) of OCD
presence of previously unidentified strokes or patients. Medical exclusion criteria included
other significant lesions documented on MRI, history of alcohol or drug abuse, head injury,
(n = 9), metabolic abnormalities or undiag- seizure disorder, cerebral vascular disease or
nosed medical illness (n = 2), or evidence stroke, current or past psychiatric disorder, or
from laboratory studies and EEG findings of any renal, hepatic, or pulmonary disease. Par-
alcohol abuse and substance intoxication ticipants included nine women and seven men,
(n =2). The final sample (n =61) included and 19% were left-handed (n = 3). Fourteen
25 men and 36 women grouped by three age were Caucasian, and two were Asian; and all
decades: 50-59 (n = 25), 60-69 (n = 21), and were ftuent in English. Two participants had a
7~79 (n = 15). All but 10 participants were history oflearning disability. Mean age was 35.8
white: four were African-American, three (13.7), mean educational level was 15.2 (2.8),
were Asian, and three were Hispanic. Mean and mean WAIS-R FSIQ was 109.1 (10.9).
educational level was 14.34 (2.63) and mean The Comalli version of the Stroop (i.e., word
WAIS-R FSIQ was 113.79 (13.51). reading, color naming, color interference) was
The Comalli version of the Stroop was administered. Mean (SD) for time in seconds
administered (i.e., word reading, color nam- to complete the color-interference portion
ing, and color interference). Mean time in of the test was 112.9 (22.5). Controls and
seconds as well as number of errors, with patients did not differ in test performance.
SDs, are presented by age grouping for each
card. A significant decline with age was Study strengths
observed for word reading and color naming, 1. Good exclusion criteria, with the excep-
with a trend toward a decline with age on tion that two participants had histories of
color interference. learning disability.
Study strengths FSIQ, gender, handedness, ftuency in
1. Overall sample size is large, although English, ethnicity, recruitment proce-
individual cell sizes are small. dures, and geographic area is reported.
2. Data are presented by age groupings. 3. Information on test stimuli and admin-
3. Good exclusion criteria. istration procedures.
4. Information regarding gender, educa- 4. Mean time in seconds and SD are pro-
tional level, IQ, geographic area, ethni- vided but only for the color-interference
city, ftuency in English, and recruitment card.
procedures is provided.
5. Test stimuli and procedures are specified. Considerations regarding use of the study
6. Means and SDs are reported for both 1. Small sample size.
time and errors. 2. Undifferentiated age range.
3. High educational level; two participants [STROOP.9] Boone, 1999 (Comalli Version)
had a history of learning disability. (Table A6.11)
4. No data reported for word-reading and The author obtained Stroop data on 155
color-naming trials. middle-aged and older individuals (age range
45-84) recruited as described by Boone et al.
[STROOP.8] Demick and Harkins, 1997 (Coma/Ji
(1990); data from the 1990 study were
Version) (Tables A6. 7-A6.1 0)
included in the 1999 publication. Mean age of
The sample consists of 231 individuals the sample was 63.07 (9.29), mean years of
recruited in Massachusetts who participated education was 14.57 (2.55), and mean WAIS-R
in a study assessing the relationship between FSIQ was 115.41 (14.11). Fifty-three were
field dependence-independence (FQI) cog- male and 102 were female. Medical and psy-
nitive style and driving behavior. Participants chiatric exclusion criteria are the same as in
were community-dwelling individuals who the 1990 publication, with the exception that
in telephone screening denied any ~tory of participants with significant white-matter hy-
major impairment in perception, cognition, or perintensities documented on MRI were
motor execution and described themselves as retained in the sample. All participants con-
having good overall health; corrected visual sidered themselves healthy, although 51 had
problems were allowed. The average educa- some evidence of vascular illness (defined
tional level of the sample was high schrol plus as cardiovascular disease and I or significant
some college courses completed. white-matter hyperintensities on MRI) based
The Comalli cards and Kaplan administra- on self-report or evidence on examination of
tion procedures (i.e., color naming. word at least one of the following: current or past
reading, color interference) were employed. history of hypertension (n = 39), arrhythmia
Means, SDs, and ranges for time in seconds, (n = 8), large area of white-matter hyper-
errors, color difficulty factor (total time on intensities on MRI (e.g., 10 cm2 ; n = 7), cor-
B/total time on A), and interference factor onary artery bypass graft (n = 3), angina
(total time on C - total time on B) are pro- (n = 2), and old myocardial infarction (n = 1).
vided for four age groupings (20--39, 40--59, Twenty-four participants were currently on
60--74, 75+ years). cardiac and/or antihypertensive medications.
The Comalli version of the Stroop was
Study strengths administered. Means and SDs for time in
1. Overall sample size is large, with in- seconds to complete the color-interference
dividual cell sizes exceeding 50. portion of the test are provided. A stepwise
2. Data are presented by age groupings. regression analysis revealed that age and
3. Probably adequate exclusion criteria. FSIQ were significant contributors to Stroop
4. Information regarding gender, overall color-interference performance, accounting
educational level, and geographic region for 15% and 13%, respectively, of test score
is provided. variance; educational level, gender, and vas-
5. Test stimuli and procedures are indicated. cular status did not account for a significant
6. Means, SDs, and ranges are prm4ded. amount of unique test score variance. Stroop
normative data are presented for color-inter-
Consideration regarding use of the study ference time in seconds stratified by IQ and
1. No information regarding inteDectual age (< 65 and ~65; average, high average, and
level. superior IQ).
Other comments
1. Theoretical issues concerning the Stroop Study strengths
(e.g., process vs. achievement measures, 1. Large overall sample size.
identification of a cognitive sty~) are 2. Presentation of the data by IQ and age
discussed. groupings.
STROOP TEST 121
3. Comprehensive medical and psychiatric Kaplan Venion

exclusion criteria including MRI brain
[STROOP.11] O'Eiia, Satz, and Uchiyama,
scans on all participants.
Unpublished Data (Kaplan Version)
4. Information regarding educational level,
(Table A6.13)
gender, geographic area, recruitment pro-
cedures, and fluency in English is provided. These data were collected in 1993 and 1994
5. Mean times in seconds and SDs for color during the course of an Federal Aviation
interference are reported. Administration/Equal Employment Opportu-
6. Test administration format is specified. nity Commission (FAA/ EEOC)-mandated
study to examine neuropsychological func-
Considerations regarding use of the study tioning of airline pilots. The sample consisted
l. Individual IQ by age groupings range in of 197 male, Caucasian airline pilots aged
size from 16 to 37. 40--59 employed by major airplane manu-
2. High IQ and educational level of the facturers in the United States. All pilots had
sample. recently passed their yearly comprehensive
3. No data reported for word-reading and FAA physical examination. Data are presented
color-naming trials. in two age groupings: 4~9 (n = 118) and
50-59 (n = 79). Those 4~9 years old had an
[STROOP.10] Boone, Swerdloff, Miller, average of 16.1 (1.9) years of education, and
Geschwind, Razani, Lee, Gaw Gonzalo, those 50-59 years old group had 15.6 (2.0)
Haddal, Rankin, Lu, & Paul, 2001 (Comalli years of education.
Version) (Table A6.12) The Comalli cards and Kaplan administra-
Stroop performance was assessed in 22 male tion procedures (i.e., color naming, word
controls as part of a study on neuropsycholo- reading, color interference) were used to
gical function in adult Klinefelter's syndrome. obtain Stroop data on the pilots as part of
Participants were recruited from newspaper a 55-minute neuropsychological screening
and radio ads and flyers in the southern battery. Means and SDs for time in seconds,
California area and paid for their participa- errors, and near-miss (i.e., self-corrected)
tion. Exclusion criteria included history of errors are provided.
learning disability, major psychiatric disorder,
substance abuse, or neurologic disorder. All Study strengths
participants were fluent in English. Mean age 1. Overall sample size is large, with
was 34.32 (14.81), mean years of education individual cell sizes exceeding 50.
was 13.36 (2.15), mean WAIS-R (Satz-Mogel) 2. Data are presented by age groupings.
VIQ was 106.46 (17.01), and mean PIQ was 3. Adequate exclusion criteria.
107.46 (16.58). 4. Information regarding gender, educa-
Means and SDs for the color-interference tional level, recruitment procedures,
trial are reported. ethnicity, and occupation and some in-
formation regarding geographic region is
1. Information regarding age, education, 5. Test stimuli and procedures are
gender, IQ, geographic area, language, indicated.
and recruitment procedures is provided. 6. Means and SDs for time and near-
2. Adequate exclusion criteria. miss (i.e., self-corrected) errors are
3. Means and SDs are reported. provided.
l. Data are not stratified by age. 1. All-male sample.
2. Small sample size. 2. High educational level of the sample.
3. All-male sample. 3. No information regarding IQ.
[STROOP.12] Schiltz, Personal Comm1111ication [STROOP.13] Strickland, D'Eiia, James, and

(Kaplan Version) (Table A6.14) Stein, 1997 (Kaplan Version)(Table A6.15)
The sample consists of 50 (28 male, 22 female) Stroop data were collected in southern Cali-
native English-speaking participants recruited fornia on 42 Mrican-American participants
from the University of California at Los (15 males, 27 females) aged 19-41 with no
Angeles undergraduate introductory psychol- remarkable history of neurologic, psychiatric,
ogy courses during 1988-1989. Th~e were cardiovascular disease or substance abuse.
healthy normal adults aged 18-20 without a Mean age for the whole sample was 30.17 (6.34)
history of head trauma or loss of cqnscious- years; mean age of males was 31.93 (5.26) years,
ness. Average years of education for t\e group and that of females was 29.19 (6.75) years.
was 13.36 (0.63, range 13-15). Mean educational level for the sample was
All students were required to participate in 14.76 (2.24) years.
ongoing research as a part of their. course- The Comalli stimulus cards and Kaplan
work, and students self-selected to the various administration procedures (i.e., color naming,
studies based on the written descriptions of word reading, color interference) were
the studies. The Comalli stimulus cards and employed to obtain Stroop data. Mean times
Kaplan administration procedures (i.e., color in seconds and SDs are provided for each of
naming, word reading, color interference) the three cards. Errors and near-miss (i.e.,
were used as part of a larger neuropsycholo- self-corrected) errors were tabulated. Women
gical battery assembled for the putpose of demonstrated significantly better performance
collecting norms. All participants were tested than men on cards 1 and 2. There was a
individually. Total battery length was 55 similar trend noted on card 3.
minutes, and the Stroop was adm4listered
about 30 minutes into the protocol. Study strengths
Means, SDs, and ranges in seconds are 1. Adequate exclusion criteria.
reported for the first half of each stimulus 2. Information regarding gender, ethnicity,
card as well as for each card in total. Perfor- age, educational level, and geographic
mance time on the second 50 items can be area is reported.
calculated by subtracting the time to complete 3. Information regarding test stimuli and
the first half of the card from the total time to test administration procedures is pro-
complete the whole card. vided.
4. Means and SDs for time in seconds,
Study strengths errors, and near-miss (self-corrected)
1. The sample size is adequate for the errors is provided for each card.
restricted age interval.
2. Exclusion criteria were minimally Considerations regarding use of the study
adequate. 1. Small sample size.
3. Data on gender composition, educational 2. Undifferentiated age range.
level, geographic area, and recrUitment 3. High educational level of the sample.
procedures are reported. 4. No information regarding IQ.
4. Test stimuli and administration proce-
dures are specified. [STROOP.14] Miller, 2003;
5. Means for time in seconds and SDs are Personal Communication (Kaplan Version)
reported. (Table A6.16)
6. Data provided for the first half of each The investigation used participants from the
card as well as total. Multi-Center AIDS Cohort Study (MACS).
The data were collected from a sample of
Consideration regarding use of the st1¥ly seronegative homosexual and bisexual males
1. No data on IQ. for the purpose of establishing normative data
STROOP TEST 123
for neuropsychological test performance ence score, are provided. Performance on

based on a large sample. There were 522 rapid word reading and color naming was
participants in the Color Naming, 521 in the significantly slower (by eight and three words,
Word Reading. and 692 in the Interference respectively) than the norms reported by
conditions. Mean age for the full sample used Golden for English-speaking individuals,
in the Interference condition was 40.57 (7.5) which the authors suggest may be due to
years, and mean education was 16.31 (2.3) longer speaking times because the Hebrew
years; 91.2% were Caucasian, 2.9% Hispanic, words were two-syllable. However, scores
5.5% black, 0.4% other. All participants were for the color-interference trial were not sig-
native English speakers. nificantly different from English-speaking
The three conditions of the Kaplan version language norms, although the interference
of the Stroop were administered according score was significantly larger. The authors
to standard instructions. The data are parti- hypothesize that Hebrew speakers may show
tioned by three age groups (25-34, 35-44, less of an interference effect because Hebrew
45-59) x three educational levels (< 16, 16, readers are accustomed to reading words
> 16 years). without vowels and determining meaning
from context, allowing them on the color-
Study strengths interference trial to "read" the words as other
1. The overall sample size is large, and than color names, thereby reducing any in-
most individual cells have more than 50 terference effect. No significant differences
participants. between men and women were detected.
2. Normative data are stratified by
age x education. Study strengths
3. Information on age, education, ethnicity, 1. Relatively large sample size for reason-
and native language is reported.
ably restricted age range.
4. Means and SDs are reported. 2. Information regarding native language,
educational level, gender distribution,
Considerations regarding use of the study and geographic area is reported.
1. All-male sample. 3. Test stimuli and procedures are
described.
3. No information on exclusion criteria. 4. Means and SDs for number of items
completed and interference score are
Golden Version reported.
5. Data provided for Hebrew version.
[STROOP.15] Ingraham, Chard, Wood, and
Mirsky, 1988 (Golden Version) (Table A6.17)
Data were gathered on 46 college students
1. Psychiatric exclusion criteria reported
and college-educated adults in Tel Aviv using
but no medical exclusion criteria.
"the general format of Golden's 1978 version
2. No information regarding recruitment
with new randomization, a bold typeface, and
procedures or IQ level.
Hebrew lettering," which is carefully detailed.
The sample consisted of 28 men and 18 wo-
men, with an average age of 28.4 (3.2) years [STROOP.16] Connor, Franzen, and Sharp,
and a range of 24-36 years. Exclusion criteria 1988 (Golden Version) (Table A6.18)
included prior psychiatric disorder, primary Stroop data were obtained on 40 college stu-
language which was not Hebrew, and prior dent volunteers in West Virginia (17 male,
familiarity with the Stroop Test. 23 female) who ranged in age from 18 to 25,
Means and SDs for number of items with the exception of one 32-year-old.
completed within 45 seconds for the three The Golden version of the Stroop was
stimulus cards, as well as Golden's interfer- administered with either standard instructions
as detailed in the test manual or •tandard Participants had no history (as judged through
instructions plus six suggestions ("l~king at medical records) of color blindness, cataracts,
no more than three words at a time; focusing or glaucoma.
on only one letter in the word; remembering The Golden Stroop Test stimuli and ad-
that the same color never occurs twlce con- ministration procedures were employed.
secutively; going at an even, steady pPce; try- Means and SDs are reported for number of
ing not to become distracted or l~e one's items completed on each trial.
place; and not repeating an already-correct Some participants (five female, three male)
answer when correcting a mistake"). had difficulty discriminating between the
Participants were administered th~ Stroop colors blue and green on the color trial.
at baseline (pretest), following five practice
sessions (post-test), and at a 1-week follow-up. Study strengths
No effect of gender or instructioq format 1. Data presented in a homogenous age
was documented. A significant effect pf prac- grouping.
tice was found between the pre- and tost-test 2. Information is given regarding mean age,
but not between the post-test and follow-up. mean educational level, gender, mean
Data are presented in means and ~Ds for Blessed Dementia Scale score, ge~
number of items completed for the jPretest, graphic area, and recruitment procedures.
posttest, and follow-up sessions. ' 3. Information is given on test stimuli and
I
administration procedures.
Study strengths 4. Means and SDs reported for number of
1. Information on the effects of practice, items completed.
gender, and alternative instruc~ons on 5. Test administration format was described.
Stroop performance is provided..
2. Information on age, gender, ~d geo- Considerations regarding use of the study
graphic area, with some inform~on on 1. Relatively small sample size.
education and recruitment prooedures, 2. No information regarding IQ.
is reported. , 3. High educational level of the sample.
3. Test stimuli and procedures are siiecified. 4. Unclear exclusion criteria.
4. Data are presented in means and SDs
for number of items completed. : [STROOP.18] Daigneault, Braun, and Whitaker,
1992 (Golden Version) (Table A6.20)
Considerations regarding use of the stfdy Stroop data were obtained on 128 French-
1. Relatively small sample size. ; speaking participants in Canada as part of
2. Undifferentiated age range, althOugh it a study investigating the effects of aging on
is somewhat restricted. I prefrontal lobe skills. Participants were
3. No information on exclusion ~riteria recruited through ads, trade union collabora-
or IQ. tion, and the help of a large sports center.
4. Data are not broken down by gender or Exclusion criteria included consumption of
education. more than 24 beers, five bottles of wine, or
15 ounces of spirits per week; consumption
[STROOP.17] fisher, Freed, and Corkin, ~990 of cocaine, LSD, or psychostimulants; any
(Golden Version) (Table A6.19) i neurological or psychiatric consultation, psy-
The authors collected Stroop data on 36 older choactive medication, head trauma with hos-
controls (typically spouses of patients) from pitalization, or major surgery (e.g., cardiac).
southern California as part of an inves,gation Participants were divided into two age
of Stroop performance in Alzheimer's cliSease. groupings: 20-35, with a mean of 27.71 (4.05)
Mean age was 72.9 (8.3) years, mean .educa- years (n = 70), and 45-65, with a mean of
Blessed Dementia Scale score was 1.'

tional level was 14.6 (2.7) years, an4 mean
(6.1).
The sample included 13 males and 23 f~males.
56.62 (5.29) years (n =58). The younger
group contained 38 men and 32 women;
they were primarily specialized blue-collar
I
STROOP TEST 125
workers, although some specialized white- recruited through newspaper ads and posted
collar and unskilled blue-collar professions advertisements. No subject had a history of
were represented. The older group contained psychiatric illness, substance abuse or de-
30 men and 28 women, and slightly more than pendence, recreational drug use in the month
half were specialized blue-collar workers, with prior to testing, schizophrenia in a first-degree
some unskilled blue-collar professions, spe- relative, sustained loss of consciousness, se-
cialized white-collar occupations, and profes- vere neurologic or medical illness, or psycho-
sional occupations represented. The mean tropic medication use. Three participants were
educational level of the younger group was excluded based on a urinalysis positive for iJ-
12.36 (2.09) years, and that of the older group licit drug use.
was 12.11 (3.63) years. Participants were divided into "psychosis-
Mean number of items completed and SDs prone" (n =26) and "non-psychosis-prone"
for the color-interference portion of the test (n = 46) groups based on MMPI criteria
are reported. The two age groups differed (Goldberg index ~60 and F ~70 or Wiggins
significantly in test performance, with the Psychoticism index ~60). Psychosis-prone
younger group outperforming the older group. individuals scored significantly below non-
psychosis-prone participants on the interfer-
Study strengths ence trial and "interference ratio." Women
1. Good exclusion criteria. performed significantly better than men on
2. Large overall sample size, and each of color naming.
the two age groupings has more than 50 Means and SDs are reported for word
participants. reading, color naming, interference, and the
3. Information regarding educational level, interference ratio for the psychosis-prone and
gender, occupations, geographic area, non-psychosis-prone groups (and for sub-
and recruitment procedures is provided. groups based on MMPI scales that determine
4. Information on test stimuli and admin- psychosis-proneness) and for men and women
istration procedures is reported. separately.
5. Means and SDs for number of items
completed on part C is provided. Study strengths
1. Excellent exclusion criteria.
Considerations regarding use of the study 2. Information provided regarding gender
1. Data were obtained on French-speaking and recruitment strategies.
participants in Canada; thus, it is unclear 3. Large overall sample size.
whether these data are appropriate for clin-
ical interpretation on English-speaking Considerations regarding use of the study
individuals in the United States. 1. No information regarding age, educa-
2. No information regarding IQ (although tion, IQ, ethnicity, or geographic area.
mean scores on the vocabulruy subtest of
the French-language WAIS analog are [STROOP.20] lvnik, Malec, Smith, Tangalos,
reported). and Petenen, 1996 (Golden Version)
3. No data provided for the first two sec- (Table A6.22)
tions of the Stroop Test. This study presents normative data for per-
4. Test administration format may have formance on the Golden (1978) version of the
been altered (i.e., participants appeared Stroop Test obtained on 356 individuals
to scan the stimulus cards across rows between the ages of 56 and 94, who partici-
from left to right). pated in the ongoing Mayo Older Americans
Normative Studies (MOANS), a project to
[STROOP.19] Swerdlow, filion, Geyer, develop normative data for elderly individuals
and Braff, 1995 (Golden Version) (Table A6.21 ) on various neuropsychological tests. The data
The authors collected Stroop data on 72 are derived from a population of "almost
"normal" controls (34 males, 38 females) exclusively Caucasian older adults who live in
an economically stable region of the United Analyses did not suggest that a performance
States" (the area surrounding Rochester, MN). correction was necessary for gender.
All participants were community-dwelling, had
no active neurologic or psychiatric disorder, Study strengths
and had undergone recent physical exams. 1. Minimally adequate exclusion criteria.
Data are reported in discrete age ranges. 2. Information regarding age, education,
Age categorization used the midpoint inter- gender, handedness, ethnicity, recruit-
val technique. The raw score distribution at ment procedures, and geographic area is
each midpoint age was "normalized" by as- reported.
signing standard scores with a mean of 10 and 3. The data are stratified by age group based
SO of 3, based on actual percentile ranks. The on the midpoint interval technique.
authors provided tables of age-corrected 4. Sample sizes for the five age groupings
norms for each age group. The procedure for spanning ~79 exceed 50.
clinical application of these data is described 5. The test version and scoring procedures
in the original article (Ivnik et al., 1996) as are specified.
follows:
first select the table that corresponds to that Considerations regarding use of the study
person's age. Enter the table with the test's raw 1. The measures proposed by the authors
score; do not use "corrected" or "final" scores for are quite complicated and might be dif-
tests that might present their own age- or education- ficult to use in clinical practice.
adjustments. Select the appropriate column in the 2. No information on IQ.
table for that test. The corresponding row in the left-
most column in each table provides the MOANS [STROOP.21] Doan and Swerdlow, 1999
Age-Corrected Scaled Score . . . for your sub- (Golden Version) (Table A6.23)
ject's raw score; the corresponding row in the right-
The Golden Stroop version translated into
most column indicates the percentile range for that
Vietnamese was administered to 30 native
same score.
Vietnamese speakers, while the standard
Mean and SO scores for performance by Golden version was given to 30 native English
age are not reported; however, the raw per- speakers. All participants resided in the
formance score can be easily translated to San Diego area. The average age of the 13
percentile performance scores (and standard males and 17 females in the Vietnamese group
scores) using the data tables. MOANS scaled was 34.4 (13.1) years, with a range of 19-68,
scores by age and education level (A&E-MSS) and the average educational level was 14.3
have to be empirically derived using the fol- (3.5) years; 47% were students, and the
lowing equation: sample averaged 16 (7) years residence in
the United States. The 12 male and 18 female
A&E-MSSstmop = K + (Wt * A-MSSstroop) native English speakers averaged 31.2 (11.9)
- (Wz * Education) years of age, with a range of 19-57, and
average educational level was 15.4 (1.6); 30%
Where K is a constant for each test, W1 is a were students. All but one subject were right-
weight to be applied to the age-corrected handed (one Vietnamese speaker was ambi-
MOANS scaled score, and W2 is a weight to dextrous).
be applied to the person's education. For the The colors employed for the Vietnamese
Stroop Test, the values are as follows: translation were blue, brown, and red because
the Vietnamese word for green is used for
K both blue and green. Means and SDs for
number of responses for the word reading,
Word 3.47 1.10 0.34 color naming, and interference sections and
Color 1.88 1.10 0.23 the Golden interference score are provided.
Interference 1.38 1.09 0.19 No significant differences in performance
STROOP TEST 127
were found between the Vietnamese speakers Limitations regarding use of the study
and English speakers or between monolingual 1. Small sample size.
Vietnamese speakers and bilingual speakers. 2. Data are not stratified by age.
3. No specific information regarding re-
Study strengths cruitment strategies.
1. Data provided for Vietnamese test ver- 4. High educational level.
sion.
(STROOP.23] Rosselli, Ardila, Santisi, Arecco,
2. Information regarding geographic area,
Salvatierra, Conde, and Lenis, 2002 (Golden
language, ethnicity (although incomplete),
age, education, handedness, and gender is
provided. Stroop Test data were obtained on 40 English
3. Test stimuli and procedures are de- monolinguals, 71 Spanish-English bilinguals
scribed. (90% with Spanish as the first language), and
4. Means and SDs for number of items 11 Spanish monolinguals in south Florida who
completed and interference score are were primarily college students, their family
reported. members, and friends. All were right-handed.
The average age of the 32 male and 39 female
Considerations regarding use of the study bilinguals was 31.98 (13.14), and average years
of education was 14.92 (2.35). The 13 male
1. No exclusion criteria are reported.
2. Data are not stratified by age. and 27 female English monolinguals averaged
3. Small sample sizes. 35.90 (13.08) years of age and 15.35 (2.45)
4. No information regarding IQ or years of education. The three male and eight
recruitment procedures. female Spanish monolinguals averaged 40.91
(15.17) years of age and 14.25 (3.49) years of
education. None had any psychiatric or neu-
[STROOP. 22] Rapport, Van Voorhis, Tzelepis,
rological conditions, and all had normal
and Friedman, 2001 (Golden Version)
MMSE scores. All bilinguals had had at least
(Table A6.24)
some formal education in English and aver-
Stroop data were collected on 32 controls aged 19 years speaking the second language.
(19 males, 13 females) who were either un- A Spanish version of the Stroop Test (Rey,
dergraduates at a large midwestern university unpublished) was administered to the mono-
or residents in the neighboring metropolitan lingual Spanish speakers and to the bilinguals,
area as part of a study of executive function who also completed the English version (or-
in adults with ADHD. Exclusion criteria der of administration of the two versions was
included history of significant neurologic dis- randomized). Number of errors and time in
order (head injury, stroke, seizure disorder), seconds to complete all stimuli (instead of the
current substance abuse, or scores greater number of items completed in 45 seconds)
than 1 SD higher than mean values on ADHD were collected.
behavior rating scales. Mean age was 33.2 Means and SDs for the three trials are
(13.2), mean years of education was 14.8 (2.5), reported for the three groups as well as for three
and mean WAIS-R FSIQ was 108.0 (7.7). subgroups of bilinguals (unbalanced-English-
Means and SDs for the word trial and color dominant, unbalanced-Spanish-dominant, and
trial are reported. balanced).
Groups did not significantly differ in Stroop
Study strengths scores, with the exception of slower perfor-
1. Generally adequate exclusion criteria. mance in the bilinguals relative to English
2. Information regarding age, education, speakers on color naming in English. Testing
gender, IQ, and geographic location is bilinguals in their second language was associ-
reported. ated with a lOo/o-15% increase in time for color
3. Test stimuli and procedures are de- naming and a 5o/o-10% increase in time for color
scribed. interference. Bilinguals who were more facile in
Spanish were significantly slower on the En- version) were included in the battery. WAIS-
glish Stroop trials, while bilinguals more fluent III Block Design raw scores are included
in English were slower on the Spanish Stroop. in Tables A6.26-A6.29. Mean performance
on the Marin Marin Acculturation Scale for
Study strengths the Los Angeles sample was 17.61 (6.19).
1. Data on Spanish-language and English- For the Los Angeles group, Picture Vocabulary
language Stroop performance in a large subscale scores from the Woodcock-Johnson-
sample of bilingual participants (n = 71) III Tests of Achievement (Mean= 5.36, SD =
as well as a smaller group of monolingual 6.01) and the Bateria Woodcock-Muiioz-R,
Spanish speakers are provided. Pruebas de habilidad cognitiva-R (Mean=
2. Adequate exclusion criteria. 29.77, SD =5.37) were used to assess level of
3. Information provided regarding age, English and Spanish word expressive abilities.
education, gender, and handedness, as The results are presented by years of edu-
well as comprehensive information on cation (0-6, 7-10), age (18-29, 30-49 years),
language characteristics. and education and age (18-29 years old, 0-6
and 7-10 years of education; 30-49 years old,
Considerations regarding use of the study 0-6 and 7-10 years of education). The authors
1. The administration format was altered found a significant difference (p< 0.05} in
(time to finish the stimuli rather than performance on the Stroop (color and color/
number of responses at 45 seconds). word interference) between the two education
2. No information regarding IQ. groups. However, the two age groups did not
3. Data are not stratified by age. differ significantly on any of the sections of the
4. High educational level. Stroop. No significant differences in scores
between individuals from Los Angeles and
[STROOP.24] Lopez-Carlos, Salazar, Mexico were noted.
Villasenor Saucedo, and Peiia, 2003
(Golden Version) (Tables A6.26-A6.29)
Study strengths
The Golden version of the Stroop was used in
1. Large sample for age and education
a study investigating the effects of demo-
groups.
graphic variables on cognitive abilities in
2. Data availability for a healthy, employable,
Spanish-speaking individuals with low educa-
monolingual Spanish-speaking group with
tion. The total sample included 115 volunteer
low education level
monolingual Latino men with $10 years of
3. The sample is stratified into two educa-
formal education, who worked at manual labor
tion groups, two age groups, and four
in the Los Angeles area (n = 65) and Jalisco,
age x education groups. Additionally,
Mexico (n =50). Volunteers were recruited
data are available for United States and
from posted advertisements in workplaces and
Mexico.
personal solicitations. The mean age for the
4. The sample composition is well
sample was 28.23 (8.74) years and mean
described in terms of age, education,
education was 6.66 (2.54) years. Exclusion
gender, geographic area, and recruit-
criteria consisted of any self-report of head
ment procedures.
injury, neurological insults, prenatal or birth
5. Adequate exclusion criteria.
complications, learning disabilities, psychiatric
problems, or substance abuse. Scores on the
7. WAIS-III Block Design subtest scores
Beck Depression Inventory-11-Spanish Ver-
are presented.
sion (Mean = 12.92, SD = 8.94) and the Beck
Anxiety Inventory-Spanish Version (Mean=
6.60, SD = 6.03) are also reported. Considerations regarding use of the study
Standard administration procedures were 1. All-male sample.
used. Participants were tested in Spanish. Se- 2. Small sample sizes for the combined age
lected subtests from the WAIS-III (Mexican and education groups.
STROOP TEST 129
[STROOP.25] Cohen, Brumm, Zawacki, Paul, Participants were recruited through the use
Sweet, and Rosenbaum, 2003 (Golden Version) of epidemiological sampling procedures. Ex-
(Table A6.30) clusion criteria included endarterectomy,
Twenty males who averaged 30.5 (10.7) years transient ischemic attack, cerebrovascular ac-
of age and 11.8 (3.3) years of education were cident, Parkinson's disease, or traumatic head
used as controls in a study of cognitive function injury with loss of consciousness and retro-
in domestic violence perpetrators at the Uni- grade amnesia; no information on psychiatric
versity of Massachusetts Medical Center; most disorders was collected. Fifteen individuals
of the controls were hospital workers. Mean were excluded on the basis of the above fac-
WAIS-R FSIQ was 100.7 (11.0), mean VIQ tors, as well as one outlier whose Stroop data
was 101.8 (10.6), and mean PIQ was 100.2 was 3 SDs from the mean of the sample,
(11.2); 16.5% admitted to at least some past which was clearly separated from the re-
drug use, and 9.3% admitted to prior alcohol- mainder of the scores.
related problems. Three reported previous The sample was divided into two age
head injury (two mi1d, one moderate), 10% groupings: 60--71 (n = 111) and 72-84 (n = 125).
had experienced learning difficulties, and Younger participants scored significantly bet-
12.5% admitted to childhood behavioral ter than older participants on all trials, with an
problems. effect size of > 0.25. In addition, education
Means and SDs for the interference trial and gender were significant predictors of
are provided. performance, with higher levels of education
and female gender associated with better
Study strengths performance. The majority of the sample
(72.5%) had<12 years of education. Data are
1. Information provided regarding gender,
stratified by age, gender, and education (< 12,
education, age, IQ, previous learning
12, >12 years), for a total of 12 separate
difficulties, childhood behavioral pro-
blems, head injury, substance use/abuse, subgroups with sizes ranging 2-56. Means and
SDs are reported. In addition, adjustments for
and geographic area.
education and gender to be applied to raw
2. Means and SDs reported for interfer-
scores are provided, as well as data on per-
ence trial.
3. Data for a sample of average IQ and centile scores for raw scores for each age
education. group.
Study strengths
Considerations regarding use of the study 1. Large sample sizes for the two age
1. Small sample size. groupings.
2. No apparent exclusion criteria; 2. Data are stratified by age, education, and
individuals with a history of substance gender, although individual cell sizes
use I abuse, learning and I or childhood ranged 2-56.
behavioral problems, and head injury 3. Good exclusion criteria for neurological
included. conditions, although psychiatric con-
3. Only males included. ditions or chronic medical illnesses
4. Data provided only for interference trial. (e.g., hypertension) were not used as
5. Data are not stratified by demographic exclusion criteria and could at least
factors. partially explain the poorer performance
observed in this sample relative to
[STROOP.26] Moering, Schinka, Mortimer, Caucasian individuals.
and Graves, 2004 (Golden Version) 4. Data provided for an African-American
(Table A6.31) population; however, most had a low
Stroop data were collected on 236 older level of education (although this was
African Americans, aged 60-84, living in pri- apparently representative of the commu-
vate residences in the Tampa, Florida, area. nities in which they lived).
5. Information on geographic atea and 4. Mean time in seconds and SDs are
recruitment strategies is provid~. reported.
6. Means and SDs are reported, as well as
percentile equivalents of raw scores and Considerations regarding use of the study
score adjustments for education and 1. No information on IQ.
gender. 2. Apparently adequate exclusion criteria,
although some controls were recruited
Considerations regarding use of the study from sheltered workshops.
1. Issues regarding exclusion criteria, 3. Undifferentiated age range.
lowered educational level, an' small
1
individual cell sizes. ; [STROOP.28] Sacks, Clark, Pols, and Geffen,
2. No data available on IQ level. . 1991 (Dodrill Version) (Table A6.33)
Stroop data were obtained on 12 male uni-
versity student volunteers in Australia, ranging
Dodrill Venion in age from 18 to 32 with a mean of 22.4 (5)
years, as a part of the development of five
[STROOP.27] Dodrill, 1978a (Dodrill Version)
alternate forms of the Dodrill Stroop. All
(Table A6.32)
participants had normal vision (20:20, as tested
Dodrill collected control data on 50 partici- with a standard Snellen wall chart) and no
pants in the state of Washington as a paft of his evidence of color blindness (assessed through
investigation of the cognitive corre~tes of Ishihara charts). Participants averaged 13.7
epilepsy. Thirty were male and 2() were (2.3) years of education. Mean abbreviated
female; and mean age and educatim~ level WAIS-R FSIQ, VIQ, and PIQ were 109.1
were 27.34 (8.41) years and 11.96 (2.01\) years, (9.5), with a range of 100-124; 108.4 (8.7),
respectively. Forty-nine were Caucasi$1, with with the range of 100-124; and 106.6 (7.1),
one listed as non-Caucasian. Nine w~re stu- with a range of97-120, respectively.
dents, six were housewives, 20 wdre un- The exact procedures used to develop the
employed, and 15 were employed. P~ipants alternate forms are specified. All participants
were recruited through employment f$:ilities, were administered all six forms of the test in
churches, a community college, a pub.c high 1 day with a 50-minute rest period between
school, a volunteer service agency, and a trials on each form. Order of completion of
semisheltered workshop. Participants under- the six forms was randomized. Participants
went a detailed neurological history, and those were halted at each error and instructed to
with diseases or other conditions affectfng the correct the mistake before proceeding. Means
nervous system were excluded. and SDs for time in seconds are reported for
The Dodrill version of the Stroop was each form. The forms were judged to be
administered. Means and SDs are r~rted equivalent, although a significant practice
for time in seconds to complete parts I ,md II. effect was still present between the first and
In addition, means and SDs are provi4ed for second test administrations. Sets of the six
part I+ part II, and part II- part I. Using a alternate forms are available from the test
cutoff of 93/94 seconds on part I, 7p% of authors.
controls were correctly classified. A cQtoff of
150/151 seconds for part II- part I res~ted in Study strengths
a 74% correct classification rate. 1. Data provided on six alternate forms and
practice effects.
Study strengths 2. Information reported on education,
1. Adequate sample size (n =50). gender, IQ, vision, age, and geographic
2. Information on age, education, ~nder, area.
occupation, geographic area, etlpricity, 3. Test stimuli development and adminis-
and recruitment procedures is pr¥ded. tration procedures are carefully de-
3. Test stimuli and procedures are s~fied. scribed.
STROOP TEST 131
4. Means and SDs for time in seconds are Study strengths

reported for each form. 1. Data are presented by narrow age
groupings.
Considerations regarding use of the study 2. Information is provided regarding mean
1. Small sample size (n = 12). age and mean educational level.
2. All-male sample. 3. Test stimuli and procedures are well
3. Data are collected in Australia; cultural described.
differences may render the data ques- 4. Means and SDs are reported for time
tionable for clinical interpretation in the and errors.
United States.
4. No exclusion criteria. Considerations regarding use of the study
1. Unclear exclusion criteria (participants
are described as "healthy").
Victoria Version
2. No information regarding IQ, gender,
[STROOP.291 Regard, 1981, cited in Spreen fluency in English, and geographic
and Strauss, 1991, 1998 (Victoria Version) recruitment area (assumed to be Canada).
(Table A6.34) 3. Small cell sample sizes.
Data were obtained on 40 right-handed young
adults of average intelligence. Average age Trenerry Version
was 26.7 (range 20--35). The Victoria Stroop
[STROOP.31] Anstey, Matters, Brown,
Test stimuli and procedures were employed.
and Lord, 2000 (Trenerry Version) (Table A6.36)
Means and SDs are reported for time and
errors. Stroop data were obtained on 369 retired
individuals residing in Anglican retirement
Study strengths villages in Australia and involved in a random-
1. Homogeneous age grouping. ized controlled trial of exercise on falls risk
2. Information regarding age, IQ, and and psychological well-being. There were 52
handedness is provided. males and 317 females, ranging in age from 62
3. Test stimuli and procedures are described. to 95, with a mean of 79.04 (6.59) years;
4. Means and SDs for time and errors are average years of education was 11.25 (2.79).
reported. Exclusion criteria included Parkinson's dis-
ease, stroke, or heart attack. Sixty-six percent
rated their health as good or very good, 18%
rated their health as excellent, and 16% rated
1. Fairly small sample size.
their health as fair or poor. The most common
2. No information regarding educational
health problems were arthritis (65%), cataract
level, gender, fluency in English, geo-
(53%), hypertension (50%), glaucoma or poor
graphic recruitment area (assumed to be
vision (38%), lung problems (19%), and dia-
Canada), or exclusion criteria.
betes (7%). Seventeen percent of the sample
had MMSE scores <24 (lowest score 17).
[STROOP.30] Spreen and Strauss, 1991 Percentile ranks for raw scores for the two
(Victoria Version) (Table A6.35) Stroop trials, stratified by four age groupings
These authors collected Stroop normative (62--69, 70-79, 80-89, and 90-95) and three
data on 86 healthy older participants aged education groupings (0-9, 10-12, 13+), are
50-94; average age was 68.5 (10.78) years. provided, as well as the means and SDs for
Mean years of education was 13.2 (3.1) years. the sample as a whole.
The Victoria Stroop Test stimuli and admin- Age and education, but not gender, were
istration procedures were used. Means and significantly related to Stroop scores. Perfor-
SDs are reported for time and errors for four mance on the Stroop Test in individuals older
age groupings: 50-59 (n = 19), 60-69 (n = 28), than 90 was particularly poor, and many could
70-79 (n = 24), and 80-94 (n = 15). not complete the task.
Study strengths five studies. Due to scarcity of the data for

1. Large overall sample size, although these conditions, they were not analyzed.
the sizes of individual cells are not re- A linear regression of the Stroop scores on
ported. age yielded an R2 of0.791, indicating that 79%
2. Stratification of data by age and educa- of the variance in scores is accounted for by
tion. the model. Based on this model, we estimated
3. Information regarding age, education, test scores for age intervals between 25 and
gender, health status, residential setting, 74 years. If predicted scores are needed for
and occupational status is provided. age ranges outside the reported age bound-
4. Test stimuli and procedures are reported. aries, with proper caution (see Chapter 3)
5. Percentiles for raw scores are provided they can be calculated using the regression
for 12 subgroupings, as well as overall equations included in the tables, which un-
means and SDs for the sample as a whole. derlie calculations of the predicted scores.
Linear regression of SDs for the Stroop
scores on age suggests that age does not
account for a significant amount of variability
1. Questionable adequacy of exclusion cri-
in SDs (R2 = 0.015). Though some increase in
teria (17% had MMSE<24).
variability with advancing age is expected, this
2. Participants age 62 or older are included.
trend was not present in the collected data.
Therefore, we suggest that the mean standard
4. Data obtained in Australia, which may
deviations for the aggregate sample be used
limit applicability in the United States.
across all age groups.
Means and SDs for the Word Reading.
Color Naming, and Interference conditions
for four studies (seven data points) that report
RESULTS OF THE META-ANALYSES
data for all three conditions are summarized
OF THE STROOP TEST DATA
in Table A6m.2.
(GOLDEN VERSION, INTERFERENCE
Examination of the effect of education on
CONDITION)
Stroop scores indicated that education did not
(See Appendix 6m)
contribute to the test scores beyond its as-
Data collected from the studies reviewed in sociation with age in the data available for
this chapter were examined. Only the data for analyses.
the Golden version had a sufficient aggregate Effects of IQ and gender on the test scores
sample to be included in the analyses. Data were not examined as data were not available
were combined in regression analyses in order for these analyses.
to describe the relationship between age and
test performance and to predict expected test
Strengths of the analyses
scores for different age groups. Effects of
1. Postestimation tests for parameter spe-
other demographic variables were explored in
cifications did not indicate problems
follow-up analyses. The general procedures
with normality or homoscedasticity.
for data selection and analysis are described
in Chapter 3. Detailed results of the meta-
analysis and predicted test scores across adult Limitations of the analyses
age groups are provided in Table A6m.l. 1. R2 of 0.791 is acceptable. However, this
After initial data editing for consistency and value indicates that only 79% of the
for outlying scores, six studies, which gener- variance in Stroop scores is accounted
ated 10 data points based on a total of for by the model.
490 participants, were included in the analyses 2. The number of studies available for the
for the Interference condition. Data for the analyses is small.
Word Reading and Color Naming conditions 3. It should be pointed out that the data-
included only eight datapoints collected from points available for the analyses are
STROOP TEST 133
distributed unevenly throughout the age 3. There is a paucity of data on the Comalli
continuum. The datapoints aggregate at version for participants less than 45 years
the younger and older extremes, with a of age.
notable lack of data in the middle part of
the age continuum (see scatterplot). Examination of the means across the
4. Review of the literature indicates that ef- Kaplan and Comalli studies suggests that there
fects of education, intellectual level, and is no difference in performance of controls on
gender on Stroop performance are equi- the two administration versions, raising the
vocal. Though we did not find an effect of possibility that these two normative data sets
education on test performance in the data can be used interchangeably. This would
available for review, we were unable to increase the total Kaplan/Comalli sample size
examine the effects of other demographic to 1,608 and remedy the lack of data on
variables due to a lack of data. younger participants in the Comalli version.
Future research is needed to determine
whether the Kaplan administration format
elicits a more pronounced color-interference
CONCLUSIONS
effect in clinical groups; it does not appear to
The Stroop has a lengthy history as an have this effect in normals, as observed here.
experimental measure in psychological studies One difficulty in interpreting Stroop scores
and more recently has been adapted for clin- has been how to parcel out the effect of slo-
ical neuropsychological use. However, the wed information processing, as reflected in
plethora of test stimuli and administration lowered scores on the first sections of the
formats has been confusing. Compilation of Stroop, from color-interference performance
the data sets suggests that the Golden version to obtain a more "pure" measure of executive
has the largest sample (n = 1,263), followed by dysfunction. Some authors have used differ-
the Kaplan (n = 981), Comalli (n = 627), ence scores (Demick & Harkins, 1997;
Trenerry (n = 525), Victoria (n = 126), and Dodrill, 1978a; Jensen, 1965), although this
Dodrill (n = 62) versions. However, within approach has been questioned (Koss et al.,
the data sets there are problems regarding 1984; Trenerry et al., 1989). Koss and collea-
representation of age groups: gues (1984) recommend an analysis of covar-
iance model, and Jensen and Rohwer (1966)
1. There does not appear to be any data on report 16 different methods for relating
the Victoria version for participants aged individual Stroop scores. Further research
36-49. is needed to determine if there is a more
2. There does not appear to be any data effective approach to Stroop interpretation
on the Dodrill version for participants than the typical independent analysis of
older than age 40. individual Stroop scores.
7
Auditory Consonant Trigrams
BRIEF HISTORY OF THE TEST formance on this measure may also be found in
frontal system dysfunction secondary to closed
The Auditory Consonant Trigram Test (ACT),
head injury (Stuss et al., 1985), discrete frontal
also referred to as the Brown-Peterson Con-
lobe lesions (Stuss et al., 1982), white-matter
sonant Trigram Memory Task or CCC (Stuss,
hyperintensities (Boone et al., 1992), Korsak-
1987), which is a variant of the Peterson and
off's syndrome (Cermak & Butters, 1972;
Peterson technique (Milner, 1970, 1972;
Leng & Parkin, 1989), presence of or risk for
Samuels et al., 1972), was originally developed
schizophrenia (Fleming et al., 1995; Rutsch-
as an experimental verbal memory procedure
mann et al., 1980), Parkinson's disease (Marie
(Brown, 1958; Peterson & Peterson, 1959).
et al., 1995), inattentive-type attention-deficit
ACT is sensitive to both deficits in short-term
hyperactivity disorder (Gansler et al., 1998),
auditory verbal memory (the ability to recall
myotonic dystrophy (Palmer et al., 1994),
rote verbal information over a distractor; Mil-
Klinefelter's syndrome (Boone et al., 2001),
ner, 1970, 1972; Samuels et al., 1972) and
and adults exposed to alcohol in utero (Connor
divided attention/working memory (amount
et al., 2001). The impaired test performance of
of information which can be processed simul-
both memory-dysfunctional patients and
taneously; Fleming et al., 1995; Marie et al.,
patients with frontal system defects may lie in a
1995; Stuss et al., 1985, 1987). Factor analysis
susceptibility to proactive interference; in the
has suggested that ACT loads primarily with
former group, the impairment may stem from
attentional (e.g., Digit Span) and verbal IQ
verbal encoding deficits, while in the latter
measures, and not with other executive tasks,
group it may be due to vulnerability to inter-
such as the Wisconsin Card Sorting Test,
fering stimuli that disrupt sustained attention
Stroop, and FAS (Boone et al., 1998).
(Stuss et al., 1982). For further information
Several studies have indicated that individ-
regarding ACT and its variants, please refer
uals with memory impairment associated with
to Lezak et al. (2004, pp. 416-418), and Spreen
left temporal lobe damage perform poorly on
and Strauss (1998).
this task (Giovagnoli & Avanzini, 1996; Milner,
1970, 1972; Samuels et al., 1972), although
there is some evidence that right temporal
lobectomy patients may also show lowered
Administration Procedures
scores (Samuels et al., 1980). In the absence of ACT involves the auditory presentation of three
a significant verbal memory deficit, poor per- consonant trigrams followed by a number. The
134
AUDITORY CONSONANT TRIGRAMS 135
patient is instructed to subtract 3s from the Schonfield et al., 1983). No other literature
number for several seconds, after which he or aside from the Boone (1999) publication could
she is asked to recall the letters. The exact be located on the impact of IQ on ACT scores.
administration instructions and test stimuli Bherer et al. (2001} and Anil et al. (2003) ob-
used by Dr. Stuss and colleagues at Ottawa served a significant contribution of educational
General Hospital are contained in Appendix level to ACT performance, in contrast to the
2c,d, and those used by Boone and colleagues findings of Stuss et al. (1987). However, the
at Harbor-UCLA Medical Center are con- findings of Boone (1999) suggest that when
tained in Appendix 2a,b. Some examiners use the effects of education and IQ are assessed
counting intervals of 3, 9, and 18 seconds simultaneously, only IQ is a significant pre-
(Boone et al., 1990, 1992; Boone, 1999; Cer- dictorof ACT performance. Stuss et al. (1987},
mak & Butters, 1972; Corsi, 1969, as reported like Boone (1999}, failed to detect a significant
in Milner, 1972; Samuels et al., 1972), while effect of gender on ACT performance.
other investigators have lengthened the sub- Of interest, there is evidence that some
traction times to 9, 18, and 36 seconds, to in- medical conditions previously thought to be
crease task difficulty and reduce ceiling effects cognitively benign may lead to decrements in
(Stuss et al., 1987). ACT performance. Specifically, the presence
of cerebrovascular risk factors (e.g., hyper-
tension, post-myocardial infarction, cardio-
Psychometric Properties
vascular disease, arrhythmias, coronary artery
Assessment of internal consistency of a Turk- bypass graft, angina, old myocardial infarc-
ish translation of ACT using 3-, 9-, and tion, white-matter hyperintensities) accounts
18-second delays revealed Cronbach's «at a for 3% of ACT test score variance (Boone,
reliable level (« = 0.8535). 1999), and once a threshold amount of total
white-matter hyperintensity volume is de-
tected on MRI (i.e., > 10 cm 2), significant
declines are detected in ACT performance
RELATIONSHIP BETWEEN ACT
(Boone et al., 1992). In fact, ACT may be one
PERFORMANCE, DEMOGRAPHIC
of the most sensitive cognitive tests to the
FACTORS, AND VASCULAR STATUS
presence of white-matter damage sustained
Relatively few studies are available on the through hyperintensities of probable vascular
impact of demographic factors and IQ on ACT origin (Boone et al., 1992) or white-matter
performance. and/ or frontal-limbic-reticular activating sys-
A recent examination of the unique contri- tem disruption secondary to acceleration-
bution of IQ and demographic factors to ACT deceleration closed head injury (Stuss et al.,
performance in a large sample of healthy, 1985).
older participants found that FSIQ accounted
for 17% of unique test score variance while age
accounted for a significant but very modest
METHOD FOR EVALUATING
amount (6%) of unique test score variance;
THE NORMATIVE REPORTS
gender and educational level did not contrib-
ute to test performance (Boone, 1999). Our review of the literature located three
The remaining literature on the relationship ACT normative reports published since 1987
between ACT performance and age has been (Anil et al., 2003; Stuss et al., 1987, 1988) as
equivocal, with some investigators failing to well as data from two studies examining the
detect an association (Bherer et al., 2001; impact of age, education, IQ, gender, and
Boone et al., 1990; Puckett & Lawson, 1989; medical illness on ACT performance (Boone
Stuss et al., 1987) and others documenting et al., 1990; Boone, 1999), which included
some deterioration with increasing age unique features such as large sample size
(Anil et al., 2003; Inman & Parkinson, 1983; (Boone, 1999) and reporting of a wide range
Parkin & Walter, 1991; Parkinson et al., 1985; of ACT scores (Boone et al., 1990).
To adequately evaluate the ACT normative Procedural Variables

reports, six key criterion variables were
deemed critical. The first four of these relate Description of the Administration
to subject variables, and the remaining two Format Used
relate to procedural variables. Minimal crite- Given that different test administration for-
ria for meeting the criterion variables were as mats involve differing lengths of distraction
follows. intervals, specific information regarding the
delays should be provided.
Subject Variables Data Reporting

Means and standard deviations, and prefer-
Sample Size
ably ranges, for total score out of 60 are im-
Fifty cases are considered a desirable sample portant. In addition, it is advantageous for
size. Although this criterion is somewhat data to be provided for each of the distraction
arbitrary, a large number of studies suggest intervals separately.
highly influenced by individual differences
population mean. SUMMARY OF THE STATUS
OF THE NORMS
Sample Composition Description In terms of subject variables, only one study
Given the evidence that ACT performance provides data by IQ level (Boone, 1999),
may be significantly impacted by medical although IQ data are reported in a second
status (e.g., vascular illness), information study (Boone et al., 1990).
regarding medical exclusion criteria is critical. Information on age, gender, education
In addition, as discussed previously, informa- level, geographic area, and recruitment pro-
tion should probably also be provided cedures is reported for all studies. In addition,
regarding educational level, gender, psychiat- medical, psychiatric, neurologic, and sub-
ric exclusion criteria, geographic region, stance abuse exclusion criteria are described
ethnicity, occupation, handedness, and re- and judged to be adequate for all studies.
cruitment procedures, even though there are Ethnic composition was indicated in two
as yet no data indicating that these factors studies (Anil et al., 2003; Boone et al., 1990).
influence test performance. Handedness data were provided only in the
investigations conducted by Stuss and col-
Reporting of Age leagues (Stuss et al., 1987, 1988). While all
studies exceeded a total sample size of 50, only
Given the equivocal and modest relationship one study reached the criterion of 50 partici-
between age and ACT performance, ACT
pants per individual grouping cell (Anil et al.,
normative data probably do not need to be 2003). In terms of procedural variables,
presented by age group intervals, but infor- information is available regarding the precise
mation on the ages of the normative samples administration formats for all studies. Means
should be provided. and SDs are reported for total score in all but
one study (Anil et al., 2003), and means and
IQ Group Intervals SDs for individual distractor delays are pro-
Given the evidence that IQ may account for vided in all but one study (Boone, 1999).
more unique test score variance than do Practice effects are investigated in the reports
demographic factors, information regarding by Stuss and colleagues (Stuss et al., 1987,
IQ level should be reported for each sub- 1988), and data on qualitative performance
group, and preferably normative data should variables (perseverations, errors in letter
be presented by IQ intervals. sequence) are provided in Boone et al. (1990).
Data are presented in ascending chrono- Test performance was not impacted by age,
logical order for two ACT versions separately: educational level, or gender. A practice effect
first for the 9-, 18-, and 36-second and then for the 9- and 18-second delays was observed
for the 3-, 9-, and 18-second delay version. The despite the alternate form.
text of study descriptions contains references ACT data are provided by six age groupings
to the corresponding tables identified by (16-19, 20-29, 30-39, 40-49, 50-59, and
number in Appendix 7. Table A7.1, the locator 60-69) for baseline testing, retesting, and the
table, summarizes information provided in the two testing sessions combined for the three
studies described in this chapter. 1 delay intervals separately. Data on gender dis-
tribution, handedness, mean age and SD, and
mean years of education, SD, range, and fre-
quency of"$high school" and ">high school"
SUMMARIES OF THE STUDIES for each age grouping are provided. In addition,
data are presented by gender and educational
Data for 9-, 18-, and 36-Second Delay level ($high school, >high school) separately.
Version
[ACT.l] Stuss, Stethem, and Poirier, 1987 Study strengths
(Tables A7.2 and A7.3) 1. Information regarding age, education,
ACT baseline and 1-week retest data were gender, and handedness for the total
collected on 60 participants in Canada, who sample and for individual age groupings
were recruited through employment agencies is provided.
and paid $10 for the two testing sessions. Par- 2. Adequate exclusion criteria.
ticipants ranged in age from 16 to 69, with a 3. Information on practice effects is
mean of 39.6 (2.62) years. Years of education reported.
ranged 8-20, with a mean of 14.5 (2.63). 4. Precise description of test administration
Thirty-three participants were male and 27 procedures.
were female. Forty-nine were right-handed. 5. Data presented by age groupings, gen-
None had a history of significant medical, der groupings, and education groupings.
neurological, or psychiatric disorder; substance 6. Data presented separately for each
abuse; or current psychotropic medication use. distraction interval.
Participants were tested in their native lan- 7. Information on geographic area and
guage (English or French). Each three- recruitment procedures is provided.
consonant combination was presented at a rate 8. Means and SDs for the test scores are
of one consonant per second followed by a reported.
three-digit number. Participants were in-
structed to count backward by 3s from the Considerations regarding use of the study
number for random delays of 9, 18, and 36 1. Small individual cell sizes (n = 10).
seconds and then to recall the trigram. Practice 2. Data collected in Canada, with some test
trials were employed until participants dem- administrations conducted in French;
onstrated understanding of the procedures. cultural and linguistic factors may limit
Five trials were conducted for each delay in- usefulness of data for clinical interpre-
terval, with intertrial delays of 2--5 seconds. tation in the United States.
The counting delays were extended from those 3. No information regarding IQ level.
employed by Cermak and Butters (1972), to
minimize ceiling effects. A total score of 15 [ACT.2] Stuss, Stethem, and Pelchat, 1988
was possible for each of the three delays. An {Tables A7.4 and A7.5)
alternate form was employed on retesting. In this publication, the authors supplement
'Norms for children and adolescents are available in the data reported in 1987 by expanding the
Baron (2004) and Spreen and Strauss (1998). number of participants, increasing cell sizes
by collapsing the data from six to three age disease, such as stroke, Parkinson's disease, or
groupings (16-29, 30-49, and 50--69), and seizure disorder. Also excluded were individ-
presenting the combined data from the two uals with laboratory findings showing serious
testing sessions in box plots, which has the metabolic abnormalities (e.g., low sodium le-
advantage of visual display of data variability. vel, elevated glucose level, or thyroid or liver
Each of three age groupings contained base- function abnormalities). Eighteen percent of
line and 1-week retest data on 30 participants, the original sample of 74 were eventually ex-
none of whom had a positive psychiatric or cluded due to the presence of previously un-
neurologic history. Data are presented on identified strokes or other significant lesions
gender distribution, handedness, mean age documented on MRI (n = 9), metabolic ab-
and SD, and mean years of education, SD, normalities or undiagnosed medical illness
and range for each age group separately. (n =2), or evidence from laboratory studies
and EEG findings of alcohol abuse and sub-
Study strengths stance intoxication (n = 2). The final sample
1. Large overall sample size (n = 90). (n = 61) included 25 men and 36 women
2. Information on practice effects. grouped by three age decades: 50-59 (n = 25),
3. Adequate exclusion criteria. 60--69 (n = 21), and 70-79 (n = 15). All but
4. Information on gender, educational 10 participants were white; four were African
level, and handedness for each age American, three were Asian, and three were
grouping is presented. Hispanic. Mean educational level was 14.34
5. Presentation of test score variability via (2.63) years, and mean WAIS-R FSIQ was
box plots. 113.79 (13.51).
6. Test administration format is the same as No significant effect of age on ACT per-
in Stuss et al. (1987). formance was documented in comparisons of
7. Means and SDs for the test scores are the three age groups.
reported. Means and SDs are presented for ACT total
score as well as for 3-, 9-, and 18-second delay
Considerations regarding use of the study for each age group separately. Total possible
1. Same as above: although the sample has was 60 (15 points for each delay interval as
been increased by 50%, the three age well as 15 points for a five-trial, 0-delay
groupings still have only 30 participants condition). Means and SDs are also reported
each. for number of perseverations and altered se-
quences. Perseveration was defined as the
reporting of an incorrect letter which was
Data for 3-, 9-, and 18-Second used as an answer on the preceding trial; a
Delay Version total of 57 perseverations were possible.
Altered sequence referred to reporting of
[ACT.3] Boone, Miller, Lesser, Hill, and D'Eiia, correct letters but in the wrong position
1990 (Table A7.6) within the trigram; a total of 20 altered
Data were collected on 61 middle-aged and sequences were possible.
older individuals ranging in age from 50 to 79,
recruited as controls in southern California
through newspaper ads, flyers, and personal Study strengths
contacts as a part of ongoing research on late- 1. Information on IQ level, years of edu-
life depression and psychosis. Participants had cation, gender distribution, geographic
no history of psychotic, major affective, or area, recruitment procedures, ethnicity,
alcohol or other drug dependence disorder and fluency in English is presented.
and spoke English fluently. (A handful of par- 2. Data are reported in terms of total score
ticipants spoke English as a second language.) but also by individual delay intervals;
Participants were excluded if there was a information is also provided on perse-
history of physical findings of neurological verations and altered sequences.
3. Comprehensive medical and psychiatric 3. Data are presented in terms of total

exclusion criteria, including MRI brain score only, with no information regard-
scans, on all participants. ing distraction intervals.
4. Test administration format is described. 4. Nearly high average IQ level.
reported. [ACT.5] Boone, 1999 (Table A7.7)
6. Data stratified by age. The author obtained ACT data on 155 middle-
aged or older individuals (ranging in age from
Considerations regarding use of the study 45 to 84 and recruited as described above for
1. Fairly small individual cell sizes (n = Boone et al., 1990; data from the 1990 study
15-25). are included in the 1999 publication). The
2. High average IQ level. mean age of the sample was 63.07 (9.29)
years, mean educational level was 14.57 (2.55)
[ACT.4] Boone, Ananth, Philpott, Kaur, and years, and mean FSIQ was 115.41 (14.11); 53
Djenderedjian, 1991 were male and 102 were female. Medical and
ACT data were obtained on 16 controls (nine psychiatric exclusion criteria are listed above,
women, seven men) as part of an investiga- with the exception that participants with sig-
tion of the neuropsychologicaJ characteristics nificant white-matter hyperintensities docu-
of obsessive-compulsive disorder (OCD). Nine mented on MRI were retained in the sample.
of the participants were siblings of OCD pa- All participants considered themselves healthy,
tients, while the remaining participants were although 51 had some evidence of vascular
recruited through newspaper ads and friends illness (defined as cardiovascular disease and/
of OCD patients in the southern California or significant white-matter hyperintensities
area. Mean age was 35.8 ( 13.7) years, and mean on MRI) based on self-report or evidence on
educational level was 15.2 (2.8) years. Mean examination of at least one of the following:
FSIQ, VIQ, and PIQ were 109.1 (10.9), 106.3 current or past history of hypertension
(13.0), and 111.8 (10.8), respectively. MedicaJ (n = 39), arrhythmia (n = 8), large area of
exclusion criteria were history of alcohol or white-matter hyperintensity on MRI (e.g.,
drug abuse, head injury, seizure disorder, ce- > 10 cm2 ; n = 7), coronary artery bypass graft
rebral vascular disease or stroke, psychosur- (n = 3), angina (n = 2), and old myocardial
gery, current or past psychiatric condition, or infarction (n = 1). Twenty-four participants
any renal, hepatic, or pulmonary disease. were currently on cardiac and/ or antihyper-
Mean ACT total score was 44.3, with an SO tensive medications.
of7.5. A stepwise regression analysis revealed that
Data are not reproduced in this book FSIQ, age, and vascular status were significant
contributors to total ACT score, accounting
Study strengths for 17%, 6%, and 3% of test score variance,
1. Information regarding age, education, respectively; educational level and gender did
gender distribution, IQ, geographic area, not account for a significant amount of unique
and recruitment procedures is reported. test score variance. ACT normative data are
2. Comprehensive psychiatric and medical presented for total ACT score stratified by IQ
exclusion criteria. and age ( < 65 and ~65; average IQ, high av-
3. Though not stated, test administration erage IQ, and superior IQ).
procedures are the same as those in
Boone et al. (1990). Study strengths
4. Means and SDs for the test scores are 1. Large overall sample size.
reported. 2. Presentation of data by IQ and age
groupings.
Considerations regarding use of the study 3. Comprehensive medical and psychiatric
1. Small sample size. exclusion criteria, including MRI brain
2. Data are not stratified by age or IQ level. scans, on all participants.
4. Information regarding educational level, 3. Data stratified by both age and educa-
gender, geographic area, recruitment pro- tion.
cedures, and fluency in English. 4. Adequate exclusion criteria.
5. Though not stated, test administration 5. Means and SDs for the test scores are
procedures are the same as dtose in reported.
Boone et al. (1990).
reported. ~
1. Test was translated into Turkish and data
were collected in Turkey, rendering
Considerations regarding use of the s#!.uly
use problematic for English-speaking
1. Individual IQ-by-age groupings have
patients.
sample sizes ranging 16-37.
2. Test administration was not standard
2. Data are presented in terms of total
(subjects counted backward by 1s rather
score rather than separately for each
than 3s).
distraction interval.
[ACT.6l Anil, Kivircik, Batur, Kabakci, ICitis,
Giiven, Basar, Turgut, and Arkar, 2003 :
(Table A7.8) CONCLUSIONS
ACT data were collected on 236 individuals in ACT has been underutilized as a clinical mea-
Turkey, who were recruited from hospttaJ staff sure of executive dysfunction despite evidence
or through personal contacts. Exclusfon cri- that it may be particularly sensitive to white-
terion included neurological or p~hiatric matter disturbance. Given emerging interest in
conditions. The sample was strati6tJ into working-memory paradigms, the consonant
three age groups (16-25, 26-45, andj46-65) trigrams task may experience an increase in
and three education groups (8-10, 11-il-4, and popularity. Most working-memory paradigms
>14 years). The youngest age group ~nsisted have been used in experimental studies, and
of 40 males and 22 females, who aver4ged 22 normative data are typically not available. The
(2. 7) years of age. The middle age grqup was fact that a normative data pool of upward of500
composed of 70 males and 55 femal.s, who participants has been collected for ACT may
averaged 34.1 (5.9) years. The oldest group make it an attractive working-memory proce-
included 28 males and 21 females, w}1o aver- dure for clinical practice.
aged 53.8 (4. 7) years. In addition, the fact that the ACT task does
The ACT was translated with the consulta- not involve a timed response makes it a
tion of a linguist, and consonants from the desirable executive measure in that test per-
Turkish alphabet showing similar phone~c char- formance is not confounded by declines in
acteristics to the original ACT were e~loyed. mental speed. For tasks such as Trails B,
Participants were instructed to count b~kward Stroop Color Interference, and word and
by 1s rather than the standard 3s. M~s and design generation, poor scores may reflect
SDs are reported for each delay intervaj. slowing in information-processing speed
Analyses revealed no gender Fffects, rather than executive dysfunction per se.
although better performance was asspciated Future research is needed to determine
with younger age and more years of ed"Qcation. which delay intervals (i.e., 3, 9, and 18 seconds
vs. 9, 18, and 36 seconds) are most sensitive
Study strengths and appropriate for clinical use. Also, norma-
1. Large overall sample size, although the tive data need to be obtained on populations
sizes of the nine subcells ate not with less than average IQs.2
reported. ~
2. Information provided for age, ~ender,
education, geographic area, l~age, 2 Meta-analyses were not performed on ACI' due to a lack
and recruitment procedures. 1 of sufficient data.

8
Paced Auditory Serial Addition Test
BRIEF HISTORY OF THE TEST at 1.2 seconds. The task of the participant is
identical for each trial, and thus, a total of
Sampson (1956) originally developed an audi- 60 correct responses per trial is possible. A
tory and a visual version of the Paced Serial practice trial of 10 digits presented at the rate
Addition Test. In a landmark study, Gronwall of 2.4 seconds precedes the four test-trials.
and Sampson (1974) used the auditory ver- Functional brain imaging studies have sug-
sion, the Paced Auditory Serial Addition Test gested that PASAT performance is associated
(PASAT), to assess information-processing with activation of right anterior and left pos-
speed and working memory in post-concussive terior cingulate, consistent with an emerging
patients. This version was further researched body of literature relating cingulate function to
and made popular by Gronwall and col- attentional mechanisms (Deary et al., 1994).
leagues (Gronwall & Wrightson, 1974; Gron- The PASAT was initially devised as a mea-
wall, 1977a). It is now believed that the PASAT sure to detect cognitive deficits in postconcus-
also measures other cognitive abilities, such as sive individuals. Several studies have shown
sustained and divided attention (Lezak. 1995; that patients with head trauma perform signifi-
Lezak et al., 2004). cantly worse than their normal control coun-
In the original version of the PASAT, a terparts on the PASAT (Bate et al., 2001;
random series of 61 digits (1-9) are presented Brooks et al., 1999; Cicerone, 1997; Gronwall
on audiotape and the participant is to add the and Sampson, 1974; Maddocks & Saling,
last digit presented to the preceding digit and 1996; Ponsford & Kinsella, 1992; Stuss, et al.,
verbalize the answer. For example, if the 1989; Tiersky et al., 1998). Cicerone and Azulay
digits 1 and 2 are presented, the participant's (2002) documented the PASAT to be among
correct response would be 3 (i.e., 1 + 2), and if one of the most sensitive neuropsychological
the digit 4 is presented next, the participant's tests for detecting impairment in patients with
correct response would be 6 (i.e., 2 + 4) and post-concussion syndrome, but Maddocks and
so on. Each trial has the same random pre- Sailing (1996), using only the 2.4-second pre-
sentation of the 61 digits; however, the pace at sentation of the PASAT, did not find the same
which the digits are presented differs for the results. The original studies with the PASAT
four trials. In trial 1, the digits are presented also found it to be a sensitive measure of
at the rate of 2.4 seconds, in trial 2 at 2.0 recovery rate and capability to return to work
seconds, in trial3 at 1.6 seconds, and in trial 4 (Gronwall & Wrightson, 1974; Gronwall, 1977a).
141
However, of concern, several subsequent stud- PASAT and magnetic resonance imaging, Sny-
ies have been unsuccessful in finding a rela- der and Cappelleri (2001) found that PASAT
tionship between PASAT scores and severity scores correlated with the total area of sclerotic
of head injury (Levin et al., 1982; ~erman brain lesions in MS patients. This correlation
et al., 1997; Stuss et al., 1989). Fos et al. (2000) was not observed when the original PASAT
found that both the auditory and ~ visual scoring method was used. It should be noted
versions of the Paced Serial Addition Test that studies have found the PASAT to be in-
significantly correlated with other tesfs of at- creasingly difficult for MS patients, and as a
tention but that neither version of this test result a number of them refuse to perform the
differentiated patients with mild traumatic task (Aupperle et al., 2002).
brain injury from normal controls in a. college Additional factors have also been shown to
population. In an interesting study }t Chan affect PASAT performance. Studies have dem-
(2001), there were no differences in TASAT onstrated a reduction in PASAT scores during
scores of those postconcussive patieJts who pain (Sjogren et al., 2000), with sleep disrup-
were considered "low symptom ref<>rters" tion (Martin et al., 1996), in solvent exposure
and those who were considered "hig~ symp- (Rasmussen et al., 1993), during hypoglycemia
tom reporters." in patients with diabetes (Gold et al., 1995), in
More recently, the PASAT has been used to HIV-positive individuals (Honn et al., 1999),
study cognitive functioning in patietlS with in individuals with schizotypal personality dis-
multiple sclerosis (MS). In fact, modified ver- order (Mitropoulou et al., 2002), in individuals
sions of the PASAT, using 3- and 2tsecond with Attention-Deficit Disorder (Katz et al.,
pacing in digit presentation, are includeP in the 1998), and in cannabis addicts (Elwan et al.,
Brief Repeatable Battery of Neuropsychological 1997). In addition, a negative effect of smok-
Tests developed by the National ~ultiple ing on PASAT performance has been reported
Sclerosis Society to be used as a scree$g tool but only in poorly educated males (Elwan
for MS (Rao et al., 1990). In a study lw Sha- et al., 1997).
waryn et al. (2002), the PASAT predictep men- Further details about Gronwall's (1977b)
tal and emotional responses of MS pa~nts on version of the PASAT and verbatim instruc-
a quality-of-life questionnaire. Johnson et al. tions can be obtained from the PASAT admin-
(1996) found that patients with MS performed istration manual and test kit (see Appendix 1
poorly on both the PASAT and the: Paced for ordering information) or Spreen and Strauss
Visual Serial Addition Test (a visual a¥og to (1998, pp. 243-251; see also Lezak et al.,
the PASAT), while patients with Chrotrlc Fa- 2004).
tigue Syndrome displayed difficulty o~y with
the PASAT. The authors postulate that ~eficits
Modifications and Alternate Formats
on both of these tasks by MS patien;s may
of the PASAT
suggest impairment of central executite sys-
tem, a view that is shared by D'Esposith et al. While the original version is the most com-
(1996). Kujala et al. (1995) reported intpaired monly administered format, modifications to
PASAT scores for a group of mildly deterio- the PASAT have been made. Several of these
rated MS patients but intact scores for~ non- modified versions and alternate formats are
deteriorated MS group. Solari et al. ;(1995) presented below.
found that the PASAT was one of two neu-
ropsychological tests that best discri~J?inated Levin Version
between MS and controls. Fisk and ~hibald Levin et al. (1987) developed a version in
(2001) used a different scoring techniq+e (the which only 50 digits (rather than 61 digits) are
"dyad" method of counting two consr:utive presented in different random order (as op-
right responses as one correct point) apd ob- posed to the same random order) for each trial
served that controls outperformed MS J.lltients using the same 2.4-, 2.0-, 1.6-, and 1.2-second
on only the first two out of four prese,tation interval presentation. This version minimizes
trials. Using the dyad scoring method !of the the practice effects observed with Gronwall's
PACED AUDITORY SERIAL ADDITION TEST 143
original version, as demonstrated by Stuss An additional PASAT administration format

et al. (1987). includes giving only one or two trials of the
original PASAT at select pacing rates (e.g., 2.4
PASAT-200, PASAT-100, and PASAT-50 or 2.0 seconds).
Shortened versions of the PASAT were also
developed by Diehr et al. (1998) and further
Psychometric Properties of the Test
modified by Diehr et al. (2003). The Diehr
et al. (1998) PASAT, also referred to as the Adequate reliability and validity have been
PASAT-200, is very similar to Levin et al.'s reported for the original PASAT. Studies have
(1987) version in that it consists of the pre- cited split-half reliability of greater than 0.90
sentation of 50 single digits (except for the (Egan, 1988), suggesting high internal consis-
number 7) in random order at four different tency, and test-retest reliability values of0.93-
pacing intervals. However, the pacing inter- 0.97 (McCaffrey et al., 1995). O'Donnell et al.
vals are 3.0, 2.4, 2.0, and 1.6 seconds per digit (1994) reported adequate construct validity of
instead of 2.4, 2.0, 1.6, and 1.2 seconds. Dif- the PASAT, demonstrating relatively strong
ferent random presentation of the digits is correlations with other tests of attention, such
used for each trial. Diehr et al. (2003) short- as Visual Search and Attention Task (r=0.55)
ened the PASAT-200 by providing normative and TrailMaking Test Part B (r = 0.58). Mod-
data on trial 1 (3-second pacing trial) only, erate correlations between PASAT and other
referred to as the PASAT-50, and trials 1 and tests of concentration, information processing,
2 combined (3- and 2.4-second pacing trials), and working memory have also been noted by
referred to as the PASAT-100. Crawford et al. (1998b), Deary et al. (1991),
Gronwall & Wrightson (1981), and Larrabee
Computerized Versions of the PASAT and Curtiss (1995). However, some authors
Holdwickand Wingenfeld (1999) and Wingen- caution that the PASAT correlates not only
feld et al. (1999) created a computerized ver- with tests of attention but also with tests that
sion of Gronwall's (1977a) original PASAT, in measure mathematical skills and overall intel-
which the auditory stimuli are presented via lectual ability (Chronicle & McGregor, 1998;
external speakers and responses are recorded Sherman et al., 1997). Sherman and colleagues
by an external microphone. However, the com- (1997) voiced concern that "PASAT perfor-
puterized administration does not record mance depends on mathematical ability,
a response as correct if it occurs after pre- at least as much as on attentional skills" and
sentation of the subsequent stimulus, while recommend that "the PASAT should not be
the traditional administration format of the interpreted as a measure of attention when
PASAT has typically given credit for correct mathematical skills are poor" (p. 43).
"late" responses. For further information on the effect of re-
Tombaugh (1999) and Royan et al. (2004) peated administration and psychometric prop-
recently developed an interesting computer erties of the PASAT, see Franzen (2000) and
version of the PASAT, referred to as the McCaffrey et al. (2000).
Adjusting-PSAT. This version measures speed
of information processing and working mem-
ory by assessing temporal thresholds versus
RELATIONSHIP BETWEEN PASAT
the traditional method of counting number of
PERFORMANCE AND DEM()(jRAPHIC
correct responses. In this version, stimuli are
FACTORS
presented through either auditory or visual
modalities, and the duration of the interval Age effects have been frequently reported for
between number presentation depends on the the PASAT. Stuss et al. (1988) found declining
correctness of the response. In other words, PASAT scores as a function of age grouping.
correct responses lead to decreased time be- Using the Levin version of the PASAT, Brittain
tween intervals and incorrect responses lead et al. (1991) also demonstrated an age-related
to increased time between intervals. decline in performance. Roman et al. (1991)
found that individuals in the 6th and 7th de- et al., 1997). Kanter (1984) observed a strong
cades of life performed significantly worse correlation between PASAT responses and
than two younger groups on all PASAT trials, speeded nonverbal intelligence tasks, and sig-
and Wiens et al. (1997) reported a steady age- nificant relationships between PASAT scores
associated decline in PASAT performance for and the Shipley tests of intelligence have also
individuals in their twenties to late forties. been reported (Brittain et al., 1991; Egan,
Further, Diehr et al. (1998) documented a 1988). Deary et al. (1991) found a signifi-
decline with age in a modified version of the cant correlation between PASAT scores and
PASAT for individuals in three age groups WAIS-R IQ in a group of diabetic patients, but
(20-34, 35-49, 50-68). A few studies, how- on closer examination, the relationship was
ever, have shown weak or no age effects. only significant between the PASAT and
Boringa et al. (2001) reported declining scores the freedom from distractibility index of the
as a function of age on the 2-second trial of WAIS- R. In terms of basic math skills and the
the PASAT but not the 3-second trial. This PASAT, Gronwall and Sampson (1974) found
would suggest that age differences emerge a weak correlation, but others have shown a
only during the more difficult portion of the stronger relationship (Sherman et al., 1997).
task. Epperson and Cripe (1985) found no Gender differences have not been found in
significant age effects for a sample of individ- most studies using the PASAT (Boringa et al.,
uals aged 18-49, and Elwan et al. (1997) found 2001; Diehr et al., 1998, 2003; Roman et al.,
no significant correlation between PASAT 1991; Stuss et al., 1987). Some studies have
scores and age in an Egyptian sample ranging found statistically significant differences in per-
from 20 to > 60 years of age. Finally, in one formance in favor of males, but the differences
study (using the 2-second delay in number were of little clinical or practical importance
presentation), PASAT scores for older indi- (Brittain et al., 1991; Wiens et al., 1997). El-
viduals (mean age= 52) were actually higher wan and colleagues (1996, 1997), administer-
than for young college students (mean age= ing the PASAT to a sample of Egyptians,
25) (Ward, 1997). found better performance in males but par-
A relatively consistent relationship between ticularly in subjects age 60 or above. Interest-
PASAT performance and education has been ingly, Wiens et al. (1997) noted that Hispanic,
reported. Stuss et al. (1987) found that indi- Asian, and Native American males in their
viduals with less than a high school education sample appeared to perform "slightly" better
performed poorer on the PASAT than those than their female counterparts, while the op-
with a college education or higher. Wiens et al. posite was true for African-American and
(1997) found education effects for trial 1 of Caucasian participants. However, cell sample
the PASAT but not the other trials. Diehr et al. sizes were too small to confirm these obser-
(1998) reported a steady increase in PASAT vations with statistical analyses.
scores as a function of higher education at- Only a few studies have examined the rela-
tainment. In contrast, Brittain et al. (1991) tionship between race/ethnicity and PASAT
and Elwan et al. (1996, 1997) could detect no performance. Brittain et al. (1991) reported a
significant relationship between education and complex interaction effect between age, IQ,
PASAT performance. and race. They found that in older "minority"
The results are mixed in terms of the rela- women, PASAT scores across all trials were
tionship between general intelligence and the associated with IQ scores. The specific racial
PASAT. Gronwall and colleagues (Gronwall & breakdown of their minority subjects was not
Sampson, 1974; Gronwall & Wrightson, 1981) provided, and this interaction effect was not
and others (Johnson et al., 1988; Roman et al., reported for their Caucasian group. Wiens et al.
1991) report weak or no correlation between (1997) found no statistically significant dif-
intelligence and PASAT, while others have ferences between African-American, Hispanic,
shown a moderate relationship between these Native-American, Asian, and Caucasian partici-
two factors (Crawford et al., 1998b; Deary pants. Diehr et al. (1998), however, reported
et al., 1991; Egan, 1988; Kanter, 1984; Wiens significantly better PASAT performance by
Caucasians relative to African Americans across regarding educational level should be re-
three age groups (20-34, 35-49, 50-68). Addi- ported for each subgroup, and preferably
tionally, using T-score conversions, Diehr et al.'s normative data should be presented by edu-
distribution of the PASAT scores of a small cational levels.
sample of Hispanic individuals more closely
resembled that of the African Americans than Reporting of Intellectual Levels
the Caucasians. Given the probable association between
PASAT performance and IQ, information re-
garding intellectual level should be reported
for each subgroup, and preferably normative
METHOD FOR EVALUATING THE
data should be presented by IQ levels.
NORMATIVE REPORTS
To adequately evaluate the PASAT normative
reports, seven key criterion variables were Procedural Variables
deemed critical. The first five of these relate to Description of Administration Procedures
subject variables and the two remaining di-
mensions refer to procedural issues. Due to variability in administration procedures,
Minimal requirements for meeting the cri- a detailed description of the procedures, in-
terion variables were as follows. cluding identification of the version of the test
administered and number of trials (with re-
ported pacing of digit presentation), is desir-
Subject Variables able. This would allow one to select the most
appropriate norms or to make corrections in
Sample Size interpretation of the data.
size. Although this criterion is somewhat ar- Data Reporting
bitrary, a large number of studies suggest that Group means and standard deviations for the
data based on small sample sizes are highly number of correct responses for each pacing
influenced by individual differences and do condition should be presented at minimum.
not provide a reliable estimate of the popu-
lation mean.
Sample Composition Description SUMMARY OF THE STATUS

Information regarding medical and psychiatric OF THE NORMS
exclusion criteria is important. It is unclear if Information presented in the studies report-
gender, geographic recruitment region, socio- ing data for the PASAT differs across studies.
economic status, occupation, ethnicity, or re- Some of these differences will be summarized
cruitment procedures are relevant. Until this below.
is determined, it is best that this information Of the studies reviewed below, nine were
be provided. essentially designed to provide normative in-
formation (Boringa et al., 2001; Brittain et al.,
Age Group Interval 1991; Diehr et al., 1998, 2003; Roman et al.,
This criterion refers to grouping of the data 1991; Stuss et al., 1987, 1988; Wiens et al., 1997;
into limited age intervals. This requirement is Wingenfeld et al., 1999). Data for "normal"
especially relevant for this test since a strong control groups from clinical comparison stud-
effect of age on PASAT performance has been ies are also included in this chapter.
demonstrated in the literature. Various test formats of the PASAT are used,
with several studies devoted to modifying test
Reporting of Education Levels versions or scoring methods. The variations in
Given the strong association between educa- testing procedure and format include the num-
tion and PASAT performance, information ber of digits used (e.g., 61 or 50), the same vs.
different random order of the digit presenta- retested with the PASAT. The retesting was
tion across trials, the number of trials admin- approximately 1 week later for head-injured
istered, and the pace at which the digits are patients; it can be assumed that it was the
presented (e.g. 3.0-, 2.4-, 2.0-, 1.6-, and/or same time delay for the controls, but there is
1.2-second pacing). no specific mention of this. There is no addi-
Among all of the clinical studies available in tional information regarding age, gender, or
the literature, we selected for review those education for this sample. No other exclusion
that used well-defined samples; presented criteria are reported.
means and SDs for more than one presenta- The 61-digit version of the PASAT was
tion condition (e.g., 2.4-second pace per digit); presented at four different pacing rates (2.4,
provided adequate description of the test ver- 2.0, 1.6, and 1.2 seconds).
sion, procedures, and format; and provided
descriptive statistics for sample demographics, Study strengths
such as age and education. In the studies re- 1. Adequate sample size.
viewed below, the test scores represent the 2. Test administration procedures are well
number of correct responses for each pacing specified.
rate or the total scores across all trials, unless 3. Means and SDs for the test scores are
indicated otherwise. reported.
Summaries of the studies are presented in
ascending chronological order for each version Considerations regarding use of the study
of the test separately. Studies using Gronwall's 1. The sample composition is not well de-
administration procedure are presented first, scribed in terms of age, education, gen-
followed by those using Levin's version, con- der, IQ, and recruitment procedures.
cluding with the PASAT-50, PASAT-100, and 2. The age range of the group is quite large,
PASAT-200 versions. and the majority of the participants are
The text of study descriptions contains between the ages 17-25 years.
references to the corresponding tables iden- 3. No exclusion criteria are provided, and
tified by number in Appendix 8. Table A8.1, the non-head-injured "accident" cases are
the locator table, summarizes information not well described.
provided in the studies described in this 4. The test-retest time frame for the nor-
chapter. 1 mal controls is not provided (but the
head-injured patients were tested 1
week apart).
5. The data were obtained on New Zea-
SUMMARIES OF THE STUDIES
landers, which may limit their usefulness
Gronwall's Administration Version for clinical interpretation in the United
States.
[PASAT.1] Gronwall, 1977a (Gronwall Version)
(Table A8.2) [PASAT.2] Stuss, Stethem, and Poirier, 1987
This is one of the first studies to use the (Gronwall Version) (Table A8.3)
PASAT in order to assess cognitive function- The authors examined age-related differences
ing in brain-damaged patients. A sample of 60 in performance on three neuropsychological
"normal" participants in New Zealand aged tests, one of which was the PASAT. The au-
14-55 years (with the majority aged 17-25), thors recruited 60 participants from Ottawa,
consisting of 10 non-head-injured accident Canada, through personal contacts or various
cases, 10 naval "ratings," and 40 first- agencies (e.g., Seniors Employment Bureau,
year university students, served as controls. Youth Employment Agency). Participants were
All subjects were initially tested and then grouped by six decades of life (16-19, 20-29,
30--39, 40-49, 50-59, 60-69). Information re-
'Nonns for children and adolescents are available in garding handedness, years of education, and
Baron (2004) and Spreen and Strauss (1998). ratio of males to females is provided for each
age group. None had a history of neurological and educational level (~high school vs.
or psychiatric illness. Educational levels of >high school). Given the significant age
males (14.36) and females (14.55) were ap- effect, these tables have not been re-
proximately the same, but significant differ- produced in this chapter but can be
ences were found between educational levels found in the original source.
of participants in the different age groups, with
the 50-59 group having the lowest educational [PASAT.l] Sluss, Stethem, and Pelchat, 1988
level. (Gronwall Version) (Table A8.4)
The original Gronwall four-trial version of This study builds on the previous normative
the PASAT was used. It should be noted that study by Stuss et al. (1987) by collapsing the
the authors report using 60 digits but also age groups (i.e., creating larger age ranges per
state that 60 correct responses are possible. group}, thus increasing the number of partic-
Thus, it is believed that the original 61-digit ipants per cell. In the current study, there
version was used. Participants were tested at were three age groups. For the 1~29 age
two different intervals, separated by 1 week. group, there were 16 males and 14 females,
The test was administered in the participants' with an average age of 22.43 (2.67) and edu-
native language of French or English. cation range of 11-18 years (mean= 14.1,
SD = 1.34); for the 30-49 group, there were
Study strengths 14 males and 16 females, with an average age
1. The sample composition is well described of 40.63 (2.97) and education range of 5-20
in terms of age, education, gender, geo- years (mean= 14.9, SD = 3.95); and for the
graphic area, and recruitment procedures. 50-69 group, there were 14 males and 16 fe-
2. The data are stratified by six age group- males, with an average age of 61.77 (3.0) and
ings. education range of ~18 years (mean= 13.2,
3. Adequate exclusion criteria. SD=2.38). See the above study (PASAT.2)
4. Test administration procedures are spec- for additional participant characteristics and
ified. recruitment procedures.
reported. Study strengths
1. The sample composition is well de-
Considerations regarding use of the study scribed in a previous study (Stuss et al.,
1. Overall sample is adequate, but individ- 1987) in terms of age, education, gender,
ual cells are very small. geographic area, and recruitment pro-
2. Educational levels are not equal across cedures.
the different age groups, and some of the 2. The data are stratified by three age
groups are highly educated. groupings.
3. The data were obtained on Canadian 3. AdeC(luate exclusion criteria are de-
subjects, sometimes in French, which scribed in a previous study (Stuss et al.,
may limit their usefulness for clinical 1987).
interpretation in the United States. 4. Test administration procedures are de-
scribed in Stuss et al. (1987).
Other comments 5. Means and SDs for the test scores are
1. Individuals in the 50-59 age group had reported.
the lowest educational level and the
lowest PASAT scores relative to the Considerations regarding use of the study
other age groups. Their PASAT scores 1. Need to access Stuss et al. (1987) study
were significantly lower than even the in order to learn about the sample re-
oldest age group (60-69). cruitment and testing procedures.
2. The authors present another table that 2. Mean educational levels for some of
collapses PASAT scores across age the age groups are relatively high;
groups, stratifying the data by gender the 1~19 and 50-59 groups have
substantially less education than the [PASAT.S] Stuss, Stethem, Hugenholtz,

other age groups. and Richard, 1989 (Gronwall Version)
3. Overall sample size is adequate, hut in- (Table A8.6)
dividual cells are small. The authors compared the performance of two
4. The data were obtained on Canadian groups of head-injured patients to controls on
subjects, sometimes in French. which three neuropsychological tests. Twenty-six con-
may limit their usefulness for · clinical trol participants (20 males, 6 females) with no
interpretation in the United StatEs. history of neurological or psychiatric disorder
were recruited. Participants were matched
[PASAT.4] Rao, Mittenberg, Bernardin,
with head-injured patients on age (± 2 years),
Haughton, and Leo, 1989 (Gronwall
education (± 2 years), and gender. Thus, con-
trol subjects ranged in age from 17 to 57, with
This study examined the effects of £~peri an average of 29.7 (12.4), and ranged in edu-
ventricular white-matter changes on 'tive cational level from 7 to 20 years, with an av-
functioning in healthy adults. The uthors erage of 13.2 (3.0). The standard 61-digit
selected 40 participants (10 males, 30 males) version using four trials (2.4, 2.0, 1.6, and 1.2
who had normal brain imaging to serve as seconds) was administered at two different
controls. Participants ranged in age ~m 25
and 60 years, with an average age of 42! (8.1),
average educational level of 14.0 (2. ), and
points in study 1 and at five different points in
study 2. Testing and retesting sessions were
separated by approximately 1 week. Data for
average Verbal IQ of 106.5 (5.8). All artici- study 1 are reported in this review.
pants were recruited from newspaper!: adver-
tisements in the Milwaukee, Wiscons~, area. Study strengths
Additional exclusion criteria were a pristory 1. The sample composition is well de-
of hypertension, cardiac or cerebro~cular scribed in terms of age, education, gen-
disease, neurological illness, head in~·, sub- der, and recruitment procedures.
stance abuse, or psychiatric illness. artici- 2. Adequate exclusion criteria.
pants underwent physical and ne logical 3. Test administration procedures are
exams. I specified.
Gronwall's 61-digit test administratiln ver- 4. Means and SDs for the test scores are
sion of the PASAT was employed, h t only provided.
two trials, at 3- and 2-second pacin rates,
were used. Total correct responses fqr both Considerations regarding use of the study
trials are reported. ' 1. The geographic location where partici-
pants were recruited is not provided;
Study strengths 1 however, it may he assumed that they
1. The sample composition is w~ de- were from the Ottawa, Canada, region,
scribed in terms of age, educatio•• gen- which may limit their usefulness for clinical
der, and recruitment procedures.! interpretation in the United States. While
2. Exclusion criteria are provided. 1 not mentioned in this study, in previous
3. Test administration procedures rj-e de- studies the authors have administered the
scribed. ! test in French or English, depending on
4. Means and SDs for the test scolies are the participant's language preference.
reported. : 2. Small sample size.
i
Considerations regarding use of the stuldy Other comments
1. Relatively small sample size. I. 1. Test data for two testing sessions (from
2. The data are not stratified by ag;, gen- study 1) have been reproduced in this
der, or education. : chapter. In addition, the authors provide
3. Data for only two pacing rates ipr the data for five testing probes (study 2),
PASAT are provided. i which can he found in the original study.
I
[PASAT.6] Rao, Leo, Bernardin, and undergraduate students from the University
Unverzagt, 1991a (Gronwall Version) of Victoria to serve as controls. Participants
(Table A8.7) ranged in age from 20 to 35, with an average
The study examined the pattern of cognitive age of23.7 (2.58) and an average education of
deficits in patients with MS using a brief 15.21 (0.79) years. No exclusion criteria are
neuropsychologicaJ battery. The authors re- provided.
cruited 100 (25 maJes, 75 femaJes) normaJ, Two triaJs of GronwaJl's 61-digit version
heaJthy adults through newspaper advertise- of the PASAT were administered at 2.0- and
ments in the Milwaukee, Wisconsin, area. 1.6-second pacing rates.
Controls were matched to MS subjects based
on age (±3 years), education (±1 year), and Study strengths
gender. Thus, control participants had an av- 1. The sample composition is well described
erage age of 46.0 (11.6) years, an average in terms of age, education, gender, geo-
education of 13.3 (2.0) years, and an average graphic location, and recruitment pro-
Verba] IQ of 107.2 (11.2). Exclusion criteria cedures.
were history of substance abuse, psychiatric 2. Means and SDs for the test scores are
illness, head injury, or other neurologicaJ reported.
disorders. All controls were given neurologicaJ
evaJuations and MRI scans. Only one partici- Considerations regarding use of the study
pant was non-Caucasian. All subjects were 1. Sample size is smaJI.
paid for their participation. 2. No exclusion criteria are described.
GronwaJJ's 61-digit administration version 3. The data were obtained on Canadian
of the PASAT was employed, but only two subjects, which may limit their useful-
triaJs, at 3- and 2-second pacing rates, were ness for clinicaJ interpretation in the
used. TotaJ correct responses for both triaJs United States.
are reported. 4. Only two triaJs of the PASATwere used.
5. Education level is high.
Study strengths
[PASAT.BJ Zalewski, Thompson, and Gottesman,
1994 (Gronwall Version) (Table A8.9)
scribed in terms of age, education, gen-
der, and recruitment procedures. The authors compared the cognitive perfor-
2. Relatively large sample size. mance of patients with Post-traumatic Stress
3. Adequate exclusion criteria. Disorder and GeneraJized Anxiety Disorder to
4. Test administration procedures are controls. The data were selected from a large
specified in a previous study (Rao et aJ. database of scores collected in the Vietnam
1989). Experience Study (VES) during 1985-1986
5. Means and SDs for the test scores are (for more description, see Decoufle et aJ.,
reported. 1991). The control group consisted of241 non-
psychiatric veterans randomly drawn from a
Considerations regarding use of study larger sample of 1,579 veterans who had never
1. The data are not stratified by age, gen- met criteria for various psychiatric disorders
der, or education. (e.g., depression, bipolar disorder, substance
2. Data for only two pacing rates are pro- abuse, personaJity disorders). No other ex-
vided. clusion criteria are provided. These partici-
pants were initiaJly recruited for the VES in
[PASAT.7] Strauss, Spellacy, Hunter, order to study the long-term heaJth effects of
and Berry, 1994 (Gronwall Version) military service in Vietnam. Participants were
(Table A8.8) Vietnam and non-Vietnam veterans who en-
The authors examined the utility of the tered the U.S. Army between 1965 and 1971.
PASAT as a tool for detecting malingering. All participants underwent comprehensive
They selected 10 (four maJes, six femaJes) medicaJ and psychologicaJ evaJuations. This
sample is most likely primarily all male, but community centers, and public service, and
there is no mention of the gender composi- were paid for their participation.
tion. They were an average of 38.0 years old The original 61-digit version of Gronwall's
and had an average of 13.6 years of education PASAT was administered in its entirety, and
(no SDs were reported). There were 189 Cau- total scores for the four trials are reported for
casians, 35 Mrican Americans, 11 Hispanics, the total sample and for three age groups.
and 6 "others" in the sample.
Two trials (2.4 and 1.2 seconds) of Gron- Study strengths
wall's version of the PASAT were adminis- 1. Large sample is used.
tered, and total correct responses for both 2. The composition is well described in
trials is reported. terms of age, education, gender, IQ,
geographic area, and recruitment pro-
Study strengths cedures.
1. Large sample size. 3. Adequate exclusion criteria.
2. Sample composition is well described in 4. Test administration procedures are
terms of age, education, and ethnicity. specified.
3. Test procedures are relatively well de- 5. Means and SDs for the test scores are
scribed. reported.
4. Means and SDs for the test scores are 6. The sample is stratified into three age
reported. groups.
1. It is unclear whether the control group 1. The data are not stratified by education
was recruited for research participation or IQ.
only or if any of the participants were 2. Total scores are reported instead of in-
referred for clinical assessment. dividual scores for each of the four trials.
2. Sample composition is not well de- 3. The data were obtained on subjects from
scribed in terms of gender or recruit- the United Kingdom, which may limit
ment procedures, but reference is made their usefulness for clinical interpreta-
to another study. tion in the United States.
3. Exclusion criteria only included psychi-
atric disorder. [PASAT.10] Prevey, Delaney, Cramer, Mattson,
4. Only two trials of the PASAT were and VA Epilepsy Cooperative Study 264 Group,
administered, and total scores were 1998 (Gronwall Version) (Table A8.11)
reported. As part of a large multicenter study of epilepsy,
the cognitive functioning of patients with
[PASAT.9] Crawford, Obonsawin, and Allan, complex partial and generalized seizure dis-
1998b (Gronwall Version) (Table A8.10) orders was examined. Control participants
The authors examined the relationship be- consisted of 45 neurologically normal individ-
tween age and PASAT performance, to obtain uals. Additional exclusion criteria were a his-
validity data on the PASAT and to provide tory of serious medical disorders, psychiatric
additional normative data. A sample of 152 disorders, or substance abuse. There is no
participants (77 males, 75 females) were mention of the gender of the participants nor
screened for neurological, psychiatric, and their IQ; however, average age was 44.4 (11.4)
systemic disorders. Participants ranged in age years and average education was 12.8 (1.9)
from 16 to 74, with an average age of 40.21 years. Participants were primarily recruited
(13.89), an average education of 12.97 (2.86) from nonmedical hospital staff at 13 different
years, and an average IQ of 105.0 (14.08). study centers across the United States.
Participants were recruited from various Only two trials (2.4 and 2.0 seconds) of
communities and organizations within the Gronwall's 61-digit version of the PASAT
United Kingdom, including recreational clubs, were administered.
Study strengths Considerations regarding use of study

1. Sample composition is relatively well 1. The sample is small.
described in terms of age, education, 2. The age, education, and gender compo-
and recruitment procedures but not sition of participants in all conditions of
gender or IQ. the study are provided but not specifi-
2. Adequate exclusion criteria. cally for the control group.
specified. [PASAT.12] Honn, Para, Whitacre, and
4. Means and SDs for the test scores are Bornstein, 1999 (Gronwall Version)
reported. (Table A8.13)
The authors examined the role of exercise in
Considerations regarding use of the study HIV-positive and -negative males and found
1. The data are not partitioned by age or
that exercise only minimally improved cogni-
education group.
tive functioning in both groups. Seventy-six
2. Only two trials of the PASAT were used.
HIV-negative homosexual or bisexual males,
[PASAT.11] Holdwick and Wingenfeld, 1999 with a mean age of 32.5 (6.3) and mean edu-
(Gronwall Version) (Table A8.12) cational level of 14.6 (2.4) years, served as
controls. Exclusion criteria were history of
The relationship between mood, anxiety, and
intravenous drug use, head injuries resulting
attention was assessed in college students. Un-
in greater than 1 hour of unconsciousness,
dergraduate participants were randomly as-
learning disability, or other neurological dis-
signed to different conditions in which various
ease. In this control sample, 32.4% (n = 13) of
mood states were induced (e.g., sad or anxious).
nonexercisers and 13.2% (n = 5) of exercisers
Twenty controls were assigned to a neutral
reported past history of marijuana abuse or
condition. There is no specific information re-
dependence. Participants were also adminis-
garding the age, education, IQ, or gender of the
tered an intelligence test (WAIS-R), the
controls. All were native English speakers, had
SCID, and various anxiety and depression
adequate hearing, and had no histmy of re-
rating measures.
peating grades in elementuy or high school.
Only three trials (2.4, 2.0, 1.6 seconds) of
Additional exclusion criteria were history of
Gronwall's 61-digit version of the PASAT
psychological problems, neurological illness af-
were administered.
fecting attention, head trauma, medication use,
substance abuse, attention problems, or learn-
ing disability. Age, gender, and ethnicity are Study strengths
described for the sample as a whole but not 1. Relatively large sample size.
specifically for the control group. 2. The sample composition is well de-
The 61-digit Gronwall version of the PASAT scribed in terms of age, education, gen-
was administered using a computer. The four der, and IQ.
trials (2.4-, 2.0-, 1.6-, and 1.2-second pacing) 3. Adequate exclusion criteria.
were delivered via synthesized computer 4. Test administration procedures are
voice, and responses were recorded by a mi- specified.
crophone. All responses were scored manually. 5. Means and SDs for the test scores are
reported.
Study strengths
1. Adequate description of participant re- Considerations regarding use of the study
cruitment procedures. 1. An all-male sample is used.
2. Adequate exclusion criteria. 2. Education levels are relatively high.
3. Test administration procedures are 3. Recruitment procedures are not speci-
specified. fied.
4. Means and SDs for the test scores are 4. A portion of the sample reports a history
reported. of marijuana abuse or dependence.
Other comments Other comments

1. 'nle exercisers scored significantly 1. Additional outcome measures, such as
higher on the 1.6-second trial of the number of errors committed and num-
PASAT relative to the nonexercisers. ber of "no" responses, are reported in
the original article, which have not been
[PASAT.13] Wingenfeld, Holdwick, Davis, reproduced in this chapter.
and Hunter, 1999 (Gronwall Version)
(Table A8.14) [PASAT.14) Bate, Mathias, and Crawford,
This study was designed to develop normative 2001 (Gronwall Version) (Table A8.1 5)
data for a computerized version of Gronwall's This study examined the relationship between
PASAT. The authors recruited 168 (80 males, the Test of Everyday Attention and various
88 females) college students between the ages neuropsychological measures in patients with
of 17 and 48 with an average age of 21 (5.1) severe head injury. The study was conducted in
years at the University of Arkansas, Fayette- Australia, where 35 controls (20 males, 15 fe-
ville. The sample was 88% Caucasian, 4% Af- males) who were native English speakers with
rican American, 4% Asian American, and 4% no history of psychiatric illness, neurological
other ethnic group. The data were first strati- disorders, intellectual disability, substance
fied by gender and then by two age groups (17- abuse, or hemiplegia of the dominant hand,
29, 30-48 years). Exclusion criteria were any were recruited. Participants were an average of
history of neurological illness, emotional prob- 30.2 (10.3) years of age, obtained an average of
lems, learning disability, attentional problems, 12.6 (2.0) years of education, and had an aver-
or uncorrected hearing difficulty. Only native age premorbid IQ of 101.1 (9.1) based on
English speakers were included. Subjects were the National Adult Reading Test-Revised
given course credit for participation. (NART-R). The exact location and procedures
The testing procedures are similar to those for participant recruitment are not specified.
of Gronwall, except that the digits are pre- Also, it is unclear whether the participants were
sented by the computer via speaker and patients with non-brain injury-related illness or
responses are recorded through an external healthy individuals from the community.
speaker. Additionally, while all four trials are 'nle Gronwall 61-digit version of the
delivered (2.4-, 2.0-, 1.6-, and 1.2-second PASATwas presented with all four trials (2.4-,
pacing), a new random series of the 61 digits is 2.0-, 1.6-, 1.2-second pacing).
presented during each trial.
Study strengths
Study strengths 1. The sample composition is well de-
1. Adequate sample sizes, except for the scribed in terms of age, education, gen-
30-48 age group. der, and IQ.
2. 'nle data are stratified first by gender 2. Adequate exclusion criteria.
and then by two age groups (17-29, 30- 3. Test administration procedures are
48 years). specified.
3. The sample composition is well de- 4. Means and SDs for the test scores are
scribed in terms of age, gender, ethni- reported.
city, and recruitment procedures.
4. Adequate exclusion criteria. Considerations reganhng use of the study
5. Test administration procedures are spec- 1. The sample size is small.
ified. 2. Recruitment procedures are not well
6. Means and SDs for the test scores are described. Controls may be non-head-
reported. injured medical patients.
3. The data were obtained on Australian
Considerations regarding use of the study subjects, which may limit their useful-
1. Cell size for the 30-48 age group is rel- ness for clinical interpretation in the
atively small (n = 12). United States.
[PASAT.15] Boringa, Lazeron, Reuling, Ader, have been reviewed in this chapter. Partici-
Hennings, Underboom, de Sonneville, Kalken, and pants were 60 (30 males, 30 females) young
Polman, 2001 (Gronwall Version) (Table A8.16) and middle-aged adults recruited from the
The sensitivity of the Brief Repeatable Battery Guy's College campus in the vicinity of Lon-
of Neuropsychological Tests, used to assess don, England, via newspaper advertisements
cognitive functioning in patients with MS, was and notices. The "young" men were an average
evaluated in Amsterdam. This battery includes of 21.1 (0.4) years of age and had an average
a modified, two-trial version of Gronwall's IQ of 113.0 (1.5), the "young" women were an
PASAT. A total of 140 healthy participants average of 20.9 (0.2) years of age and had an
(62 males, 78 females) between the ages of average IQ of 112.4 (1.7), the "middle-aged"
22 and 73, with an average age of 45.8 years, men were an average of 57.5 (1.3) years of age
were recruited from the community. None and had an average IQ of 117.7 (1.8), and the
had central nervous system disease, psychiat- "middle-aged" women were an average of 60.3
ric illness, learning disability, history of sub- (0.7) years of age and had an IQ of 113.3 (2.2).
stance abuse, serious head injury, or other All participants were screened for physical
major medical illness. In terms of education, illness in the past week, use of any medication,
31 participants had< 9 years, 55 had 9 or 10 history of psychiatric disorders, and high
years, and 53 had> 10 years (one participant scores on a depression or anxiety scale.
did not state his education). All four trials (2.4, 2.0, 1.6, and 1.2 sec-
Gronwall's 61-digit version of the PASAT onds) of Gronwall's version of the PASAT
was administered using only two trials (3- and were used, and scores for each trial are pre-
2-second pacing). sented.

1. Large sample size. 1. The sample composition is well described
2. The sample composition is well de- in terms of age, gender, IQ, geographic
scribed in terms of age, education, and area, and recruitment procedures.
gender. 2. The data are stratified by two age groups
3. Adequate exclusion criteria. (young and middle-aged) x gender.
4. Test administration procedures are 3. Adequate exclusion criteria are used.
specified. 4. Test administration procedures are
5. Means and SDs for the test scores are specified.
reported. 5. Means and SDs for the test scores are
reported.
1. Over half of the sample has < 10 years of Considerations regarding use of the study
education. 1. Overall sample size is adequate, but in-
2. The data were obtained on individuals dividual cells are relatively small.
from Amsterdam, which may limit their 2. Intelligence level for the sample is rela-
usefulness for clinical interpretation in tively high.
the United States. 3. Educational levels are not reported.
4. The data were obtained on individuals
[PASAT.16] Fluck, Fernandes, and File, 2001 from London, England, which may limit
(Gronwall Version) (Table A8.17) their usefulness for clinical interpreta-
The study had two goals: (1) to examine the tion in the United States.
effects of two dosages of lorazepam on atten-
tion in healthy individuals and (2) to investi- [PASAT.17] Snyder, Cappelleri, Archibald,
gate the effects of age and gender on selected and Fisk, 2001 (Gronwall Version)
tests of attention. More comprehensive norms (Table A8.18)
are presented for the part of the study that Using two different scoring methods for the
examined age and gender; thus, those data PASAT, the authors examined the classification
rates of patients with secondary progressive Levin's Administration Version

and relapsing-remitting types of MS. The au-
thors reanalyzed data from MS patients and [PASAT.18] Brittain, Ia Marche, Reeder, Roth,
35 (9 males, 26 females) healthy controls col- and Boll, 1991 (Levin Version) (lables A8.19
lected in an earlier study (Fisk & Archibald, and A8.20)
2001). Staff, volunteer workers, and $tudents In this normative study using the Levin et al.
from the Queen Elizabeth II Health ·Science (1987) version of the PASAT, the authors
Centre, Dalhousi University, and MS~ Society present data for 526 healthy participants (aged
in Nova Scotia, Canada, served as qontrols. 17-88 years). The data were stratified by four
The average age of the participants ~ 37.97 age groups (< 25, 25-39, 40-54, and > 55
(12.94) years, average education ~ 14.06 years). In the< 25 age group, there were 145
(2.27) years, and average raw WAIS-R!Vocab- (55 male, 90 female) participants, 79 Cauca-
ulary subtest score was 54.5 (7.0). Eiclusion sians and 66 "other" race, with an average of
criteria were history of drug or alcohol abuse, 13.0 (1.3) years of education and an average
major psychiatric illness, learning disability, Shipley IQ of 105.0 (9.1). In the 25-39 age
seizures, head trauma, or other neurological group, there were 164 (67 male, 97 female)
disorder. Additional exclusion criteria were participants, 114 Caucasians and 50 "other"
use of specific medications, such as· neuro- race, with an average of 14.0 (2.2) years of ed-
leptics, benzodiazepines, antiepileptic drugs, ucation and an average Shipley IQ of 103.0
or sedatives. (10.4). In the 40-54 age group, there were
All four trials (2.4, 2.0, 1.6, and 1.2 s+conds) 95 (50 male, 45 female) participants, 79 Cauca-
of Gronwall's version of the PASAT ~read sians and 16 "other" race, with an average of
ministered. Two mean outcome measqres are 13.0 (3.1) years of education and an average
reported: (1) the mean number of ~rrect Shipley IQ of 101.0 (12.6). In the >55 age
responses across the four trials (i.e., the sum group, there were 122 participants, 119 Cau-
of the correct responses for all trials divided casians and 3 "other" race, with an average of
by 4) and (2) the dyad score, in which ~airs of 12.0 (2.5) years of education and an average
correct responses were counted as one correct Shipley IQ of 106.0 (15.1). For the >55 age
point. group, the authors report 82 males and 82
females, but this appears to be a misprint
Study strengths since there were only 122 participants in total
1. The sample composition is well de- for this age group. Exclusion criteria were a
scribed (in an earlier study by Fisk & history of psychiatric or neurological problems,
Archibald, 2001) in terms of age, edu- as well as concussions or loss of consciousness.
cation, Vocabulary subtest perfollllance, A detailed description of this modified ver-
geographic area, and recruitment pro- sion of the PASAT is presented. Error rates
cedures. (rather than correct responses) and seconds
2. Adequate exclusion criteria. taken for each response are used as the out-
3. Test administration procedures are come measures.
specified.
4. Means and SDs for the test scotes are Study strengths
reported. 1. The sample composition is well de-
Considerations regarding use of the study der, and Shipley IQ.
1. The sample size is relatively small 2. The data are stratified by age and IQ
2. The data were obtained on Canadian level.
subjects, which may limit their .seful- 3. Adequate exclusion criteria.
ness for clinical interpretation In the 4. Test administration procedures are well
United States. specified.
3. The educational level is relatively high 5. Means and SDs for the error scores are
(14.1 years). reported.
Considerations regarding use of the study gender, ethnicity, IQ, geographic loca-
1. The data are not stratified by educational tion, and recruitment procedures.
level. 3. The data are presented for three age
2. Overall sample is adequate, but some of groups.
the individual cells are small. 4. Test administration procedures are
specified.
Other comments 5. Means and SDs for the test scores are
1. Number of errors rather than correct reported.
responses are reported.
2. Data for number of seconds taken to Considerations regarding use of the study
respond are reported in the original ar- 1. Educational levels are high in the
ticle, but since these data are rarely used middle-aged and older adult age groups.
in clinical evaluations, they have not
been reproduced in this chapter. Other comments
1. IQ was estimated using only the Vocab-
[PASAT.19] Roman, Edwall, Buchanan, and ulary and Block Design subtests of the
Patton, 1991 (Levin Version) (Table A8.21) WAIS-R.
The authors conducted this study in order
to provide additional normative data for the [PASAT.20] Cicerone, 1997 (Levin Version)
Levin et al. (1987) version of the PASAT. (Table A8.22)
They recruited 143 white adults in three dif- The author compared the attentional abilities
ferent age groups (18-27, 33-50, and 60-75). of mildly head-injured patients and normal
IQ was prorated with the Block Design and controls on four neuropsychological tests.
Vocabulary subtests from the WAIS-R. In the Forty control participants between the ages of
18-27 age group, there were 62 (58% female) 18 and 59, with an average age of 33.3 (12.4)
participants, with an average education of 12.0 years and average educational ·level of 14.9
(0.77) years and an average IQ of llO (12.3). (2.2), were enrolled. Participants had no his-
In the 33-50 age group, there were 40 (50% tory of head injury, neurological disease, or
female) participants, with an average educa- psychiatric illness and were recruited from the
tion of 15.0 (2.6) years and an average IQ of Edison, New Jersey, community. They were
110 (12.3). In the 60-75 age group, there were administered the Levin et al. (1987) version of
41 (51% female) participants, with an average the PASAT.
education of 15.0 (3.2) years and an average
IQ of 107.0 (11.0). Participants were under- Study strengths
graduate students and employees of Baylor 1. Adequate sample size.
University, students from a local business col- 2. The sample composition is well described
lege, members of service clubs and retired in terms of age, education, geographic
professional groups, employees of local busi- area, and recruitment procedures but not
nesses, individuals from senior citizen orga- gender.
nizations, and individuals in retirement 3. Adequate exclusion criteria.
communities. Only one-fourth of the partici- 4. Test administration procedures are
pants were paid ($5 each). Exclusion criteria specified.
were a history of head injury with loss of 5. Means and SDs for the test scores are
consciousness, other neurological disorders, reported.
substance abuse, psychiatric disorders, or
current use of psychoactive medication. Considerations regarding use of the study
1. Wide age range among participants.
Study strengths 2. Educational level is relatively high.
1. Relatively large sample. 3. Total PASAT scores, rather than indi-
2. The sample composition is well de- vidual scores for each of the four trials,
scribed in terms of age, education, are reported.
[PASAT.21] Wiens, Fuller, and Crossen, 1997 [PASAT.22] Tierslcy, Cicerone, Natelson, and
(Levin Version) (Tables A8.23 and A8.24) Deluca, 1998 (Levin Version) (Table A8.25)
This is a normative study for Levin et al. 's Information-processing speed was compared
(1987) version of the PASAT. The authors among patients with chronic fatigue syn-
selected 821 (672 male, 149 female) partici- drome, mild head injwy, and normal controls.
pants aged 20-49 years who were adminis- All 20 normal control participants were
tered neuropsychological and psychological females, who were recruited from advertise-
tests as part of a civil service job selection ments in the local community of New Jersey
process. There were 699 Caucasians, 46 Afri- and paid for their participation. Participants
can Americans, 31 Hispanics, 32 Asians, and were an average of37.1 (2.4) years of age, with
13 Native Americans in the sample. The data an average education of 15.0 (0.55) years.
were stratified by gender. Male participants Exclusion criteria were current medical ill-
were an average of 29.2 (6.1) years of age, with nesses, a history of loss of consciousness > 5
an average education of 14.6 (1.5) years and minutes, psychiatric illness, use of medication,
an average WAIS-R full-scale IQ (FSIQ) of or participation in a regular exercise program.
106.6 (11.0). Female participants were an av- The Levin et al. (1987) version of the PASAT
erage of 29.2 (5.6) years of age, with an av- was used, and the total number of correct
erage education of 14.5 (1.6) years and an responses for all four trials was reported.
average WAIS-R FSIQ of 105.4 (11.1). They
were all from the Pacific Northwest of the Study strengths
United States. All participants had passed 1. The sample composition is well de-
physical and medical health screening prior to scribed in terms of age, education, gen-
test administration. All had passed a test of der, geographic area, and recruitment
basic academic skills, and none had alcohol or procedures.
substance abuse. All four trials of Levin's 2. Adequate exclusion criteria.
version of the PASAT were administered. 3. Reference is provided for test adminis-
tration procedures.
Study strengths 4. Means and SDs for the test scores are
1. The sample composition is well de- reported.
der, IQ, ethnicity, geographic location, Considerations regarding use of the study
and recruitment procedures. 1. Small sample size.
2. The data are stratified by gender and by 2. Female participants only.
age x IQ. 3. Education level is high.
3. Adequate exclusion criteria. 4. Total scores are reported instead of in-
dividual scores for each of the four trials.
specified. [PASAT.23] Stein, Kennedy, and Twamley, 2002
5. Means and SDs for the test scores are (Levin Version) (Table A8.26)
reported.
The authors examined the difference in neu-
ropsychological test performance of female
Considerations regarding use of the study victims of partner violence with posttraumatic
1. Overall sample size is adequate, but stress disorder (PTSD) compared to victims
some of the individual cells are relatively without PTSD and nonvictimized controls.
small. Twenty-two female control participants were
recruited through posted advertisements and
Other comments personal contacts in the San Diego, California,
1. The authors found differences between community. They were an average of 29.4
the ethnic groups, but the sample sizes (10.7) years of age, had an average of 13.9
were too small to make any definitive (1.5) years of education, and had an average
conclusions. raw WAIS-111 Verbal subtest score of 45.9
(7.4). All participants were ftuent English normal Mini-Mental Status Exam scores.
speakers and had at least an 8th grade reading Participants were recruited from either the
ability. Further exclusion criteria were meet- Kessler Institute in West Orange, New Jersey,
ing DSM-IV criteria for PTSD; use of psy- or the local community.
chotropic medication within the last 6 weeks The authors report using a 50-digit version
of the study; use of oral or intramuscular of the PASAT at four pacing intervals (2.4, 2.0,
steroids within the last 4 months of the study; 1.6, and 1.2 seconds). However, it is unclear
learning disability; history of attention-deficit whether the standard version of Levin et al.'s
disorder, substance abuse, seizure disorder, (1987) procedures were used.
schizophrenia, or other psychotic disorders; or
neurological illness. The Levin et al. (1987) Study strengths
version of the PASAT was used, and the total I. The sample composition is relatively well
number of correct responses for all four trials described in terms of age, education, IQ,
was recorded. geographic area, and recruitment pro-
cedures but not gender.
Study strengths 2. Adequate exclusion criteria.
1. The sample composition is well 3. Means and SDs for the test scores are
described in terms of age, education, reported.
Vocabulary subtest performance, geo-
graphic area, and recruitment proce- Considerations regarding use of the study
dures. 1. The sample size is small.
2. Adequate exclusion criteria. 2. It is unclear whether the digits were
3. While test administration is not de- presented in a different random order or
scribed, appropriate reference is made in a fixed random order across trials.
to the version of the PASAT used. 3. The educational level is relatively high.
reported.
PASAT-50, PASAT-100, and PASAT-200
Considerations regarding use of this study Administration Versions
1. The sample is small.
2. An all-female sample is used. [PASAT.25) Diehr, Heaton, Miller, and Grant,
3. Summary scores across all trials are re- and the HIV Neurobehavioral Center, 1998
ported, rather than correct responses for (PASAT-200 Version) (Table A8.28)
each individual trial. The authors present normative data for a large
sample of Caucasian and African-American
[PASAT.24] Diamond, Deluca, Kim, and Kelley, males and females, using a modified version of
1997 (Levin Version) (Table A8.27) the PASAT (i.e., PASAT-200; see section on
This study compared performance on the Modifications and Alternate Formats of the
PASAT and the visual analog version of the PASAT). A total of 566 participants were used
PASAT (the PVSAT) of patients with MS and from four separate studies. One hundred fifty
controls. The authors recruited 22 participants of the participants were HIV-!-seronegative
to serve as controls on the PASAT task. There controls recruited from a research center in
is no information about the gender of the par- San Diego, California; 277 participants were
ticipants. They ranged in age from 31 to 56, African-American volunteers recruited for a
with an average age of 40.9 (8.9), average normative study from the San Diego, Cali-
educational level of 15.4 (2.2), and average fornia community; 78 served as controls for a
North American Adult Reading Test (NAART) study examining the effects of alcohol on
premorbid IQ of 113.6 (13.0). None of the cognitive performance; and 60 were controls
participants had a history of psychiatric or for a study examining the effects of eosino-
neurological disorders, drug or alcohol abuse, philia myalgia syndrome. Exclusion criteria for
or loss of consciousness. All participants had all studies were history of neuropsychiatric
conditions such as substance abuse or de- of education, 21% had a high school education,
pendence, head injury, and developmental and 12% had lower than a high school educa-
disability. Participants ranged in age from 20 tion. Forty-five percent of the sample were
to 68, with an average age of39.7 (12.1) years, Caucasian, while the remaining 55% were Af-
and ranged in education from 9 to 20, with an rican American.
average education of 14.2 (2.6) years; 39% All participants were screened for psychi-
were female and 55% were African American. atric illness, developmental disabilities, sub-
Briefly, the PASAT-200 is very similar to stance abuse, and head injuries. A more
Levin et al.'s (1987) version in that it consists detailed description of the sample is provided
of the presentation of 50 single digits (except above (PASAT.25) and in Diehr et al. (1998).
for 7) in random order at four different pacing Brie8y, the PASAT-50 consists of one trial
intervals. However, the pacing intervals are of 50 digits (excluding 7) presented in random
3.0-, 2.4-, 2.0-, and 1.6-seconds per digit, in- order at a pace of 3 seconds. The PASAT-100
stead of 2.4-, 2.0-, 1.6- and 1.2-seconds. consists of the same 50 digits presented over
two trials, 3-second pace and 2.4-second pace.
Study strengths
1. Large sample size. Study strengths
2. The sample composition is well de- 1. Large sample size.
scribed in terms of age, education, eth- 2. The sample composition is well de-
nicity, gender, geographic area, and scribed in terms of age, education, eth-
recruitment criteria. nicity, gender, geographic area, and
3. Test administration procedures are recruitment criteria.
specified. 3. Test administration procedures are
4. Adequate exclusion criteria are used. specified.
5. Means and SDs for the test scores are 4. Adequate exclusion criteria are used.
6. Data are stratified by ethnicity and by reported.
educational level.
Considerations regarding use of the study 1. The average education level of the sam-
1. Total PASAT-200 scores, rather than ple is relatively high.
individual scores for each of the four 2. Total scores, rather than individual
trials, are reported. scores for each trial, are reported.
[PASAT.26] Diehr, Cherner, Wolfson,
Miller, Grant, Heaton, and the HIV
Neurobehavioral Research Center Group, 2003 CONCLUSIONS
(PASAT-50, -100, -200 Versions) Studies have documented the utility of the
(Table A8.29) PASAT as a measure of attention/concentra-
The authors present demographically corrected tion, working memory, and information pro-
normative data for two shortened versions of cessing. In fact, the National Multiple
the PASAT-200, namely, the PASAT-50 and the Sclerosis Society included a version of this
PASAT-100. The authors used 560 (61% male) test in their Brief Repeatable Battery of
participants from a pool of archival data on Neuropsychological Tests. However, the ma-
which the PASAT-200 normative information jor drawback of the original version of the
was based (Diehr et al., 1998). Participants PASAT is that it can be a lengthy, difficult,
ranged in age from 20 to 68, with an average age and stressful test. In fact, several studies have
of 39.7 (12.1), and 24% of the sample was over noted participant frustration and attrition.
50 years. Their education level ranged from 9 to Fortunately, there are alternatives to the
20, with an average education of 14.2 (2.6) original version of the test. Clinicians can
years. Most (33%) had between 13 and 15 years administer only one or two trials rather than
all four or use alternative, shortened versions further normative studies partitioning the ef-
of the PASAT. fects of age, education, and IQ are needed.
A review of the literature reveals that there Significant practice effects have been re-
are no significant gender effects for the PA- ported for the Gronwall (1977a,b) version of
SAT but that scores are strongly affected by the PASAT, presumably because the digits are
age, education, and intellectual functioning. presented in the same random order during
As would be expected for most tests involving each pacing trial. This problem has been ad-
speed, PASAT performance significantly de- dressed to some degree with Levin et al.'s
clines with age, particularly as the pacing time (1987) version, in which digits are presented
for the digits is reduced, requiring more cog- in a different random order during each trial.
nitive resources. Likewise, inspection of the The effects of culture, ethnicity, and lin-
data clearly reveals an improvement in per- guistic background on the PASAT have re-
formance with higher educational levels. While ceived very little attention. Only one study
not all studies have found strong correlations explicitly examined the role of ethnicity in
between the PASAT and intellectual func- PASAT performance (Diehr et al., 1998). It is
tioning, the data reviewed in this chapter in- clear that future PASAT normative studies
dicate that it is an important factor to consider need to examine factors such as culture, eth-
when administering this test. It is clear that nicity, and bilingualism. 2
•Meta-analyses for the PASAT were conducted using data

reported in this chapter for each of the four presentation
rates separately. Although the R2 and significance level for
the resulting regression were minimally acceptable, we
felt that the solution was greatly inHuenced by only few
data points which had a considerable weight. Therefore,
the results of meta-analyses are not presented in this
chapter.
J
9
Cancellatiori Tests
BRIEF HISTORY OF THE TESTt commonly used cancellation tests with the
most available literature and have been se-
A number of cancellation tests have been
lected for review in this chapter.
developed over the years. Such tests te pri-
marily designed to assess aspects of atlt!ntion,
such as sustained and selective a~ntion.
Sustained attentiOn "refers to the abllity to RUFF 2&7 SELECTIVE ATTENTION TEST
maintain a consistent level of perfoemance Brief Overview of the Ruff 2&7
over an extended period of time,''; while
selective attention entails selection of ~levant The Ruff 2&7 Selective Attention Test was
target stimuli while avoiding distracto~ (Ruff developed by Ruff and colleagues and is
& Allen, 1996). Some cancellation te$ts are included in the San Diego Neuropsychologi-
also referred to as "vigilance tests" (tezak, cal Test Battery (Baser & Ruff, 1987; Ruff &
1995; Lezak et al., 2004) and typically fivolve Crouch, 1991). The test is designed to exam-
measures of both speed and accuracy of per- ine both sustained and selective attention
formance. A number of cancellation testJ using using two distractor conditions. The test con-
letters, numbers, or symbols as target stimuli sists of 20 blocks, each containing three lines
are available to clinicians. The Ruff 2&7 (Ruff of 50 characters. Within each line, 10 target
et al., 1986a), Digit Vigilance (Le;vis & digits (2s and 7s) are intermixed with either
Rennick, 1979), Digit Cancellation Test (Della other number distractors or capital let-
Salla et al., 1992, 1998}, Visual Searcih and ter distractors. Ruff distinguished two test
Attention Test (Trenerry et al., 1990), Yerbal conditions: (1) blocks in which the tar-
and Nonverbal Cancellation Tasks (Mt$ulam, get numbers are embedded among letters,
1985}, Letter and Symbol Cancellatio+ Task referred to as the "Automatic Detection"
(Caplan, 1985), and Star Cancellation (Halli- condition, and (2) blocks in which the target
gan et al., 1991; Wilson et al., 1987) are $mong stimuli are embedded among other numbers,
the many cancellation tests available t~ clin- referred to as the "Controlled Search" con-
icians and researchers (see Lezak, 19~. and dition. The presentation of the conditions
Lezak et al., 2004, for more details on these (blocks of all digits or blocks of digits and
tests). letters) is alternated. Following brief practice
The Ruff 2&7 Selective Attentioq Test trials, the examinee is given 15 seconds to
and Digit Vigilance Test are the tw~ most complete each of the 20 blocks. He or she is
160
CANCELLATION TESTS 161
prompted to move to the succeeding block (but whose symptoms persisted for at least 3
when the examiner says "next." months), found the Ruff2&7 test to be among
Ruff and Allen (1996) state that in the Auto- the most sensitive and specific measures. They
matic Detection condition, because the num- concluded that this test "can be used with
bers belong to a different stimulus category confidence" since those without concussions
from the letters, the selection process is auto- were unlikely to display impairments on the
matic (i.e., "single-step retrieval of categorical Ruff 2&7. Finally, Ruff et al. (1993) found
information"). However, in the Controlled that the Ruff 2&7 was among the neu-
Search condition, since the targets and dis- ropsychological tests that most strongly
tractors belong to the same category, a more predicted head-injured patients' ability to re-
effortful search involving aspects of working turn to work after 1-6 months postinjury.
memory is required. Ruff (1994) observed relatively mild
Three outcome measures can be obtained impairment in depressed patients on the Ruff
for each of the two conditions: (1) speed is 2&7. The percentile ranking of the majority of
measured with total number of target letters patients fell within the average range for
crossed out, (2) errors consist of the total speed and accuracy. In fact, none of the de-
number of commissions and omissions, and pressed patients was impaired on the accuracy
(3) detection accuracy is calculated by divid- measures, and only three patients exhibited
ing the speed value by the sum of the speed slowed speed.
plus error values (Ruff & Allen, 1996). Weiss (1996) reported that schizophrenic
A number of clinical studies have been patients had more difficulty with speed (only
conducted with the Ruff 2&7 test. Ruff et al. 23% of patients scored in the normal range)
(1992) found that patients with right hemi- than with accuracy (67% scored in the normal
sphere cerebral lesions performed at far range) on the Ruff 2&7. Additionally, patients
slower rates than those with left-sided lesions were better able to detect a target stimulus
and normal controls. Interestingly, those with when it was embedded in letters (Automatic
right anterior lesions were also far less accu- Detection condition) rather than within other
rate in their performance, while patients with digits (Controlled Search condition).
left anterior lesions performed similar to Finally, Schmitt et al. (1988) discovered
controls. that AIDS patients and patients with AIDS-
Ruff et al. (1989a) examined the effects of related complex who were on medication
cognitive rehabilitation on Ruff 2&7 perfor- displayed improved performance on the Ruff
mance in patients with head injury. They 2&7 relative to those who were receiving a
found that teaching cognitive strategies, such placebo.
as focused, sustained attention, as well as Further details about the Ruff 2&7 testing
teaching spatial relationships and memory materials, administration procedures, and
strategies actually improved test performance scoring can be obtained from the test manual
over time. Specifically, on the Ruff 2&7, and kit (see Appendix 1 for ordering infor-
patients in the cognitive strategy condition mation; also Lezak et al., 2004).
made fewer errors relative to those in the
control condition. Bate et al. (2001) found that
Psychometric Properties of the Ruff 2&7
patients with severe traumatic brain injury
(TBI) crossed out fewer target stimuli (i.e., Ruff et al. (1986a) performed a test-retest
were slower) than normal controls. Addition- reliability study of the Ruff 2&7 for four age
ally, while significance values are not re- groups, ranging between 16 and 70 years of
ported, the TBI patients who were within 1 age. Testing probes were separated by 6
year postinjury were slower than those who months. The correlation coefficients for the
were at least 2 years postinjury. Cicerone and four age groups by the two conditions (i.e.,
Azullay (2002), in their examination of the automatic or controlled) ranged 0.84-0.97.
sensitivity and specificity of various neu- The r values were in approximately the same
ropsychological tests in patients with mild TBI ranges for the four age groups; however,
slightly better performance was noted· for the the effects of intellectual functioning, ethnicity,
automatic condition (letter distractors) rela- and motor functioning on the Ruff 2&7.
tive to the controlled condition (dipt dis- For further normative information regard-
tractors). While an improvemept of ing the Ruff 2&7, see the professional manual
approximately 10 points on the retest was reproduced by Ruff and Allen (1996).
ported, the two conditions showed similar
rates of practice effects (Ruff et al., 1986a).
Baser and Ruff (1987) conducted factor DICIT VIGILANCE TEST
analysis on the Ruff 2&7 along with a jhost of
Brief Overview of the DVT
other neuropsychological tests and fouhd that
in normal controls the Ruff 2&7 best [loaded The Digit Vigilance Test (DVT) was devel-
on a factor they termed "complex lintelli- oped by Lewis and Rennick (1979) as part of a
gence." This factor also contained suclt mea- larger test battery, the Repeatable Cognitive-
sures as Controlled Oral Word Ass<>fiation, Perceptual-Motor Battery. The DVT is a test
Full Scale IQ, Vocabulary, Block ~esign, of vigilance and sustained attention, which
Digit Span, and Digit Symbol. Howdver, in also measures aspects of rapid visual tracking
the same study, using a mixed clinical ~ample ability and psychomotor speed. This test
(e.g., psychiatric and head-injured pai;ients), consists of two pages, with 35 single digits
the Ruff 2&7 outcome measures loade4 on an appearing within 59 rows. The digits on the
"arousal" factor (which also included :Finger first page are printed in red ink, and the digits
Tapping, mean designs on the Ruff Figural on the second page are printed in blue ink.
Fluency Test, Digit Symbol) and a "pikning For the standard administration, the task is to
and 8exibility" factor (which also ittluded cross out the number 6, which is randomly
outcome measures from the Wiscons~ Card dispersed throughout the page of digits. The
Sorting Test, perseverative score frdm the alternate administration procedure requires
Ruff Figural Fluency Test, and Ruff-Light that the participant cross out the number 9,
Trail Learning Test). I which also randomly appears throughout the
page of digits. The time in seconds taken to
complete the task, the number of omissions
Relationship Between Ruff 2&7
(target numbers not crossed out), and the
Performance and Demographic Factors
number of commissions (numbers other than
Ruff et al. (1986a) examined diff~ences the target crossed out) are recorded.
between genders, four age groups, and three There are relatively few clinical or norma-
educational levels on the two Ruff 2&!1 con- tive studies on this test. In a study of mildly
ditions. They found no gender effec~. with hypoxemic patients with chronic obstructive
males and females performing similarly :across pulmonary disease (COPD), Prigatano et al.
the two conditions. Clear age effect~ were (1983) observed that patients required a sig-
found across the two conditions, with a: linear nificantly greater amount of time to complete
decline in performance as age increased! Simi- the DVT relative to normal controls. In a
larly, they found that performance im~roved study by Bardwell et al. (2001), DVT was the
as educational level increased up to 15~years; only neuropsychological test score to signifi-
Ruff 2&7 performance plateaued at > 1$ years cantly improve in obstructive sleep apnea
of education. They also found that on ayerage patients who were given continuous positive
individuals performed approximately 15 ~ints airway pressure relative to those who were
better on the Automatic Detection (letter given placebo treatment (Grant et al., 1987).
distractors) relative to the Controlled Search These studies suggest that the DVT, and
(digit distractors) condition. perhaps similar cancellation tests, is sensitive
Clearly, more normative studies are qeeded to detecting neuropsychological deficits in
to better understand the relationship hEttween patients with even mild forms of hypoxemia.
key demographic factors and Ruff 2&7 gerfor- Smith et al. (2001) reported better perfor-
mance. Additional studies should also elilmine mance on the DVT in postmenopausal women
who were on hormone replacement therapy neuropsychological tests that did not dis-
(HRT) relative to their age-matched counter- criminate between mild, moderate, and severe
parts who were not taking HRT. Shean et al. hypoxemic COPD patients but did discrimi-
(2002) found that coaching or providing test- nate between the COPD group as a whole and
taking instructions significantly improved DVT normal controls. Overall, these authors con-
performance in a group of patients with clude that the DVT clusters with tests of
schizophrenia. Additionally, these authors attention and psychomotor speed and that it is
detected that negative symptoms and degree a sensitive test for discriminating COPD
of disorganized thought significantly corre- patients from controls but not for discrimi-
lated with lack of ability to benefit from nating patients at various stages of COPD.
coaching on the DVT. These findings essen-
tially replicated an earlier study by Eckman
Relationship Between DVT Performance
and Shean (2000).
and Demographic Factors
As noted earlier, there are very few normative
Psychometric Properties of the DVT
studies available for the DVT. Heaton et al.
Kelland and Lewis (1994) reported a test- (1991) included the DVT in their comprehensive
retest (probes separated by 1 week) coefficient normative book on various neuropsychological
of 0.87, with a 95% confidence interval of tests, making this the largest normative study to
0.71--0.95, for the standard form test admin- date on the DVT. Heaton et al. (1991) detected
istration of the DVT and a coefficient of 0.89, that in a group of 210 participants, age and years
for the alternate form administration, with a of education accounted for 24% and 13% of
95% confidence interval of 0.75--0.96. Unfor- variability in the time to complete the test, re-
tunately, these data are based on a sample of spectively, and for 15% and 16% of variability in
only 20 individuals. In a subsequent study, the number of errors committed. However,
Kelland and Lewis (1996) reported practice gender alone accounted for only 2% of the var-
effects on the DVT, with test speed improving iability in DVT outcome measures. Kelland
on the second week of test administration and Lewis (1996) also found no gender effect for
relative to the first (initial) testing session. total time required to complete the task or for
However, no improvements were noted total number of errors in a group of college
between the third week of testing relative to students.
the second.
Kelland and Lewis (1996) also assessed the
convergent validity of the Repeatable Cogni-
tive-Perceptual-Motor Battery, which contains METHOD FOR EVALUATING
the DVT, by evaluating its sensitivity to diaz- THE NORMATIVE REPORTS
epam. While the overall score for the battery
To adequately evaluate the Ruff 2&7 and
discriminated between individuals on diaze-
DVT normative reports, five criterion vari-
pam and placebo, no differences were found
ables were deemed critical. The first four of
between the two groups for the DVT. How-
these are related to subject variables, and the
ever, this was also a small sample, with each
last one refers to procedural issues.
group containing only 20 individuals.
Grant et al. (1987) conducted a factor
analysis on tests from the Halstead-Reitan Subject Variables
Neuropsychological Test Battery and several
other neuropsychological tests, including the Sample Size
DVT, in COPD patients and healthy controls. Fifty cases are considered a desirable sample
They observed the DVT to cluster with tests size. Although this criterion is somewhat
of "alertness-psychomotor speed," such as arbitrary, a large number of studies suggest
Trails B and Digit Symbol. In the same study, that data based on small sample sizes are
they noted that the DVT was one of only three highly influenced by individual differences
and do not provide a reliable estimate of the Only one study was designed to provide
population mean. normative information on the Ruff 2&:7 (Ruff
et al., 1986a). Other data on the Ruff 2&:7
Sample Composition Description come from control groups in clinical com-
Information regarding medical and p~hiatric parison studies. Ruff et al. (1986a) partition
exclusion criteria is important. It is unclear normative data for the two conditions by four
if gender, intellectual level, handedness, geo- age groups and three educational levels; the
graphic recruitment region, socioeconomic other studies report demographic information.
status, occupation, ethnicity, or recrUitment Another study by Ruff et al. (1992) provides
procedures are relevant. Until this is deter- normative data for speed and accuracy for
mined, it is best that this information be normal controls. Finally, Bate et al. (2001)
provided. provide Ruff 2&7 data on a small sample of
healthy controls. Most of these studies report
Age Group Interval either speed or speed and accuracy data
This criterion refers to grouping of the data summed across the two Ruff 2&:7 conditions.
into limited age intervals. This requireJnent is Additional normative information, particularly
especially relevant for this test since a strong tables for converting raw scores into T scores
effect of age on cancellation test perfo~ance and percentiles, based on age and educational
has been demonstrated in the literatu~. level, are provided in the Ruff 2&:7 profes-
sional manual (Ruff &: Allen, 1996).
Reporting of Educational levels There are very few normative studies on the
Given the possible association betw~ edu- DVf. Most of the studies have small sample si-
cation and cancellation test scores, informa- zes (10--40), with the exception of Heaton et al.'s
tion regarding educational level shotild be (1991, 2004) normative manuals, which include
reported for each subgroup. data for 210 participants with standardized
scores adjusting for age, education, and gender
presented for African-American and Caucasian
Procedural Variable
participants separately in the 2004 edition.
Data Reporting In this chapter, we review studies which use
Ruff2&:7, followed by DVf studies. Published
For the Ruff 2&7, group means and standard
manuals are reviewed first, followed by nor-
deviations for the number of items correctly
mative studies and control groups from clini-
cancelled should be reported for the Au~matic
cal comparison studies presented in ascending
Detection and Controlled Search co~tions
chronological order for each test separately.
separately. For the ovr, the mean and SD for
The text of study descriptions contains
time in seconds taken to complete the task
references to the corresponding tables identi-
should be reported. Additional useful irforma-
fied by number in Appendix 9. Table A9.1, the
tion for the cancellation tests includes the
locator table, summarizes information pro-
number of omissions (target numbers not can-
vided in the studies described in this chapter. 1
celled) and the number of commissions: (num-
bers other than the target digits cancelltil).
Ruff 2&7 Manual
OF THE NORMS [Ruff 2&7.1] Ruff and Allen, 1996
Information presented in the studies fleport- The normative information in this manual is
ing data for the cancellation tests differslacross primarily based on previous studies by Ruff
studies. Some of these differences Jill be 'Children's norms for various cancellation tests are avail-
summarized below. able in Baron (2004) and Spreen and Strauss (1998).
and colleagues (Ruff et al., 1986a; Baser & Normative Studies and Control Groups
Ruff, 1987; Ruff & Crouch, 1991). A total of in Clinical Comparison Studies for the
360 (180 male, 180 female) healthy volunteers Ruff 2&7
between the ages of 16 and 70 years and with
[RUFF 2&7.2] Ruff, Evans, and Light, 1986a
7-22 years of education participated in the
(Table A9.2)
study. The sample was initially stratified by
four age groups (16-24, 25-39, 40-45, and The authors recruited 259 healthy participants
55-70 years) and three education groups (107 male, 152 female) as part of this
(:512, 13-15, 16 years) but not gender since normative study. Nearly half of the sample
this was not a significant factor in test per- was recruited from California and the rest,
formance. The authors mention that the from Michigan. The investigators selected
sample "roughly approximated the 1980 U.S. individuals with a wide age range and educa-
census proportions with regard to race," but tional attainment in order to examine the ef-
no specific ethnicity data are provided. Data fects of these demographic factors on test
are available for speed and accuracy for each performance. Participants were aged 16-70.
condition individually, as well as total scores The authors report that their sample had 7-72
for speed and accuracy for the two conditions years of education, but it is unclear whether
combined. Thus, a total of six outcome vari- the upper limit reported is a misprint. The
ables are available. sample was stratified by four age groups
Raw score to T score conversion and per- (16-24, 25-39, 40-54, and 55-70 years) and
centiles are available by age and educational three educational levels (:512, 13-15, ~16
level. Sixty-five percent of the sample was years). Standard administration procedures
recruited from California, 30% from Michi- were used.
gan, and the rest from the eastern seaboard.
The normative data contained in Ruff and Study strengths
Allen's manual are not reproduced here, and 1. The sample composition is well de-
the interested reader is referred directly to scribed in terms of age, education, gen-
this publication for further information. der, and geographic area.
reported.
Study strengths 3. Data are stratified by four age x three
1. The sample composition is well described education groups.
geographic area. Considerations regarding use of the study
2. The testing procedures and scoring are 1. Overall sample is adequate, but individ-
well described in the manual. ual cells are relatively small (e.g., some
3. Means and SDs are reported for some of cells contain only 10 participants).
the Ruff 2&7 outcome measures. 2. No exclusion criteria and recruitment
4. Raw scores can easily be converted to T procedures are reported.
scores and percentiles for four age
groups and three educational levels. [RUFF 2&7.3] Ruff, Niemann, Allen, Farrow, and
Wylie, 1992 (Table A9.3)
Considerations regarding use of the study This study examined the effects of cerebral
1. Overall sample is adequate, but some lesions on Ruff 2&7 performance. The authors
individual cells are relatively small (e.g., selected 60 normal controls from a larger
fewer than 20 participants in the 55-70 standardization sample of 259 reported by
year age group who have 13-15 years of Baser and Ruff (1987). The larger sample was
education). recruited from California, Michigan, and New
2. No exclusion criteria and recruitment York. Participants were screened for chronic
procedures are provided. medical illness, "extensive" substance abuse,
or loss of consciousness due to a heacJ injury. 2. Adequate exclusion criteria.

The ethnic breakdown is reported by Baser 3. Means and SDs for the test scores are
and Ruff (1987) for the larger subject pool but reported.
not for the subsample that setved · in this
study. The 60 participants in the currept study Considerations regarding use of the study
were an average of31.2 (4.1) years of~ge and 1. The sample size is relatively small.
had an average of 12.9 (1.5) years ol educa- 2. Recruitment procedures are not well
tion. There is no information on the gender described. Controls may be non-
distribution for this sample. Standard, admin- head-injured medical patients.
istration procedures were used. i 3. The data were obtained on Australian
Study strengths fulness for clinical interpretation in the
1. Sample size is adequate. . United States.
2. The sample composition is well d~cribed
in terms of age and education. i DVT Manual
3. Adequate exclusion criteria. ,
4. Means and SDs for the total sceres for [DVT.1J Heaton, Grant, and Matthews, 1991;
both conditions are reported. Heaton, Miller, Taylor, and Grant, 2004
The DVf manual (Lewis, 1995) refers the
Consideration regarding use of the study reader to the comprehensive normative book
1. The data are not partitioned by age. published by Heaton et al. (1991). Heaton et al.
(1991) gathered a large sample of data on
[RUFF 2&7.4] Bate, Mathias, and Crawfdrd, various neuropsychological tests over a 15-year
2001 (Table A9.4) period using several studies. The DVf is
This study examined the relationship ~tween among the tests for which normative data are
the Test of Everyday Attention and ~arious presented. The total sample used in this nor-
neuropsychologicaJ measures in patierlts with mative book was recruited from various areas
severe head injury. The study was cor!ducted across the United States, including California,
in Australia, where 35 controls (20 nfde. 15 Washington, Colorado, Texas, Oklahoma,
female), who were native English SP,eakers, Wisconsin, Illinois, Michigan, New York, and
with no history of psychiatric illness, (neuro- Virginia, as well as Canada. It is unclear which
logicaJ disorders, intellectual disability. sub- specific regions were used for DVf data col-
stance abuse, or hemiplegia of the dominant lection. All participants reportedly completed
hand, were recruited. The exact locatibn and structured interviews, and those with a history
procedures for participant recruitment are not of learning disabilities, neurologicaJ illness,
specified. Also, it is unclear whether tite par- "significant" head injury, "serious" psychiatric
ticipants were patients with non-brain injury- illness (e.g., schizophrenia), or substance
related illness or healthy individuals frf>m the abuse were excluded from the normative data
community. Participants were an avefage of set. The DVf normative data were gathered
30.2 (10.3) years of age, had an average of 12.6 on a total of 280 participants, who were an
(2.0) years of education, and had an tfverage average of 44.9 (20.0) years of age and ob-
premorbid IQ of 101.1 (9.1), as estim.ed by tained an average of 14.0 (3.2) years of edu-
the National Adult Reading Test-Revised cation. The manual provides regression-based
(NART-R) (Crawford, 1992). S~ndard raw to T score and percentile conversion for
administration procedures were used. · the DVf (and other neuropsychological tests)
based on gender, 10 age groups (20-34, 35-39,
Study strengths 40-44,45-49,50-54,55-59,60-64,65-69,70-
1. The sample composition is: well 74, and 75--80 years) and six education groups
described in terms of age, edti:ation, (6-8, 9-11, 12, 13-15, 16-17, and 18+ years).
gender, and premorbid IQ. l The average DVf raw score reported for the
entire sample of 280 participants for time ta- dard administration procedures were used in
ken to complete the task is 388.5 (86.5), and both manuals.
that for errors committed is 7.1 (8.7). Other
data from the manual are not reproduced here. Study strengths
Interested readers are referred to the original 1. The sample composition is well described
publication. in terms of age, gender, ethnicity, and
In their recently updated normative manual, education.
Heaton et al. (2004) have gathered additional 2. Adequate exclusion criteria.
normative data for the DVf (and other neu- 3. Means and SDs are reported for Cauca-
ropsychological tests). Their sample consists of sian and African-American participants
860 normal participants, of whom 466 are separately and for the entire sample. Ad-
Caucasian and 394 are African American. The ditionally, T scores and percentiles cor-
average age of the Caucasian sample was 47.0 rected for age and education are reported
(20.2) years, and average educational level was for different demographic groups.
14.0 (2.9) years; approximately 57.3% of the
sample were male. The average age of the Af- Considerations regarding use of the study
rican-American sample was 38.7 (12.2) years, 1. Specific sample sizes used per cell are
and average educational level was 13.5 (2.5) not reported.
years; approximately 49.7% of the sample were 2. Recruitment procedures are not well
male. The authors report that the data were described.
gathered from various individual and multi-
center collaborative research projects over a 25-
Other comments
year period. Participants were from various
1. The interested reader is referred to the
U.S. states and Canada, including California,
Fastenau and Adams (1996) critique of
Washington, Colorado, Texas, Oklahoma, Wis-
the Heaton et al. (1991) norms, and
consin, Illinois, Michigan, New York, Virginia,
Heaton et al.'s (1996a) response to this
and the province of Manitoba, Canada.
critique.
All participants reportedly completed struc-
tured interviews, and those with a history of
learning disabilities, neurological illness, "sig- Normative Studies and Control Groups in
nificant" head injury, "serious" psychiatric ill- Clinical Comparison Studies for the DVT
ness (e.g., schizophrenia), or substance abuse
were excluded from the normative data set. [DVT.2] Prigatano, Parsons, Levin, Wright, and
The manual provides regression-based raw Hawryluk, 1983 (Table A9.5)
to T score and percentile conversion for the The authors examined the neuropsychological
DVf (and other neuropsychological tests) test performance of mildly hypoxemic patients
based on gender, 11 age groups (20-34, with COPD. Twenty-five healthy controls
35--39, 4~. 45-49, 50-54, 55-59, 60-64, were matched to the COPD patients based on
65-69, 70-74, 75-79, and 80-85 years), six age, education, handedness, and gender.
education groups (7-8, 9-11, 12, 13-15, Control participants were an average of 59.6
16-17, and 18-20 years), and two ethnic (9.0) years of age and obtained an average of
groups. The average DVf raw score reported 10.5 (3.3) years of education. Participants
for the entire sample of 860 participants for were excluded if they had an "illness that
time taken to complete the task is 390.87 might interfere with their neuropsychological
(57.59). The average time taken to complete testing (e.g., physical handicap, emotional
the DVf for the Caucasian sample is 394.59 problems, alcoholism or psychosis)," had
(88.92), and that for African-American sample COPD, were taking medications for heart or
is 380.41 (86.63). Other data from the manual lung disease, or had diabetes. Fifteen of the
are not reproduced here. Interested readers participants were selected from Winnipeg,
are referred to the original publication. Stan- Manitoba, Canada, and 10 were selected from
Oklahoma City, Oklahoma. Standard· admin- 2. Data are not partitioned by age.
istration procedures were used. 3. Low educational level.
Study strengths . [DVT.4] Kelland and Lewis, 1994

1. The sample composition is well descnbed (Table A9.7)
in terms of age, education, geOgraphic
location, and recruitment procedjues. This study was designed to assess the test-
2. Adequate exclusion criteria. : retest reliability and validity of the DVf, as
3. Means and SDs for the test scores are well as to measure the single-dose effects of
reported. diazepam in groups of college students. The
authors selected 20 college students (10 male,
Considerations regarding use of the study 10 female) from a "large urban university" to
1. Small sample size. . serve as controls (who were administered a
2. Wide age range for the sample. Data are placebo rather than diazepam). Participants
not presented by age group. ranged in age from 18 to 30, with an average
3. The data for over half of the sample were age of 20.0 (2.8) and an average educational
obtained on Canadian part:ipipants, level of 13.1 (1.3) years. Participants were
which may limit their use~ss for excluded from the study if they reported tak-
clinical interpretation in the :United ing medications; had a history of subs~ce
States. ' abuse· had a medical history that reqwred
4. Low educational level. centr~ nervous system~epressant medica-
tion use; had a history of neurological, cardiac,
[0Vl.3] Grant, Prigatano, Heaton, McS~eeny, renal, or hepatic disease; or drank more than
Wright and Adams, 1987 (Table A9.6) . two cups of coffee a day. The DVf, along ~.th
other neuropsychological tests, was adminis-
The authors examined neuropsycbPlogical
tered two times to each participant, with each
functioning in COPD patients wi~ mild, session separated by 1 week. Standard
moderate, and severe hypoxemia. They selected administration procedures were used. Data
99 "nonpatient" participants (75 m!Ie, 24
are reported for both the standard (crossing
female) who did not have COPD, a ru.tory of out 9s) and the alternate (crossing out 6s)
"significant" head injury, a history of substance
administrations.
abuse, heart disease that required treaa,ent, or These data were later reanalyzed by Kel-
neurological or metabolic illnesses. Partfipants
land and Lewis (1996), who found a practice
were an average of 63.1 years of age apd had
effect from week 1 to week 2 of test admin-
obtained an average of 10.2 (3.6) yead of ed-
istration but no differences between week 2
ucation. The authors do not specify !testing and week 3. The Kelland and Lewis (1996)
procedures but do mention the larger ~ttery data for weeks 1 and 2 are the same as those
from which the Dvr is drawn (i.e., the reported in this study and, thus, will not be
Rennick-Lafayette Repeatable Battery). reproduced in this chapter.
Study strengths
1. Relatively large sample size. Study strengths . . .
2. The sample composition is well 1. The sample composition 1S well descnbed
described in terms of age, edtfation, in terms of age, gender, education, and
and gender. I
: recruitment procedures.
3. Adequate exclusion criteria. 2. Adequate exclusion criteria.
4. Means and SDs for the test sco!es are 3. Means and SDs for the test scores are
reported. ' reported.
4. Test-retest data are reported.
1. Test administration procedures are not Consideration regarding use of the study
specifically described. 1. Small sample size.
[DVT.S] Bamcord and Wanlass, 1999 scores for the Vocabulary, Block Design, and
(Table A9.8) Wide Range Achievement Test (WRAT)
The authors compared the performance of Reading for the HRT group were 14.2 (3.3),
college students on six neuropsychological 12.7 (2.4), and 108.6 (5.5), respectively; values
tests administered in the standard, paper-and- for the non-HRT group were 13.9 (3.7),
pencil format vs. a more ecological format of 11.8 (3.5), and 108.8 (12.7), respectively.
using plastic sheet protectors so as to not The women on HRT made significantly fewer
create paper waste. For the purposes of this errors on the DVT than those who were not
chapter, the participants in the standard test- onHRT.
ing format were considered the "normal"
controls. Ten college students (five male, five Study strengths
female) were recruited. Participants were an 1. The sample composition is well described
average of 19.8 (3.95) years of age, with an in terms of age, education, gender, and
average of 12.8 (0.63) years of education. recruitment procedures, with limited IQ
data available.
1. The sample composition is well described 3. Means and SDs for the test scores are
in terms of age, education, and gender. reported.
2. Means and SDs for the test scores are 4. Data are reported for postmenopausal
reported. women on HRT and those not on HRT.
Considerations regarding use of this study

Considerations regarding use of the study 1. The sample is small.
1. The sample is small.
2. Educational level is relatively high.
2. No exclusion criteria are provided.
3. An all-female sample is used.
3. Test administration procedures are not
specified. [DVT.7] Stein, Kennedy, and Twamley, 2002
(Table A9.10)
[DVT.6] Smith, Giordani, Lajiness-O'Neill, and The authors compared neuropsychological
Zubieta, 2001 (Table A9.9) test performance of female victims of partner
The neuropsychological effects of HRT were violence with PTSD to victims without PTSD
examined in 29 healthy postmenopausal and nonvictimized controls. Twenty-two fe-
women. Participants were recruited through male control participants were recruited
advertisements and selected if they were 60 through posted advertisements and personal
years or older, had received HRT without contacts in the San Diego, California, com-
interruption after menopause, or had never munity. They were an average of 29.4 (10.7)
been treated with HRT. Exclusion criteria years of age, had an average of 13.9 (1.5) years
included participants who had stopped and of education, and had an average raw WAI S-
restarted HRT for more than 1 month at a Ill Vocabulary subtest score of 45.9 (7.4). All
time; had a significant general medical, neu- participants were fluent English speakers and
rological, or psychiatric illness; had a history of had at least an 8th-grade reading ability.
head trauma leading to loss of consciousness; Further exclusion criteria were presence of
had substance dependence; or were taking PTSD (DSM-IV criteria), use of psychotropic
medications affecting the central nervous medication within the last 6 weeks of
system. Standard administration procedures the study, use of oral or intramuscular steroids
were used. Participants taking HRT were an within the last 4 months of the study, learning
average of 65.0 (4.0) years of age, with an disability, history of attention-deficit disorder,
average of 15.0 (2.0) years of education; and history of substance abuse, seizure disorder, a
those not on HRT were an average of 67.0 history of schizophrenia or other psychotic
(6.0) years of age, with an average of 16.0 (3.0) disorders, or neurological illness. Standard
years of education. Average WAIS-R standard administration procedures were used.
Study strengths visual tracking ability. Two tests were selected

1. The sample composition is well ~scribed for discussion in this chapter, the Ruff 2&.7
in terms of age, education, geographic Selective Attention Test and the DVI'.
area, and recruitment procedw.'es, with A review of the literature indicates that there
limited Verbal IQ data (i.e., Vocabulary •are no gender differences on either of these
raw scores were available). tests but that performance clearly declines
2. Rigorous exclusion criteria. with age. Performance on such tests appears to
3. Means and SDs for the test scbres are improve with higher levels of education.
reported. Additionally, there appear to be some crit-
ical gaps in the existing normative data for the
Considerations regarding use of this study cancellation tests reviewed in this chapter. For
1. The sample is small. i example, for the Ruff 2&7, when the data are
2. An all-female sample is used. partitioned by age, sample sizes are vecy small
(fewer than 20), particularly for individuals
older than 40 years. For the DVT, most par-
CONCLUSIONS ticipants over 50 years of age tend to have
Clinicians and researchers use canqellation lower educational levels (<12 years).
tests to assess various aspects of atkention, Unfortunately, very few large-scale norma-
including vigilance and sustained and ~lective tive studies have been conducted on either of
attention. There are numerous such te$ts from these cancellation tests. Aside from age and
which to choose, and most involve p~ 1 r-and- education, future studies should examine
pencil administration. Such tests also equire factors such as ethnicity and intellectual
aspects of psychomotor responding, , well as functioning when developing normative data. 2
• Meta-analyses were not perfonned due to a lack of suffi-

cient data for the two tests discussed in this chapter.
Ill
LANGUAGE
10
Boston Naming Test
BRIEF HISTORY OF THE TEST 82 aphasic patients partitioned by aphasia

severity level.
The Boston Naming Test (BNT) is a test of In 2000, Kaplan et al. published the second
confrontation naming consisting of simple edition of the BNT, which includes a 15-item
line-drawn pictures. Its experimental version short form, four multiple-choice options for
includes 85 drawings (Kaplan et al., 1978). The each of the same 60 items that were used in the
modified version of the BNT, published in previous edition of the test, and error codes to
1983, is limited to 60 of the original 85 draw- categorize incorrect responses. The ceiling was
ings, arranged in order of ascending difficulty changed to eight items. The normative data
(Kaplan et al., 1983). Participants are allowed included in the record booklet are partitioned
20 seconds to name each item. Stimulus cues by 15 age groups for children, spanning in age
are offered to correct for misperception errors. between 5-0 and 12-5 years, and five age
They are followed by phonemic cues, which groups for adults between 18 and 79 years.
provide the first phonemes of the word, facil- In addition to being used as a stand-alone
itating lexical retrieval. The total score on the test, the BNT, second edition, is included in
test is the number of correct responses pro- the Boston Diagnostic Aphasia Examination
duced spontaneously (SR) and with the aid of (BDAE) published by Psychological Assess-
stimulus cues (SC). The basal rule is eight ment Resources (Goodglass et al., 2001).
consecutive pictures correctly named without All studies listed in this chapter are based
any assistance, and the discontinuation rule is on the original 60-item version of the BNT
six consecutive failures. (For detailed admin- since no studies based on the second edition
istration and scoring instructions, see Lezak of the test were published by the time this
et al., 2004; Spreen & Strauss, 1998; and book went into production.
instructions in the test stimulus booklet.) Thompson and Heaton (1989) and Heaton
The authors provide normative data on the et al. (1991) reported high correlation
60-item version for children 5.5-10.5 years of between the 85-item and the 60-item versions
age, broken down into six age groups based on (r=0.96). However, the mean percent of
five participants in each group; for normal correct responses was somewhat lower for the
adults aged 18--59 years, broken down into original version (85.1% vs. 87.8%) in their
two educational groups and five age groups sample of clinical referrals for neuropsycho-
based on a total of 84 participants; and for logical evaluation.
173
174 LANGUAGE
Studies Using BNT Error Quality Analyses on the higher frequency of circumlocutory de-
scriptions and semantically related responses in
Several authors have studied the errors made
the elderly, the authors cited difficulty with
on the BNT by different clinical vs. normal
perception and semantic identification as un-
groups (Albert et al., 1988; LaBarge et al.,
likely contributors to the naming difficulty.
1992; Nicholas et al., 1985; Reiter, 2000; Smith
They concluded that the major age-related
et al., 1989; Tombaugh & Hubley, 1997).
difficulty lies in the label (lexical) retrieval stage.
Approaches to the classification of naming
Using the same taxonomy of naming mech-
errors are typically based on the presumed
anisms, LaBarge et al. (1992) identified 17
underlying mechanisms: perceptual (analysis
types of error, which were classified into three
of the visual features of the picture), semantic
categories: no content, linguistically related,
(access of the underlying conceptual ~presen
and perceptually related (Table 10.2). The
tation), and lexicol (retrieval of the appropriate
authors hypothesized that linguistic errors
name for the stimulus) (Snodgrass, 1$84).
reflect a loss in lexical (and potentially se-
Following Borod et al.'s (1980) study, which
mantic) information; no-content errors are
indicated an impaired lexical retrievtl mech-
representative of a loss in semantic content;
anism underlying naming difficultieJ in the
and perceptual errors are indicative of a
normal elderly, Nicholas et al. (1985) explored
breakdown in the perceptual mechanism.
the integrity oflexical retrieval in normal aging
Based on the analysis of errors produced
through qualitative analyses of naming,errors in
by 49 elderly with very mild or mild senile
the 85-item version of the BNT. Th6 authors
dementia of Alzheimer's type (SDAT), the
identified several error types, which are out-
authors identified loss of lexical information as
lined in Table 10.1.
well as some disruption in specific semantic
Using this system in the analysis of BNT
attributes as processes underlying confronta-
errors for a group of 162 healthy prupcipants
tional naming difficulty in early SDAT. With
aged 30-79 years, the authors concluded that
progression of the disease, increasing involve-
confrontation naming requires sever~ stages
ment of core semantic structures is implicated.
of information processing: (1) perception of the
Hodges et al. (1991) developed a different
object, (2) semantic identification, (3) retrieval
error classification system, which, in reference
of the label that corresponds to that semantic
to the item "beaver," can be illustrated as
"concept," (4) encoding the articulatory pro-
follows: category names ("animal"), within-
gram, and (5) correct articulation of that label
category semantic errors ("skunk"), semantic
or name.
associates ("dam"), and semantic circumlocu-
The authors reported a decline in naming
tions ("an animal that builds dams").
ability with age, especially after age 70. Based
To further refine the process of error clas-
sification along a semantic dimension, Nicho-
las et al. (1996) proposed a system of rating
Table 10.1. Types of Naming Error Identified by errors on a 5-point scale of semantic related-
Nicholas et al. (1985)• ness to the target name, with 1 being not at all
similar in meaning (for single-word responses)
Coding Category Example from BNT
and poor, incomplete definition or description
No response {comment) ··1 have one of those on my (for multiword descriptions) and 5 being very
porch" similar in meaning (for single-word responses)
Augmented-correct ''Propeller on an airplane"
and good, complete definition or description
Semantically related '"Harness" for yole
Phonologically related ''Prong·· for tong$ (for multiword responses).
Perceptually related ''Flower'' for pinyheel Tombaugh and Hubley (1997) exa~ne~
Whole-part, part-whole '"Clock" for pendiilum the frequency of errors in responses to mdi-
Off-target utterance ..Artistic thing fo$ flower" vidual items and distribution of errors across
{circumlocution) for trellis
seven categories, such as no response, cir-
•For a description of each type of error, see Ni~olas et al. cumlocution, semantic, phonemic, visual, per-
{1985). severation, and miscellaneous, in a sample of
BOSTON NAMING TEST 175
Table 10.2. Types of Error Described by LaBarge et al. (1992)

Type of Error Examples
No Content
Empty phrase I don't know
Can't think of it
No interpretation possible No response or jargon
LinguiaticaU, Belated
Phonological
Phonologically related Pelican= pentagon
Unicorn= hornicorn
Rhinoceros= nostros
Sphinx= phoenix
Semantic
Same category Latch=hasp
Super- or subordinate Camel= animal
Asparagus= vegetable
Function Funnel= used for pouring
Compass= makes circles
Attribute Beaver= eats wood or builds dams
Volcano= fire
Context Stethoscope= doctors use it
Sphinx= found in Egypt
Description Noose =a rope with a slip knot
Acoustic
Meaningful sound Whistle =make a whistling sound or blow noiselessly
Volcano= make a whooshing sound
Pantomime or Gesture
Gesture Comb= gesture to head like combing
Accordion= swing arms and hands like playing
Perceptually Belated (Viaually)
Whole Whistle= trailer hitch or pacemaker
Knocker= chandelier
Igloo=turtle or spider's web
Part Dart= feather
Rhinoceros= two big horns
Perspective Harmonica= windows or apartment
building or file drawers
Function Dart= nurse to give shot
Broom= wash my clothes
Attribute Beaver= fella who goes underground
Mask= a bad picture
Context Wreath=see 'em at a wedding
Whistle= hanging on a tree limb
Copyright © 1992 by the Educational Publishing Foundation. Adapted with permission.
219 cognitively intact adults 25-88 years of made by adult aphasic patients. They found
age. They also addressed errors resulting from that semantic and circumlocution errors
the ambiguity of items, cultural or regional accounted for 82% of the errors made by boys,
terms, commonly confused words, and syno- whereas adult aphasics' performance was
nyms and suggested specific probes to be used characterized by phonemic errors in addition
to clarify ambiguous responses. to the two error types demonstrated by the
Kirk (1992a) compared quality of errors boys. The authors proposed 'A Revised Chil-
made by 212 boys 5-13 years of age to those dren's BNT' based on item analyses of the
176 LANGUAGE
boys' responses and provided normative data Frank et al., 1996; Henderson et al., 1990;
for the BNT collected on a sample of 382 Hodges et al., 1991; Huff et al., 1986b; Mar-
schoolchildren. golin et al., 1990; Martin and Fedio, 1983). A
number of studies suggest that this contribu-
tion increases as a function of dementia sever-
Current Views on the Mechanisms
ity (Hodges et al., 1991; Huff et al., 1986b;
Underlying Confrontation Naming
LaBarge et al., 1992; Shuttleworth & Huber,
Deficits
1988). As a result, in the early stages of AD, a
Naming ability in normal aging has received naming deficit might manifest itself through
close attention in the literature. According to lexical access difficulties (LaBarge et al., 1992;
LaBarge et al. (1986), uncomplicated aging is Neils et al., 1988), which resembles the pat-
not associated with naming deficit. However, a tern characteristic for normal aging.
large body of evidence points to an age-related Data challenging the common view of
decline in naming ability. Several investigators semantic disruption as the cause of naming
suggest that deficient access to the lexical difficulties in AD were presented by Nebes
network is the leading mechanism of naming and colleagues (Nebes et al., 1984; Nebes,
difficulty (Bowles & Poon, 1985; Nicholas et 1989; Nebes & Brady, 1990), who viewed
al., 1985). However, other investigators argue lexical retrieval as the source of naming diffi-
that age-related naming difficulties cannot be culties. Similarly, Nicholas et al. (1996) poin-
fully attributed to disruption in lexical access. ted to the breakdown in lexical access in AD
Barresi et al. (2000) suggest that impaired and referred to the previous findings of
lexical access is the leading mechanism of semantic breakdown as an artifact of the meth-
naming failures in individuals under the age of odologies of the previous studies. This view is
70. However, in those older than 70, naming further supported by results using the General
difficulties are in part related to semantic Processing Tree approach, which is a type of a
degradation. Similarly, Au et al. (1995) viewed multinomial model that estimates the proba-
perceptual and semantic processing deficits as bilities of cognitive processes that are pre-
partially responsible for naming difficulties in sumed to underlie performance on a cognitive
the elderly. Moberg et al. (2000) showed that task, as measured by categorical data (Reiter,
lexical access was not affected by the aging 2000). The author found that lexical access,
process in their study; however, they pointed perceptual analysis, and phonological realiza-
to methodological limitations as a possible tion abilities decline with increase in dementia
cause of this finding. Ferraro et al.'s (1998) severity, with the most consistent decline
results indicate that decline in speed of pro- being in lexical access.
cessing might contribute to age-related decline Studies supporting the lexical deficit hy-
in BNT performance. pothesis suggest a breakdown in the retrieval
Understanding of faulty processes in Alz- stage, which is based on the following findings:
heimer's disease (AD) also remains controver- (1) incidence of errors in low-frequency words
sial. Whereas the majority of the more recent is higher than that in high-frequency words:
investigations rule out disruption in the per- This mechanism is modulated by lexical
ceptual stage as a primary cause of this break- processing (Kirshner et al., 1984; Skelton-
down (Bayles & Tomoeda, 1983; Frank et al., Robinson & Jones, 1984); (2) facilitation of
1996; Huff et al., 1986b; LaBarge et al., 1992; lexical access by phonemic cues (Martin & Fe-
Martin and Fedio, 1983; Smith et al., 1989), the dio, 1983); (3) semantic relatedness of the words
relative contributions of lexical vs. semantic produced by the participant to the target word
dysfunction are highly debated in the literature. (Bayles & Tomoeda, 1983; Smith et al., 1989).
Regarding the semantic deficit hypothesis, In spite of the controversy regarding the
disruption in the content and organization of mechanisms accounting for naming difficulties
semantic information is implicated as the in AD vs. normal aging, the majority of studies
primary source of naming difficulties in AD have demonstrated utility of the BNT in dis-
(Bayles & Tomoeda, 1983; Flicker et al., 1987; tinguishing between AD and age-related
decline in naming ability (Beatty et al., 2002; observed that analysis of misperception errors
Huff et al., 1986b; Margolin et al., 1990; allows identification of perceptual fragmenta-
Storandt & Hill, 1989). Goldman et al. (1998) tion and inattention to a part of the visual
reported decline in naming ability as a func- field, which are associated with nondominant
tion of severity of Parkinson's disease. hemisphere dysfunction.
Several studies have explored the mechan-
isms of naming deficits in different types of
Modifications and Short Versions
aphasia. According to Nicholas et al. (1985),
of the BNT
aphasic participants (across all major aphasic
groups, except for anomies) have difficulty in An attempt to create two shorter equivalent
the phonological encoding of words. Kohn and forms of the BNT for repeated testing was un-
Goodglass (1985) support this finding by dertaken by Huff et al. (1986a). Based on the
demonstrating considerable similarity in error experimental 85-item version, these authors
types across different aphasic groups. In addi- developed two 42-item versions that proved to
tion, they provide a more specific analysis of be reliable (r=0.71-0.81 for controls and
anomie errors associated with different types of r=0.97 for AD patients) and equivalent in
aphasia: "Negated responses were associated difficulty. Both versions were standardized on
with Broca's aphasia, whole-part errors (hose normal and brain-damaged participants.
for nozzle) were associated with frontal anomia, The different forms of the test were com-
and poor phonemic cuing was associated with pared by Thompson and Heaton (1989). They
Wernicke's aphasia" (p. 266). The authors also administered an 85-item version of the BNT to
reported that anomie aphasics produced the a clinical group of participants; data were then
highest frequency of multiword circumlocu- rescored according to the criteria for the 60-
tions and the lowest number of phonemic er- and 42-item forms. Although certain differ-
rors, which they relate to minimal word ences between forms were found, there were
production difficulty in anomie aphasia relative high correlations among different versions of
to other aphasia syndromes. On the other hand, the test (ranging 0.82-0.96) and between BNT
Lewis and Soares (2000) showed that pre- scores and other language measures.
language conceptual organization deficit might Heaton et al. (1991, 2004) published nor-
underlie naming difficulties in aphasic patients mative data for the 85-item version. The
who present with semantic paraphasias as a revised set of norms (2004) is based on a
leading feature of their language disturbance. sample of over 1,000 normal adults, stratified
Investigation of the neuroanatomical sub- by age, education, gender, and race/ethnicity
strates of naming may shed light on its cog- (African American and Caucasian).
nitive mechanisms. The hippocampus of the Farmer (1990) proposed modifications of
dominant hemisphere has been widely impli- the administration, response coding, and scor-
cated in naming function (Davies et al., 1998; ing procedures for the full version of the BNT,
Martin et al., 1999; Sawrie et al., 2000; Sei- which were used by the author to assess non-
denberg et al., 1998). Other aspects of the brain-damaged adults.
dominant temporal lobe are also involved in Eight short versions of the test were com-
naming function, according to Ojemann et al. pared by Mack et al. (1992) and Williams et al.
(1993) and Wiggs et al. (1999). Bell et al. (1989) in patients suffering from AD and
(2000) showed that a decline in naming ability neurologically intact elderly. The short forms
is a frequent sequela of left anterior temporal were four 15-item versions developed by these
lobectomy. These findings consistently point authors, one 15-item version used by the Con-
to the involvement of the dominant temporal sortium to Establish a Registry for Alzheimer's
area in naming function. In addition, areas of Disease (CERAD), and three 30-item versions.
the superior parietal and frontal cortices are Scores on each version could be extrapolated
implicated by Wiggs et al. (1999). to a complete 60-item BNT score.
In addition to the obvious use of the BNT Franzen et al. (1995) compared different
in assessing word retrieval, Kaplan (1988) short forms on a sample of 320 individuals
178 LANGUAGE
with various neuropsychiatric diagnoses. The ability comparable to the full 60-item version,
authors report adequate internal consistency using stepwise discriminant analysis.
for all forms and reasonable correlations Two 30-item versions, comprised of the odd
between forms. Based on their analysis of item and even items from the 60-item version, were
difficulty, the authors identified the CERAD administered by Fisher et al. (1999) to 30
version as least desirable. normal elderly and 32 patients with probable
Similarly, Larrain and Cimino (1998) AD. The forms were found to be equivalent
reported poor criterion validity of the CERAD and discriminated well between the control
15-item version as its agreement with the full and patient groups. The combined mean was
version of the BNT was only 70% in classify- consistent with that derived by retrospective
ing patients with probable AD as impaired. extraction in the original odd/even test con-
Fastenau et al. (1998) administered four struction study.
15-item versions (Mack et al., 1992) and two One of the 15-item versions (version 2)
30-item versions in counterbalanced order to developed by Mack et al. (1992) was used by
108 normal adults between 57 and 85 years of Calero et al. (2002) to assess normal and
age. Fifteen-item versions 1 and 2 were com- demented elderly with low educational level.
bined by Fastenau et al. in the first 30-item The authors found a high degree of equiva-
version, and 15-item versions 3 and 4 were lence between the full and the short versions.
combined in the second 30-item version. The second edition of the BNT (Kaplan
Alternate-form reliability coefficients for the et al., 2000) includes one of the 15-item short
six versions ranged 0.53-0.76, and Cronbach's versions (version 4) developed by Mack et al.
ex coefficients ranged 0.37-0.75. Validity coef- (1992).
ficients of 0.93 and above support the hypoth-
esis that the short forms sample the same
Cultural Adaptations and Culture-Specific
domain as the long form. The authors pointed
Normative Data for the BNT
out that short forms 3 and 4 have slightly better
psychometric properties than forms 1 and 2. The literature supports effects of demograph-
Tombaugh and Hubley (1997) derived eight ics, including language and culture, on the
short versions from the performance of 219 BNT. Cruice et al. (2000) and LeDorze and
healthy volunteers on the full version. The Durocher (1992) identified the cultural rele-
authors concluded that the 30-item versions vance of items, such as word length, frequency,
are preferable to the 15-item forms and cor- and familiarity, as important in selecting items
relate highly with the full test. for a naming task. Several adaptations of the
Ferraro and Barth (2003) compared four BNT for different cultural and ethnic groups,
15-item versions and the CERAD version with which use culturally appropriate subsets of
the full 60-item test and found the short ver- items or modified sets of items, are available.
sions to be as reliable as the full version in Ponton et al. (1996, 2000) described the
regard to lexicality issues. Ponton-Satz BNT, which is an adaptation of
Saxton et al. (2000) developed two empiri- the standard version for assessment of His-
cally derived (based on item difficulty) equiv- panic patients. It consists of 30 items derived
alent 30-item short forms, which were strongly from the original test, which are presented in
related to each other and to the total BNT different order from the original. Some items
score. have several possible correct responses listed
Lansing et al. (1999) compared scores of on the answer sheet, depending on the country
719 normal elderly and 325 AD patients on of origin of the examinee. Normative data for
eight previously reported short forms derived a sample of 300 Hispanic participants strati-
from participants' performance on the 60-item fied by gender, age, and education are pro-
version. The short forms differed in their vided (Ponton et al., 1996).
ability to discriminate between patients and Allegri et al. (1997) reported normative data
controls. The authors derived a new, 15-item, for the Spanish BNT published in Madrid
gender-neutral short form with discrimin- (see Appendix 1), which were collected on
200 residents of Buenos Aires across the adult A Korean version, K-BNT, was developed
age span. The standard administration proto- following the procedures for development of
col was used. The authors proposed an alter- the American version of the test. Based on
native order of items, based on analysis of the item difficulty and concreteness of concepts,
frequency of correct responses in their sam- 60 items were selected from a pool of 175
ple. The Spanish version is an adaptation of items (Kim & Na, 1999). Normative data for a
the original test, with some items substituted sample of 600 normal participants, stratified
with culture-fair items. by eight age groups (15-75+ years) and five
Several studies have also addressed the educational levels (0-13+ years), are reported.
impact of culture on the English version of
the BNT. A modified Australian version of the
BNT was used by Cruice et al. (2000) and
Worrall et al. (1995). Items "beaver" and Review of the literature suggests that valida-
"pretzel" from the standard version were tion studies for the BNT have focused on its
replaced with "platypus" and "pizza." diagnostic and predictive properties in dis-
Barker-Collo (2001) addressed cultural bias criminating between normal and clinical
in the performance of 58 New Zealand uni- groups (Cahn et al., 1995; Jacobs et al., 1995;
versity students on the standard BNT. Item Knesevich et al., 1986).
analysis indicated that overall performance was Concerns regarding asymmetry (negative
hindered by unfamiliarity with several items on skew) and "peakedness" (extreme kurtosis) in
the test. The authors suggested strategies for the distribution of BNT scores and resulting
adaptation of the test to New Zealand culture. limited score variability and small SDs in nor-
Stewart et al. (2001) administered the mative samples restricted to high-functioning
CERAD short version of the BNT to 285 participants have been widely discussed in the
African Caribbean participants 55-75 years of literature (Fastenau, 1998; Hamby et al., 1997;
age, who were residents of south London. Hawkins & Bender, 2002; Killgore & Adams,
Normative data are stratified by two age 1999). In essence, the test does well what it is
groups, gender, three education groups, and designed to do. It is designed to measure
three occupational classes. deficit in naming ability or severity of aphasia,
Performance on a shortened version of the rather than level of skill or proficiency within
BNT has been reported for Native Americans the range of normality. It is sensitive to the
(Ferraro & Bercier, 1996; Ferraro et al., severity of the naming deficit, with the non-
2002). Fillenbaum et al. (1997, 2002) report anomie end of the distribution falling within
performance of elderly white and African- the "normal" category. As such, the test is not
American community residents on the CER- expected to discriminate between average,
AD version of the test. above average, and superior naming ability.
Marien et al. (1998) reported normative Therefore, intact performance on this test
data for the standard BNT on a sample of 200 would be most accurately described as "within
native Dutch-speaking Flemish elderly and normal limits." Use of z scores and percentiles
recommended cutoff scores according to age, in describing impaired performance should be
education, and gender. Error analysis was done with full understanding that the distri-
performed. The authors compared American bution of test scores is not normal.
English-, Australian English-, and Dutch- High test-tretest reliability over a 1-2 week
speaking elderly performance based on a lit- interval in a group of elderly participants was
erature review and inferred that linguistics do documented by Flanagan and Jackson (1997).
not affect the overall score but do impact the Dikmen et al. (1999) reported test-retest
error distribution in different languages. reliability of 0.92 over an 11-month interval
A French version of the BNT was intro- in a mixed sample. Mitrushina and Satz (1995)
duced by Colombo and Assai (1992). This reported test-retest reliability ranging 0.62-
report provides data for 420 normal French- 0.89 over three annual probes in a sample of
speaking Swiss adults 20--89 years old. neurologically intact elderly. Data on repeated
180 LANGUAGE
administration are also presented by McCaf- Goulet et al., 1994; Kimbarrow et al., 1996;
frey et al. (2000). LaBarge et al., 1986; Lansing et al., 1999;
High correlation of BNT performance with Mackay et al., 2002; Marien et al., 1998; Neils
verbal fluency was reported by Locascio et al. et al., 1995; Nicholas et al., 1989; Randolph
(1995), with r=0.5 for AD patients and et al., 1999; Saxton et al., 2000; Van Gorp et al.,
r = 0.52 for normal control participants. In 1986; Worrall et al., 1995). Several inves-
contrast, a comparison of BNT performance tigators suggest that the most pronounced
with measures of different aspects of memory decline occurs only after age 70 (Albert et al.,
suggested that BNT scores are unrelated to 1988; Mitrushina & Satz, 1995; Nicholas et al.,
learning and memory scores (Albert et al., 1985). Furthermore, Welch et al. (1996) sug-
1988). gest that individuals with >12 years of educa-
Two reports showed that a lack of agree- tion retain intact naming ability into their 80s.
ment between clinicians in the administration On the other hand, Goodglass (1980) as well as
and scoring of the BNT affects the usefulness LeDorze and Durocher (1992) suggest that a
of published norms. Ferman et al. (1998) significant drop in naming ability may occur in
pointed out that the total score achieved on the sixth decade of life. Schmitter-Edgecombe
the test is affected by variability in the inter- et al. (2000) point to the cohort effect or gen-
pretation of the discontinuation rule of six con- erational familiarity with individual items that
secutive failures. According to the lenient may account for some of the age-related dif-
criteria, correct responses aided by phonemic ferences in performance identified in cross-
cues are not counted toward the discontinua- sectional studies. Goulet et al. (1994), in their
tion rule. The authors compared the rigorous review of 25 studies, point to inconsistent
and lenient interpretations of the discontinu- reports on the effect of age on BNT perfor-
ation rule and found that the final scores mance, with some studies suggesting no age-
varied in 3% of the sample of 655 normal related decline in naming. They attribute these
elderly and in 31% of 140 patients with AD, differences to methodological issues and sub-
depending on the interpretation. Lopez et al. ject characteristics. Moberg et al. (2000) exam-
(2003) examined three scoring methods that ined the effect of age on lexical properties of
might be viewed as correct interpretations of the word representations for the BNT stimuli,
the test instructions and found discrepancies such as familiarity, number of letters, fre-
in the resulting scores and impairment levels. quency of occurrence, and number of sylla-
It is unknown if comparable rates of mis- bles, and found that the process of lexical
interpretation would be found upon replica- access was similar in young and older adults.
tion of this inquiry in different settings with Coffey et al. (2001) did not find a relationship
different groups of clinicians as follow-up between BNT performance and age-related
studies are not available to date. changes evident on the MRI in 320 nonclinical
For further information on the psychomet- volunteers aged 66-90. This conclusion is
ric properties of the BNT, see Franzen (2000), consistent with their review of the relevant
Lezak et al. (2004), and Spreen and Strauss neuroimaging literature.
(1998). Education was found to be related to BNT
scores in several studies (Allegri et al., 1997;
Borod et al., 1980; Deloche et al., 1996;
Hawkins et al., 1993; Hawkins & Bender,
RELATIONSHIP BETWEEN BNT
2002; Heaton et al., 1999; Henderson et al.,
1998; Kimbarrow et al., 1996; Lansing et al.,
FACTORS
1999; Le Dorze & Durocher, 1992; Marien
There is abundant evidence of the effect of age et al., 1998; Neils et al., 1995; Nicholas et al.,
on BNT performance, particularly declining 1985; Ponton et al., 1996; Randolph et al.,
scores and increasing performance variability 1999; Ross et al., 1995; Saxton et al., 2000;
with advancing age (Au et al., 1995; Farmer, Thompson & Heaton, 1989; Welch et al.,
1990; Fastenau et al., 1998; Feyereisen, 1997; 1996; Worrall et al., 1995). Higher variability
in BNT performance was observed in groups et al. (1999), Marien et al. (1998), Ran-
with lower educational levels. In contrast, dolph et al. (1999), Saxton et al. (2000), and
Cruice et al. (2000), Farmer (1990), Fastenau Welch et al. (1996) reported males outper-
et al. (1998), Ivnik et al. (1996), and LaBarge forming females in normal samples. Randolph
et al. (1986) did not find any association et al. (1999) suggested that the gender effect
between BNT performance and educational is due to performance on specific items that
level, which might be related in part to are more familiar to men.
restricted ranges of educational levels in some Reports of the effect of ethnicity and culture
samples. A combined effect of age and edu- on BNT performance yield contradictory
cation should be taken into consideration, findings. In the study by Henderson et al.
according to Heaton et al. (1999), as older (1998), comparing healthy African-American
individuals with lower educational levels are and Caucasian participants, ethnicitywas unre-
more likely to be misidentified as dysnomic. lated to BNT performance. Similarly, Manly
Similarly, an interaction of age and education et al. (2002) did not find a notable difference in
was reported by Borod et al. (1980), Farmer naming ability between African-American and
(1990), and Welch et al. (1996). Manly et al. Caucasian elders. In contrast, Lichtenberg
(1999) showed that illiterates scored signifi- et al. (1994) and Ross et al. (1995) report
cantly lower than literate participants with up higher scores for Caucasian compared to
to 3 years of education in their sample of Mrican-American participants in a group of
Spanish-speaking, non-demented elders. medical inpatients. Similar findings are re-
Albert et al. (1988), Killgore and Adams ported by Kimbarrow et al. (1996) on a sample
(1999), Thompson and Heaton (1989), and of geriatric rehabilitation patients and by
Tombaugh and Hubley (1997) found that Whitfield et al. (2000) on a sample of healthy
verbal intelligence, as measured by WAIS-R elderly. These findings are discussed by Hen-
Vocabulary score, in their samples of neuro- derson et al. (1998) in the context oflower edu-
logically normal participants strongly affected cational level of Mrican-American participants.
BNT performance. Similarly, Hawkins et al. Kimbarrow et al. (1996) emphasize socio-
(1993) found reading vocabulary score to be economic status, ethnicity, and cultural factors
strongly correlated with BNT performance in and Whitfield et al. (2000) point to the cul-
a sample of psychiatric and normal partici- tural appropriateness of the material as ex-
pants. The authors presented BNT perfor- planatory variables. Furthermore, Manly et al.
mance expectation guidelines based on the (1998) found that participants who reported
Gates-MacGinite Reading Vocabulary Test for less acculturation obtained lower scores on
use as a complement to the published norms. the BNT in their sample of medically healthy
Based on a review of the relevant literature, Mrican Americans. Similarly, Touradji et al.
Hawkins and Bender (2002) emphasized the (2001) reported lower naming performance in
contribution of premorbid vocabulary to BNT foreign-hom Caucasian elders compared to
performance and made recommendations on those born in the United States. Further
moderator variables to be considered in fur- moderating variables were acculturation level
ther research. and language use.
Gender was shown in several studies to be Qualitative differences in BNT performance
unrelated to naming efficiency in normal as a function of ethnic background and geo-
samples (Cruice et al., 2000; Fastenau et al., graphical region were reported by Goldstein
1998; Henderson et al., 1998; Ivnik et al., et al. (2000). Participants in their study tended
1996; LaBarge et al., 1986). However, based to use alternative responses to several BNT
on an analysis of BNT performance, Ripich items that are specific to their region, with
et al. (1995) suggest that naming skills are further differences between black and white
poorer for women than for men with similar participants in which alternative responses
clinical dementia rating (CDR) scores and were used.
demographic characteristics in their sample of Utility of the standard version of the BNT in
60 early AD participants. Similarly, Lansing assessment of monolingual Spanish-speaking
182 LANGUAGE
and Spanish/English bilingual individuals and/ age on BNT performance has been demon-
or item analysis in assessing the difficulty level strated in the literature.
and cultural appropriateness of each item were
reported by Kahnert et al. (1998), Roberts et al. Reporting of Educational Levels
(2002), and Rosselli et al. (2000). Given a strong association between education
and BNT scores, information regarding edu-
cational level should be reported for each sub-
group, and preferably normative data should
METHOD FOR EVALUATING 11-IE
be presented by educational levels.
NORMATIVE REPORTS
The normative reports reviewed were. limited Reporting of Intellectual Levels
to those employing the standard English Given the strong evidence of a relationship
60-item version. ; between BNT performance and IQ, especially
To adequately evaluate the BNT n~ative verbal IQ, information regarding intellectual
reports, six key criterion variablea were level!Vocabuhuy score should be reported for
deemed critical. The first five of these ;elate to each subgroup, and preferably normative data
subject variables, and the remaining refer to should be presented by IQ levels.
procedural variables.
Minimal requirements for meeting the cri-
terion variables were as follows. · Procedural Variables
Data Reporting
Subject Variables To facilitate interpretation of the data, group
means, standard deviations, and ranges for the
Sample Size total score achieved on the BNT should be pre-
Fifty cases are considered a desirable :sample sented at a minimum. However, SDs should
size. Although this criterion is somewhat arbi- be used with caution in evaluating the relative
trary, a large number of studies suggest that standing of an individual score because BNT
data based on small sample sizes arq highly scores are not normally distributed. More
influenced by individual differences Fd do detailed reporting of statistics for the number
not provide a reliable estimate of tile pop- of correct responses produced (1) sponta-
ulation mean. · neously, (2) with stimulus cues, (3) with pho-
nemic cues, as well as (4) the number of
Sample Composition Description missed items, is recommended.
As discussed previously, information regard-
ing medical and psychiatric exclusion criteria
is important. In addition, emerging litErature
indicates that ethnicity, acculturation, and SUMMARY OF THE STATUS
native language impact test perfonnance; OF THE NORMS
therefore, it is preferable that this infoljmation According to a survey of the participants of
be reported for the normative samples. It is the 1988 and 1989 Clinical Aphasiology Con-
unclear if gender, geographic rec~tment ference, the BNT was identified as one of the
region, socioeconomic status, occupaticpn, and two tests most frequently used to supplement
recruitment procedures are relevant, Until comprehensive aphasia batteries (Jackson &
this is determined, it is best that thii infor- Tompkins, 1991). Similarly, the BNT is used
mation be provided. in many studies to explore the efficiency of
confrontational naming in normal and clinical
Age Group Intervals samples across various demographic groups
This criterion refers to grouping of the data and diagnostic categories. These studies vary
into limited age intervals. This requiretnent is from several perspectives: (1) versions of the
relevant for this test since a strong elect of BNT utilized (experimental 85-item version,
standard 60-item version, as well as a variety of Appendix 10. Table A10.1, the locator table,
short versions; see above), with the second edi- summarizes information provided in the
tion of the test being published more recently, studies described in this chapter. 1
which will undoubtedly attract the attention
of clinicians and researchers; (2) administra-
tion procedures, particularly in respect to
SUMMARIES Of THE STUDIES
provisions for the stimulus cues; and (3) aspects
of performance reported (total score and/or [BNT.1l Van Gorp, Satz, Kiersch, and Henry,
error analysis [error classification systems also 1986 (Table A10.2)
differ between studies], percent of correct The article provides normative data on BNT for
responses per item, and recommended cutoff 78 normal, independently living elderly residing
criteria for impaired performance rating). in southern California (29 males, 49 females)
Russell and Starkey (1993) developed the aged 59-95 with a mean FSIQ of 122 (range
Halstead-Russell Neuropsychological Evalua- 87-150). Participants were screened for neu-
tion System (HRNES), which includes the rological disorders based on their self-report.
BNT among 22 tests. In this system, individual The data are presented in five age groupings.
performance is compared to that of 576 brain- The correlation of BNT scores with age was
damaged participants and 200 participants who r = -0.33, with more variability demonstrated
were initially suspected of having brain damage by older groups. The authors provide suggested
but had negative neurological findings. Data cutoff criteria for impaired performance, which
were partitioned into seven age groups and are based on a score falling over 2 SDs below
three educational/IQ levels. The authors pub- the mean for the respective age group.
lished an appendix to the manual (HRNES-R;
Russell & Starkey, 2001), which contains tables Study strengths
of scale scores based on the original HRNES 1. Demographic characteristics of the
norms, demographic corrections, and regression- sample are well described in terms of
based predicted scores. These data will not be age, education, IQ, gender, and geo-
reviewed in this chapter because the "normal" graphic area.
group consisted of Veterans Administration 2. Adequate exclusion criteria were used.
patients who presented with symptoms re- 3. The data are partitioned into five age
quiring neuropsychological evaluation. (For groups.
further discussion of the HRNES system, see 4. The authors provide suggested cutoff
Lezak et al., 2004, pp. 676--677). criteria.
Demographically adjusted BNT norms 5. Means and SDs for the test scores are
based on a sample of over 1,000 normal adults reported.
between 20 and 85 years of age, stratified by
age, education, gender, and race/ethnicity Considerations regarding use of the study
(African American and Caucasian), are pre- 1. SDs for the IQ indices describing the
sented by Heaton et al. (2004). whole sample and the individual age
Among all the studies available in the lit- groups are not provided.
erature, we selected for review those based on 2. Sample sizes for age groups are small.
well-defined samples. Only studies using the 3. Mean education and intelligence levels
60-item version were reviewed. In all articles are high.
reviewed below, the score represents the total
number of correctly named drawings (spon- [BNT.2J Farmer, 1990 (Table A10.3)
taneously or with a stimulus cue) out of 60.
The author provides data on the BNT for 125
In this chapter, normative publications and
normal male participants aged 20-69 (M =43.9,
control data from clinical studies are reviewed
in ascending chronological order. The text of
'Nonns for children for the BNT and BNT-Second Edi-
study descriptions contains references to the tion are available in Baron (2004), Kaplan et al. (2000),
corresponding tables identified by number in and Spreen and Strauss (1998).
184 LANGUAGE
SD = 14.3) recruited in California. Education procedures, and Ruency in English is

ranged 8-22 years (M = 14.6, SD = 2.2). All reported.
participants were native English speakers and 4. Means and SDs for the test scores are
had vision and hearing within normal limits. reported.
None of the participants reported a history of 5. Good exclusion criteria.
brain injury or disease. The BNT was admin-
istered according to standard instructions. Considerations regarding use of the study
Data are presented for five age decades, 1. Education and intelligence levels of the
with 25 participants in each age group. Anal- samples are high.
ysis of errors is discussed. According to the 2. Undifferentiated age range.
results, age was significantly correlated with
BNT but educational level was not. [BNT.4] Mitrushina and Satz, 1995
(Tables A10.5, A10.6)
Study strengths The article provides BNT data based on a
1. Demographic characteristics of the sample of neurologically intact, highly func-
sample are described in terms of age, tioning, independently living participants
education, gender, geographic area, and residing in southern California, who were
Ruency in English. tested over three longitudinal annual probes.
2. The data are partitioned into five age The sample of 156 participants who partici-
groups. pated in the first probe included most of the
3. Minimally adequate exclusion criteria. sample of 78 participants described by Van
4. Means and SDs are reported. Gorp et al. (1986) [BNT.1]. Due to attrition
over a period of 3 years, only 122 participants
Considerations regarding use of the study participated in all three probes. Participants in
1. Mean educational level is high. this sample (49 males, 73 females) had Mini-
2. Individual cell sizes are relatively small. Mental State Exam (MMSE) scores >24 and
3. The sample is all male. ranged in age from 57 to 85 years, with a mean
4. No information regarding IQ level is age of 70.4 (5.0) years, at the first testing
provided. probe. Mean education was 14.1 (2.7) years,
and mean FSIQ was ll8.2 (13.0). The sample
[BNT.3] Boone, Lesser, Miller, Wohl, Berman, was partitioned into four age groups, which
Lee, Palmer, and Back, 1995 (Table A10.4) did not differ in level of education.
The authors compared 73 outpatient, de- Participants were screened for a history of
pressed elderly and no controls on a battery neurological or psychiatric disorder. All par-
of neuropsychological tests. All participants ticipants were native English speakers. The
were Ruent English speakers over age 45 re- BNT was administered according to standard
siding in southern California, who were re- instructions as part of a large neuropsycholo-
cruited through newspaper ads. Both clinical gical battery.
and control groups underwent physical and Some decline in scores after age 70 was
neurological examinations and psychiatric in- apparent from cross-sectional age group com-
terviews. Strict exclusion criteria were used. parisons. The pattern of correlations with
Data for the control group are reproduced in various neuropsychological measures suggests
Table A10.4. a predominantly verbal mode of information
processing in BNT performance on the first
Study strengths probe, as opposed to a visuospatial mode by
1. The overall sample sizes are large. the third probe.
2. A comparison of normal controls with A comparison of BNT scores across the
depressed elderly is provided. three probes revealed adequate stability of
3. Information regarding VIQ, age, educa- scores over time, with test-retest correlations
tion, gender, geographic area, recruitment ranging r = 0.62-0.89.
Study strengths Data across wide ranges of different de-

1. Infonnation regarding age, education, mographic characteristics are presented.
gender, geographic area, IQ, and fluency 2. Strict selection criteria were used for
in English is reported. neurological disorders and cognitive
2. Adequate exclusion criteria were used. dysfunction.
3. The data are partitioned into four age 3. Overall very large sample size.
groups. 4. The data are presented in an age-by-
4. Test-retest data are provided. education-by-living environment matrix.
5. Overall sample size is large, with some 5. Means and SDs for the test scores are
cells approaching 50 while some cells reported.
being rather small.
6. Means and SDs for the test scores are Considerations regarding use of the study
reported. 1. No infonnation regarding intellectual
level.
Consideration regarding use of the study 2. Sample sizes in individual cells are small.
1. Mean education and intelligence levels 3. The administration procedure somewhat
are high. differed from standard instructions.
[BNT.S] Neils, Baris, Carter, Dell'aira, Nordloh, [8NT.6] Ross, Lichtenberg, and Christensen,
Weiler, and Weisiger, 1995 (Table A10.7) 1995 (Table A10.8)
The study addresses the effects of demo- This article represents an expansion on the
graphic factors on BNT perfonnance. Parti- previously reported data in Lichtenberg et al.
cipants were 323 nonnal elderly (244 females, (1994).
79 males) aged 65-97 residing in northern In study 1, the authors provide data for 123
Kentucky and the greater Cincinnati, Ohio, geriatric medical inpatients at an urban reha-
area; 167 participants were living indepen- bilitation hospital in Michigan (60% African
dently and 156 were institutionalized in American, 40% Caucasian, 62% female, 38%
extended-care facilities for at least 1 month. male). Mean age was 75.87 (7.42), with mean
All participants were carefully screened for education of 11.05 (3.38). Rigorous exclu-
neurological disorders and had adequate vision criteria for neurological disorders and
sion, language comprehension, and attention. depression were used. Mean Mattis Dementia
The administration procedure differed Rating Scale (DRS) score for the sample was
from standard in that the stimulus cues were 132.76 (4.93). Patients treated for hyperten-
offered after any error was made, irrespective of sion, diabetes, and hypothyroidism were in-
whether it was a visual-perceptual error. cluded if their conditions were well controlled
The data are presented in an age-by- with medications and without neurologi-
education-by-living environment matrix. cal complication. Some participants were tes-
The combination of age, education, and liv- ted 2-3 weeks after orthopedic surgery and
ing environment accounted for 32% of the were not on narcotic medications at the time
perfonnance variance. The results suggest that of assessment.
scores for low-education and high-education In study 2, participants from study 1 were
groups are less affected by age and living envi- compared as a "nonnative" group to a "cog-
ronment than scores for participants with 10-12 nitively impaired" group of 151 participants
years of education. Correlation between BNT with Mattis DRS scores below 123 (61% Af-
score and education was r=0.38, whereas the rican American, 39% Caucasian, 30% male,
correlation of BNT with age was r = -0.33. 70% female). Mean age for this group was
79.7, with mean education of 8.9 years. Par-
Study strengths ticipants from this group presented with a
1. Infonnation regarding age, education, wide variety of physical disorders which are
gender, and geographic area is provided. likely to affect cognitive status. Twenty-four
186 LANGUAGE
percent of these participants had scores above The BNT was administered according to
10 on the Geriatric Depression Scale (GDS). standard instructions, followed by a trial of
The results of study 1 indicated significant seven alternative items as potential substitutes
correlations of BNT scores with age, educa- for low-frequency original items. In addition
tion, and ethnicity (-0.308, 0.375, and 0.326, to standard scoring, an analysis of errors was
respectively). The combined effects of demo- conducted according to current systems (e.g.,
graphic variables accounted for 21% of the Nicholas et al., 1989).
BNT variance. The results revealed that the mean BNT
In study 2, a discriminant function analysis score was 2-5 points below that reported for
based on the BNT and demographic data North American samples. Interrater reliabil-
discriminated between cognitively intact and ities for the total score and for error scoring
impaired participants with an accuracy of were high (94.89% and 98.17% agreement,
72.75% (sensitivity 63%, specificity 80%). respectively). Age, education, visual acuity,
The authors underscore the importance of and backward digit span were signi&cantly
using a demographically appropriate set of related to BNT scores (r=0.23-0.33). The
normative data and suggest use of their data in analysis of errors indicated that semantically
urban medical settings. related errors and "don't lmow" responses
were most frequent.
Study strengths The authors emphasized an effect of culture-
1. Means and SDs for the test scores are related word frequency on BNT performance.
reported. The proposed alternate items for "beaver" and
2. Data are presented by age group. "pretzel" were "platypus" and "pizza."
3. A comparison of BNT performance for The longitudinal follow-up data for 91 par-
clinical and medical control groups is ticipants from this sample are reported in
presented. Cruice et al. (2000).
ethnicity, gender, and geographic area is Study strengths
reported. 1. Minimally adequate exclusion criteria
5. Individual cell sizes approach 50. are reported.
2. Data are presented by age group.
3. Authors recommend cutoff scores.
Considerations regarding use of the study 4. Analysis of errors was performed.
1. "Normal" participants were geriatric 5. Information regarding age, gender, geo-
inpatients, many of whom had physical graphic area, and recruitment proce-
illnesses potentially affecting cognitive dures is reported.
status.
2. The age range for the oldest group is not Considerations regarding use of the study
reported. 1. Education and intellectual level are not
3. No information on intellectual level. reported.
2. Sample sizes for most of the age groups
[BNT.7] Worrall, Yiu, Hickson, and Bamett, are small.
1995 (Table A10.9) 3. Participants were recruited in Australia,
The authors assessed the validity of the BNT and it is unclear if these norms are suitable
as part of a large educational project on 136 for clinical interpretation in the United
independently living older Australians. Parti- States given that this sample scored 2-5
cipants were a recruited through advertise- points below North American samples.
ments. Participants with a reported history of
neurological disease and non-native English [BNT.8] Lafleche and Albert, 1995
speakers were excluded. The mean age for the (Table AlO.lO)
sample was 70.43 (SD = 7.8) years, and 74.3% The BNT was administered to 20 volunteers
were female. who comprised a control group in a study on
executive function deficits in mUd AD. The percenttle ranks. The authors provided tables
control group included nine men and 11 of age-corrected norms for each age group.
women, with a mean age of 76.2 years, mean The procedure for clinical application of these
education ofl4.7 years, and mean MMSE score data is described in the original article (Ivnik
of 29.4 (0.8). Participants were screened for et al., 1996) as follows:
severe head injury, alcoholism, major psychiat-
first select the table that corresponds to that per-
ric illness, epilepsy, and learning disabtlities.
son's age. Enter the table with the test's raw score;
They did not show evidence of a dementing
do not use "corrected" or "final" scores for tests
process, either on testing or by history.
that might present their own age- or education-
adjustments. Select the appropriate column in the
Study strengths table for that test. The corresponding row in
the left-most column in each table provides the
MOANS Age-Corrected scaled score . . . for
reported.
your subject's raw score; the corresponding row in
Considerations regarding use of the study the right-most column indicates the percentile
1. The sample is small. range for that same score.
2. SDs and ranges for age and education Further, linear regressions should be ap-
are not provided. plied to the normalized, age-corrected
3. Recruitment procedures are not re- MOANS scaled scores (A-MSS) derived from
ported. the tables, to adjust patient scores for educa-
4. Education level for the sample is high. tion. Age- and education-corrected scores for
5. No information on IQ is reported. the BNT (A&E-MSS) can be calculated as
follows:
[BNT.9] lvnik, Malec, Smith, Tangalos, and
Petersen, 1996 (Table A10.11)
A&E-MSSsNT = K+(W, •A-MSSsNT)
The study provides age-specific norms for the - (W2 *Education)
BNT obtained in Mayo's Older Americans
Normative Studies (MOANS), which produce
normative data for elderly individuals on dif- where the following indices are specified for
ferent neuropsychological tests. The total the BNT:
sample consisted of 746 cognitively normal K 3.32
volunteers residing in Minnesota, over age 55, w. 1.07
663 of whom took the BNT. Mean MAYO w2 o.34
FSIQ (which differs somewhat from standard
WAIS-R FSIQ) for the whole sample was Education should enter the formula as years
106.2 (14.0), and mean Mayo General Mem- of formal schooling.
ory Index on the Wechsler Memory Scale- The tables of scaled scores per age group
Revised (WMS-R) was 106.2 (14.2). For a provided by the authors should be used in the
description of their samples, the authors refer context of the detailed procedures for their
to their earlier publications. Participants were application, which are explained in Ivnik et al.
independently functioning, community- (1996). Therefore, they are not reproduced in
dwelling persons who were recently examined this book. Interested readers are referred to
by a physician and had no active neurological the original article. Table AIO.ll summarizes
or psychiatric disorder with the potential to sample sizes for different demographic
impact cognition. groups.
Age categorization utilized the midpoint
inteiVal technique. The raw score distribution Study strengths
for each test at each midpoint age was "nor- 1. Information regarding age, education,
malized" by assigning standard scores with a IQ, gender, ethnicity, handedness, and
mean of 10 and SD of 3, based on actual geographic area is reported.
188 LANGUAGE
2. The data are stratified by age group exclusion criteria were employed. Participants
based on the midpoint inte~ tech- with well-controlled hypertension or who had
nique. adequate corrected vision were included.
3. The innovative scoring system is well The data were presented for five age groups
described. The authors developed new and then further stratified into five age groups
indices of performance. by two educational levels and into five age
4. The sample sizes for each group are groups for males and females separately. The
large. table for five age groups includes suggested
5. Restricted age range in each cell. cutoff scores.
The results indicated that the interaction of
Considerations regarding use of the ~dy age and education is a better predictor of
1. The measures proposed by the ~uthors BNT performance than age alone. Perfor-
are quite complicated and might be dif- mance variability was higher in the older age
ficult to use in clinical practice. • and lower education groups. In the ~12th
2. Participants with prior history ol neuro- grade education group, BNT performance
logical, psychiatric, or chronic ptedical remained stable until 80 years, while in
illnesses were included. the <12 years of education group the decre-
ment was evident at 70 years. Gender differ-
Other comments ences were also reported, with males
1. The theoretical assumptions un4erlying outperforming females.
this normative project have lJe;n pre-
sented in Ivnik et al. (1992a,b). Study strengths
2. The authors cautioned that the )validity 1. Information regarding age, education,
of the MAYO indices depends he,vily on gender, ethnicity, occupation, recruit-
the match of demographic featUres of ment procedures, and geographic area is
the individual to the normative :Sample reported. The sample is representative
presented in this article. of the regional population along most
3. Correlation of the BNT with age was demographic parameters.
-0.46, whereas correlations with edu- 2. Strict exclusion criteria were used.
cation and gender were 0.26 and .-0.19, 3. Data are presented by age group and by
respectively. · different combinations of demographic
variables (education, gender).
[BNT.10] Welch, Doineau, Johnson, and king, 4. Authors recommend cutoff scores.
1996 (Tables A10.12-A10.14) 5. Means and SDs for the test scores are
The study provides data on BNT perfofiJlance reported.
for 176 normal older adults from middle
Tennessee (74 males, 102 females), ranging in Considerations regarding use of the study
age from 60 to 93, with a mean age of 74 years. 1. Sample sizes for some of the age groups
Education ranged from third grade to lf years, are small.
with a mean of 12.28 years. The sample con- 2. No information regarding IQ is provided.
sisted of 61% urban and 39% rural partici-
pants; 29% professional, 28% skilled, m;d 43% [BNT.11] Hoff, Riordan, Morris, Cestaro,
labor workers; 71% white, 28% ~rican Wieneke, Alpert, Wang, and Volkow,
American, and 1% other. Participan~ were 1996 (Table A10.15)
recruited mostly from senior-citizen organiza- The authors used the BNT in a study explor-
tions and retirement centers, to ensure tample ing the relationship of cocaine use to perfor-
representation approximating the gene¥ pop- mance on neuropsychological tests tapping
ulation for the following parameters: yarious functions of frontal and temporal brain re-
occupational levels (skilled, professio~al, or gions. Performance of crack cocaine users was
manual labor), race and living characf:J'ristics compared to that of the control group, which
(urban vs. rural). Strict medical and psyruatric consisted of 54 paid male volunteers with
a mean age of 32.1 (9.7) years and mean ed- between Marin and Marin (1991) accultura-
ucation of 15.4 (2.4) years. The sample in- tion scale scores and neuropsychological
cluded 48 white, 4 black, and 2 Hispanic variables are provided.
participants. Exclusion criteria were a history The Ponton-Satz BNT is an adaptation of
of medical, neurological, or psychiatric pro- Kaplan's BNT, consisting of 30 items that are
blems; more than moderate use of alcohol (12 based on the original test but presented in
oz/week), history of intravenous drug use, and different order. The selection of items was
self-reported history of learning disability based on the ratings of expert judges in terms
(with enrollment in special education classes). of cultural appropriateness and difficulty. In
the follow-up study on the factor structure of
Study strengths the NeSBHIS (Ponton et al., 2000), which
1. Relatively large sample size. extracted five factors, the BNT primarily loa-
2. The sample composition is described in ded on the Language factor, with a varimax-
terms of age, education, and ethnicity. rotated factor loading of 0.84.
4. Means and SDs for the test scores are Study strengths
reported. 1. Large overall sample with acceptable
size for most of the cells.
Considerations regarding use of the study 2. The sample composition is well described
1. Wide age and education range. No in- in terms of age, education, gender, ac-
formation on IQ or gender distribution. culturation information, geographic area,
2. Recruitment procedures are not reported. and recruitment procedures.
3. Educational level for the sample is high. 3. Adequate exclusion criteria.
[BNT.12] Ponton, Satz, Herrera, Ortiz, Urrutia,
Young, D'Eiia, Furst, and Namerow, 1996
specified.
(Table A10.16)
reported.
The Ponton-Satz BNT was administered to 6. Data are partitioned by gender x age x
Spanish-speaking volunteers as part of a larger education.
battery in a project designed to provide a
standardization of the Neuropsychological
Screening Battery for Hispanics (NeSBHIS). Considerations regarding use of the study
Volunteers were recruited through fliers and 1. Thirty-item version of the test adapted
advertisements in community centers of the for use with Spanish-speaking pop-
greater Los Angeles area over a period of 2 ulation was administered.
years. Exclusion criteria were a history of 2. No information on IQ is reported.
neurological or psychiatric disorder, drug or 3. It is unclear which of the two educa-
alcohol abuse, and head trauma. Data for a tional groups included participants with
sample of 300 participants, with a median 10 years of education.
educational level of 10 years, were analyzed.
Participants ranged in age from 16 to 75 years, [BNT.13] Tombaugh and Hubley, 1997
with a mean of 38.4 (13.5) years. Education (Tables A10.17, A10.18)
ranged 1-20 years, with a mean of 10.7 (5.1) The study provides age- and education-
years. Male to female ratio was 40/60%. The stratified norms for 219 community-dwelling,
average duration of residence in the United cognitively intact volunteers, who participated
States was 16.4 (14.4) years. Seventy percent in a large study on the effect of aging on ac-
of the sample were monolingual Spanish- quisition and retention of information. They
speaking, and 30% were bilingual. The pro- were recruited through booths at shopping
portion of the sample respective to their centers, social organizations, places of em-
country of origin closely approximates the ployment, psychology classes, and word of
1992 U.S. Census distribution. Correlations mouth. The sample included participants aged
190 LANGUAGE
25-88 years (M = 59.0, SD = 16.9). Average [8NT.14] Henderson, Frank, Pigatt, Abramson,
educational level was 12.9 (2.3) years; 46% and Houston, 1998 (Table A10.19)
were male. Mean WAIS-R Vocabulary scaled The authors examined effects of race, gender,
score was 11.6 (2.4). Participants were and educational level on BNT scores. The
screened based on a self-reported history of sample included 50 African-American and 50
medical and psychiatric problems, including a Caucasian participants, with 25 males and fe-
list of currently prescribed medications. Per- males in each group, who ranged in age from
sons with a known history of neurological 17 to 87 years and in education from 10 to
disease, psychiatric illness, head injury, or >17 years. Volunteers from local churches in
stroke were excluded. Participants with Richland and Florence counties in South
MMSE score <25 and GDS score >13 were Carolina; students, faculty, and staff from the
also excluded. University of South Carolina (USC); and
Participants were administered all items for participants in a lexical function study at the
20 seconds, starting with item 1. Specific admin- USC were included in the sample. Exclusion
istration procedures are described by the au- criteria were a history of mental retardation,
thors in detail. The standard scoring procedure dementia or developmental language disorders,
was used. The authors recorded rates of cor- traumatic brain injury, cerebrovascular acci-
rect spontaneous responses (SR), number dent, treatment for alcoholism, or current
correct after a stimulus cue (SC), and number psychiatric illness including depression. Par-
correct after a phonemic cue (PC). Rates of ticipants with scores above 3 on the Hachinski
SR + SC (according to the original scoring Ischemia Rating Scale, above 0.5 on the Zung
procedure) and the sum of all correct re- Depression Scale, and <130 on the DRS were
sponses (SR + SC +PC) are provided. excluded from the study.
The test was administered by trained staff
Study strengths who followed the standard administration pro-
1. Administration procedure is well out- cedure. However, all60 items were presented.
lined. Credit was given for spontaneously correct
2. Sample composition is well described in responses and for correct responses with a
terms of age, education, gender, WAIS-R stimulus cue. Participants and examiners ad-
Vocabulary score, and recruitment pro- ministering the test were matched by race.
cedures. The results revealed a significant effect of
3. Strict selection criteria were used. education on BNT scores but no significant
4. The table of means is stratified by age effect of race or gender. An item analysis was
group, education, and gender; the table used to examine the vocabulary base under-
of percentiles is presented in age x lying performance on BNT items.
education cells.
5. Means, SDs, and percentiles for the test Study strengths
scores are reported. 1. Large sample.
Considerations regarding use of the study scribed in terms of age, education, gen-
1. Administration procedure deviated from der, and geographic area.
the standard procedure described in the 3. Rigorous exclusion criteria.
test manual. 4. Test administration procedures are
2. Sample sizes for age groupings are rela- specified.
tively small, although overall sample size 5. Means and SDs for the test scores are
is quite large. reported.
3. Data were collected in Canada and,
therefore, might be of limited use for Considerations regarding use of the study
clinical interpretation in the United 1. Overall sample is adequate, but individ-
States. ual cells are relatively small.
2. Demographic characteristics of the post-baccalaureate training, with a mean of 14.1

sample are provided in different combi- (2.4) years. MMSE scores ranged 24--30, with a
nations of subgroups. However, they do mean of 28.1 (1.5). The sample was predomi-
not match the breakdown of the sample nantly Caucasian. English was the primary
used for normative data reporting. language for all participants. They were finan-
3. Recruitment procedures are not re- cially compensated for participation. Individ-
ported. uals with a history of cerebrovascular insult,
4. No information on IQ is reported. head injury with loss of consciousness exceed-
ing 5 minutes, and chronic substance abuse,
[BNT.15] Stuss, Alexander, Hamer, Palumbo, according to the structured interview, were
Dempster, Binns, Levine, and lzulcawa, 1998 excluded. All participants lived independently
(Table A10.20) in the community and denied uncorrected vi-
The study addresses the effect of brain lesion sion, hearing, or motor impairment.
location and etiology on verbal fluency. The The BNT items were reorganized according
BNT, among other tests, was administered to to the order of items in the short forms which
assess basic language competence. The control were validated in this study. Test was admin-
group included 37 participants (19 males, 18 istered by trained graduate students. The data
females) without neurological or psychiatric are stratified into three age groups, with equal
disorder, with a mean age of 54.4 (14.4) years numbers of males and females in each group.
and mean education of 13.9 (2.3) years. Mean The authors reported an age effect on
National Adult Reading Test (NART)- BNT performance but no education or gender
estimated IQ was 113.8 (6.1). effects.

1. The sample composition is well de- 1. Large sample.
scribed in terms of age, education, gen- 2. The sample composition is well de-
der, and estimated IQ. scribed in terms of age, education, gen-
2. Adequate exclusion criteria. der, ethnicity, MMSE scores, primary
3. Means and SDs for the test scores are language, incentive for participation,
reported. sampling procedure, geographic area,
and recruitment procedures.
Considerations regarding use of the study 3. Rigorous exclusion criteria.
1. Recruitment procedures are not re- 4. Detailed description of test administra-
ported. tion procedure and personnel is provided.
2. The data were obtained on Canadian 5. Means and SDs for the test scores are
participants, which may limit their use- reported.
fulness for clinical interpretation in the 6. Data are partitioned into three age
United States. groups.
[BNT.16] Fastenau, Denburg, and Mauer, Considerations regarding use of the study
1998 (Table A10.21) 1. Administration was not standard: test
The authors provided data for the full BNT in items were reorganized.
the context of a study addressing the validity of 2. Education and intelligence level for the
short forms of the BNT. Participants were re- sample are high.
cruited through newspaper advertisements and 3. No information on IQ is reported.
at local churches and older adult organizations
in urban areas. A stratified sampling procedure [BNT.17] Ross and Uchtenberg, 1998
was used to produce a sample of 108 partici- (Table A10.22)
pants (47% female) with a mean age of 72.2 Normative data for 233 neurologically in-
(7.0) years. Education ranged from 8 years to tact medical patients are provided for use in
192 LANGUAGE
urban medical settings. This study represents 4. Means and SDs for the test scores are
an extension of the earlier study by Ross et al. reported.
(1995) and is part of a continuing geriatric 5. Data are stratified by age and education.
research protocol, the Normative Studies
Research Project (NSRP) for older urban Considerations regarding use of the study
adults. Participants were tested during inpa- 1. Overall sample is large, but individual
tient postacute rehabilitative services at an cells are relatively small.
urban university-affiliated hospital. All con- 2. No information on IQ is reported.
secutively admitted patients were tested ap-
proximately 1 week after their admission and [8NT.18] Kohnert, Hernandez, and Bates, 1998
2-3 weeks prior to their hospital discharge. (Table A10.23)
The medical problems included arthritic con- The BNT was administered to 100 bilingual
ditions, fractures, limb amputations, back in- adults of Mexican-American descent in both
juries, spinal cord injuries, and admissions Spanish and English. Participants had > 12
for physical reconditioning following illness or years of education and were recruited from
surgery. Patients were on a variety of medica- University of California San Diego, University
tions. The sample included 56% African- of California Santa Barbara, and the San
American and 44% Caucasian participants, Diego community. Spanish was the first lan-
73% of whom were female, ranging in age guage and the primary home language for all
from 65 to 95. Their mean age was 76.1 (7.1) participants, with English acquisition prior to
years, and mean education was 11.1 (3.2) the age of 8. Proficiency in both languages was
years. Selection criteria were normal neuro- assessed with a self-rating questionnaire.
logical examination upon hospital admission; Participants were given course credit or paid
absence of medical history suggesting past $5 for their participation. Exclusion criteria
neurological dysfunction, psychiatric illness, were left-handedness; a history of speech,
or alcohol abuse; a score above 123 on the language, or hearing impairment; uncorrected
DRS; and a GDS score <10. Participants visual deficits; proficiency or prolonged expo-
with current hypertension, diabetes, and hy- sure to languages other than Spanish and
pothyroidism were included only when these English; or a medical history potentially re-
conditions were well controlled with medica- sulting in compromised neurological status
tions and did not cause neurological compli- (based on responses on the health question-
cations. naire). The sample included 41 males and
Normative data are stratified by age and 59 females, with a mean age of 20.82 (2.6)
education. years, mean education of 14.4 (1.7) years, and
Results of regression analysis suggest that mean age of English acquisition of 4.6 (3.0)
education accounts for approximately 14% of years.
the variance. Age, ethnicity, and gender each The BNT was administered two times to
account for an additional2% of the variance in each participant in both languages in the
test performance. The authors argue that the counterbalanced order. Instructions were gi-
selection of the normative data for clinical use ven in the language of test administration. All
should reflect the degree to which the demo- 60 pictures were shown to all participants, with
graphic characteristics of the normative sam- no basal or ceiling values set. Spanish protocol
ples match those of their patients. was developed based on the original version of
the test. The test was administered by trained
Study strengths personnel. Total correct scores were defined as
1. Large overall sample. spontaneously correct responses plus those
2. The sample composition is well de- aided with semantic cues. Three group per-
scribed in terms of age, education, gen- formance scores were derived: English only,
der, ethnicity, medical condition, and Spanish only, and a composite score indicating
recruitment procedures. the total number of items correctly named,
3. Rigorous exclusion criteria. independent of language.

1. Large sample. 1. Information regarding age, education,
2. The sample composition is well described IQ, gender, ethnicity, handedness, and
in terms of age, education, gender, geographic area is reported in the origi-
factors inHuencing language profi- nal articles based on this study (Ivnik
ciency, incentives for participation, and et al., 1992a,b).
setting. 2. The data were stratified by age group
3. Adequate exclusion criteria. based on the overlapping midpoints
4. Test administration procedures are technique.
specified in detail. 3. The sample sizes for each group are
5. Means and SDs for the test scores are large.
reported.
6. Comparison of performance in Spanish Considerations regarding use of the study
and English is provided. 1. To interpret the data presented in age
groups broken down by the overlapping
Consideration regarding use of the study midpoints technique, the reader is re-
1. No information on IQ is reported. ferred to the original articles by Ivnik et al.
2. Participants with a prior history of psy-
[BNT.19] Randolph, Lansing, lvnik, Cullum, chiatric or chronic medical illnesses
and Hermann, 1999 (Table A10.24) were included.
The effects of age, education, gender, and
diagnostic group with respect to overall BNT [BNT.20] Killgore and Adams, 1999
performance; the inHuence of phonemic cu- (Table A10.25)
ing; and performance on individual items The study investigates the relationship be-
were examined on samples of neurologically tween BNT performance and WAIS-R Vo-
normal elderly, AD patients and temporal cabulary score and derives regression-based
lobe epilepsy patients. The control group in- expected BNT scores from Vocabulary scaled
cluded 719 paid and unpaid volunteers for scores. The sample consisted of patients con-
studies on neuropsychological function in secutively referred for neuropsychological
normal aging. Procedures for recruitment and evaluation at a large midwestern medical
sample description are provided in earlier center over a 26-month period who were
articles based on this study. The follow-up found to be without demonstrable neurolog-
articles provide additional information (Lan- ical impairment. All patients had negative
sing et al., 1999). The sample was almost ex- neurological evaluations and negative neuroi-
clusively white and 60% female, with a mean maging and laboratory studies. Patients were
age of73.6 (10.3) years and mean education of excluded if there was evidence of mild de-
13.4 (2.9) years. mentia or a history of alcohol abuse, learning
Spontaneously correct responses or re- disability, or seizure disorder. The sample
sponses correct with stimulus cue were scored consisted of 28 males and 34 females, ranging
as correct. The test was discontinued after six in age from 17 to 85 years, with a mean age of
consecutive failures. The authors present 45.7 (15.1) years, mean education of 13.1 (2.7)
means and SDs for the data broken down by years, and mean WAIS-R FSIQ of95.1 (12.0);
age groups using the overlapping midpoints 34% had a psychiatric diagnosis such as Major
technique. They also present data for three Depressive Episode or Adjustment Disorder.
education groups: <12, 12, and> 12 years. BNT was administered by trained person-
The authors found that age and education nel, according to the standardized instructions.
systematically inHuenced BNT scores. Gender The results did not suggest a relationship
had a significant effect across diagnostic between BNT performance and age. How-
groups, with males outperforming females, ever, the regression of Vocabulary scaled
which was interpreted by the authors as an scores on the BNT scores accounted for 42%
item-related effect. of the variance in BNT scores and was used to
194 LANGUAGE
derive predicted BNT scores for different 3. Rigorous exclusion criteria.

Vocabulary levels. 4. Means and SDs for the test scores are
reported.
Study strengths
1. The sample composition is well described Considerations regarding use of the study
in terms of age, education, gender, IQ, 1. T scores are stratified by demographic
geographic area, and recruitment proce- groups; however, the raw score is pre-
dures. sented for the entire sample.
2. Exclusion criteria are well defined. 2. No information on IQ is reported.
specified. [BNT.22] Rosselli, Ardila, Araujo, Weekes,
4. Means and SDs for the test scores are Caracciolo, Padilla, and Ostrosky-Solis,
reported. 2000 (Table A10.27)
The authors examined the impact of bilingual-
Considerations regarding use of the stfldy ism on verbal fluency and repetition tests in
1. Though the overall sample size is ade- older Hispanic bilinguals. The BNTwas used to
quate, it represents a wide age np1ge. assess naming proficiency. The sample included
2. The data are not partitioned by age 82 right-handed south Florida residents (28
group. male, 54 female) who volunteered to partici-
3. The data are based on a clinical ~ample pate. Participants ranged in age from 50 to 84,
of patients referred for neuropsfc:holo- with a mean age of 61.76 (9.3) years and mean
gical evaluation, although n~logical education of 14.8 (3.6) years. They indepen-
damage was ruled out by multipe mo- dently had MMSE scores >27, had Beck De-
dalities. pression Inventory-IT scores <5, and were
4. Patients with psychiatric diagnoses were screened for a history of neurological or psy-
included in the sample. chiatric problems using a structured interview.
Forty-five participants were monolingual in
[BNT.21] Heaton, Avitable, Grant, and English (born in the United States and only
Matthews, 1999 (Table A10.26) spoke English), 18 were Spanish monolinguals
The authors respond to the criticism of their (Latin American immigrants living in the city
regression-based norms by providing updated of Hialeah, Florida, who migrated to the
norms for the BNT, based on a santple of United States after age 50 and had been living
healthy community residents between 60 and in the United States an average of 5 years, had
80 years of age, who volunteered to participate no formal education in English or previous
as normal controls in studies of late-life psy- employment in which English was required,
chosis at the San Diego Veterans Adminis- and used Spanish in all daily activities), and 19
tration Medical Center. The exclusion eriteria were bilingual (as determined by their re-
were history of neurological illness, significant sponses on a bilingual questionnaire, self-
head trauma, recent or current substance use rated language proficiency in both languages,
disorder, major psychiatric illness, or ~temic and an acceptable score on the Spanish and
illness likely to affect central nervous $ystem English versions of the BNT norms, with a
function. correction for age).
The test was administered by trained tech-
nicians. Raw scores and demographically Study strengths
corrected T scores for the BNT are preSented. 1. The sample composition is well de-
Study strengths der, geographic area, and incentive for
1. Large sample. participation.
2. The sample composition is well deacribed 2. Adequate exclusion criteria.
in terms of age, education, gender,~c 3. Means and SDs for the test scores are
composition, and geographic area. : reported.
Considerations regarding use of the study 4. Means and SDs for the test scores are
1. Overall sample is adequate, but individ- reported.
ual cells are relatively small.
2. Recruitment procedures are not re- Considerations regarding use of the study
ported. 1. Overall sample is adequate, but individ-
3. Level of education for the monolingual ual cells are relatively small.
English group is high. 2. Recruitment procedures are not re-
4. No information on IQ is reported. ported.
3. SDs for age are not reported.
[8NT.23] Schmitter-Edgecombe, Vesneski, and 4. Educational levels for the two older
Jones, 2000 (Table A10.28) groups are high.
The study compared word-finding abilities
across three age groups. The sample included [BNT.24] Saxton, Ratcliff, Newman,
26 participants in each of three age groups: Belle, Fried, Yee, and Kuller, 2000
18-22 (M=18.93), 58-74 (M=66.29), and (Table A10.29)
75-93 (M = 79.19) years. Each group in- The BNT was administered as part of the
cluded 7 males and 19 females. The FSIQ of Memory and Aging Study (MAS), conducted
participants was estimated based on their as an ancillary project to the Cardiovascular
performance on the WAIS-R Vocabulary, Health Study (CHS), a multicenter observa-
Block Design, Similarities, and Arithmetic tional study of heart disease and stroke in
subtests. The young group was comprised Washington County, Maryland, and Pitts-
of undergraduate students, who received burgh, Pennsylvania. No selection criteria
course credit for their participation. The were used. Data were analyzed for a sample of
older groups included healthy, community- 989 participants (444 males, 545 females) who
dwelling volunteers. All participants were completed all of the cognitive tests included in
native English speakers. Exclusion criteria the battery. The mean age for the sample was
were a history of substance abuse, brain 73.63 (4.45) years, with mean education of
surgery, cerebrovascular or cardiovascular 13.23 (2.85) years; 93.9% of the sample were
accident, non-normative levels of cognitive white. This sample was divided into two clin-
decline, brain damage sustained earlier from ical groups and a "no disease" group, based on
a known cause, psychiatric disorder, or seri- cardiovascular status.
ous health problems, per self-report. In- Scores on the BNT for the "no disease"
dividuals taking more than two drugs rated to sample of 357 participants are reproduced in
have more than minimal effects on attention Table A10.29. Demographic characteristics
were excluded. Color-blind individuals or for this sample are not reported by the au-
those who could not see the cards were also thors. However, we assume that they are
excluded. similar to the demographics for the entire
The BNT was administered according to sample described above.
standard instructions.
An item analysis revealed a cohort effect Study strengths
(generational familiarity) in performance on 1. Large sample size.
four BNT items. 2. The sample composition is described in
terms of age, education, gender, setting,
Study strengths geographic area, and recruitment pro-
1. The sample composition is well de- cedures.
scribed in terms of age, education, gen- 3. Means and SDs for the test scores are
der, estimated FSIQ, and incentive for reported.
participation.
3. Test administration procedures are 1. No exclusion criteria.
specified. 2. The data are not partitioned by age group.
196 LANGUAGE
3. No information on IQ is reported. participants were proficient in English and

learned it as children. French participants had
[BNT.2S] Bell, Hermann, Woodard, Jones, lived in Canada. Participants had no history of
Rutecki, Sheth, Dow, and Seidenberg, communication problems, head injury, or
2001 (Table A10.30) drug or alcohol abuse.
The study examined object-naming ability and All 60 items of the BNT were administered.
depth of semantic knowledge in healthy con- Participants were not instructed to use a single
trols and patients with early-onset temporal word for each picture. Two types of score
lobe epilepsy. The control group included 29 were used: strict {responses listed in the BNT
friends, relatives, and spouses of temporal booklet) and lenient (allowing synonyms and
lobe epilepsy patients (72% female), aged 1~ variants for 18 items).
60 years, with a mean age of 34.4 (12.5) years; Performance of bilingual groups was sig-
FSIQ (as measured with the WAIS-III seven- nificantly lower than that of monolingual En-
subtest short form) of 69-110, with a mean glish speakers. Item difficulty differed across
FSIQ of 97.7 (6.4); and mean education of groups.
13.0 (1.7) years. Exclusion criteria were cur-
rent substance abuse, psychotropic medica- Study strengths
tion use, medical or psychiatric condition that 1. The sample composition is described in
could affect cognitive functioning, episode of terms of age and education.
loss of consciousness longer than 5 minutes, 2. Factors inHuencing language proficiency
developmental learning disorder, and repeti- are well described.
tion of a grade in school. 3. Adequate exclusion criteria.
All 60 BNT items were administered to all 4. Test administration procedures are
participants. Integrity of conceptual knowl- specified.
edge was assessed by asking participants to 5. Means and SDs for the test scores are
define a subset of items from the BNT. reported.

1. The sample composition is well de- 1. No information on gender and IQ is re-
scribed in terms of age, education, gen- ported.
der, FSIQ, and recruitment criteria. 2. Recruitment procedures are not re-
2. Adequate exclusion criteria. ported.
3. Test administration procedures are spec- 3. Educational level for the sample is
ified high.
4. Means and SDs for the test scores are 4. The data for French-speaking partici-
reported. pants were obtained on Canadian re-
sidents, which may limit their usefulness
Considerations regarding use of the study for clinical interpretation in the United
1. The sample is small and includes a wide States.
age range. 5. Administration procedures were altered.
group. [BNT.27] Coffey, Ratcliff, Saxton, Bryan, Fried,
and Lucke, 2001 (Table A10.32)
[BNT.26] Roberts, Garcia, Desrochers, The authors examined cognitive correlates of
and Hernandez, 2002 (Table A10.31) age-related physiological changes in brain
The authors compared BNT performance and structure, as evident on quantitative MRI.
order of difficulty for the individual items for Participants were 320 elderly {38% male, 62%
three language groups: 42 monolingual En- female) between 66 and 90 years of age, who
glish speakers, 32 Spanish/English bilinguals, were selected from Pittsburgh, Pennsylvania,
and 49 French/English bilinguals. All bilingual and Hagerstown, Maryland, centers of the
multicenter, population-based Cardiovascular self-report and per Schedule for Affective

Health Study. Data for a slightly larger sub- Disorders and Schizophrenia Interview,
sample from this project are presented in physical examination, and urinalysis. Mean
Saxton, 2000 (see review above). Volunteers age for the group was 25.2 (6.07) years, mean
were excluded from the study if they were not education 15.0 (1.48) years, mean WAIS-R IQ
right-handed, had a lifetime history of psy- 109.3 (11.51), and male/female ratio 21/10.
chiatric illness or any illness or injury re- The sample is further described in the articles
ferrable to the brain, MR images revealing by Bilder et al. and Lieberman et al., pub-
structural abnormalities, or MMSE scores lished between 1991 and 2000.
of <24. Mean age for the sample was 74.85
(4.95) years, mean education 12.98 (2.87) Study strengths
years, mean MMSE score 28.29 (1.5), and 1. The sample composition is well de-
mean WAIS-R Vocabulary score 47.52 scribed in terms of age, education, gen-
(13.26). Of this sample, 71% were taking der, FSIQ, geographic area, and
medications for one or more medical condi- recruitment procedures.
tions. None was taking medication known to 2. Stringent exclusion criteria.
affect brain size (e.g., steroids). 3. Means and SDs for the test scores are
The BNT was administered according to reported.
standard instructions. Data are presented for
the whole sample and for males and females Considerations regarding use of the study
separately. 1. The sample is small.
2. Educational level for the sample is high.
Study strengths
RESULTS OF THE META-ANALYSES OF
THE BOSTON NAMING TEST DATA
der, MMSE score, WAIS-R Vocabulary
(See Appendix 1OM)
score, geographic area, and research
setting. Data collected from the studies reviewed in
3. Adequate exclusion criteria. this chapter were combined in regression
4. Test administration procedures are spec- analyses in order to describe the relationship
ified. between age and test performance and to
5. Means and SDs for the test scores are predict expected test scores for different age
reported. groups. Effects of other demographic vari-
ables were explored in follow-up analyses. The
Consideration regarding use of the study general procedures for data selection and
1. The data are not partitioned by age analysis are described in Chapter 3. Detailed
group. results of the meta-analysis and predicted test
scores across adult age groups are provided in
[BNT.28] Giovannetti, Goldstein, Schullery, Appendix lOrn.
Barr, and Bilder, 2003 (Table A10.33) After initial data editing for consistency and
The BNT was administered to 31 control for outlying scores, 14 studies, which gener-
participants in order to assess basic language ated 42 data points based on a total of
skills associated with temporal lobe functions 1,684 participants, were included into the
in a study on the mechanisms of verbal Hu- analyses. Normative data from Kaplan's stan-
ency deficits in first-episode schizophrenia. dardization sample were not included in the
Participants were recruited from the hospital database.
community through announcements in local A quadratic regression of the BNT scores
newspapers and within the medical center. on age yielded an R2 of 0.850, indicating that
They had no history of substance abuse or 85% of the variance in BNT scores is ac-
neurological/psychiatric/medical illness, per counted for by the model. Based on this
198 LANGUAGE
model, we estimated BNT scores for ;age in- sample support the curvilinear relation-
tervals between 25 and 84 years. If p¥cted ship between age and the test scores
scores are needed for age ranges outslde the apparent in the normative data, indicat-
reported age boundaries, with proper taution ing improvement in test performance up
(see Chapter 3) they can be calculatef using to the fourth decade of life, with a sub-
the regression equations included in ~e ta- sequent decline, which is considerably
bles, which underlie calculations of ~e pre- accelerated in the seventh and eighth
dicted scores. decades. The advantage of the predic-
According to the test design, BNT: scores tions presented in this book is that they
for healthy young to middle-aged sam~es are are provided for 12 narrow age intervals
not expected to be normally distribUted. It spanning ages 25-84 in 5-year incre-
should be noted that the mean age ~r the ments, whereas the normative data pro-
aggregate sample is 67.91 (15.26). As rtflected vided by the test authors are grouped
in the scatterplot of the BNT data aroqnd the in five wide age intervals spanning ages
regression line, the majority of the .tudies 18-79.
available for review contained data fot older
age groups. The mean score for the agjregate Limitations of the analyses
sample is 52.25 (3.26), reflecting ~ age- 1. The number of studies that report data
related decline from the optimal perfoqnance for older age ranges considerably ex-
expected in younger samples. Thus, thei distri- ceeds that of studies concerned with
bution of scores in our sample is more tormal younger age ranges. Thus, the mean age
than expected in younger samples 1ue to for the aggregate sample is 67.91 (15.26)
variability in both directions from thetean, years, and scores are more normally
avoiding scores being skewed due to a iling distributed than would be expected in
effect. case of more even representation of age
A quadratic regression of SDs on a yiel- groups.
ded an R2 of 0.583, indicating an incr ase in 2. A strong effect of education, intellectual
variability with advancing age, consisteJtt with level, and fund of vocabulaly on BNT
the literature. Predicted SDs, based ~ this performance has been reported in the
model, are reported. literature. However, our data did not
Examination of the effects of demo!fciphic support this association, which is likely
variables on BNT scores indicated th~ edu- due to a restricted range of education
cation and IQ did not contribute to test scores (11-16.6 years) and IQ (113.8-119.9) in
in the data available for analyses. : the data available for analyses. We did
The difference in mean scores f<*" two not have sufficient data to explore the
genders across six studies reporting scores effect of verbal vocabuLuy. Mean edu-
for males and females separately was; 3.450 cational level of the aggregate sample is
in favor of males, which was not sta~cally 13.79 (1.50) years, and mean IQ is
significant. 116.10 (2.60). (IQ is available for only six
data points.) Higher scores on the BNT
Strengths of the analyses , are associated with increased educa-
1. Total sample size of 1,684 partici~ts. tional and IQ levels. Therefore, the
2. R2 of 0.850, indicating a good m<>fel fit. predicted values are likely to overesti-
3. Postestimation tests for paramet~ spe- mate expected performance for individ-
cifications did not indicate pr~lems uals with lower educational levels and/or
with normality or homoscedasticitf. average and lower than average intel-
4. Predicted scores and SDs presented in lectual levels.
Appendix 10m parallel the nor:!_lative 3. Superiority of males in BNT perfor-
data provided by Kaplan et al. (2~) in mance has been reported in several
the BNT-11 test booklet. Furthe~ore, studies. Our data are consistent with
predicted values based on the aggfegate this finding in that a difference in
mean scores of 3.450 in favor of males It should be noted that the distribution of
was found, which, however, was not performance scores for the BNT is far from
statistically significant. Only six data normal. The performance of the majority of
points for each gender were available for intact individuals falls at the upper range of
analyses. the score distribution. As a result, this test
does not discriminate well at the high-
performance levels (e.g., the distinction be-
tween "well above average" and "superior"
CONCLUSIONS
performance levels cannot be well defined).
A review of the literature suggests that con- On the other hand, this test discriminates well
frontation naming ability is affected by many at the lower range of performance. It is sen-
factors that need to be considered in inter- sitive to identifying outliers whose perfor-
preting BNT performance. Inspection of the mance falls below the expected range.
data sets suggests that educational level is as However, it is important that the effect of low
important as, if not more important than, age education and vocabulary be taken into con-
in BNT performance. This indicates that fu- sideration in clinical interpretation of low
ture normative studies need to present data by scores, as small SDs resulting from extreme
education groupings. Although the results of kurtosis of the score distribution cause lower
our meta-analysis did not reveal an association test scores consistent with limited vocabulary
of BNT performance with educational or in- to fall within the impaired range.
telligence level, this is likely due to a restricted Because of the negative skew in the score
range of education and IQ in the aggregate distribution, investigators examining an effect
sample used for the analyses. In addition, a of demographic variables on test performance
majority of the studies were limited to older should ensure adequate representation of
samples. Additional normative studies are participants with lower levels of education and
needed on younger populations, especially to intellectual level/vocabulary fund. It is the
ascertain if the same relationships between lower tail of the distribution that contributes
BNT scores and educational level are found in most to the relationship of the test scores with
younger groups. demographic variables. Unrepresentative sam-
BNT performance is also directly affected by ples with scores clustered around the mean
one's culturally determined linguistic back- will wash out this relationship.
ground and accumulated vocabulary. Therefore, The involvement of different information-
uncritical use of cutoff criteria might result in processing mechanisms in confrontation nam-
false-positive misclassification errors. To avoid ing has been extensively researched. Due to
unsubstantiated determination of a naming im- conRicting findings, however, further research
pairment, the BNT score needs to be inter- in this area would enhance our understanding
preted within the context of a patient's linguistic of the processes underlying naming ability and
background and cultural/educational exposure. their relation to visual-perceptual/recognition
In addition, qualitative analysis of performance ability. Consequently, the differential me-
(frequency of "don't know" responses, mis- chanisms determining age-related decline in
perceptions, tip-of-the-tongue errors, readiness naming ability vs. anomia associated with de-
to give up, response latencies, etc.) could con- generative brain conditions would be further
tribute to interpretation accuracy. illuminated.
11
Verbal Fluency Test
BRIEF HISTORY OF THE TESTi S, with 60 seconds for each letter (Bechtoldt
et al., 1962; Borkowski et al., 1967). This ver-
There are several types of task that ntasure sion became part of the Neurosensory Center
verbal fluency (VF). Their histori~ roots Examination for Aphasia (Benton, 1967;
stem from the Thurstone Word Fluen4Y Test Spreen &: Benton, 1969).
(Thurstone &: Thurstone, 1962), whi<fl is a Later, this test was included in the Multilin-
component of Thurstone's Primary ~ental gual Aphasia Examination Battery (Benton &:
Abilities Tests. The major disadvant4ge of Hamsher, 1978; Benton et al., 1994a) under a
Thurston's task was the format of thb test, new name, Controlled Oral Word Association
which required the examinees to write 'words Test (COWA), to eliminate a potentially mis-
beginning with the letter S in 5 minut.s, fol- leading reference to a fluent/nonfluent aphasia
lowed by writing four-letter words be~ning (see Ruff et al., 1996). The COWA is based on
with the letter C in 4 minutes (Heaton! et al., two sets ofletters, CFL and PRW. Whereas the
1991; Lezak et al., 2004; Pendleton Ft al., former stimuli (FAS) were chosen at random,
1982). Psychometric properties of this t+st are the selection of letters for the COWA was
reviewed by Cohen and Stanczak ~000). based on analysis of word difficulty as deter-
Normative data and T-score equivaleqts for mined by number of words in the English
this version of the test, based on a Saiif?le of language that begin with that particular letter.
486 participants grouped into 10 agel cate- As a result, the CFL and PRW versions of the
gories by six education categories for !males COWA are of equal difficulty and can be used
and females separately, are provided b~ Hea- interchangeably.
ton et al. (1991). In the 2004 revision of Hea- According to the analysis of letter difficulty
ton et al.'s manual, the sample is incre~ed to (Borkowski et al., 1967), which was based on
over 1,000 participants and the data ar~ addi- frequency of associations for 24 different let-
tionally stratified by race/ethnicity (African ters, the letters F, A, S, C, P, and W were
American and Caucasian). I classified as "easy," whereas Land R fell in the
Because of the confounds introduced ~y the category of "moderately difficult" letters. In
writing format of this test, Benton dev~oped spite of unequal difficulty levels for letters
an oral version, the Controlled Verbal Flpency included in the FAS vs. CFL and PRW sets, a
Task (CVFT), in which examinees orallj gen- study of equivalency between the three letter
erated words beginning with letters F, 4, and sets conducted on 106 patients with various
200
VERBAL FLUENCY TEST 201
neuropsychological dysfunctions yielded cor- supermarket (Barr & Brandt, 1996; Kozora &
relation coefficients of 0.87-0.94 for different Cullum, 1995; Monsch et al., 1992; Troyer,
samples (Lacy et al., 1996). The authors con- 2000); fruits and vegetables, foods, and things
cluded that these intercorrelations even sur- people drink (Acevedo et al., 2000; Miller,
pass correlations between CFL and PRW. In 2003; Monsch et al., 1992; Randolph et al.,
spite of such an optimistic view, the norms for 1993; Simkins-Bullock et al., 1994); first names
the FAS test should be used with caution in (Kozora & Cullum, 1995; Monsch et al., 1992);
application to the COWA sets (CFL and tools and clothing (Huff et al., 1986b); U.S.
PRW) due to different levels of letter difficulty states (Kozora & Cullum, 1995); and inanimate
(Ruff et al., 1996). objects (Fama et al., 2000). Fuld (1981) used
Other combinations of letters have been category naming tasks, such as proper names
used in several studies. Cavalli et al. (1981) of people (same gender as the examinee),
used P, F, and L in a study on lateralized foods, vegetables, things that make people
deficits in linguistic processing. Nielsen et al. happy, and things that make people sad, as
(1989) used S, N, and F on a large neurolog- distractor trials for delayed recall of the origi-
ically intact Danish sample. S and P were used nally presented stimuli in her Fuld Object-
by Barr and Brandt (1996) in a study on flu- Memory Evaluation (see also Marcopulos
ency deficits in dementia, Ganguli et al. (1991, et al., 1997). Food, clothing, animals, and things
1993, 1996) in a study on cognitive impair- to ride categories are included in the McCar-
ment in an elderly rural population, Goldman thy Scales for Children's Abilities (McCarthy,
et al. (1998) in a study examining cognitive 1972). A version of the VF task tapping se-
deficits associated with Parkinson's disease, mantic switching is included in the Delis-
and Coffey et al. (2001) in a study exploring Kaplan Executive Functions System (Delis
cognitive correlates of human brain aging. et al., 2001). This test assesses letter fluency (F,
Lannoo and Vingerhoets (1997) used N, A, A, S), category fluency (animals and boys'
and K in a project collecting normative data names), and category switching (fruits and
for a Dutch version of the phonemic fluency furniture). The test has a standard and alter-
test. Lopez-Carlos et al. (2003) used P, M, and nate forms. Two parallel forms of a semantic
R in their project, collecting normative data fluency test are included in the Repeatable
for monolingual Spanish-speaking individuals Battery for the Assessment of Neuropsycho-
with a low level of education. The authors logical Status (RBANS; Randolph, 1998).
noted that illiterate individuals tend to make One version of a category naming test is the
errors in the words that start with A and S, Set Test (Isaacs & Kennie, 1973), which in-
given that many words that start with a silent volves generating items from four successive
H begin with the A sound and words that start categories: colors, animals, fruits, and towns.
with a C sound like S. According to this version, examinees are to
Other versions of the fluency tests involve recall up to 10 items from each category, after
generation of words from certain semantic which they are instructed to shift to the next
categories (Category Naming), such as animal category. The score is the total number of
naming (Acevedo et al., 2000; Barr & Brandt, items recalled for all categories. The versions
1996; Beatty et al., 1997; Brady et al., 2001; proposed by Newcombe (1969), used in assess-
Crossley et al., 1997; Epker et al., 1999; Fama ing patients with lateralized missile wounds,
et al., 2000; Ganguli et al., 1991, 1993; Gio- involved naming objects and animals and al-
vannetti et al., 2003; Kempler et al., 1998; ternating between naming birds and colors
Kozora & Cullum, 1995; Lopez-Carlos et al., over 1 minute for each of the three trials. The
2003; Monsch et al., 1992; Morris et al., 1989; number of correctly generated items for the
Rosen, 1980; Rosselli et al., 2002a; Seines et al., first two conditions and correct alternations
1991; Simkins-Bullock et al., 1994; Tombaugh for the third condition are recorded. Villardita
et al., 1999; Troyer, 2000; Ylikoski et al., 1993); et al. (1985) used a modification of the Set
types of transportation and parts of a car Test to assess a group of normal elderly,
(Weingartner et al., 1984); items found in a employing the categories proper names of
202 LANGUAGE
persons, foods, and animals over 1-minute inhibition capacity, speeded mental proces-
trials. The score was the total number of items sing, and long-term vocabulary storage (Boone
for all categories. et al., 1998; Cauthen, 1978b; Crowe, 1998a;
For test administration instructions and Estes, 1974; Lafleche & Albert, 1995;
further discussion of the VF tasks, see Lezak Martin & Fedio, 1983; Parks et al., 1992;
et al. (2004) and Spreen and Strauss (1998). Perlmuter et al., 1987; Ruff et al., 1997).
Parkin and Lawrence (1994) further refined
the notion of cognitive flexibility and suggested
that word generation relies on spontaneous
Analysis of internal consistency for the COWA flexibility, rather than on reactive flexibility, as
version (CFUPRW), reported by Ruff et al. defined by Eslinger and Grattan (1993). Such
(1996) revealed a high coefficient oc (r=0.83) multidimensionality of cognitive processes in-
for the three letters, indicating high test ho- volved in word generation is reflected in the
mogeneity. Interrater reliability is reported to results of factor-analytic studies on large bat-
be near perfect in several studies (Spreen & teries of neuropsychological. tests. High load-
Strauss, 1998). Norris et al. (1995) reported an ings of phonemic and semantic fluency tests
interrater reliability of r = 0.98. Test-retest on factors comprised of verbally mediated
reliability for different versions of this test is tasks, such as Vocabulary, Boston Naming
quite high (see Spreen & Strauss, 1998). Test- Test, Similarities, and Digit Span, were re-
retest reliability reported by Ruff et al. (1996) ported by Lamar et al. (2002) and Ponton et
for a 6-month retest with an alternate set of al. (2000). On the other hand, Boone et al.
letters yielded a coefficient of r = 0. 74. A gain (1998) demonstrated high loading of a pho-
of about three words on the retest was inter- nemic fluency test on the factor representing
preted by the authors as a practice effect. speeded mental processing, along with Stroop
Similarly, test-retest reliability reported by and Digit Symbol. These findings confirm the
Dikmen et al. (1999) for an 11-month retest premise that word generation depends on
with an alternate set of letters was r=0.72. vocabulary storage and speeded mental pro-
The magnitude of practice effect in their study cessing in addition to executive functions. Use
was 1.22 (7.85) words. Data on repeated ad- of vocabulary storage in word generation was
ministration are also presented by McCaffrey explored by Crowe (1998a), who demon-
et al. (2000). strated a decrease in the rate of word gener-
For further information on the psychomet- ation with each 15-second increment of time,
ric properties of the COWA and other ver- indicating storage of high-frequency words
sions of VF tests, see Franzen (2000), Lezak accessed during the early phases of the task
et al. (2004), and Spreen and Strauss (1998). performance, with a gradual exhaustion of this
storage.
A qualitative analysis of the behavioral as-
Cognitive Mechanisms Underlying
pect of word generation offers further insight
Word Generation
into the strategy used in accessing lexical or
VF has been commonly viewed as a compo- semantic storage. Several systems of qualita-
nent of executive function, which is subserved tive scoring of VF output have been proposed
by the prefrontal cortex. However, a number (Raskin et al., 1992; Stuss et al., 1998; Troyer
of studies suggest that cognitive mechanisms et al., 1997), which are based on the premise
underlying efficient organization of verbal re- that word generation involves phonemic anal-
trieval and recall in word generation are ysis or semantic categorization (clustering)
multidimensional and involve auditory atten- and shifting from one subcategory to another
tion, short-term memory (in keeping track of (switching) (Gruenewald & Lockhead, 1980;
words already said), ability to initiate and Troyer et al., 1997). Whereas clustering is an
maintain word production set, cognitive flexi- automatic process that relies on memory
bility (in rapidly shifting from one word to the storage for words, switching is an effortful
next within the selected category), response process that requires speed and cognitive
flexibility, which determine the effectiveness information is involved in cueing word re-
of the search process (Troyer, 2000; Troyer trieval and keeping track of recent responses.
et al., 1997). According to Abwender et al. In contrast, category fluency utilizes a vi-
{2001), switching can be divided into subtypes suospatial sketchpad, allowing manipulation of
that reflect different cognitive processes in visuospatial information. Both the phonologi-
phonemic vs. semantic fluency. cal loop and visuospatial sketchpad storage
The criteria for clusters and switching are systems are passive slave systems to the cen-
defined differently in different scoring sys- tral executive (Baddeley, 1986).
tems. Qualitative scoring across different sys- Contributions of different cognitive mech-
tems includes the following components: anisms to word generation have also been
cluster size, number of switches, a pattern of demonstrated in the neural network model
word production in IS-second increments, and of phonemic VF proposed by Parks et al.
errors (perseverations, use of proper names, (1992). This approach uses parallel distributed
nonwords, and alternate forms of the same processing (PDP) models, which stem from
word). Psychometric properties of the Troyer earlier associationist (Hebbian) models and
et al. (1997) scoring system are evaluated by capture mental representations depicted as
Ross (2003). patterns of activity distributed across networks
Qualitative scoring was employed in an in- of simple processing units. The PDP model of
vestigation of the effect of age on strategies VF is dependent on attentional biases based
used in word generation (Hughes & Bryan, on reinforcement. Attention in this case is di-
2002) as well as in studies exploring the cog- rected at stored memories of previous stimuli.
nitive mechanisms of VF deficits in different Two reinforcement paths are identified: the
clinical groups (Beatty et al., 1997, 2002; selective positive reinforcement of words be-
Elvevag et al., 2002; Epker et al., 1999; Ho ginning with a given letter and the selective
et al., 2002; Mayr, 2002; Rich et al., 1999; negative reinforcement of words that are ex-
Troester et al., 1998; Troyer et al., 1998a,b; cluded according to the rules of the task and
York et al., 2003). Of note, several studies previously produced words. An examination of
challenge the specificity of clustering and a hierarchical relationship between features
switching variables and prompt further in- (individual letters) and categories in the case
vestigation in this area (Abwender et al., 2001; of phonemic word generation suggests that
Demakis et al., 2003; Ho et al., 2002). the categories are mutually exclusive and
Interactions between different cognitive based on a single relevant feature {letter).
mechanisms contributing to word generation
are best understood in the context of Badde-
Biochemical and Anatomical
ley's (1986) model of the central executive
Correlates and Effect of Brain Pathology
component of working memory, which con-
on Verbal Fluency
trols and regulates cognitive processing by
activating appropriate information from long- Recent advancements in neuroimaging and
term memory, selectively attending to relevant lesion studies provide validation to the em-
stimuli and filtering out irrelevant ones, pirically derived theories that point to an in-
switching between mental sets, and monitor- terplay of different cognitive mechanisms
ing incoming stimuli (Baddeley, 1996). Rende contributing to word generation. Functional
et al. (2002) suggest that in addition to the neuroimaging techniques that were used in
central executive control that underlies both the studies summarized below to image brain
phonemic and category word generation, sep- activation in response to silent or overt word
arate mechanisms are deployed in each of generation include proton magnetic resonance
these tasks. Phonemic fluency relies on a pho- spectroscopy eH-MRS), functional magnetic
nological loop, which, according to Baddeley's resonance imaging (fMRI), regional cerebral
model, is a temporary storage system for lin- blood flow (rCBF), positron emission tomog-
guistic information. According to Abrahams raphy (PET), and single-photon emission
(2000), short-term memory of phonological computed tomography (SPECT).
204 LANGUAGE
Jung et al. (1999) found word generation, in the verbal fluency performance and, possibly,
comparison to other neuropsychological tests, in the retrieval of words that are freely asso-
to be most closely related to concentrations of ciated in the temporal cortex, whereas the
the neurometabolite N-acetylaspartate (NAA) anterior cingulate gyrus may be responsible
in their study examining biochemical markers for directing the temporal cortex as to which
of cognition measured by 1H-MRS in normal associations should be attended to or sup-
human brain. The authors related the associ- pressed.
ation between variability in NAA levels and Several studies have suggested involvement
neuropsychological functioning to mitochon- of both left and right frontal lobes in word
drial function of the neuron. generation (Fama et al., 2000; Parks et al.,
A review of recent studies of brain areas 1988; Philpot et al., 1993). Elfgren and Ris-
involved in verbal fluency indicates that the berg (1998) and Stuss et al. (1998) hypothe-
left dorsolateral prefrontal cortex, inferior fron- sized that areas activated in the course of word
tal cortex (Broca's area), and anterior cingulate generation may reflect differences in cognitive
gyrus are primarily responsible for both pho- strategy. Furthermore, Crosson et al. (2003)
nemic and semantic word generation (Aude- proposed bilateral involvement of basal gan-
naert et al., 2000; Baldo et al., 2001; Elfgren & glia in word generation, where the left basal
Risberg, 1998; Elfgren et al., 1996; Fama et al., ganglia-ventral anterior thalamic loop is in-
2000; Frith et al., 1995; Gaillard et al., 2000; volved in retrieval of words from preexisting
Leggio et al., 2000; Levin et al., 2001; Parks lexical stores, whereas the right basal ganglia
et al., 1988; Phelps et al., 1997; Pihlajamaeki activity serves to suppress activity of right
et al., 2000; Pujol et al., 1996; Ravnkilde et al., frontal structures, preventing them from in-
2002; Stuss et al., 1998; Tucha et al., 1999; terfering with language production.
Warkentin & Passant, 1997; Schloesser et al., Martin et al. (2000) found that verbal ftu-
1998). Ruff et al. (1997) proposed a distinction ency impairment might be secondary to an
between two conditions that cause a reduction epileptogenic lesion in the left or right ante-
in verbal fluency: (1) low word fluency sec- rior temporal lobes due to disruption of distal
ondary to deficient verbal attention, word extratemporal regions, particularly the dorsal
knowledge, and/or verbal long-term memory, lateral prefrontal cortex. This finding is con-
which are due to dysfunction of diffuse mul- sistent with the nociferous cortex hypothesis
tifocal or nonfrontal brain areas; (2) low word and is supported by improvement in verbalftu-
fluency without the deficits listed above, which ency after anterior temporal lobectomy.
is likely to be associated with prefrontal lobe Levin et al. (2001) investigated the effect of
impairment. Further refinement of the role of closed head injury severity, frontal brain le-
the left frontal cortex in word generation comes sions, and age at injury in a longitudinal study
from Elfgren and Risberg (1998), who, based of children. Although verbal ftuency recovery
on the results of an rCBF study, suggested that after severe head injury was slower in younger
the left frontal cortex is engaged in the gener- compared to older children, left frontal lesions
ation of internally driven responses, which is a had a more adverse effect on verbalftuency in
major cognitive component of word genera- older children. This dissociation was inter-
tion. Piatt et al. (1999a,b) found that action preted by the authors as a reflection of the
fluency was sensitive to frontostriatal patho- more established functional commitment of
physiology. the left frontal region to expressive language
More extensive involvement of the left in older children. This finding is consistent
hemisphere in word generation, to include with the results of Gaillard et al.'s (2000)
temporoparietal regions, has been proposed in fMRI investigation comparing brain activation
several studies (N'Kaoua et al., 2001; Pihlaja- in children and adults in response to a phone-
maeki et al., 2000; Schloesser et al., 1998; mic task. Cortical activation in younger chil-
Stuss et al., 1998). Ravnkilde et al. (2002) dren was wider than in adults and involved
hypothesized that the left prefrontal cortex the right inferior frontal gyrus in addition
may be involved in the initiating component of to the left inferior frontal cortex. However,
predominantly left frontal localization of acti- et al. (1987), is based on the ability to access
vation appeared to be established by middle and retrieve semantic knowledge, than in
childhood. phonemic Ruency, which is based on phono-
A number of studies suggest that in addition logical/lexical retrieval mechanisms, is viewed
to the left frontal structures commonly viewed as being mostly due to disruption in the
as neural bases for word generation, phone- structure of semantic memory early in the
mic vs. semantic Ruency might involve dif- course of dementia.
ferent neural systems. Right (or bilateral) Coen et al. (1996) related distinction in the
cerebellum has been found to participate in efficiency of phonemic vs. semantic Ruency
phonemic, but not in semantic, processing to the rate of cognitive decline. The authors
(Leggio et al., 2000; Ravnkilde et al., 2002; showed that shorter duration of illness in their
Schloesser et al., 1998), whereas hippocampi sample of patients suffering from dementia of
were found to play a role in semantic, but not Alzheimer's type (DAT) is associated with
in phonemic, Ruency performance (Gleissner greater impairment in letter Ruency, whereas
and Elger, 2001). Similarly, Pihlajamaeki et al. longer duration resulted in predominance of
(2000) suggested that the left medial temporal the category fluency impairment. Consider-
lobe (hippocampal formation or posterior ing that there was no difference between the
parahippocampal gyrus) is activated in re- groups in dementia severity, these findings
trieval by category. Stuss et al. (1998) sug- were viewed as a function of the rate of disease
gested that in addition to the left hemisphere progression, with greater impairment in pho-
centers participating in phonemic processing, nemic Ruency being associated with more ra-
semantic processing involves right dorsolateral pid cognitive decline.
and inferior medial regions. This view is sup- Effects of different types of brain pathol-
ported by N'Kaoua et al.'s (2001) finding that ogy, including brain injuries, aphasias, some
phonemic processing involves the left tempo- amnesiac conditions, and degenerative de-
ral lobe, whereas semantic processing involves menting conditions on verbal production, are
the left and right temporal lobes. addressed in a number of studies, many of
Differential rates of deterioration in pho- which provide information allowing compari-
nemic vs. semantic Ruency in dementia sup- son of word generation in clinical and control
port the notion that these two types of Ruency groups (Barr & Brandt, 1996; Carew et al.,
are subserved by different neural mech- 1997; Cerhan et al., 2002; Clark et al., 1997;
anisms. Animal naming has been shown in Coen et al., 1996; Dalrymple-Alford et al.,
some studies to be performed at higher levels 1994; Elvevag et al., 2001; Eslinger et al.,
than word-generation tasks (Ober et al. 1986; 1984; Geffen et al., 1993; Goethe et al., 1989;
Rosen, 1980). Similar findings are reported Goldman et al., 1998; Gurd, 2000; Huff et al.,
with reference to other semantic category 1986b; Joyce et al., 1996; Klimczak et al.,
naming tests (e.g., fruits: Ober et al., 1986; 1997; Lafieche and Albert, 1995; Locascio
Randolph et al., 1993). In contrast, Bayles et et al., 1995; Margolin et al., 1990; Miller,
al. (1989), Monsch et al. (1994), and Sherman 1985; Piatt et al., 1999a; Poreh et al., 1995;
and Massman (1999) did not find differences Robert et al., 1998; Shoqeirat et al., 1990; Zec
in efficiency of word generation for semantic et al., 1999).
vs. phonemic tasks. Furthermore, greater im-
pairment of semantic Ruency in comparison to
Assessment of Verbal Fluency
phonemic Ruency tasks in clinical samples
in Different Languages
was documented by Barr and Brandt (1996),
Butters et al. (1987), Cahn et al. (1995); Cer- Spanish versions of the verbal Ruency test are
han et al. (2002); Crossley et al. (1997), available, which are based on different sets of
Mickanin et al. (1994), Monsch et al. (1992), letters (Artiola i Fortuny et al., 1999; Lopez-
Rosser and Hodges (1994), and other authors. Carlos et al., 2003; Rey & Benton, 1991).
This pattern of greater deterioration in se- Artiola i Fortuny et al. (1999) used letters
mantic Ruency, which, according to Butters P, M, and R to assess phonemic Ruency as part
206 LANGUAGE
of a standardized and validated battery of on the animal and transportation versions of

neuropsychological tests culturally adapted for the category fluency test is reported by Chan
Spanish-speaking individuals. Lopez-Carlos and Poon (1999). Normative data for the
et al. (2003) provided data for monolingual Animal Naming and a combined category of
Spanish-speaking participants with ~10 years Fruits and Vegetables for Cantonese-speaking
of education for the PMR version (see study Hong Kong Chinese adolescents and adults
VF. 43, below). Ponton et al. (1996) provided are provided by Lee et al. (2002). Chiu et al.
normative data for Spanish-speaking partici- (1997) assessed the validity of the modified
pants for an FAS version of the test (see study Fuld Verbal Fluency Test as a screening
VF.17, below). Acevedo et al. (2000) provided instrument for differential diagnosis of de-
comparative data for English- and Spanish- mentia on a sample of 53 normal and 56 de-
speaking elderly for category flusency (see mented Chinese (Hong Kong) participants,
study VF.36, below). Rosselli et al. (2000, aged 59-97 years. Scores for all semantic ca-
2002a) compared oral fluency strategies in tegories (animals, fruits, vegetables) differen-
FAS and animal word generation of Spanish tiated well between normal and demented
and English monolingual participants with groups. To improve test-retest reliability, the
bilingual participants in both languages (see authors recommended use of a composite
study VF.40, below). La Rue et al. (1999) measure.
provided normative data for verbal fluency, Flemish normative data for two category
measured as generation of same-sex first fluency trials, animals and professions, and for
names, for a sample of Hispanic older adults a Dutch version of the phonemic fluency test,
(65-97 years of age), stratified by two age le- which used letters N, A, and K. are reported
vels and four educational levels. by Lannoo and Vmgerhoets (1997).
Gollan et al. (2002) provided comparative Rodriguez-Aranda (2003) reported norma-
data for Spanish/English bilingual and English tive data for a Norwegian version of the oral
monolingual participants on 12 semantic, fluency (letters F, A, S; categories Animals,
10 letter, and two proper name fluency cate- Fruits, and Professions) and written fluency
gories. Word fluency for all categories was (letters S, I<; categories Vegetables, Sports,
lower for bilingual participants, with espe- and Farm Animals) tests. The data are strati-
cially pronounced differences in semantic ca- fied by age group.
tegories. Bilinguals' fluency performance did
not improve when words in both languages
were used. The authors discussed mechanisms
RELATIONSHIP BETWEEN
of the bilingual lexical system.
VFT PERFORMANCE AND
Benita-Cuadrado et al. (2002) provided
DEMOGRAPHIC FACTORS
age- and education-adjusted normative data
for the Animal Naming test collected on a The original normative data for the CVFI'
sample of 445 participants between 18 and (based on the F, A, S letter set) presented in
92 years of age in Barcelona, Spain. Dellatolas the Neurosensory Center Examination for
et al. (2003) explored the effect of illiteracy on Aphasia manual (Spreen & Benton, 1969)
cognition on 97 normal Brazilian illiterate were based on a rural sample with low edu-
adults and 41 schoolchildren 7-8 years old. cational background and, therefore, have lim-
The authors provided normative data for se- ited clinical relevance (Spreen & Strauss,
mantic fluency (animals, clothes) and letter 1991). Norms compiled more recently on
fluency (P, F, M). demographically advanced samples yield con-
Psychometric properties and clinical utility sistently higher scores across different studies.
of Hebrew versions of phonemic and semantic Benton and Hamsher's (1978) manual for the
fluency tests are discussed by Axelrod et al. Multilingual Aphasia Examination (based on
(2001). C, F, L and P, R, W letter sets) provides
Performance of Chinese native speakers re- corrections for age, gender, and education,
siding in Hong Kong, ranging in age 7-95 years, implicating effects of these variables on test
performance. (See Lezak, 1995, and Lezak factors. Schaie (1983) suggested that a re-
et al., 2004, for the correction table and the quirement for motor response is a factor
percentile rank table.) contributing to age-related decline. Norris
A decline in verbal fluency with age was et al. (1995) hypothesized that the discrepancy
documented by Acevedo et al. (2000), Brady in the findings across different studies re-
et al. (2001), Furry and Baltes (1973), Kempler garding effect of age on VF performance
et al. (1998), Norris et al. (1995), Rodriguez- might be due to several factors: (1) sampling
Aranda (2003), Stuss et al. (1998), and Wie- differences, with selection of high IQ partici-
derholt et al. (1993). Schaie and Parham pants reducing a correlation that might be
(1977) and Schaie and Strother (1968b) re- observed in a broader sample; (2) use of
ported decline associated with advancing age clinical samples may mask the unique contri-
in cross-sequential comparisons. Benton et al. bution of age, which is attenuated by the ef-
(1981) reported a decline in verbal fluency fect of brain damage; (3) cohort effect might
only after age 80 based on a sample of 65- influence the results, with smallest VF per-
84 years old. Parkin and Lawrence (1994) formance differences among middle-aged
identified significant decline with age only in adults to young-old adults and greater differ-
older cohorts with low educational level. Par- ences across older samples; (4) diversity in the
kin and Java (1999) reported age-related de- elderly populations from which participants
cline in semantic, but not in phonemic, are drawn (community vs. institutionalized
fluency. Similarly, Troyer (2000) found age to elderly).
be more strongly related to semantic fluency Acevedo et al. (2000), Anstey et al. (2000),
than to phonemic fluency. Chan and Poon Axelrod et al. (2001), Chan and Poon (1999),
(1999) reported an increase in category fluency lvnik et al. (1996), Kempler et al. (1998),
from childhood to adulthood, with peak per- Lannoo and Vingerhoets (1997), Loonstra
formance in adults 19-30 years of age and et al. (2001), Norris et al. (1995), Ponton et al.,
a subsequent decline with advancing age. (1996), and Troyer (2000) found a significant
Loonstra et al. (2001) presented aggregate effect of education on VF performance; how-
statistics for phonemic fluency suggestive of a ever, in the studies by Axelrod and Henry
progressive age-related decline. (1992), Bolla et al. (1990), and Stuss et al.
In contrast, Anstey et al. (2000), Axelrod (1998), education was not significantly related
and Henry (1992), Bolla et al. (1990), Boone to efficiency of word generation. Ruff et al.
(1999), Boone et al. (1990), Cauthen (1978b), (1996) suggested that gender moderated the
Daigneault et al. (1992), Mittenberg et al. effect of education.
(1989), Ponton et al. (1996), Ruff et al. (1996), Benito-Cuadrado et al. (2002) noted a sig-
Seines et al. (1991), and Tomer and Levin nificant effect of age and education on se-
(1993) did not find age-related differences in VF mantic fluency in a Spanish sample. Based on
performance. Relative stability in VF perfor- regression analysis, they developed a predic-
mance with advancing age was also reported by tive function for the semantic fluency score:
Miller (1984), Perhnuter et al. (1987), and F(x) = 23.89 +age( -0.144) + education(0.39).
Yeudall et al. (1987). Hultsch et al. (1993) Tombaugh et al. (1999) found, in a large
demonstrated that, after controlling for self- sample of cognitively intact individuals aged
reported health status and activity levels, age no 16-95 years, that phonemic fluency was more
longer significantly contributed to the variance sensitive to the effects of education than age
in word generation. Similarly, Coffey et al. (18.6% and 11.0% of variance, respectively),
(2001) did not find a relationship between verbal whereas semantic fluency was more affected
fluency and age-related changes evident on MRI by age than education (23.4% and 13.6% of
in a sample of 320 nonclinical volunteers aged variance, respectively).
66-90. This conclusion is consistent with their A positive relationship between verbal in-
review of the relevant neuroimaging literature. telligence and verbal fluency was documented
The discrepancy between different studies by Bolla et al., (1990), Boone (1999), Cauthen
can be explained by numerous confounding (1978b, for ~60-year-old subgroup only),
208 lANGUAGE
Borkowski et al. (1967), and Miller (1984). In In addition to demographic variables,

contrast, Axelrod and Henry (1992) did not functional status was found to contribute to
find an effect of verbal intelligence on VF the efficiency ofVF performance, especially in
performance. Bolla et al. (1990) an~ Boone geriatric samples. An effect of depression on
(1999) suggest that the effect of edu~tion is VF performance is well documented (Boone
mediated by verbal intelligence. Bo"r et al. et al., 1995; Caine, 1986; Norris et al., 1995).
(1990) concluded that data based OJ\ verbal Effects of levels of physical and mental activity
intelligence rather than on educatiot( would are documented by Craik et al. (1987). Neu-
be more accurate in differentiating tietween ropathological changes in the brain associated
normal and abnormal performance. i with cardiovascular disease and cerebrovas-
An effect of gender, with superiority of fe- cular risk factors also contribute to decline in
male performance, was documented 'y Ace- VF efficiency (Boone et al., 1993b; Breteler
vedo et al. (2000), Bolla et al. (1990), paddes et al., 1994). Brady et al. (2001) showed that
and Crockett (1975), Loonstra et al. k2001), the relationship between stroke risk and de-
Ruff et al. (1996), and Veroff (1980)b' How- cline in semantic fluency was 80% as large as
ever, no gender differences in verbal ruency the relationship between age and fluency de-
in normal and clinical samples were fwnd by cline. Similarly, Kilander et al. (2000) reported
Boone (1999), Cauthen (1978b), Ch~n and a strong relationship between diastolic blood
Poon (1999), Ponton et al. (1996), frupich pressure at age 50 and performance on the
et al. (1995), Saxton et al. (2000), To~baugh FAS 20 years later in a population-based study
et al. (1999), Troyer (2000), and Yeu~ et al. conducted in Sweden on a sample of 502 men.
(1987). .
Johnson-Selfridge et al. (1998) demon-
strated differences in FAS and Animal Nam-
ing word generation among groups ofi white, METHOD FOR EVALUATING THE
black, and Hispanic male veterans based on a NORMATIVE REPORTS
sample of 200 participants in each gro;.p (see To adequately evaluate the VF normative
VF.24, below). It should be pointed Oflt that reports, eight key criterion variables were
the income, level of education, ancJ; Wide deemed critical. The first six of these relate to
Range Achievement Test-Revised (~T-R) subject variables, and the two remaining refer
Reading subtest scores were significantly re- to procedural issues.
lated to verbal fluency in this study and, Minimal requirements for meeting the cri-
therefore, might be mitigating factors :in this terion variables were as follows.
comparison.
An effect of native language was noted by
Kempler et al. (1998), who compared samples Subject Variables
of Chinese, Hispanic, and Vietnamese parti-
Sample Size
cipants, with Vietnamese-speaking ,artici-
pants demonstrating the highest rate of animal Fifty cases are considered a desirable sample
generation and Spanish-speaking, the lowest. size. Although this criterion is somewhat ar-
The authors attributed this difference :to the bitrary, a large number of studies suggest that
fact that in the Vietnamese language animal data based on small sample sizes are highly
names are short (predominantly one syllable), influenced by individual differences and do
while in Spanish they are longer than ,n any not provide a reliable estimate of the popu-
other language used in this study (two ot three lation mean.
syllables).
Rodriguez-Aranda (2003) reported a 15trong Sample Composition Description
relationship between readinglhandvmting As discussed previously, information regard-
speed and verbal fluency in a sample of 101 ing medical and psychiatric exclusion criteria
Norwegian participants 20-88 years of ~ge. is important. It is unclear if socioeconomic
status, occupation, ethnicity, native language, SUMMARY OF THE STATUS

geographic recruitment region, or recruitment OF THE NORMS
procedures are relevant. Until this is deter-
There are a number of studies exploring the
mined, it is best that this information be
efficiency of word generation in normal and
provided.
clinical samples across various demographic
Age Group Intervals groups and diagnostic categories. A consider-
able variability between studies, obscuring their
nus criterion refers to grouping of the data comparability, includes the following aspects:
into limited age intervals. In spite of the con- time allotted for each category, type of se-
troversy in the literature regarding the effect of mantic or phonemic category, relative difficulty
age on VF performance, accuracy of data in- of letters within a phonemic category, pre-
terpretation is facilitated by using a narrow- sence/absence of feedback on intrusion or
range age group as a reference sample. repetition errors, instructions for item exclu-
Reporting of Educational levels
sion (inconsistency regarding exclusion of num-
bers), inconsistent administration ofan example
Given consistent evidence of effects of edu- with practice trial prior to the first test trial.
cation on VF performance, normative data Some authors do not specify which letters
should be grouped by educational level. were used in their phonemic categories.
Among all the studies available in the lit-
Reporting of Intellectual levels
erature, we selected for review those based on
Given consistent evidence of effects of intel- well-defined samples and test versions. In the
lectual level on VF performance, normative majority of studies reviewed below, the test
data should be grouped by IQ level. scores represent a total number of words
generated over three 60-second trials in the
Reporting of Gender Composition
phonemic category or a number of words
Given the probable association between gen- generated in one 60-second trial in the se-
der and VF performance in favor of females, mantic category. Deviations from this format
information regarding gender composition are identified in each table.
should be reported for each subgroup, and In this chapter, normative publications and
preferably normative data should be pre- control data from clinical studies are reviewed
sented by gender. in ascending chronological order. The text of
study descriptions contains references to the
Procedural Variables corresponding tables identified by number in
Appendix 11. Table All.l, the locator table,
Descriptio~ of Administration Procedures summarizes information provided in the
Due to V!lPability in administration proce- studies described in this chapter. 1
dures (see below), a detailed description of
the procedures, including identification of the
version of the test administered and restric-
tions in the types of word to be used, is de- SUMMARIES OF THE STUDIES
sirable. nus would allow one to select the [Vf.1] Cauthen, 1978b (Table A11.2)
most appropriate norms or to make correc- The author administered the Wechsler Adult
tions in interpretation of the data. Intelligence Scale (WAIS, Satz-Mogel short
form) and the VF test as part of a large study
Data Reporting
in Canada aimed at normative data collection
Group means and standard deviations for the for age and IQ levels on neuropsychological
number of words generated for each condition
and the total score for all letters or categories 'Norms for children are available in Baron (2004) and
should be presented at minimum. Spreen and Strauss (1998).
210 LANGUAGE
tests. The VF test consisted of eight letters 4. Participants from the older group lived in
administered in the same order to all institutional settings; however, the mean
participants-S, G, U, N, F, T, ], P-fqllowing FSIQ was quite high. It is unclear from
the standard procedure (i.e., !-minute oral the description of the sample composi-
production). The sample was divided ihto two tion why they were institutionalized.
age groups: 20-59 and 60-94 yeais. The 5. Data collected in Canada, which may
younger group consisted of 12 males and 39 limit usefulness for clinical interpreta-
females gathered from a variety of ~urces, tion in the United States.
with full-scale IQ (FSIQ) ranging lp0-140 6. Wide age ranges within each age group-
(M = 115.6, SD = 8.7). The older ~up ining.
cluded 28 males and 36 females, li~g pri-
marily in institutional settings, wi~ FSIQ [VF.2l Yeudall, Fromm, Reddon, and Stefanyk,
ranging 80-140 (M = 111.5, SD = 13.1~ 1986 (Tables A11.3-A11.5)
Analysis of the results for the ~unger The authors obtained VF data on 225 Cana-
group did not indicate any relation ~tween dian volunteers (127 males, 98 females)
VF performance and FSIQ. Therefore, the VF recruited from posted advertisements in work-
data were presented for the total sa_m;ple. In places and personal solicitations. Participants
contrast, VF performance differed signifi- included meat packers, postal workers, transit
cantly for the older group across three FSIQ employees, hospital lab technicians, secretar-
ranges: 80-106, 107-118, 119-140. 'lite au- ies, ward aides, student interns, student nur-
thors hypothesized that speed of perfo\mance ses, and summer students. In addition, high
is a determining factor in the relationship school teachers identified for participation
between VF scores and IQ (specifica$y per- average students in grades 10-12. Exclusion
formance IQ [PIQ] scale). Further 4oalysis criteria were evidence of "forensic involve-
suggested that IQ level did not interaft with ment," head injury, neurological insult, pre-
letter difficulty. No relationship betw~n VF natal or birth complications, psychiatric
performance and age was evident. · problems, or substance abuse.
The authors concluded that the use cf these Experienced testing technicians gathered
norms in the 20-59 age group is inapp~priate VF data and "motivated the participants to
for those with FSIQ <100. achieve maximum performance," partially
through the promise of detailed explanations
Study strengths of their test performance. Rigorous exclusion
1. Data are presented in age groupings. criteria were used.
2. Efficiency of word generation for eight Data are presented for four age groups for
letters was compared. males and females combined and separately.
3. Data for the older group were stiatified The data are reported for the F AS and for
by FSIQ level. Written Word Fluency. Norms for the oral ver-
4. Information provided regarding sender sion only are reproduced in this book. No sig-
and geographic recruitment area. i nificant effects of gender or age were evident.
5. Sample sizes for each age group are ad-
equate, but individual cell sizes are small. Study strengths
6. Test administration procedures were 1. The sample size is large, with individual
specified. cells approximating 50.
7. Means and SDs for the test sco~s are 2. The sample is stratified into four age
reported. groups.
3. Data are presented for males and fe-
Considerations regarding use of the study males separately.
1. Participants' level of education ~ not 4. Data availability for a 15-20 year age
reported. group.
2. Younger participants' FSIQ was hfgh. 5. Adequate medical and psychiatric ex-
3. Exclusion criteria were not speci~d. clusion criteria.
6. Information regarding handedness, ed- characteristics since they are drawn from
ucation, IQ, gender, occupation, re- the same population.
cruitment procedures, and geographic 2. Subject selection criteria are not reported.
area is provided. 3. It is unclear which version of the test was
7. Means and SDs for the test scores are administered.
reported. 4. No information on IQ is reported.
Considerations regarding use of the study [VF.4] Bolla, Lindgren, Bonaccorsy, and
1. Both the written and oral versions of this Bleecker, 1990 (Table A11.7)
test were administered, but the order of The authors examined the effect of demo-
administration is not specified. Issues of graphic factors and influence of different cog-
practice effect are not addressed. nitive processes on VF (FAS) performance in
2. Education and intelligence level for the healthy elderly.
sample are high. Participants were 199 Caucasian volunteers,
3. The data were obtained on Canadian 80 men and 119 women, enrolled in the Johns
participants, which may limit their use- Hopkins Teaching Nursing Home Study of Nor-
fulness for clinical interpretation in the mal Aging, who were recruited through news-
United States. paper advertisements. Participants' ages ranged
39-89 (M = 64.3, SD = 13.5); education ranged
[VF.3] Gordon and Lee, 1986 (Table A 11.6) 8-22 years, with a mean of 14.7 years (SD = 3).
The authors studied the relationship between Rigorous exclusion criteria were used.
gonadotropins and visuospatial function. A The F AS version of the VF test was ad-
word-generation test was given to experimen- ministered as part of a comprehensive neu-
tal participants of a blood-monitoring study ropsychological battery. Standard instructions
and to 250 control participants who were were used. Participants were instructed to
university students, drawn from the same po- exclude proper nouns. Series of numbers and
pulation as the experimental participants. proper nouns were not scored.
The standard administration procedure was To examine the effect of verbal intelligence
used. on FAS performance, WAIS-R Vocabulary test
The data for 90 males and 160 females from scores were used in a regression analysis along
the control group are given in Table A11.6. with demographic variables. Verbal intelli-
Total scores for three trials are reported. gence and gender accounted for a significant
The results suggest considerable gender proportion of the variance in F AS perfor-
differences in the rate of word generation, with mance. Age and education were not related
superiority of females in both experimental significantly to performance. Therefore, the
and control groups. authors grouped their data by verbal intelli-
gence for males and females separately. Based
Study strengths on their raw WAIS-R Vocabulary scores, par-
1. Sample size is large. ticipants were divided into three verbal intel-
2. Test administration procedures are ligence groups: average (30-53), high (54--60),
specified but not test version. and superior (61-68).
4. Data are presented for males and fe- 1. The sample composition is well de-
males separately. scribed in terms of age, gender, educa-
tion, verbal IQ, geographic area, and
Considerations regarding use of the study recruitment procedures.
1. Participants are identified as university 2. The data are presented for three verbal
students. Age range for the experimental intelligence groups for males and fe-
group is provided. Control participants males separately.
are assumed to have similar demographic 3. Adequate exclusion criteria.
212 LANGUAGE
4. Test administration procedures are spec- 4. Means, SDs, as well as scores for per-
ified. centiles 5 and 10 are presented.
5. Means and SDs for the test scores are 5. Minimally adequate exclusion criteria.
reported.
Considerations regarding use of the study 1. All-male sample.
1. Education and intelligence level for the 2. No information on IQ is reported.
sample are high. 3. High educational level of the sample.
2. It is unclear whether participants were
instructed to avoid numbers in the pro- [VF.6] Axelrod and Henry, 1992 (Table A11.9)
cess of word generation. The authors compared the performance of
3. Overall sample is adequate, but individ- 80 healthy, independently living individuals
ual cells are relatively small. aged 50-89 on tests tapping executive func-
4. Data were collected in Canada, and it is tioning and WAIS-R (Satz-Mogel abbreviation).
unclear if they are appropriate for use in Participants were recruited from a university-
the United States. related project and the community. Rigorous
exclusion criteria were used. The FAS version
[VF.S] Seines, Jacobson, Machado, Becker, of the VF test was administered according to
Wesch, Miller, Visscher and McArthur, standard criteria Items excluded were proper
1991 (Table A11.8) names and variations of the same word In ad-
The investigation used participants from the dition, participants self-rated their health status
Multi-Center AIDS Cohort Study (MACS). on a 1-5 scale and reported the number of
The article presents results of 733 seronega- physician appointments in the past 12 months
tive homosexual and bisexual males for the as an objective measure of health status. Verbal
purpose of establishing normative data for intelligence was measured with WAIS-R Vo-
neuropsychological test performance based on cabulary scores. No relationship was found
a large sample. The majority of the sample between VF performance and intellectual
consisted of Caucasian participants. Partici- competence, educational experience, or general
pants with a history of head injury with loss of health status.
consciousness > 1 hour and who reported
drinking ~21 drinks per week in the previous Study strengths
6 months were excluded. Percent of African- 1. Administration procedures are well out-
American participants ranged 3.4%-4.1% for lined.
different age groups. Percent of left-handers 2. Sample composition is well described in
ranged 11.3%-14.9%. terms of IQ, age, education, gender, and
The F AS version was administered accord- ethnicity.
ing to standard instructions. The investigators 3. Strict subject selection criteria were
also utilized an animal category task with a used.
1-minute interval. 4. Data are stratified by age group.
1. The overall sample size is large, and in-
dividual cell sizes are large. Considerations regarding use of the study
2. Normative data are stratified by age and 1. The sample sizes for each age group are
education. small.
3. The demographic composition of the sam- 2. High educational level of the sample.
ple is described in terms of age, gender,
sexual orientation, handedness, ethnicity, [VF.7] Monsch, Bondi, Butters, Salmon,
and geographic area; demographic com- Katzman, and Thai, 1992 (Table A11.1 0)
position is described for each age and The authors compared performance of
education cell separately. 89 OAT patients and 53 demographically
matched control participants on four measures [VF.8] Simkins-Bullock, Brown, Greiffenstein,

of VF: category, letter, first names, and su- Malik, and McGillicuddy, 1994 (Table A 11.11)
permarket fluency. Control participants, who The study addresses the relative utility of
were 52-86 years of age and had 7-19 years of various tests in investigating cognitive aspects
education, were recruited through newspaper of memory and executive functioning in pa-
advertisements or were spouses of patients. tients with anterior communicating artery
Participants with a history of alcoholism, other aneurysm and normal control participants.
drug abuse, learning disability, and/or a seri- The control group consisted of 10 males and
ous neurological or psychiatric illness were 9 females with no history of neurological or
excluded. major psychiatric illness, mean age of 52.6
The standard administration procedure was (15.6), mean education of 8--19 years, and
used for the F AS version of the test. The WAIS-R FSIQ of 85--124.
category fluency task included three trials, VF measures included semantic fluency
60 seconds each, for animals, fruits, and (animals and fruits or vegetables) and F AS.
vegetables. First names and supermarket flu-
ency consisted of one 60-second trial for each
Study strengths
condition. The reported scores were the total
numbers of correct responses across all trials
scribed in terms of age, gender, educa-
for each test and corresponding SOs.
tion, and FSIQ.
A comparison of performance for two groups
across four measures of VF using receiver
3. Means and SOs for the test scores are
operating characteristic curves demonstrated
reported.
superiority of category fluency in discriminat-
ing between OAT and normal participants,
whereas letter fluency was the least accurate. Considerations regarding use of the study
The authors proposed that the superiority of 1. Test administration procedures are not
category fluency is due to its dependence on well identified. It is not clear whether
the structure of semantic knowledge, which the semantic fluency condition included
deteriorates early in the course of OAT. two or three trials.
Analysis of gender differences for the con- 2. Recruitment procedures a not reported.
trol group indicated that females outperformed 3. The sample is small, which does not al-
males on all measures, with significance levels low partitioning of data into age groups.
ranging from p < 0.001 to p = 0.03.
[VF.9] Parkin and Lawrence, 1994
Study strengths (Table A11.12)
1. The sample composition is described in The F AS word fluency test was administered
terms of age, gender, and education. to 22 elderly participants in a study investi-
2. Rigorous exclusion criteria. gating the relationship between frontal lobe
3. Test administration procedures are well function and age-related memory decline.
specified. The sample included 18 females and 4
4. Means and SDs for the test scores are males, with mean age of 71.9 (4.8), mean ed-
reported. ucation of9.4 (1.3), and National Adult Read-
ing Test (NART) FSIQ of 106.1 (12.6), who
Considerations regarding use of the study were living independently and in good health,
1. Overall sample size is adequate; how- per self-report. Volunteers with diabetes or a
ever, the data are presented across a history of neurological illness were excluded.
wide range of age and education. All participants passed screening for dementia
2. The authors reported statistically signif- using the Blessed Scale.
icant gender differences; however, the A standard administration procedure was
data are not grouped by gender. used, with the exception that participants
3. No information on IQ is reported. were not instructed to avoid numbers.
214 LANGUAGE

1. The sample is small. 1. The sample composition is described in
2. The sample composition is well de- terms of age and recruitment proce-
scribed in terms of age, gender, educa- dures.
tion, and estimated IQ. 2. Rigorous exclusion criteria.
3. Adequate exclusion criteria. 3. Means and SDs for the test scores are
4. Test administration procedu~s are reported.
specified. :
5. Means and SDs for the test s~res are Considerations regarding use of the study
reported. ; 1. The sample is small.
6. Relatively low educational level· of par- 2. The sample for this arm of the study
ticipants. Scarce data are avail4ble for represents a subset of a larger sample,
this educational group. whose demographics were only cursorily
described.
Considerations regarding use of the stJJ,dy 3. It is unclear whether the administration
1. Small sample size. procedure required restrictions in the
2. Recruitment procedures were not re- types of word to be used in the process
ported. of word generation.
3. The data were obtained in the i United
Kingdom, which may limit theirj useful- [VF.11] Kozora and Cullum, 1995
ness for clinical interpretation . in the (Tables A11.14, A11.15)
United States. The authors compared category and letter
fluency in normal aging individuals. Partici-
[VF.1 0] Friedman, Kenny, Jesberger, Chqy, pants were volunteers (n = 174) 50-89 years
and Meltzer, 1995 (Table A 11.13) of age, who were recruited through local
The F AS word fluency test was admi4istered media announcements as part of an ongoing
to 24 adult volunteers recruited by adjtertise- aging study. Participants were screened using
ment primarily from hospital staff to serve as a semistructured neuromedical interview.
part of a control group in a study of the rela- None of the participants selected for the study
tionship between smooth pursuit eye-tf:'acking had a known history of substance use, major
and cognitive performance in schizopbrenia. psychopathology, uncontrolled hypertension,
The control group used in this arm of~ study other major medical illnesses, learning dis-
represents a subset of a larger control group ability, or neurological disorder or was taking
described in an earlier publication byi Fried- medications known to affect central nervous
man et al. (1991), which included 45 partici- system (CNS) functioning. Participants with
pants screened for health problems $sing a Mini-Mental State Exam (MMSE) score <24
health questionnaire and a structure<f inter- were excluded. The sample was divided into
view. Exclusion criteria were a history !of psy- four age groups by decade, which were
chiatric or neurological disease, major frog or equated for educational level, gender distri-
alcohol abuse, a significant ophthalmcplogical bution, and verbal intellectual level (as mea-
condition, having a first-degree relati'{e with sured by WAIS-R Vocabulary subtest).
psychiatric illness, or receiving medicatifm with Five different VF tasks were administered
was 35.8 (11.0). Educational level and t

known psychotropic, neurological, or ophthal-
mological effects. The mean age for the sample
not reported; however, it was noted at the

were
patients had significantly fewer years f edu-

as part of a larger study. Letter fluency was
measured using the F AS task, with instruc-
tions not to use proper names, numbers, or
the same word with different suffixes. Cate-
gory fluency was measured with four tasks: (1)
cation and lower IQ than the control gf<>up. the supermarket item list from the Dementia
The score for the F AS is the total dumber Rating Scale (DRS; Mattis, 1988), which was
of words generated over three 60-second administered according to standard instruc-
conditions. tions except that the total number of items
generated in 1 minute was used as the total care facilities), and had MMSE scores ~.20.
score; (2) animal naming; (3) state naming. The third group included undergraduate stu-
listing U.S. states; and (4) first name genera- dents from a large southwestern university
tion, listing both male and female names. aged 18-28 years (M = 19.4, SD = 1.8).
Performance within 1 minute was recorded Participants were assessed with the F AS
for each task. version of the VF test. The standard proce-
Qualitative aspects of VF were assessed by dure was used. Participants were instructed to
calculating the hierarchical structure of words exclude proper names, numbers, and different
generated on the supermarket ftuency task extensions of the same word. 11te letter T was
and by examining the frequency of persever- used as an example. Two scorers rated all
ative responses and intrusion errors made on protocols, with interrater reliability of r = 0.98.
each task. Depression was assessed with the Geriatric
The authors concluded that category ftu- Depression Scale (GDS; Yesavage et al.,
ency appears to be disproportionately reduced 1983). Functional status was measured with
compared with letter ftuency in normal aging, the Functional Assessment Scale, which is a
which would be consistent with some degra- minor modification of functional scales de-
dation of semantic memory systems. veloped by Lawton and Brody (1969).
A hierarchical regression was used to ex-
Study strengths amine the incremental effects of age, educa-
1. Administration procedures are well out- tion, depression, and functional status on VF
lined. performance. Age alone explained 15.8% of
.2. Sample composition is well described in the variance in VF scores, while age and ed-
terms of age, education, gender, and ucation together accounted for .25 ..2% of the
verbal intelligence (based on Vocabulary variance. Depression was associated with de-
score). creased scores on VF only in functionally in-
3. Strict subject selection criteria were used. dependent adults. 11te role of this finding in
4. Data are stratified by age group. differential diagnosis of depression in older
5. Means and SDs for the test scores are adults is underscored by the authors.
reported.
6. Sample sizes for each age grouping ap- Study strengths
proximate 50. 1. Data for two older groups and a young
group are presented.
Consideration regarding use of the study .2. Sample composition is well described in
1. High educational level of the sample. terms of age, native language, recruit-
ment procedures, and education.
[VF.12] Norris, Blankenship-Reuter, 3. Sample sizes for each group approach
Snow-Turek, and Finch, 1995 (Table A11.16) 50.
The study addressed the effect of depression 4. Test administration procedures are spec-
on cognitive deficits in the elderly. Partici- ified.
pants were 54 community-living elderly, 5. Means and SDs for the test scores are
35 institutionalized elderly, and 40 young reported.
adults who were paid or received course
credit for their participation and whose first Considerations regarding use of the study
language was English. 1. There is a considerable difference in
The first group included independently liv- educational level between the two elderly
ing individuals aged 60-86 (M = 73.1, SD = groups, which might partially account for
6.1), who were solicited through ads and the differences in VF performance.
personal references. Participants comprising .2. Data are of limited clinical use due to
the second group were aged 6.2-89 years overinclusive age ranges for each group.
(M = 75.3, SD = 7.5), were living in institu- 3. Subject exclusion criteria were not
tional settings (skilled-care and intermediate- specified.
216 LANGUAGE
4. No data on intellectual level or gender pants' group assignment. Time to completion

are provided. was reported for the entire sample. In addition,
5. High educational level of the two control the authors provided optimal cutoff scores and
(noninstitutionalized) groups. sensitivity/specificity for the diagnosis of OAT:
90/83% for Category Fluency at a cutoff of
[VF.13] Cahn, Salmon, Butters, Wiederholt, 32 words and 76169% for Letter Fluency at a
Corey-Bloom, Edelstein, and Barrett-Connor, cutoff of 31 words.
1995 (Table A 11.17)
The study examines the accuracy of neu- Study strengths
ropsychological measures in detecting De- 1. Large sample size.
mentia of the Alzheimer's Type (OAT) in a 2. The sample composition is well described
community-dwelling elderly sample. Partici- in terms of age, education, gender, DRS
pants are stable, upper middle-class, retired score, geographic area, history of the
older adults who entered the Rancho Ber- project, and recruitment procedures.
nardo Study, surveying for heart disease risk 3. Rigorous exclusion criteria.
factors, between 1972 and 1974. The initial 4. Test administration procedures are
sample included 5,052 adults 30-79 years old, specified.
who have been followed until the present. 5. Means and SDs for the test scores are
Participants over the age of 65 who returned reported.
for reexamination in 1988 and later and 6. Sensitivity and specificity for optimal
screened positive for cognitive impairment cutoff scores for the two parts of the test
were seen in clinic for diagnostic purposes are reported.
(n = 199). A matched control sample of 203
normal elderly participants who screened Considerations regarding use of the study
negative for cognitive impairment was ran- 1. The data are not partitioned by age
domly selected for comprehensive evaluation, group.
which included neurological examination, 2. No information on IQ is reported.
neuropsychological assessment, standard
medical history and examination, and, in some [VF.14] lvnik, Malec, Smith, Tangalos, and
cases, CT scans of the brain. Petersen, 1996 (Table A11.18)
On the basis of the diagnostic evaluation, The study provides age-specific norms for the
the group composition was re-assessed. The COWA obtained in Mayo's Older Americans
final sample of normal elderly included Nonnative Studies (MOANS), which aim at
238 participants (97 males, 141 females) with obtaining nonnative data for elderly individu-
a mean age of 78.4 (6.8), mean education of als on different neuropsychological tests. The
13.8 (2.6), and mean DRS score of 136.8 (5.4). total sample consisted of 746 cognitively nor-
The Letter Fluency Test was administered mal volunteers over age 55,743 of whom took
according to the procedures described by Bor- the COWA. Mean MAYO FSIQ (which differs
kowski et al. (1967). We infer that the FAS somewhat from standard WAIS-R FSIQ) for
version of this test was administered. The Cat- the whole sample was 106.2 (±14.0), and mean
egory Fluency Test was not described by the Mayo General Memory Index on the Wechsler
authors. In their earlier publication (Wie- Memory Scale-Revised (WMS-R) was 106.2
derholt et al., 1993), the authors used data for (±14.2). For a description of their samples, the
Animal Naming. However, the values for Cat- authors refer to their earlier publications.
egory Fluency provided by the authors in Calm Participants were independently functioning,
et al. (1995) seem to be too high for one 60- community-dwelling persons who were re-
minute Animal Naming trial. Therefore, data cently examined by a physician and had no
for the Category Fluency Test will not be reactive neurological or psychiatric disorder with
produced in this book. The tests were admin- the potential to impact cognition.
istered as part of a larger battery by a trained Age categorization used the midpoint in-
psychometrist who was blind to the partici- terval technique. The raw score distribution
for each test at each midpoint age was "nor- Study strengths
malized" by assigning standard scores with a 1. Information regarding age, education,
mean of 10 and SD of 3, based on actual gender, ethnicity, occupation, recruit-
percentile ranks. The authors provide tables ment procedures, and geographic area is
of age-corrected norms for each age group reported.
(see below). The procedure for clinical appli- 2. The data were stratified by age group
cation of these data is described in the original based on the midpoint interval tech-
article (Ivnik et al., 1996) as follows: nique.
3. The innovative scoring system was well
first select the table that corresponds to that per-
described. The authors developed new
son's age. Enter the table with the test's raw score;
indices of performance.
do not use corrected or final scores for tests that
4. The sample sizes for each group are
might present their own age- or education-adjust-
large.
ments. Select the appropriate column in the table
5. Restricted age range in each cell.
for that test. The corresponding row in the left-
most column in each table provides the MOANS
Age-Corrected Scaled Score . . . for your sub-
1. The measures proposed by the authors
ject's raw score; the corresponding row in the right-
are quite complicated and might be dif-
most column indicates the percentile range for that
ficult to use in clinical practice.
same score.
2. Participants with prior history of neuro-
Further, linear regressions should be ap- logical, psychiatric, or chronic medical
plied to the normalized, age-corrected illnesses were included.
MOANS scaled scores (A-MSS) derived from 3. It is assumed that the authors used the
the tables to adjust the patient's score for CFL set of letters based on specification
education. Age- and education-corrected that the MAE COWA was used. How-
scores for the COWA {A&E-MSS) can be ever, due to frequent reporting of FAS
calculated as follows: as "COWA," this assumption is only
tentative.
A&E-MSScowA = K + (Wt * A-MSScowA)
- (W2 • Education) Other comments
1. The theoretical assumptions underlying
where the following indices are specified for this normative project have been pre-
the COWA: sented in Ivnik et al. (1992a,b).
K 3.50
2. The authors cautioned that the _validity
Wt 1.16
of the MAYO indices depends heavily on
w2 o.40 the match of demographic features of
the individual to the normative sample
Education should enter the formula as years presented in this article.
of formal schooling. 3. Correlations of COWA with age and
The tables of scaled scores per age group gender were -0.15 and 0.12, whereas
provided by the authors should be used in the correlation with education was 0.38. The
context of the detailed procedures for their authors underscore the effect of educa-
application, which are explained in Ivnik et al. tion on test scores.
(1996). Therefore, they are not reproduced
in this book. Interested readers are referred [Vf.15] Ruff, Light, Parker, and Levin, 1996
to the original article. Table A11.18 summa- (Tables A11.19, A11.20)
rizes sample sizes for different demographic The authors summarized the history of VF
groups. tasks. Their normative study was based on
A follow-up article by this group (Lucas 360 native English-speaking normal volun-
et al., 1998) provides MOANS normative data teers 1~70 years of age and ranging in edu-
for category fluency. cation 7-22 years, who resided in mostly
218 LANGUAGE
urban/suburban areas of California, Michigan, participants, who were retested after a

and the eastern seaboard. Participan1S with a 6-month delay with the alternate version. The
positive history of psychiatric hospitalization, order of versions administered was the same
chronic polydrug abuse, or neurological dis- for all participants. The results yielded an
orders were excluded from the sample. acceptably high test-retest reliability coeffi-
The COWA (letters CFL and PRW) was cient (r=0.74). However, a gain of about
administered as part of a comprehensive neuro- three words on the retest was interpreted by
psychological battery. The standard adminis- the authors as a practice effect.
tration procedure was used. The investigators The authors pointed out that the raw scores
instructed participants to exclude proper for the FAS and COWA versions of the test
names and same words with different ~ndings. are not comparable. However, percentiles or
Numbers of correctly generated wofds and standard scores are comparable, based on a
perseverative errors were recorded. : comparison with other studies.
Total numbers of words for the thre' letters In a follow-up publication based on the
are reported for three educational grqups for same sample (Ruff et al., 1997), the authors
males and females separately. ~ provided rates of perseveration for different
Analyses revealed that age did not' have a age groups and proposed that a reduction in
significant effect on word generation. Gender word Huency is not only linked to left pre-
moderated the effect of education, ~d edu- frontal damage but also can be determined by
cation alone accounted for 8% of total vari- diffuse, multifocal, and nonfrontal damage.
ance. The authors proposed correctim~ factors
computed for each cell in a gender-lly-edu- Study strengths
cation group matrix. A table of percentile 1. The sample composition is well de-
ranks and normalized T scores from the scribed in terms of geographic area, age,
360 participants is provided (Table AU.20). education, gender, and IQ.
According to the design of the t~t. the 2. Test instructions are given in detail.
three letters differ in terms of diffic~. The 3. The data are presented in an education-
authors confirmed that the mean pr~uction by-gender matrix.
for letters C (14.1, SD=4.15), F: (13.3, 4. Correction factor and T-score/percentile
SD=4.10), and L (12.7, SD=4.0) was sig- equivalents for the raw scores are pro-
nificantly different. vided.
Analysis of errors (repetitions or : perse- 5. Exclusion criteria are adequate.
verations) revealed an effect of age on the 6. The sample size is sufficiently large for
error rate, with those 16-24 years old! perse- the elaborate analyses conducted by the
verating at a much lower rate th~ those authors. The data cover a wide age span.
25-79 years old. Based on their analysis of 7. Means and SDs for the test scores are
error rate, the authors proposed the following reported.
cutoff scores:
0 perseverations represent intact . perfor-
1. The raw data for separate age groups are
mance (56% of the total sample)
not presented.
1 low average (26%)
2. Data for the retest are not provided.
2 borderline (11%)
3. High intellectual (WAIS-R FSIQ 11~
3 deficient (5%)
111) and educational (14 years) levels.
;?:4 seriously deficient (2%)
Analysis of internal consistency revealed a Other comments
high coefficient a (r=0.83) for the thfe let- A raw score for a given individual must be
ters, indicating high test homogeneity. education-adjusted according to Table A11.19.
Test-retest reliability assessment w~ based Then the percentile and T-score ranking can
on five or more randomly selected parti,ipants be obtained by comparing an education-ad-
from each cell, resulting in a total ;f 120 justed score to Table A11.20.
[VF.16] Hoff, Riordan, Monis, Cestaro, logical or psychiatric disorder, drug or alco-
Wieneke, Alpert, Wang, and Volkow, 1996 hol abuse, and head trauma. Data for a
(Table A11.21) sample of 300 participants with a median ed-
The authors used the COWA in a study of the ucational level of 10 years were analyzed.
relationship of cocaine use to performance on Participants ranged in age 16-75 years, with a
neuropsychological tests tapping functions of mean of 38.4 ( 13.5) years. Education ranged
frontal and temporal brain regions. Perfor- 1-20 years, with a mean of 10.7 (5.1) years.
mance of crack cocaine users was compared to Male to female ratio was 40%/60%. The av-
that of a control group, which consisted of 54 erage duration of residence in the United
paid male volunteers with a mean age of 32.1 States was 16.4 (14.4) years. Seventy percent
(9. 7) years and mean education of 15.4 (2.4) of the sample were monolingual Spanish-
years. The sample included 48 white, 4 black, speaking, and 30% were bilingual. The pro-
and 2 Hispanic participants. Exclusion criteria portion of the sample respective to their
were a history of medical, neurological, or country of origin closely approximates the
psychiatric problems; more than moderate use 1992 U.S. Census distribution. Correlations
of alcohol (12 oz/week); history of intravenous between Marin and Marin (1991) accultura-
drug use; and self-reported history of learning tion scale scores and neuropsychological var-
disability (with enrollment in special educa- iables are provided.
tion classes). The FAS test was administered in the par-
ticipants' native language, Spanish.
In the follow-up study on the factor struc-
Study strengths
ture of the NeSBHIS (Ponton et al., 2000),
1. Sample size is >50.
which extracted five factors, the FAS primarily
loaded on the Language factor, with a var-
terms of age, education, and ethnicity.
imax-rotated factor loading of 0.71.
1. Large overall sample, with acceptable
Considerations regarding use of the study sample size for most of the cells.
1. It is unclear which version of the test was 2. The sample composition is well described
administered. in terms of age, education, gender, ac-
2. Wide age and education range. No in- culturation information, geographic area,
formation on IQ is reported. and recruitment procedures.
3. Recruitment procedures were not re- 3. Adequate exclusion criteria.
ported. 4. Test administration procedures are
4. Educational level for the sample is specified.
high. 5. Means and SDs for the test scores are
reported.
[VF.17] Ponton, Satz, Herrera, Ortiz, Urrutia, 6. Data are partitioned by gender xage x
Young, D'Eiia, Furst, and Namerow, 1996 education.
(Table A11.22)
The F AS version was administered to Spanish- Considerations regarding use of the study
speaking volunteers as part of a larger battery 1. It is unclear whether the administration
in a project designed to provide standardiza- procedure required restrictions in the
tion of the Neuropsychological Screening types of word to be used in the process
Battery for Hispanics (NeSBHIS). Volunteers of word generation.
were recruited through fliers and advertise- 2. No information on IQ is reported.
ments in community centers of the greater 3. It is unclear which of the two educa-
Los Angeles area over a period of 2 years. tional groups included participants with
Exclusion criteria were a history of neuro- 10 years of education.
220 LANGUAGE
[Vf.18] Crossley, D'Arcy, and Rawson, 1997 as determinants of hierarchical organization of

(Table A11.23) semantic memory. Performance of an Alzhei-
The authors compared performance on letter mer's group was compared to that of an el-
and category fluency in a sample of cognitively derly control group, which consisted of
normal seniors (n=635) and in samples ofDAT 38 volunteers: 18 males and 20 females. None
and vascular dementia patients participating of the participants had a history of major
in the Canadian Study of Health and Aging. psychiatric or medical illness, drug or alcohol
The control sample included community- abuse, head injury, learning disability, or other
dwelling individuals who were screened for neurological disease. Standard procedures for
cognitive impairment using the Modified administration of the FAS and Animal Nam-
Mini-Mental State Examination (3MS). All paring versions were used. Responses were re-
ticipants were fluent in either English or corded on audiotape and later analyzed.
French. A detailed overview of the study par- In the follow-up studies on VF mechanisms in
ticipants, methods, and findings is provided by Alzheimer's and Parkinson's diseases (Tr6ester
the Canadian Study of Health and Aging et al., 1998; Piatt et al., 1999a), the authors ap-
Working Group (1994). parently used the same control sample (at least
Letter fluency was assessed with the F AS in part). Therefore, the data from these articles
task, administered in three 60-second trials. will not be reproduced in this book.
Participants were instructed to avoid proper
nouns and the same word with a different Study strengths
suffix. Category fluency was assessed with the 1. The sample composition is described in
animal name generation task, within a 60- terms of age, gender, and education.
second interval. 2. Rigorous exclusion criteria.
The data are reported by age group, gender, 3. Administration procedure is well de-
and educational level. scribed.
1. Administration procedures are well out-
lined. Considerations regardtng use of the study
2. Sample composition is well described in 1. The sample is relatively small.
the previous reports. 2. Recruitment procedures were not re-
3. Subject selection criteria are outlined. ported.
4. Data are stratified by age group, gender, 3. No information on IQ is reported.
and education.
[VF.20] Nybers, Winocur, and Moscovitch,
1997 (Table A11.25)
reported.
6. Sample sizes for each demographic The FAS word fluency test was administered
grouping are very large. as part of a test battery sensitive to medial-
temporal and frontal lobe function in a study
Considerations regarding use of the study investigating age-related differences in the
1. Data were collected in Canada and, effect of lexical priming on memory. The
therefore, might be of limited use in the sample included 39 healthy elderly partici-
United States. pants who ranged in age 66-87 years, with a
2. It is unknown to what extent having mean age of 77.3 years. Education ranged 8-
some data collected in French impacted 22 years, with a mean of 13.6. Performance on
the overall results. the WAIS Vocabulary test was used as a
screening measure.
[VF.19] Beatty, Testa, English, and Winn, 1997
(Table A11.24) Study strengths
The authors used FAS and Animal Naming to 1. The sample composition is described in
investigate clustering and switching strategies terms of age and education.
2. Test administration procedures are spec- furniture, and vegetable categories were used
i£i.ed. in the category fluency test. The data are re-
3. Means and SDs are reported for the ported for each trial separately.
FAS.
Study strengths
Considerations regarding use of the study 1. Sample size is large.
1. The sample is relatively small. 2. The sample composition is well de-
2. Exclusion criteria are not described. It is scribed in terms of age, education, gen-
unclear which version of the WAIS was der, and various health indices.
administered and what performance on 3. Recruitment procedures are speci£i.ed.
Vocabulary served as a cutoff for inclu- 4. Data are partitioned into three age
sion into the study. groups.
3. Recruitment procedures are not re- 5. Test administration procedures are spec-
ported, and gender distribution is not i£i.ed.
specified. 6. Means and SDs for the test scores are
4. It is unclear whether the administration reported.
procedure required restrictions in the
types of word to be used in the process Considerations regarding use of the study
of word generation. 1. Exclusion criteria are not well identified.
5. SDs for age and education are not re- 2. High educational level for each age group.
ported.
6. The data were obtained on Canadian [VF.22] Kempler, Teng, Dick, Taussig, Davis,
and/or Swedish participants, which may 1998 (Table A 11.27)
limit their usefulness for clinical inter- The Animal Naming test was administered to
pretation in the United States. 317 Chinese, Hispanic, and Vietnamese im-
migrants, speaking primarily their native lan-
[VF.21] Salthouse, loth, Hancock, and guage, and to white and African-American
Woodard, 1997 (Table A11.26) English speakers 54-99 years old. Participants
The authors examined controlled and auto- generated animal names in their native lan-
matic processes underlying memory and guage. The test was administered as part of a
attention using the process-dissociation pro- normative study for the Cross-Cultural Neuro-
cedure, as well as age-related influences on psychological Battery. Volunteers who had a
these processes. Participants were 115 healthy history of stroke, head injury, or psychiatric,
adults (47% male, 53% female) aged 18-78 speech, language, or memory problems, as re-
years, who were recruited from appeals to ported on a self-rated health history ques-
groups and acquaintances. They were in- tionnaire, were not included in the study.
cluded in the study if reported to be in "rea- The standard administration procedure was
sonably good health,'' to not be a current used.
student, and to have at least 11 years of edu- The results indicated an inverse relation-
cation. No other exclusion criteria are reported. ship of word fluency with age and a positive
Participants were administered a battery of relationship with education. A pronounced
neuropsychological tests in their homes. The effect of native language was also noted (see
data were stratified into three age groupings: above).
18-39 years [mean age= 29.0 (4.8); mean
education= 15.5 (1.7)], 40-59 years [mean Study strengths
age=49.1 (5.1); mean education= 15.2 (2.5)], 1. Large sample.
and 60-78 [mean age= 69.2 (5.1); mean ed- 2. The sample composition is well described
ucation= 15.3 (2.6)]. in terms of age, education, gender, eth-
Letters C, F, and L were used in the letter nicity, and information on acculturation
fluency test, with the constraint that none of level for the immigrant groups.
the words should be proper nouns. Animal, 3. Adequate exclusion criteria.
222 LANGUAGE
4. Test administration procedures are spec- males scoring considerably higher than
ified. younger females on the letter fluency
5. Means and SDs for the test scores are task.
reported, grouped by age, education, 2. Recruitment procedures are not reported.
gender, and ethnicity. 3. The data were obtained on Canadian
Consideration regarding use of the study fulness for clinical interpretation in the
1. No information on IQ is reported. United States.
[VF.23] Stuss, Alexander, Hamer, Palumbo, [VF.24] Johnson-Selfridge, Zalewski, and

Dempster, Binns, Levine, and lzukawa, 1998 Aboudarham, 1998 (Table A11.29)
(Table A11.28) The authors examined the effect of ethnicity
The study addresses the effect of brain lesion on word fluency, measured with the F AS
location and etiology on VF. The control and Animal Naming versions. The sample in-
group included 37 participants (19 males, cluded white, black, and Hispanic male ve-
18 females) without neurological or psychiat- terans, with 200 participants in each group,
ric disorder, with mean age of 54.4 (14.4) who were randomly drawn from a larger
years and mean education of 13.9 (2.3) years. sample of 4,462 veterans participating in the
Mean NART-estimated IQ was 113.8 (6.1). Vietnam Experience Study. Hispanic partici-
The letter fluency task (FAS) was adminis- pants were not differentiated by country of
tered according to Benton and Hamsher's origin or primary language. However, the au-
(1978) instructions (numbers were not ex- thors stated that <2% of the test results were
cluded according to the instructions). Se- considered questionable due to language or
mantic fluency was measured with the animal other problems. The recruitment procedures
name generation task. Number of target are not described, though a reference is made
words generated, different error types, and to earlier articles describing this study. It is
measures of clustering were recorded. Mea- unclear if any exclusion criteria were used.
sures of VF correlated with age but not with The sample includes participants with a his-
education or NART IQ. tory of motor vehicle accidents, loss of con-
Normative data for letter and semantic flu- sciousness, seizures, medical conditions, and
ency tasks for three age groups (21-39, 40-64, lifetime psychiatric disorders. However, the
65-81 years) stratified by gender are provided. statistics provided by the authors suggest that
The authors reviewed the results in light of the three ethnic groups did not differ signifi-
the relationship between different cognitive cantly in the number of participants meeting
processes and brain regions. criteria for various conditions.
Participants ranged in age 31-46 years, with
Study strengths a mean of 37.9 (2.61) years, and in education
1. The sample composition is well de- 2-18 years, with a mean of 13.2 (2.25) years.
scribed in terms of age, education, The authors also provided information on
gender, and estimated IQ. handedness, place of service, income, and
2. Adequate exclusion criteria. WRAT-R reading scores.
3. Test administration procedures are de- The F AS and Animal Naming tests were
scribed. administered according to standard instruc-
4. The data are stratified by three age tions. Data are reported for each ethnic group
groups and by gender. separately in raw scores, scores adjusted for
5. Means and SDs for the test scores are demographics, and T scores. Only raw scores
reported. and demographically adjusted scores are re-
produced in this chapter.
Considerations regarding use of the study The authors found that income, educational
1. Small sample size and data are incon- level, and WRAT-R Reading subtest scores
sistent across age groups, with older fe- were significantly related to test performance.
Differences between ethnic groups in VF 81 of whom were tested twice. The partici-
were small and influenced by other variables, pants in this group had no histOI)' of recent
but they may have important clinical im- trauma and were friends of head-injured pa-
plications, according to the authors. tients. Their mean age was 28.5 (12.2) years
and mean education was 12.2 (1.9) years; 60%
Study strengths of the sample were males, and the test-retest
1. Large sample size. interval was 11.1 (.6) months. Participants
2. The sample composition is well de- were tested at the University of Washington
scribed in terms ofage, education, gender, under the direction of one of the authors.
handedness, ethnicity, income, WRAT-R Twenty percent of the sample had preexisting
Reading subtest scores, setting, place of conditions that might affect test performance,
military service, and number of partici- the most significant being alcohol abuse and a
pants with various medical and psychi- significant traumatic brain injul)'. The rest of
atric conditions in each group. the participants denied any histOI)' of condi-
3. Test administration procedures are tions that might be expected to affect brain
specified. function. The other two groups used in this
4. Data for three ethnic groups are pro- article do not have data for the F AS and
vided. therefore will not be described in this chapter.
5. Means and SDs for the test scores are The mean WAIS FSIQ (Wechsler, 1955) on
reported. initial testing for the three groups combined
was 108.8 (12.3).
Considerations regarding use of the study The F AS was given according to standard
1. It is unclear if any exclusion criteria were instructions (Lezak, 1995). An alternate com-
used. bination of letters (B, D, T) was used on the
2. Recruitment procedures are not re- retest. The authors provide raw scores for
ported, though a reference is made to performance at two time probes, as well as
earlier publications. various measures of test-retest reliability and
3. All-male sample of veterans. magnitude of practice effect. The test-retest
4. No information on IQ is reported, reliability for the FAS over 11.1 months was
though WRAT-R Reading subtest scores r=0.72
are provided.
5. The sample includes participants with a Study strengths
histol)' of various medical (including 1. Relatively large sample.
neurological) and psychiatric conditions. 2. The sample composition is well de-
However, the authors specified that the scribed in terms of age, education, gen-
ethnic groups did not differ significantly der, IQ, geographic area, and setting.
in the number of participants meeting 3. Test administration procedures are spec-
criteria for the various conditions. This ified.
lends support to the validity of the com- 4. Means and SDs for the test scores are
parative analyses of test performance reported.
across the three groups in this study but 5. Information on test-retest reliability is
raises concerns about the use of these provided.
data with nonveteran, medically and
psychiatrically healthy individuals. Considerations regarding use of the study
1. Exclusion criteria are not clearly de-
[VF.25] Dikmen, Heaton, Grant, and Temkin, scribed. As authors pointed out, twenty
1999 (Table A 11.30) percent of the sample had preexisting
The F AS was used in a study on the psycho- conditions that might affect test perfor-
metric properties of a broad range of neuro- mance, the most significant being alcohol
psychological measures based on a sample of abuse and a significant traumatic brain
138 normal or neurologically stable adults, injul)'.
224 LANGUAGE
I
2. The data are not partitioned · by age Three category fluency conditions were
group. used-animals, food, and clothing-with stan-
dard administration procedures for the Boston
[VF.26] Manly, Jacobs, Sano, Bell, Mer+ant, Diagnostic Aphasia Examination (BDAE).
Small, and Stem, 1999 (Table A 11.31) ; The score represents the number of words
Category fluency tests were used in :a study averaged over the three conditions.
comparing neuropsychological test 1 perfor- The authors concluded that category flu-
mance among nondemented literate d illit- ency is not affected by literacy status.
erate elders. The sample was select d from
participants in a community-based e demio- Study strengths
logical study of normal aging and 1. Large overall sample size.
in the ethnically diverse neighbor 2. The overall sample is described in terms
northern Manhattan, New York. The pula- of age, education, gender, ethnicity, geo-
tion from which the sample was :wn is graphic area, setting, recruitment pro-
comprised of individuals from sever~ differ- cedures, and sampling methods.
ent countries, representing three ethtfic cate- 3. Adequate exclusion criteria.
gories: Hispanic, African Americ~ and 4. Test administration procedures are
white. specified.
The sample was restricted to elders ged 65 5. Means and SDs for the test scores are
and above with 0-3 years of formal ed cation. reported.
Exclusion criteria were Parkinson's · ease, 6. Data for illiterate and low-education
stroke, or head injury. All particip~ were samples are provided.
found to be nondemented by a neuro gist.
The final sample included 123 lite te and Considerations regarding use of the study
64 illiterate elders. Separate analys~s were 1. Demographic characteristics for three
performed for the following groups: : out of four groups are not provided.
I 2. The data are not partitioned by age
1. Stratified mndom sample: Gr~ps of group.
education-matched literate and Q.literate 3. No information on IQ is reported. How-
participants (n = 43 for each group) were ever, WAIS-R Similarities performance
created using a stratified randobt sam- is reported.
pling method. The literate grou~ had an
average age of 76.2 (6.1) years pod in- [VF.27] Boone, 1999 (Table A11.32)
cluded 81 %-Hispanic, 9% AfricaD!Ameri- The chapter summarizes the results of a study
can, and 10% non-Hispanic 1 white on the effect of aging, demographic factors,
participants. The illiterate groug had a and medical conditions on executive functions,
mean age of 74.8 (5.7) years ~d in- which were presented in earlier publications
cluded 91% Hispanic and 9% ;African (Boone et al., 1990, 1995). Participants are
American participants. Both grout>s were 155 healthy elderly volunteers (53 males, 102
74% female. · females) aged 45-84 years, with a mean age of
2. Uneducated sample: Data for lterates 63.07 (9.29), mean education of 14.57 (2.55),
(n = 26) and illiterates (n = 47) l-ith no and mean FSIQ of 115.41 (14.11). All parti-
formal education were analyze4 sepa- cipants were fluent English speakers and were
rately. ; recruited through newspaper ads. Participants
3. Stratified random Spanish-sfleaking underwent physical and neurological ex-
sample: Stratified random sanjple of aminations and psychiatric interviews. Rigor-
education-matched literate and illiterate ous exclusion criteria were used, including
elders (n = 32 for each group). ; history of psychosis, major affective disorders,
4. Uneducated Spanish-speaking 4ample: alcohol dependence, neurological disorders,
Uneducated literate (n = 17) an4 illiter- and serious metabolic abnormalities. Fre-
ate (n=43) elders. quency of vascular illnesses and intake of
cardiac and/or antihypertensive medications The standard administration procedure was

was recorded. used.
The F AS version of the test was used.
Normative data are stratified by IQ level (av- Study strengths
erage, high average, superior) based on per- 1. The sample composition is described in
formance on the Satz-Mogel abbreviation of terms of age, education, gender, and geo-
the WAIS-R. graphic area.
The results identified the FSIQ as the only 2. Test administration procedure is speci-
significant predictor of F AS performance, re- fied.
sponsible for 15% of test score variance, based 3. Means and SDs for the test scores are
on stepwise regression analysis. reported.

1. The sample size is large. 1. The sample is small.
2. Composition of the sample is well de- 2. Exclusion criteria are not clearly de-
scribed in terms of IQ, age, fluency in scribed.
English, education, gender, and recruit- 3. It is unclear which version of the test was
ment procedures. administered.
3. Rigorous exclusion criteria. 4. Recruitment procedures were not re-
4. Normative data are stratified by IQ. ported.
5. Means and SDs for the test scores are 5. No information on IQ is reported.
reported.
[VF.29] Epker, Lacritz, and Cullum, 1999
Considerations regarding use of the study (Table A11.34)
1. Age and education for each of the three The authors used F AS and Animal Naming in a
IQ groups are not provided. study of the diagnostic utility of a qualitative
2. Education and intelligence levels of the scoring technique for fluency tasks in Alzhei-
sample are high. mer's and Parkinson's diseases. The control
3. Data are not presented by age group- group included 65 elderly participants with a
ings. mean age of 70.6 years (4.7), mean education
of 14.3 (2.9) years, and a male/female ratio of
[VF.28] Demakis, 1999 (Table A11.33) 22/43, who participated in an investigation of
The authors used the COWA as part of a cognitive function in aging. They were screened
battery in a study of response consistency for health problems using a semistructured
across a 3-week interval in an analog malin- neuromedical interview. Participants did not
gering design. Data are presented for control have a known history of substance abuse, major
and dissimilation groups. All participants mental illness, learning disability, neurological
were students from undergraduate psychology disease, or major psychopathology.
courses at a small midwestern liberal arts Standard administration procedures were
college. The control group consisted of used.
21 participants with a mean age of 22.5 years
(7.99) and mean education of 13.6 (1.46) Study strengths
years; 67% were female. Control participants 1. Relatively large sample.
were told that they were in a car accident but 2. The sample composition is well de-
that they had not suffered any injuries and scribed in terms of age, education, gen-
were instructed to perform to the best of their der, and MMSE score.
ability. Participants were retested 3 weeks af- 3. Adequate exclusion criteria.
ter the initial testing. Control participants 4. Test administration procedures are well
demonstrated a practice effect on the retest. specified.
Only data for the initial testing probe for the 5. Means and SDs for the test scores are
control group are replicated in this book. reported.
226 LANGUAGE
Considerations regarding use of the siludy was not found to affect performance on either
1. Recruitment procedures were ~ot retest.
ported. '
2. Educational level for the sample ?is high. Study sfrengths
3. No information on IQ is reportetl. 1. Large sample.
[VF.30] Tombaugh, Kozak, and Rees, 1gf9 scribed in terms of age, education, gen-
(Tables A11.35-A11.37) der, and recruitment procedures.
The article provides normative data fur FAS 3. Adequate exclusion criteria.
and Animal Naming stratified by three 1evels of 4. Test administration procedures are spec-
age (16-59, 60-79, 80-95) and three ~vels of ified.
education (0-8, 9-12, 13-21), as weQ as for 5. Means and SDs for the test scores are
nine age groups, four education gro~s, and reported.
the two genders separately. The total!sample 6. Data are stratified by age, education,
included participants from two differelt stud- gender, and age x education.
ies. Participants were recruited throu~ booths
at shopping centers, social organizatic:fs, pla- Considerations regarding use of the study
ces of employment, psychology classfs, and 1. The data were obtained on Canadian
word of mouth. Volunteers with a kn~ his- participants, which may limit their use-
tory of neurological disease, psychiatri~ illness, fulness for clinical interpretation in the
head injury, or stroke were excluded ~m the United States.
study. A subsample of participant$ were 2. No information on IQ is reported.
judged to be cognitively intact on the ~asis of
history, clinical and neurological exa~ation, [VF.31] Basso, Bomstein, and Lang, 1999
and an extensive battery of neuropsycht>logical (Table A11.38)
tests. All participants stated that Engijsh was The study examined the practice effect on
their first language. : repeated administration of several tests over a
The subset of the sample for the F~S test 12-month interval. The baseline sample con-
included 895 participants aged 16-9~ years, sisted of 82 men recruited through newspaper
with a mean age of 60.7 years (19.9), 4J1d ed- advertisements, who were not paid for their
ucation ranging 0-21 years, with a me+n edu- participation. Fifty men out of this sample
cation of 12.1 (3.2). The male-to-female ratio returned for the repeated testing 12 months
was 559n4L I later. The composition of the latter sample
The subset of the sample for the Animal was 48 Caucasian, 1 African American, and
Naming test included 735 participanis aged 1 Hispanic, with a mean age of 32.5 (9.27)
16-95 years, with a mean age of 67.0 years years, mean education of 14.98 (1.93) years,
(19.8), and education ranging 0-21 yeats, with and mean FSIQ of 109.30 (12.29) at baseline.
a mean education of 11.4 (3.4). The n.ale-to- At each probe, participants were screened
female ratio was 310/425. . for neurological disease, head injury, learn-
The standard administration pr~dures ing disabilities, or other medical illnesses
were used, with the exception that n*mbers based on an informal interview. They were
were allowed on the F AS test. Meru} num- also screened for psychiatric disorders through
bers of words are presented for four edpcation a structured clinical interview. None was ex-
groups, nine age groups, and the two genders cluded based on these screens.
separately. Percentile scores and mea+ num- The F AS was administered according to
ber of words are also presented in thtee age standard procedures by thoroughly trained
(16-59, 60-79, and 80-95) by three edrcation and supervised technicians. The authors
(0-8, 9-12, and 13-21) cells. . compared FAS performance at baseline and
FAS was found to be more sens~ve to on the retest using reliable change indices and
the effects of education than age. For f\nimal concluded that FAS scores did not change on
Naming, the relationship was opposite. (;ender the retest.
The number of words generated on the disorder, other major psychiatric illness, cur-
FAS for the two probes, with age, gender, and rent substance dependence or abuse within
education corrections applied, is reported for the last 6 months, or primary language other
the entire sample. than English.
F AS and Animal Naming were adminis-
Study strengths tered. According to the F AS instructions,
1. Adequate sample size. proper names and plurals were excluded.
2. The sample composition is described in Total number of words generated for three
terms of age, education, gender, ethni- FAS trials and for Animal Naming are re-
city, FSIQ, and recruitment procedures. ported for the sample stratified by three age
3. Adequate exclusion criteria. groups (20-34, 35-49, 50-101 years) and
4. Test administration procedures are three education groups (0-11, 12-15, 16-
thoroughly described. 20 years). Data stratified by age are also pre-
5. Means and SDs for the test scores are sented for African Americans and Caucasians
reported. separately. In addition, multiple regression
analyses were used to develop equations for
Considerations regarding use of the study demographic corrections. Tables for conver-
1. The data are not partitioned by age sion of raw scores to demographically cor-
group. rected T scores were provided by the authors.
2. Educational level for the sample is high. Raw scores for FAS and Animal Naming are
reproduced in this chapter.
[VF.32] Gladsjo, Schuman, Evans, Peavy, Miller,
and Heaton, 1999 (Tables A11.39, A11.40) Study strengths
The authors provided normative data and 1. Large sample size.
demographic corrections for age, education, 2. The sample composition is well de-
and ethnicity, for letter and category fluency scribed in terms of age, education, gen-
tasks, based on a sample of 768 normal adults der, ethnicity, geographic area, setting,
aged 20-101 years, with education of 0- and recruitment procedures.
20 years; 55% are Caucasian and 45% African 3. Adequate exclusion criteria.
American; 52% are male. Mean age is 50.4 4. Test administration procedures are
(19.4) years; mean education is 13.6 (3.1) specified.
years. The sample consists of volunteers who 5. Normative data are stratified by age x
were enrolled as normal comparison partici- education for the whole sample and by
pants in various clinical studies at the UDi- age for African Americans and Cauca-
versity of California San Diego. Caucasian sians separately.
participants were recruited through local me- 6. Means and SDs for the test scores are
dia announcements and personal contacts. reported.
Mrican-American participants were part of a
federally funded study (African American Consideration regarding use of the study
Norms Project) and were recruited to match 1. No information on IQ is reported.
the census representation of Mrican Amer-
icans within the larger San Diego area. Parti- [VF.33] Binder, Storandt, and Birge, 1999
cipants were screened with the Structured (Table A11.41)
Clinical Interview for DSM-III-R or based on The authors examined the relationship be-
self-report of no past history of diagnosis or tween performance on psychometric tests and
treatment for an Axis I disorder. Exclusion a modified Physical Performance Test (modi-
criteria were history of significant head trauma fied PPT) in a sample of 125 adults aged
with loss of consciousness for >20 minutes or 75 years and older, who participated in trials
persisting neurological sequelae, neurological of exercise or hormone replacement therapy.
illness, conditions expected to affect neu- The study was approved by the Washington
ropsychological test performance, psychotic University School of Medicine, St. Louis. The
228 LANGUAGE
mean age for the sample was 82.3 (4.4), mean on the relationship between regional brain
education was 13.5 (3.0), 25% were male, and volume and semantic, phonological, and non-
87% were Caucasian. Indices of physical verbal fluency. The control group included
health, Blessed score, and Geriatric Depres- 51 participants with a mean age of 66.7 (7.4)
sion Scale score are reported. Preliminary years and mean education of 16.4 (2.3) years.
screening included a medical history; physical Exclusion criteria were significant history of
examination; the Short Blessed Test of mem- psychiatric or neurological disorder, past or
ory, concentration, and orientation; blood and present alcohol or drug abuse or dependence,
urine chemistries; a chest X-ray; and a cross- or other serious medical condition, as identi-
validated self-report regarding health prob- fied on a psychiatric interview and medical
lems in the previous 12 months. Exclusion examination.
criteria were inability to walk 50 feet inde- The standard administration procedure was
pendently, active medical problems that used for the FAS, with the exception that
would contraindicate performance of a graded participants were not instructed to avoid num-
exercise stress test, inability to complete the bers. Semantic fluency was measured with two
graded exercise stress test or the modified !-minute trials, in which participants were
PPT, a score >8 on the Short Blessed Test, instructed to generate names of animals and
inability to provide informed consent due to names of inanimate objects, respectively.
cognitive impairment, and inability to follow These data were used in a previous article
the directions for the psychometric tests due by Fama et al. (1998) in calculations of stan-
to visual or auditory impairments. dardized z scores for Alzheimer's participants
The test was administered according to that corrected the raw scores for age.
standard instructions.
The authors found that VF was not signifi- Study strengths
cantly associated with total modified PPT 1. Relatively large sample.
score. 2. The sample composition is described in
terms of age and education.
Study strengths 3. Rigorous exclusion criteria.
1. Large sample size. 4. Test administration procedures are spec-
2. The sample composition is well de- ified.
scribed in terms of age, education, gen- 5. Means and SDs for the test scores are
der, ethnicity, indices of physical health, reported.
Blessed score, Geriatric Depression Scale
score, geographic area, and research Considerations regarding use of the study
setting. 1. Recruitment procedures a not reported.
3. Adequate exclusion criteria. 2. Gender distribution is not reported
4. Test administration procedures are spec- 3. The data are not partitioned by age
ified. group.
5. Means and SDs for the test scores are 4. Educational level for the sample is high.
reported. 5. No information on IQ is reported.
Considerations regarding use of the study [VF.35] Troyer, 2000 (Table A11.43)
1. The data are not partitioned by age The study addressed clustering and switching
group. on phonemic and semantic VF tasks in a total
2. No information on IQ is reported. sample of 411 healthy adults aged 18-91. This
is a follow-up on previous publications by
[VF.34] Fama, Sullivan, Shear, Cahn-Weiner, these authors (Troyer et al., 1997, 1998a,b).
Marsh, Lim, Yesavage, Tinklenberg, and The mean age for the sample was 59.8 (20. 7)
Piefferbaum, 2000 (Table A11.42) years, and mean education ranged 5-21 years,
Fluency tests were administered to Alzhei- with a mean of 13.9 (2.9). The male/female
mer's patients and normal controls in a study ratio was 30%nO%. All participants were
fluent in English. Participants were screened fulness for clinical interpretation in the
for neurological or psychiatric disorders. Par- United States.
ticipants aged 2::60 were screened for cogni-
tive decline. Only those participants who [VF.36] Acevedo, Loewenstein, Barker,
obtained MMSE score 2::25 or scores within Harwood, Luis, Bravo, Hurwitz, Aguero,
the normal range on an episodic memory test Greenfield, and Duara, 2000
were included. (Tables A11.44-A11.47)
The F AS version of the phonemic fluency The authors provided normative data for
test was administered to 257 participants and three conditions of the Category Fluency test,
the CFL version, to 154 participants. Standard Animals, Vegetables, and Fruits, for 424
administration procedures were used, with the English-speaking and 278 Spanish-speaking
exception that participants were not instructed participants over the age of 50. The sample was
to avoid numbers. Two 60-minute semantic drawn from a larger pool of community-
fluency trials were administered: animal flu- dwelling individuals who presented for free
ency version was administered to 407 partici- memory screening sessions offered by the Wien
pants; 156 participants from this sample were Center for Alzheimer's Disease and Memory
also administered supermarket fluency. Disorders between 1994 and 1999. Participants
Based on the results of regression analyses, in the English-speaking group spoke English as
the author inferred that age had a greater ef- their prirruuy language and were born in the
fect on semantic than on phonemic fluency. United States. Participants in the Spanish-
Education affected both semantic and pho- speaking group spoke Spanish as the primary
nemic fluency. Gender was not related to VF language and were hom in a country where
performance. Spanish is the primary language. All partici-
pants were screened in their primary language
Study strengths using the MMSE, Hamilton Depression Rating
1. Large sample. Scale (Hamilton, 1960), and questionnaires re-
2. The sample composition is well de- lated to demographic information, medical and
scribed in terms of age, education, gen- psychiatric history, and cognitive status. Only
der, and native language. participants who had MMSE score 2::27 and a
3. Test administration procedures are spec- score 2::10 on four delayed recall trials of the
ified. three words used in the MMSE (based on the
4. Means and SDs for the test scores are cutoff identified in Loewenstein et al., 2000)
reported. were included in the study. For English
speakers, the mean age was 69.1 (6.9) years,
Considerations regarding use of the study mean education was 14.4 (2.5) years, male/fe-
1. Recruitment procedures are not remale ratio was 26%n4%, and mean MMSE
ported. score was 28.9 (1.0). For Spanish speakers, the
2. Participants were screened for neuro- mean age was 64.9 (7.7) years, mean education
logical or psychiatric disorders; however, was 13.4 (3.2) years, male/female ratio was 30.8/
medical exclusion criteria are not re- 69.2%, and mean MMSE score was 28.7 (1.0).
ported. Among English speakers, 99% were classified
3. Demographic characteristics for subsets by the examiner as white, <1% as African
of participants in each condition are not American, and <1% as Asian American. Among
provided. Spanish speakers, 97% were classified as
4. Phonemic fluency norms are provided as white, <1% as black, and 3% as "other."
the mean for FAS/CFL. Three 60-second trials on Animals, Vege-
5. The data are not partitioned by age tables, and Fruits categories, were adminis-
group. tered to all participants.
6. No information on IQ is reported. Normative data are stratified by age, edu-
7. The data were obtained on Canadian cation, and gender. Cells with low sample si-
participants, which may limit their use- zes are not included in the tables.
230 LANGUAGE
The authors concluded that age, ewcation, The authors identified the assessed domains
and gender affect Category Fluency· perfor- as category fluency and initial letter fluency,
mance, with gender being the best predictor with references to Borkowski et al. (1967) and
after adjusting for age and education,· regard- Benton and Hamsher (1976). However, we
less of the primary language. Performances of gathered from the earlier publications based
English and Spanish speakers were shpilar for on this study that category fluency was as-
the Animals and Fruits categories. However, sessed using the Animals and Fruits categories
English speakers generated more words for and initial letter fluency was measured with
the Vegetables category. generation of words starting with the letters P
and S. Evidently, the scores reported by the
Study strengths i authors represent the total number of words
1. Large samples for most of the c+lls. for two conditions within each domain.
2. The sample composition is well d~scribed
in terms of age, education, gender, eth- Study strengths
nicity, language, geographic area., and re- 1. Large sample size.
cruitment procedures. 2. The sample composition is well described
3. Adequate exclusion criteria. in terms of age, education, gender, history
4. Test administration procedures are of the project, and geographic area.
specified. ' 3. Rigorous exclusion criteria.
5. Means and SDs for the test scqres are 4. Means and SDs for the test scores are
reported. reported.
6. Normative data are stratified ~y age,
education, and gender. Considerations regardtng use of the study
Consideration regarding use of the stlf:ly group.
1. No information on IQ is reporteil. 2. No information on IQ is reported.
3. Number of participants with less than
[Vf.37] Chen, Ratcliff, Belle, Cauley, Dekosky, a high school education is reported.
and Ganguli, 2000 (Table A 11.48) ' However, mean education and SD are
A control sample of 483 elderly nondemented not reported.
individuals was derived from a com"unity- 4. The tests and administration procedures
based multiwave prospective study, thei Mono- used are not specified. This information
ngahela Valley Independent Elders ·Survey was gathered from earlier articles de-
(MoVIES), in southwestern Pennsylvania. scribing this study.
The purpose of the study was to 1<1entify
cognitive measures that are most acc*'ate in [VF.38] Anstey, Matters, Brown, and Lord,
discriminating between individuals with pre- 2000 (Tables A 11.49, A 11.50)
symptomatic Dementia of Alzheimer\; Type The authors report normative data for a
(DAT) and nondemented individuals. ~ontrol sample of old and very old Australian adults
participants remained nondemented :over a living in retirement villages and hostels. A
10-year follow-up period. The study Jtotocol sample of 369 participants was drawn from
included a standardized general medifal his- 13 residential sites located throughout the
tory and physical examination; a detailed Sydney and Illawarra regions. The criterion
neurological and mental status e~ation; for inclusion into the subsample of 280 parti-
hematological, metabolic, and sertlogical cipants described in this article was a mini-
tests; and neuroimaging when appr9priate. mum of four neuropsychological tests
Relevant medical records were abstracted. completed. The exclusion criteria were a his-
The sample included 302 females atd 181 tory of Parkinson's disease, stroke, or heart
males, with a mean age of 74.9 (4.4~ years; attack. Participants ranged in age 62-95 years,
31.9% of participants had less than fl high with a mean age of79.04 (6.59) years, and had
school education. mean education of 11.25 (2.79) years; there
were 52 males and 317 females; the majority Normative Aging Study at the Department of
of the sample rated their health as good, very Veterans Affairs Medical Center in Boston. All
good, or excellent. Participants with MMSE participants were free of cardiac disease, hy-
scores < 24 (the lowest recorded score was 17) pertension, cataracts, loss of hearing, or ab-
were included in the sample. normal laboratory tests on entering the study.
The FAS was administered according to At each visit, participants received a thorough
standard instructions. Repetitions and nonmedical exam.
words were excluded. The data are reported A 60-second trial of the Animal Fluency test
in raw scores for a subsample of 280 partici- was administered at each longitudinal session
pants and in percentile distribution for a to 235 stroke- and dementia-free men with a
subsample of 260 participants stratified by mean age of 66.41 (6.73) years and mean ed-
age x education. ucation of 14.03 (2.62) years.
The authors found that performance on The results for the initial testing and for
F AS correlated with education but not with the retest 3 years later are reported in Table
age (which is probably due to the restricted A11.51. The test-retest difference in word
age range). generation was significant at 0.05 level.

1. Large overall sample. 1. Large sample.
2. The sample composition is well de- 2. The sample composition is described
scribed in terms of age, education, gen- in terms of age, education, gender, and
der, health status, range of MMSE geographic area. Recruitment proce-
scores, geographic area, setting, and re- dures are reported in previous publica-
cruitment procedures. tions.
3. Exclusion criteria are reported. 3. Adequate exclusion criteria.
4. Test administration procedures are spec- 4. Test administration procedures are
ified. specified.
5. Means and SDs for the test scores are 5. Means and SDs for the test scores are
reported for the entire sample. reported.
6. The percentile distributions are re-
ported for the sample stratified by age x Considerations regarding use of the study
education. 1. All-male sample.
Considerations regarding use of the study group.
1. Overall sample is large, but sample sizes 3. No information on IQ is reported.,
for half of the age x education cells are
less than 10. [VF.40] Rosselli, Ardila, Salvatierra, Marquez,
2. Participants with MMSE scores <24 (the Matos, and Weekes, 2002a (Table A 11.52)
lowest recorded score was 17) were in- The authors compared oral fluency strategies
cluded in the sample. of Spanish English bilinguals in Spanish and
3. The data were obtained on Australian English with Spanish and English monolingual
participants, which may limit their use- participants tested in their native language, in
fulness for clinical interpretation in the their performance on the F AS and animal
United States. name generation tests. This article expands on
4. No information on IQ is reported. the data analyses presented in an earlier pub-
lication by these authors (Rosselli et al., 2000).
[VF.39] Brady, Spiro, McGiinchey-Berroth, The sample included 28 males and 54 females
Milherg, and Gaziano, 2001 (Table A11 .51 ) 50-84 years old, with a mean age of 61.76
The Animal Fluency test was used in a study (9.30) and education ranging 2-23 years, with
on the effect of stroke risk factors on cognitive a mean of 14.8 (3.6) years. Based on the ques-
functioning, which is part of the longitudinal tionnaire assessing participants' bilingualism,
232 lANGUAGE
45 participants were English monolinguals, 18 Considerations regarding use of the study

were Spanish monolinguals, and 19 were 1. Recruitment procedures are not reported.
Spanish English bilinguals. All English mono- 2. It is unclear whether the administration
lingual participants were born in the United procedure required restrictions in the
States. All Spanish monolingual participants types of word to be used in the process
were Latin American immigrants living in the of word generation.
city of Hialeah, Florida, had been living in the 3. Demographic characteristics for each
United States for an average of 5 years, and linguistic group are not provided.
had migrated after age 50. The latter partici- 4. The data are not partitioned by age group.
pants claimed Spanish as their first language, 5. Education for the sample is high.
mean age of exposure to English of 18.85 6. No information on IQ is reported.
(14.24) years, and mean number of years of
exposure of 35.95 (13.37). [VF.41] Grady, Yaffe, Kristof, Un, Richards,
Participants were screened for a history of and Barrett-Connor, 2002 (Table A11.53)
neurological or psychiatric problems, as well Data on Animal Naming were collected for a
as for current dementia and depression, us- subsample of 1,063 older women in a multi-
ing a structured interview, the MMSE, and center study examining the effect of hormone
the Beck Depression lnventory-11 (BDI-11). replacement therapy on cognitive functioning
All participants performed above 27 on the in postmenopausal women. Participants were
MMSE and below 5 on the BDI-11. Groups younger than 80 years, with established coro-
were similar in their naming ability, as mea- nary disease and an intact uterus. They were
sured by the English and Spanish versions of randomly assigned to a treatment or placebo
the Boston Naming Test. group in a double-blind experiment. They
Three 60-minute trials for the F AS and one were followed for 4.2 (0.4) years. At the end of
trial for Animal Naming were used. The order the trial, cognitive functioning was measured
of presentation for bilingual participants was in both groups. The data are reported for 517
counterbalanced. To avoid errors in scoring, participants in the treatment group and 546 in
the examiners read to the participants their the placebo group, separately. Mean ages for
responses after they were finished. the two groups at the time of testing were 66.3
The authors concluded that bilinguals pro- (6.4) and 67.3 (6.3) years, mean education was
duced significantly fewer words than English 12.7 (2.7) years for both groups; approxi-
monolinguals in the Animal Naming, but not mately 90% of the sample were white. There
in the F AS, condition. The use of grammatical are no notable differences between the groups
vs. content words and crosslinguistic differ- on any demographic variables or physicaJ
ences in the recall of alphabetical words were indices.
discussed. The Animal Naming test was administered
according to standard procedures.
Study strengths The authors concluded that there were no
1. Large sample. differences between the treatment and pla-
2. The sample composition is well de- cebo groups on any cognitive measures. More-
scribed in terms of age, education, gen- over, on the VF test, performance for the
der, acculturation level, and geographic placebo group was higher than for the treat-
area. ment group.
4. Test administration procedures are spec- Study strengths
ified. 1. Large sample size.
5. Means and SDs for the test scores are 2. The sample composition is well de-
reported. scribed in terms of age, education, gen-
6. Data for the Spanish-speaking sample der, physicaJ findings, clinicaJ setting,
are provided. and selection criteria.
3. Test administration procedures are spec- [VF.43] Lopez-Carlos, Salazar, Villasenor

ified. Saucedo, and Peiia, 2003
4. Means and SDs for the test scores are (Tables A11.55-A11.58)
reported. The PMR version of the phonemic VF and
Animal Naming tests were used in a study
Considerations regarding use of the study investigating the effects of demographic vari-
1. Participants had established coronary ables on cognitive abilities in Spanish-speak-
disease. It is unclear if any neurological ing individuals with low educational level. The
exclusion criteria were used. total sample included 115 volunteer mono-
2. All-female sample. lingual Latino men with ::;10 years of formal
3. The data are not partitioned by age education, who work in manual labor in the
group. Los Angeles area (n = 65) or Jalisco, Mexico
4. No information on IQ is reported. (n =50). Volunteers were recruited from
posted advertisements in workplaces and
[VF.42] Giovannetti, Goldstein, Schullery, and personal solicitations. The mean age for the
Barr, 2003 (Table A11.54) sample was 28.23 (8.74) years, and mean ed-
Animal Naming was administered to 31 con- ucation was 6.66 (2.54) years. Exclusion cri-
trol participants in a study on the mechanisms teria were self-report of head injury,
of VF deficits in first-episode schizophrenia. neurological insults, prenatal or birth compli-
Participants were recruited from the hospital cations, learning disabilities, psychiatric prob-
community through announcements in local lems, or substance abuse. Scores on the BOI-
newspapers and within the medical center. 11-Spanish version (M = 12.92, SO= 8.94) and
They had no history of substance abuse or the Beck Anxiety Inventory (M = 6.60,
neurological!psychiatridmedical illness, per SO= 6.03) are also reported.
self-report and per Schedule for Affective Standard administration procedures were
Disorders and Schizophrenia Interview, phys- used. Participants were tested in Spanish. The
ical examination, and urinalysis. Mean age for PMR version of the VF test was selected over
the group was 25.2 (6.07) years, mean edu- the F AS, to minimize the effects of education.
cation was 15.0 (1.48) years, mean WAIS-R IQ Two common errors among individuals with
was 109.3 (11.51), and male/female ratio was limited education in the latter measure are
21110. The sample is further descibed in the naming words that begin with an "A" sound
articles by Bilder et al. and lieberman et al., but are preceded by a silent "Ha" and naming
published between 1991 and 2000. words that begin with an "S" sound but their
The standard administration procedure was spelling begins with the letters C or Z (Artiola
used. i Fortuny et al., 1999).
Selected subtests from the WAIS-III (Mexi-
Study strengths can version) and the (EIWA) were included in
1. The sample composition is well described the battery. WAIS-111 Vocabulary raw scores
in terms of age, education, gender, FSIQ, are included in Tables A11.55-All.58. Mean
geographic area, and recruitment proce- performance on the Marin Marin Acculturation
dures. Scale for the Los Angeles sample was 17.61
2. Stringent exclusion criteria. (6.19). For the Los Angeles group, Picture Vo-
3. Test administration procedures are cabulary subscale scores from the Woodcock-
specified. Johnson-III Tests of Achievement (M =
4. Means and SDs for the test scores are 5.36, SO= 6.01) and the Bateria Woodcock-
reported. Mufioz-R, Pruebas de Habilidad Cognitiva-R
(M = 29.77, SO= 5.37) were used to assess
Considerations regarding use of the study level of English and Spanish expressive ability.
1. The sample is small. The results are partitioned by education
2. Educational level for the sample is high. group (0--6, 7-10 years), by age group (18-29,
234 LANGUAGE
30-49 years), and by education x age groups. [VF.44] Miller, 2003; Personal Communication
The authors reported that the diffe~ce be- (Table A 11.59)
tween the two education groups in perfopnance The investigation used participants from
on the PMR was significant atp < 0.05l«f.'el. No MACS. The data were collected from 728 se-
difference between the two education :groups ronegative homosexual and bisexual males for
was evident on Animal Naming. The tiwo age the purpose of establishing normative data for
groups did not differ significantly on ei~er test. neuropsychological test performance based on
The mean PMR score for the Los !ngeles a large sample. These data represent an update
group was 30.00 (11.09) and that for the (Mexico on the data provided by Seines et al. (1991).
group was 35.62 (10.30). The mean !Animal Mean age for the sample was 37.5 (6.9) years,
Naming score for the Los Angeles ~p was and mean education was 16.3 (2.3) years; 93.7%
17.16 (4.46), and that for the Mexico ~up was were Caucasian, 2.6% Hispanic, 2.9% black,
18.52 (5.53). Differences in scores ijltween 0.8% other. All participants were native En-
individuals from Los Angeles and Mexi4o were glish speakers.
noted on PMR, with the Mexican ~ample The F AS and Foods VF tests were admin-
scoring slightly higher. The authors atliibuted istered according to standard instructions.
this difference to using Spanish more fre- The data are partitioned by three age groups
quently in Mexico and/or a slightly higll.er ed- (25-34, 35-44, 45--59) times three educational
ucational level of 7.76 (2.18) years 4>r this levels (<16, 16, >16 years).
sample, as opposed to 5.82 (2.49) years~or the
Los Angeles sample. No statistically :;:i"ficant Study strengths
differences were noted on Animal N g.
1. The overall sample size is large, and
This study provided data for a ealthy
most of the individual cells have more
working monolingual Spanish-speaking sam-
than 50 participants.
ple with low educational level, whi~ is a
2. Normative data are stratified by age x
valuable addition to the available normative
education.
data primarily based on English-~aking,
3. Information on age, education, ethnicity,
highly educated participants.
and native language is reported.
Study strengths
reported.
1. Large sample for age and education
groups.
2. Data provided for a healthy employable Considerations regarding use of the study
monolingual Spanish-speaking group 1. All-male sample.
with low educational level. 2. No information on IQ is reported.
3. The sample is stratified into two educa- 3. No information on exclusion criteria.
tion groups, two age groups, and age x
education groups. [VF.45] Ravdin, Katzen, Agrawal, and Relkin,
4. The sample composition is well defcribed 2003 (Table A 11.60)
in terms of age, education, gend$-, goo- The authors examined the effects of mild de-
graphic area, and recruitment p~ures.' pressive symptoms on letter and semantic flu-
5. Adequate exclusion criteria. ency in a sample of 188 community-dwelling
6. Means and SDS for the test scotes are adults aged 60-92 years, who were recruited
reported. ; from community-based lectures and health
7. WAIS-111 Vocabulary subtest sco,-es are fairs. All participants lived independently and
available. were in self-reported good health. Exclusion
criteria were history of neurological disease,
Considerations regarding use of the study head injury with loss of consciousness >5
1. All-male sample. minutes, substance abuse, or psychiatric
2. The sample sizes for the combined age treatment. Participants were also excluded if a
and education groups are small. more extensive neuropsychological evaluation
revealed evidence of cognitive decline, de- results of the meta-analyses and predicted test
fined as any score falling more than 2 SD scores across adult age groups for the F AS and
below the mean for their age. Animal Naming versions of the VF tests are
Data for 149 participants who were not provided in Appendix 11m.
depressed, based on GDS scores of< 10, were The age range represented in the original
divided into three groups: young-old (60-69 data set is 17.76-87.5 years. However, there
years), middle-old (7~79 years), and old-old are no data available for FAS between ages
(8~92 years). The mean age for the nonde- 74.3 and 87.5 years. Therefore, the 87.5 year
pressed sample was 74.82 (6.63) years, mean data point was dropped from the analyses.
education was 15.57 (2.67) years, AMNART- Examination of the distribution of education
estimated Verbal IQ was 120.44 (5.74), and ranges suggests relatively smooth increments
male/female ratio was 321117. from 9.4 to 17 years, whereas data are lacking
Letter fluency was measured with the between 4 and 9.4 years. Therefore, three data
COWA-CFL version, and semantic fluency points representing participants with 4 years
with the Animals, Fruits, and Vegetables ca- of education were excluded for both tests, to
tegories. Scores for each letter and category, avoid extrapolation of a prediction rule over
as well as total scores for three trials for letters ranges that are not supported by existing data.
and semantic categories, respectively, were After further data editing for consistency and
reported. for outlying scores, 18 studies for FAS and 11
for Animal Naming, which generated 30 and
Study strengths 25 data points for the two tests, based on totals
1. Adequate sample size. of 3,469 and 2,823 participants, respectively,
2. The sample composition is well described were included in the analyses.
in terms of age, education, gender, IQ, Quadratic r~ssion of the FAS scores on
setting, and recruitment procedures. age yielded an R2 of 0. 711, indicating that 71%
3. Adequate exclusion criteria. of the variance in F AS scores is accounted for
4. Data are stratified by age groups. by the model. Based on this model, we esti-
5. Means and SDs for the test scores are mated FAS scores for age intervals between
reported. 18 and 74 years. Linear regression of Animal
Naming scores on age yielded an R2 of 0.764.
Considerations regarding use of the study Based on this model, we estimated Animal
1. It is unclear whether participants were Naming scores for age intervals between 25
instructed to avoid numbers or proper and 87 years. If predicted scores are needed
names in the process of word generation. for age ranges outside the reported bound-
2. Educational and intelligence levels for aries, with proper caution (see Chapter 3)
the sample are high. they can be calculated using the regression
equations included in the tables, which un-
derlie calculations of the predicted scores.
It appears that age-related changes in rates
RESULTS Of THE META-ANALYSES
of phonemic fluency have a curvilinear pat-
OF THE VERBAL FLUENCY DATA
tern, with increase in fluency up to the third
(See Appendix 11m)
decade of life, followed by a gradual decline.
Data collected from the studies reviewed in In contrast, the relationship between Animal
this chapter were combined in regression Naming and age is linear. This difference
analyses in order to describe the relationship might be attributable to gains in vocabulary
between age and test performance and to fund facilitating phonemic fluency up to the
predict expected test scores for different age late 30s, which is counteracted by age-
groups. Effects of other demographic vari- related anatomical changes thereafter. Evi-
ables were explored in follow-up analyses. The dently, the increase in vocabulary fund does
general procedures for data selection and not contribute to the generation of animal
analysis are described in Chapter 3. Detailed names.
236 LANGUAGE
Regressions of SDs for the FAS and Animal Inclusion of education into the analysis of
Naming scores on age suggest that age does Animal Naming also indicated a considerable
not account for a significant amount of vari- reduction in tau2 and improvement in R2 .
ability in SDs (R2 = 0.018 and 0.319, respec- However, education did not significantly ac-
tively). Though some increase in variability count for the variability in test scores (t = 1.93,
with advancing age is expected, this trend p = 0.083). Thus, a correction table for edu-
was not present in the collected data. There- cation was not provided for Animal Naming.
fore, we suggest that the mean SDs for the IQ had a significant effect on FAS scores.
aggregate sample be used across all age However, the limited number of studies that
groups. reported IQ does not allow close examination
Examination of the effects of demographic of this relationship. The effect of IQ on Ani-
variables on FAS scores revealed that educa- mal Naming was not examined due to a lack of
tion is a significant predictor of test perfor- data.
mance. Values of estimated between-study With regard to the effect of gender, analysis
variance (tau2 ) for regression of test means was conducted only for the F AS. The differ-
with education were considerably lower than ence in mean scores for the two genders
the corresponding values for regression with- across six studies reporting scores for males
out education. This suggests that education and females separately was negligible, 0.088 in
explains a considerable amount of the hetero- favor of females.
geneity in the outcome variable. Inclusion of
education into the regression ofFAS means on Strengths of the analyses
age considerably improved the R2 (see Ap- 1. Total sample size of 3,469 for the F AS
pendix 11m). It should be noted that in this and 2,823 for the Animal Naming test.
analysis, regression with and without educa- 2. Postestimation tests for parameter spec-
tion was rerun on a subset of studies that had ifications did not indicate problems with
education reported for each data point normality or homoscedasticity.
(n = 17). The t value for education was 2.47 3. Effects of education and IQ on F AS
(p = 0.025). The coefficient for education of performance were evident, which is
0.498, rounded to 0.50, indicates that with a consistent with the literature. A signifi-
1-year increment in education we expect a cant effect of education on F AS called
0.50-unit increment in word production. This for corrections for education.
suggests that the table of predicted FAS scores
is accurate for individuals with 14.31 years of Limitations of the analyses
education, rounded to 14 years (which is the 1. R2 of 0.711 and 0.764 for the two tests
mean education for the original data set) in are acceptable. However, these values
the education correction table. With every indicate that only 71% and 76% of vari-
year of education above or below this level, ance in the FAS and Animal Naming
we suggest correcting the obtained score by scores, respectively, is accounted for by
adding or subtracting 0.50 to or from the the models.
predicted score given in the table for the 2. The number of studies on the Animal
relevant age group (see Chapter 3 for an ex- Naming test that report data for older age
ample). Correction factors for different edu- ranges considerably exceeds the number
cational levels are included in Table A11m.1 in concerned with younger age ranges.
Appendix 11m. This correction should be ap- Thus, the mean age for the aggregate
plied within the education range of 10-17 sample is 62.60 (18.13) years.
years since this is the range available in the 3. Educational levels of the aggregate sam-
original data set. Unfortunately, data for lower ples are over 14 years for both tests, and
educational levels were not available in the the IQ level for the F AS sample is 113.91
literature. Any extrapolation of scores outside (7.72). (IQ is available for only seven data
the reported boundaries should be made with points.) Higher rates of word generation
caution. are associated with higher educational
and IQ levels. Therefore, the predicted racy of interpretation. Due to different levels
values are likely to overestimate expected of letter difficulty, the norms for the F AS
performance for individuals with lower should be used with caution in application to
educational levels and/or average and the COWA sets (CFL and PRW) and vice
lower than average intellectual levels. A versa. Other issues of concern are time allot-
correction table for education for the ted for each category, type of semantic or
FAS is provided in the appendix (only for phonemic category administered, presence/
the educational ranges represented in absence of feedback on intrusion or repeti-
our data). Although a significant effect of tion errors, instructions for item exclusion
IQ on the FAS was found in our data, (whether or not numbers were excluded), and
which is consistent with the literature, administration of an example with a practice
due to a scarcity of data on IQ reported in trial prior to the first recorded trial. These
the reviewed studies, close examination procedural aspects are inconsistent across
of this effect was not possible. studies. Attention to these aspects of admin-
4. Possible gender differences in word istration and data reporting is recommended
generation in favor of females have been for future studies involving VF tasks.
reported in the literature. Our compari- Other issues of interest to future in-
son of six data points for the FAS that vestigators of this test might include the fol-
were broken down by gender indicated lowing: (1) comparative efficiency of word
no relationship between gender and generation for semantic vs. phonemic cate-
word-generation fluency. Clearly, the gories in normal and clinical samples; (2)
scarce data available for review were not comparison of psychometric properties for
sufficient for a close examination of the different types of VF task, aiming at the pos-
effect of gender on VF. sibility of interchangeable use of the normative
data across tasks; (3) effect of demographic
variables on test performance across different
age ranges, educational/intelligence levels,
CONCLUSIONS
genders, ethnic, cultural, and linguistic back-
Different versions of the VF task are widely grounds, and clinical diagnoses.
used in clinical practice as measures sensitive Of concern, the majority of normative
to executive dysfunction. Use of the normative studies were confined to participants of high
data assuring accurate interpretation of the educational and/or IQ levels. Because a
test results is obscured by variability in pro- number of studies, as well as the results of the
cedural and reporting aspects of the studies. meta-analyses performed on the aggregate
There is confusion across studies in identify- sample described in this chapter, indicate that
ing the version of the test administered: the education (and/or IQ) is highly predictive of
CVFr test (FAS) is commonly presented as VF performance, the scarcity of normative
the COWA (see Brief History of the Test, data for lower IQ and educational groups is a
above). This confusion undermines the accu- problem for clinical practice.
IV
PERCEPTUAL ORGANIZATION:
VISUOSPATIAL AND TACTILE
12
Rey-Osterrieth Complex Figure
BRIEF HISTORY OF THE TEST systems, as well as alternate versions of the

complex figure, have emerged to meet the
The Rey-Osterrieth Complex Figure (ROCF), needs of specific clinical groups.
also known as the Rey Figure and the Com-
plex Figure Test (CFT), consists of a complex
Administration Procedures
two-dimensional line drawing containing
18 details, including crosses, squares, trian- Rey's (1941) original instructions are as follows.
gles, and a circle, arranged around a central The subject is presented the stimulus figure
rectangle (Fig. 12.1). with the isosceles triangle and circle oriented to
The patient is instructed to carefully copy the right and is instructed to "make a copy of
the design with pencil on paper. Rey (1941) this design as best as possible" on a plain sheet
asserts that the figure does not require a high of paper with a colored pencil. The subject is
level of graphic aptitude; each of the details is told that "the copy can be an approximation as
simple to reproduce separately, and the diffi- far as the proportions are concerned but that
culty of the task is due to the arrangement of care should be taken not to forget any detail."
the elements. Organizational strategy is docu- The subject is handed a different-colored pencil
mented by having the patient use different- with which to draw "each time an element is
colored pencils when executing the task (Rey & determined;" typically, five to six pencils are
Osterrieth, 1993). Some studies report alter- utilized. The order in which the colored pencils
nate methods for recording the strategy of are used is noted on the side of the paper. Time
reproduction, such as numbering lines on a to complete the copy is recorded. The subject is
copy of the ROCF or replicating the subject's not allowed to change the orientation of the
drawing with indication of the directionality of stimulus figure but may reposition the drawing
lines, as the subject proceeds (Binder, 1982; sheet. When the subject stops drawing, he
Kirk & Kelly, 1986; Waber & Holmes, 1986). or she is asked if he or she is finished, and the
The popularity of the ROCF in assessing copy sheet and the model are removed from
visuospatial constructional and visual memory view. After 3 minutes, the subject is handed a
deficits with different clinical groups has been new sheet of paper and a regular pencil and
steadily increasing (see Knight et al., 2003, for asked to draw the design from memory.
survey results on current use). A number of dif- L. B. Taylor's (1969) instructions are a
ferent administration procedures and scoring commonly used variation: "Copy this drawing
241
242 PERCEPTUAL ORGANIZATION: VISUOSPATIAL AND TACTILE
of recall (Berry & Carpenter, 1992). Similarly,

the pilot data collected by Wood et al. (1982)
indicated that there are no differences in
Taylor figure scores obtained on 30- vs. 60-
minute delayed recall in patients or normals.
Delayed recall is, however, affected by ad-
ministration of the Immediate Recall trial or
repeated Delayed Recall trials, which have a
facilitating effect on delayed performance
(Chiulli et al., 1989; Loring et al., 1990).
Loring et al. (1990) reported an increase in
accuracy of 30-minute delayed recall of about
' 6 points due to interposing a 30-second re-
Figure 12.1. Rey-Osterrieth Complex Figure (O!(terrieth, call trial.
1944). ~ The interested reader is referred to Knight
(2003), Lezak et al. (2004), and Spreen and
Strauss (1998) for additional information on
as well as you can. Make sure you do n~ leave ROCF administration procedures.
out anything." No time limit is impos~, but
the length of time required to complQI:e the
Alternate Versions
copy is recorded. Forty minutes laterJ a re-
production of the drawing from me~'ry is Repeated administration of the ROCF results
requested (p. 278). In a second publi ation, in inflation of the retest scores due to the prac-
L. B. Taylor (1979) provided additio al in- tice effect. According to Spreen and Strauss
formation regarding instructions: "EI'
osure
of the figure for copying is limited to min-
(1991), the inflation reaches about 10% of the
original score on the 1-month retest. The
utes. Then 45 minutes later the pa . nt is Taylor figure (L. B. Taylor, 1969; Hubley &
asked to reproduce as much of the figur; as he Tomhaugh, 2003; Lezak, 1995; Lezak et al.,
can remember" (p. 167). According ~ Os- 2004) was produced as an alternate version of
terrieth (1944), participants are not allo"ed to the complex figure to avoid the practice effect
make erasures. in repeated testing situations. The assumption
Copy and recall administration proc~dures of comparability of the two figures stems from
have varied across investigations (see Ruffolo the fact that both have an equal number of
et al., 2001). According to the standard pro- details of assumed equal complexity.
cedure, participants are asked to rec411 the To validate this assumption, several studies
figure without being forewarned. In .· many have compared scores obtained by partici-
studies, the copy condition is followed by the pants on the ROCF and Taylor figure in nor-
immediate recall. The reported interVal for mal and clinical populations (Berry et al., 1991;
the delayed recall varies between 3 rrinutes Delaney et al., 1992; Duley et al., 1993; Hamby
and 24 hours. According to Knight ~ al.'s et al., 1993; Hubley, unpublished honor the-
(2003) survey of practicing neuropsychoJogists sis; Hubley & Tombaugh, 2003; Kuehn &
(International Neuropsychological ~iety Snow, 1992; Peirson & Jansen, 1997; Strauss
membership), the most frequently use<l com- & Spreen, 1990; Tombaugh & Hubley, 1991;
bination of trials is Copy, Immediate iecall, Tombaugh et al., 1990, 1992a; Vingerhoets
and Delayed Recall, with a 0--5 second:inter- et al., 1998). The results indicated that the
val between Copy and Immediate RecaD, and figures yield equivalent copy scores. However,
a 16-30 minute interval between Imm~diate recall scores or percentage of the copy score
and Delayed recall trials. retained on the recall condition was much
According to some investigators, varying higher for the Taylor figure, irrespective of the
the delay interval between 15 and 60 minutes delay interval, learning paradigm (incidental
for the ROCF has minimal effect on thb
rate or intentional), or scoring system. According
REY-OSTERRIETH COMPLEX FIGURE 243
to Casey et al. (1991), this might be attributed figure. This system replaced that of Rey
to the fact that the Taylor figure is more likely (1941) for scoring copy and 3-minute delayed
than the ROCF to be encoded verbally. This recall, which was based on a 47-point scale.
finding limits interchangeable use of these Translations of Rey's (1941) and Osterrieth's
figures in test-retest situations. (1944) original articles describing scoring cri-
Loring and Meador (2003) describe the teria, along with critical commentaries and
psychometric properties, scoring criteria, and recommendations for administration, are of-
normative data for four complex figures de- fered by Corwin and Bylsma (1993).
veloped earlier (Meador et al., 1991, 1993), Scoring criteria published by E. M. Taylor
which are similar to the ROCF and Taylor (1959, adapted from Osterrieth, 1944) are
figure. Some of the same elements used in the reproduced in Figure 12.2 and Table 12.1.
previous figures were incorporated but placed Ratings of the accuracy of reproduction for
in different locations. New components were each unit are presented in Table 12.2.
added. A 36-point scoring system is used. To improve the quantitative objectivity of
Hubley and Tremblay (2002) developed a scoring and to address organizational aspects
modified version of the Taylor Complex Fig- of performance, several other scoring systems
ure (MTCF) by removing or downplaying have been proposed.
distinctive elements which are likely to be
verbally encoded, adding more lines to in- 1. Visser (1973) developed a system which
crease the complexity of the visual array, and quantifies figure accuracy (based on the
modifying the placement of some elements. presence or omission of cettain details)
The resulting figure was compared by the au- and organization (based on interruption
thors to the ROCF on 62 university students and sequence scores). The method of test
using three experimental paradigms. The re- administration has been altered to record
sults suggest comparability of the two figures the sequence of segments reproduced by
on learning, memory, and copy scores. the patient. Three aspects of perfor-
Hubley et al. (2003) further described the mance are scored: (1) omission of a detail
MTCF as well as the development of new or its portion, (2) interruption of a line
simplified figures for assessment of older before being completed, and (3) sequence
adults. Information on psychometric proper- of reproduction. This system was devel-
ties and normative data for these modified oped for research with brain-damaged
figures is provided. patients.
A recent attempt to design a complex figure 2. Binder (1982) used the original Os-
that is equivalent to the ROCF is described by terrieth scoring criteria to quantify the
Frazier et al. (2001). The authors introduced accuracy of reproduction but developed
the Mack Complex Figure Test (Mack CFT), his own system to quantify organization.
and presented a comparative study of the copy Five structural elements of the figure
and 45-minute delayed recall accuracy scores were identified, which were drawn as a
for the two figures on a sample of 245 adults. single unit by non-brain-damaged parti-
The authors report a high degree of equiva- cipants in the pilot study: the horizontal
lence between the figures. midline, vertical midline, two diagonals,
and the vertices of the pentagon. The
organizational score is based on the
Scoring Systems
number of structural units drawn as a
Osterrieth's (1944) scoring system, adapted by single unit vs. fragmented units and on
E. M. Taylor (1959), is most commonly used the number of missing units. This system
in scoring the copy and recall reproductions of was used to evaluate reproductions of
the ROCF. It is based on the accuracy of a patients with unilateral cerebrovascular
subject's reproduction and assigns 0-2 points brain pathology.
based on placement and presence of distor- Savage et al. (1999) developed a
tion for each of 18 structural elements of the system for quantifying the organizational
~.
------~1 2~,------~~
18
Figure 12.2. Scoring units for the Rey-Osterrieth Complex Figure. (Reproduced by permission from
0. Spreen and E. Strauss, 1998.)
Table 12.1. Rey-Osterrieth Complex Figure Scor- Table 12.2. Ratings for Each Rey-Osterrieth Com-
ing Criteria plex Figure Unit
Unit.: Correct Placed properly 2 points
l. Cross upper left comer, outside of rectangle Placed poorly 1 point
2. Large rectangle Distorted or Placed properly 1 point
3. Diagonal cross incomplete Placed poorly 112 point
4. Horimntal midline of 2 but recognizable
5. Vertical midline
6. Small rectangle, within 2, to the left Absent or not recognizable 0 points
7. Small segment above 6 Maximum 36 points
8. Four parallel lines within 2, upper left
9. Triangle above 2, upper right
10. Small vertical line within 2, below 9
3. Klicpera (1983) used the original Os-
ll. Circle with three dots within 2
12. Five parallel lines with 2 crossing 3, lower right terrieth scoring to assess the accuracy of
13. Sides of triangle attached to 2, on the right reproduction. In addition, he introduced
14. Diamond attached to 13 organizational criteria, such as presence of
15. Vertical line within triangle 13, parallel to the right parts of the configuration (main rectangle,
vertical of 2
internal structural components, external
16. Horimntal line within 13, continuing 4 to the right
17. Cross attached to 5, below 2 and internal details, intersections, and
18. Square attached to 2, lower left segments forming the large rectangle, di-
agonals, and perpendiculars), organization
Scoring: Consider each of the 18 units separately.
Appraise accuracy of each unit and relative position
(intersections, alignment and arrangement
within the whole of the design. Rating for each unit is of details), and approach to drawing (se-
presented in Table 12.2. quence of construction, continuit;y of lines,
segmentation of key parts). Klicpera used
this system to explore planning abilities in
approach based on the structural units dyslexic children.
identified by Binder plus the base 4. Denman (1984) developed an itemized
rectangle, which they used to evaluate scoring system for the ROCF as part of
organizational strategies in patients with the Denman Neuropsychology Memory
obsessive-compulsive disorder (OCD). Scale, which assigns 0-3 accuracy points
to each of the 24 "designs" of the figure, Additional normative data and clinical
which are divided into eight sectors. examples based on this scoring system are
Rapport et al. (1995) developed several provided in Bernstein (2003).
measures of hemispatial deficit that may 7. Kirk and Kelly (1986) adapted the scoring
be incorporated in this scoring system. method for accuracy and error from the
Tombaugh (unpublished research) de- system developed by Waber and Holmes
veloped a similar system for scoring the (1985). In addition, they developed a
Taylor complex figure. Both systems con- system designed to provide an objective
sist of a total possible of 72 (69) points basis for recording, classifying, and eval-
and are more detailed than the original uating starting strategy (configurational
36-point scoring system to provide more vs. piecemeal) and the level of organiza-
objective interpretation. tion (structured vs. nonstructured) in
5. Bennett-Levy (1984) proposed scoring ROCF reproductions by children. Start-
criteria to evaluate strategy in addition ing strategies were also evaluated from
to Osterrieth's original accuracy criteria. a part/whole perspective. In addition, pro-
"Strategy total" included scores for "good gression strategies were assessed. This
continuation" (requiring a line to be drawn system allows evaluation of the relation-
as one piece and continued until in- ship between strategy, accuracy, and er-
tersection with another line) and for rors in ROCF reproductions.
"symmetry" (reflecting construction of 8. A modified scoring system was used by
symmetrical units and their components). Becker (Becker, 1988; Becker et al., 1988)
6. Waber and Holmes (1985, 1986) intro- to evaluate visuoconstructional skills and
duced the Developmental Scoring Sys- visual memory in Alzheimer's patients.
tem (DSS-ROCF) to quantify goodness According to this system, 12 "bits" of the
of organization and style in the context of drawing were scored for accuracy and
developmental changes in children, which placement (2 points) and for accuracy
was published by Psychological Assess- only (1 point).
ment Resources (Bernstein & Waber, 9. Loring et al. (1988, 1990) developed an
1996). Administration of the DSS-ROCF 11-point method for scoring qualitative
requires five colored pencils, which are errors, which reflects the accuracy of
switched after specific time periods. The reproduction of each of 18 original de-
figure is separated into the smallest seg- tails in recall of the ROCF. Scoring al-
ments, which are categorized as part of lowed a certain degree of "reproduction
the four major components: base rectan- tolerance" due to the focus on memory
gle, main substructure, outer configura- functioning, rather than on construc-
tion, and internal detail. The presence tional ability. These criteria were applied
of intersections and alignment of key to reproductions produced by patients
components (base rectangle, main sub- with temporal lobe epilepsy and found to
structure, and outer configuration) are be effective at discriminating between
scored. right and left temporal lobe epilepsies.
In addition, ratings of organization and However, application of this method
style of reproduction are obtained. Qual- to scoring reproductions produced by a
ity of reproduction of 24 organizational sample of college students resulted in
features results in placement of the pro- over 95% of the sample scoring 36 out of
tocol in one of the five organizational le- 36 points.
vels ranging from poor (I) to excellent 10. Simplified versions of the ROCF and
(5). The style rating is based on reproduc- Taylor figure were developed by Hubley
tion of 18 "criteria! juncture features," and colleagues (Hubley, unpublished honors
which place the protocol into part- thesis; Hubley et al., 2003; Tombaugh
oriented, intermediate, or configurational et al., 1990) to assess elderly and neurolog-
categories. ically impaired individuals. Fifty-point
246 PERCEPTUAl ORGANIZATION: VISUOSPATIAl AND TACTilE
itemized scoring systems were devel- original system and similar to the system
oped to score the figures. These systems presented by Loring et al. (1990). How-
are similar to the itemized scoring sys- ever, the authors introduced a 114" rule
tems developed for the full versions of for "misplacement" and a 118" rule for
the figures, with the exception of scoring drawing errors, which reduce the effect
only 19, rather than 24 (23), "designs" of of subjective judgment on scoring.
the figures. The Recognition subtest is composed
11. Beny et al. (1991) modified the original of 12 components of the ROCF pre-
scoring criteria to improve the sensitivity sented in their proper size, shape, and
of the score to distortion and displace- orientation mixed randomly with 12 dis-
ment of details. Scores for each of the 18 tracters. The subject is to circle all the
details ranged ~2 in half-point incre- parts recognized from the original de-
ments. A score of one-quarter point was sign. Scoring of the Recognition subtest
also allowed, to denote gross distortion is based on the number of correct, false-
or severe displacement of the detail. positive, and false-negative responses.
These criteria were applied to scoring Meyers and Lange (1994) compared
of the ROCF for a sample of elderly par- scores on the Recognition subtest for
ticipants. The modified system scores are different clinical groups and a normal
on the average 2 points higher than group. The results suggested that this
those obtained using the original scoring subtest discriminates best between brain-
system; however, the correlations be- injured and normal participants or parti-
tween the two systems are high (ranging cipants with minor brain injuries. Dawson
0.82-0.96 for different conditions). and Grant (2000) used remarkable im-
12. Chervinsky (see Chervinsky et al., 1992) provement on the Recognition trial in
developed the "organizational scoring sys- comparison to poor free recall to dem-
tem," which is designed to assess the onstrate impaired retrieval in the face of
organization of reproductions and is min- intact acquisition in recently detoxified
imally affected by reproduction accu- alcoholics.
racy. According to this system, the subject The professional manual for the Rey
is offered a different-colored pencil when Complex Figure Test and Recognition
it is judged by the examiner that he or Trial (RCFT), published by Psychologi-
she has completed a "chunk" of the fig- cal Assessment Resources (Meyers &
ure. The figure is separated into six Meyers, 1995b), contains test descrip-
sections, consisting of different combi- tion, administration, scoring and profiling
nations of conceptual "chunks." A total procedures, interpretation guidelines,
score reflecting the organizational qual- information on psychometric properties,
ity of reproduction is based on the scores normative data expressed in raw scores,
assigned for completeness of the con- and demographically corrected norma-
ceptual "chunks" (reproduced with the tive data compiled on a sample of 394
same-colored pencil) minus penalty scores. individuals 18-90 years of age. Supple-
(One conceptual detail is reproduced mental norms for children and adoles-
with different-colored pencils.) cents ~18 years of age (Meyers &
13. Meyers and Meyers' (1992) modified Meyers, 1996) are also available.
administration procedure includes four 14. Duley et al. (1993) developed explicit
conditions: Copy (recording the time to scoring criteria for the ROCF and Taylor
completion), 3-Minute Delayed Recall figure, which substantially increase in-
(from the time of completing the copy), terrater reliability. These criteria define
30-Minute Delayed Recall (from the scorable distortions and misplacements
time of completing the copy), and a Rec- and outline clear rules for scoring de-
ognition subtest. Scoring of the Copy viations in drawings for each of 18 ele-
and Recall conditions is based on the ments of the original 36-point scoring
system for both figures. This system was sample of 433 individuals aged 18-94. In
used by the authors in a study with addition, Folbrecht et al. (1999) reported
participants infected with HIV. good to excellent indices of interrater re-
15. Hamby et al. (1993) developed an orga- liability and internal consistency. Som-
nizational quality scoring system for the erville et al. (2000) demonstrated modest
ROCF and Taylor figure. Administration correlations between BQSS indices of
of the figures requires colored pens, executive functioning and performance
which are switched at equal points in the on other tests measuring this domain.
construction of the figure. The scoring is The main disadvantage of the BQSS
based on 18 standard elements for each is the complexity and considerable time
figure. It focuses, however, on the re- commitment required to learn it (Boone,
production strategy through determining 2000). However, the authors report a
the order of placement of configura! ele- rapid reduction in total scoring time to
ments (base rectangle or square, horizon- 10-15 minutes for all three trials. In ad-
tal and vertical midlines), the appropriate dition, the Quick Scoring Guide (QSG)
continuation of lines, and the order of was developed by the authors, which al-
placement of details. Ratings of organi- lows rating BQSS scores without refer-
zational quality are made on a 5-point ence to the elaborate criteria. A majority
scale and reflect the presence of three of the ratings correlate with compre-
types of mistake (configura!, diagonal, and hensive scores above 0.7. Further review
detail), with higher scores indicating of the advantages and disadvantages of
better organization. This system was used the BQSS compared to the 36-point
by the authors to discriminate between system is provided by Hartman and
asymptomatic and symptomatic HIV- Potter (1998).
positive patients. Akshoomoff and Stiles (1995a,b, 2003)
16. R. A. Stem et al. (1994, 1999) proposed applied the BQSS (with some modifica-
the Boston Qualitative Scoring System tions) to explore strategies used by chil-
(BQSS), which divides the ROCF into dren in ROCF copy and recall. Cahn
three hierarchical sets of elements: con- et al. (1996) and Schreiber et al. (1999)
figural elements, clusters, and details. These used the BQSS to examine qualitative fea-
sets are scored for different combinations tures ofROCF performance in attention-
of the following features: presence, accu- deficit hyperactivity disorder (ADHD).
racy, fragmentation, and placement. Javorsky and Stem (1999) demonstrated
Reproductions are also evaluated with superiority of the BQSS to the 36-point
respect to planning, organization, size scoring system in discriminating between
distortion, perseveration, confabulation, dementia of Alzheimer's type and ische-
rotation, neatness, symmetry, and im- mic vascular dementia. Usefulness of the
mediate and delayed retention. These BQSS in discriminating between clinical
qualities yield 17 initial scores for each of groups and controls is described in the
the three reproductions (Copy, Imme- manual.
diate Recall, and 20-30 Minute Delayed 17. Fastenau (1996, 2002b, 2003) developed
Recall). In addition, six summary scores Recognition and Matching trials for the
are calculated. According to the profes- original version of the complex figure,
sional manual for the BQSS (Stem et al., which are to follow the Copy, Immediate
1999), the authors were pursuing dual Recall, and Delayed Recall trials. This elab-
goals: to provide quantitative summary orated version was named the Extended
scores and to formulate a system of Complex Figure Test (ECFT). The Rec-
qualitative ratings. A description of the ognition trial contains 30 multiple-choice
BQSS and information on its psycho- items, which are classified into Global and
metric properties are provided in the Detail scales. The Left-Detail and Right-
manual, based on a standardization Detail subscales are incorporated into the
Detail scale. Ten matching items repre- and Carpenter (1992), Berry et al. (1991),
sent a subset of the items used in the Boone et al. (1993a), Carr and Lincoln (1988),
Recognition trial. Each item is presented Delaney et al. (1992), Rapport et al. (1997),
with five multiple-choice options. and Stem et al. (1994) report respectable in-
Psychometric properties of the scales terrater reliabilities of 0.80-0.99.
and normative data for adults 30-85 years Liberman et al. (1994) assessed interrater
of age are presented by Fastenau (1996, and intrarater reliabilities of ROCF scoring in
2003) and Fastenau et al. (1999). Devel- a large sample of male boxers. The authors
opment of child normative data is pre- report very high intrarater reliability and good
sented by Sasher and Fastenau (2001). interrater reliability for Copy, Immediate
Applicability of 30-50 year norms for Recall, and Delayed Recall.
younger adults is discussed by Fastenau Tupler et al. (1995) established excellent
(2002a). The ECFT has been published inter- and intrarater reliabilities for total scores
by Western Psychological Services (Fas- (ranging 0.85-0.97) in a sample of95 memory-
tenau, 2002b). The test manual provides impaired elderly participants. However, cor-
further information on psychometric responding reliabilities for the 18 individual
properties and normative data for this test. items were highly variable, ranging 0.14-0.96.
The authors recommended amplified delin-
A review and critical analysis of different eation of relevant decision criteria for scoring
scoring systems are provided in several sour- of individual items.
ces (Chervinsky et al., 1992; Hamby et al., Interrater reliability for the Savage et al.
1993; Knight, 2003; Lezak et al., 2004; Troyer (1999) scoring method, established on a sub-
& Wishart, 1997). sample of 15 drawings obtained by Deckersbach
In summary, a number of different scoring et al. (2000a) in a study on organizational ap-
systems for the ROCF and Taylor figure were proach to the complex figure in OCD individ-
proposed, which focus on different aspects of uals, was moderate to high, with Cohen's K
reproduction: accuracy, organization, strategy, coefficients ranging 0.69-0.92 for different or-
and style. Due to the variations in scoring ganizational elements of the figure.
systems and differences in operationalization Berry et al. (1991) and Rapport et al. (1997)
of conceptual aspects of reproduction, results reported adequate internal consistency reli-
of different studies and normative data re- abilities for the standard scoring system. The
ported in these studies should be interpreted latter study also reported higher internal
with caution. consistency for the Denman system in com-
parison to the standard system.
Data on test-retest reliability of the ROCF
Reliability
over a 1-year period were provided by Berry
Of some concern in the use of the ROCF is et al. (1991). Based on performance of a
the issue of interrater reliability. Most clin- sample of 41 elderly "normal" participants,
icians use the E. M. Taylor (1959) scoring they reported low reliability for the Copy
criteria, and those familiar with the system are condition but moderate reliability for Imme-
aware of the subjective judgment involved in diate Recall and Delayed Recall. Similarly,
determination of a "distortion." According to test-retest reliability coefficients for ROCF
Bennett-Levy (1984), pilot study data have performance of elderly participants, reported
indicated that only by use of very strict or very by Mitrushina and Satz (1991a), were quite
lenient scoring criteria can adequate interrater low. Assessed over three annual probes, the
reliability be attained. He states that Taylor coefficients ranged 0.56-0.68 for Copy and
conveyed to him in a personal communication 0.57...{).77 for 3-Minute Delayed Recall. Data
that he, in fact, employs very strict criteria on repeated administration are also presented
involving quality of draughtsmanship as well as by McCaffrey et al. (2000).
presence, distortion, and misplacement of For further information on the psychomet-
figures. However, Bennett-Levy (1984), Berry ric properties of the ROCF, see Franzen
(2000), Knight (2003), Lezak et al. (2004), and of the integrated, multisystem neural me-
Spreen and Strauss (1998). chanisms, which are not limited to the right
hemisphere and involve all four lobes of the
brain. Specifically, the left hippocampus was
Clinical Utility
shown to make a major contribution to the
The ROCF assesses visuoperceptual/con- efficiency of ROCF recall. The author hy-
structional skills, spatial organizational skills, pothesized that this finding might reflect use
and visual memory. It has frequently been of verbal mediation in recall strategies.
found to be sensitive to nondominant hemi- Regard and Landis (1994) analyzed clinical
sphere functioning and right temporoparietal findings in 37 patients who produced a happy
area integrity in particular (Binder, 1982; face ("smiley") instead of the detail involving a
Milner, 1975; Pimental & Ross, 2003; Taylor, circle with three dots. The authors concluded
1969; Wood et al., 1982), although some in- that the "smiley" is rare but that its presence
vestigators have failed to document this rela- is associated with dysfunction of the anterior
tionship (King, 1981). Some authors have part of the right hemisphere.
suggested that right hemisphere- and left Seidman et al. (1995, 1997) found that de-
hemisphere-damaged patients may show dif- velopmental analysis of the ROCF identifies
ferent types of ROCF copy and recall fail- organizational difficulties related to ADHD,
ure. On copy, right hemisphere-damaged which is associated with developmentally lower
patients have exhibited distortions, while left levels of copy organization and recall style.
hemisphere-damaged patients have produced However, Reader et al. (1994) did not find
the design in a piecemeal and fragmented lower ROCF performance to be associated
fashion but still frequently produced an ade- with ADHD when using a traditional scoring
quate copy (Binder, 1982). On recall, right system. Fujii et al. (2000) reported visuospa-
hemisphere-damaged patients have shown dis- tial skills to be more salient in predicting copy
tortions and loss of the general organization of performance in ADHD adults with high IQ,
the figure (based on Taylor's version), while whereas organizational skills were more pre-
left hemisphere-damaged patients have shown dictive in those with low IQ. The authors
preservation of the four-square quadrant but discussed their findings in light of the overall
loss of details (Wood et al., 1982). Poreh and efficiency of information processing. ROCF
Shye (1998) showed that right-sided local productions in ADHD children are further
elements of the ROCF are most useful discussed in Conners et al. (2003) and Teknos
for discriminating between right and left et al. (2003).
hemisphere-damaged patients. Kixmiller et al. (2000) investigated the role of
Kaplan (1988) emphasized the importance perceptual and organizational factors in ROCF
of qualitative interpretation of the drawing performance in three amnesic groups. They
strategy in the context of the process analysis found that the Korsakoff patients' impairment
of a patient's performance. According to on the copy task is due to the superimposing of
Kaplan, both frontal damage and posterior visual-organizational difficulties upon visuo-
damage might result in poor reproduction of perceptual processing deficiencies, whereas
the ROCF. In case of left frontal pathology, diminished copy accuracy in patients with an-
patients retain the outer contour and the terior communicating artery aneurysm is re-
major structural lines. In case of a right pari- lated to organizational and behavioral control
etal lesion, the breakdown in contour and inefficiencies. As for Korsakoff patients' poor
organization is expected on the side contra- recall, visuoperceptual deficits play a role in
lateral to the lesioned hemisphere (i.e., the their immediate recall deficits. A combination
left side of the design). of visuoperceptual deficits and severe amnesia
Using quantitative MRI morphometric may explain their poor delayed recall. Diamond
analysis in patients with localized lesions due and DeLuca (1996) reported a profound loss of
to traumatic brain injuries, Bigler (2003) showed information between the copy and immediate
that ROCF reproduction requires involvement recall conditions in anterior communicating
artery amnesiacs, despite normal limits G:>r copy normal elderly controls, using a scoring sys-
scores. The authors attributed this losi of in- tem developed by the authors, that divides the
fonnation to inadequate encoding, ir1apaired ROCF into six perceptual categories: right,
consolidation, or accelerated rates of forgetting. left, upper, lower, basic gestalt, and inner
Deckersbach et al. (2000a,b) and Savage detail. They found that demented patients
et al. (1999, 2000) found problems with orga- had deficits in all six categories compared to
nization of the drawing in spite of ~urate normal controls. In addition, left hemispatial
reproduction of the geometric figure ip their inattention contributed to impaired perfor-
sample of individuals with OCDqwhich mance in DAT. Use of the ROCF in dementia
implicates frontostriatal dysfunction, b ed on is further described by Libon et al. (2003) and
neuroimaging findings. Organization uring Saxton et al. (2003).
the copy condition was a strong pr · tor of The ROCF has been extensively used with
subsequent memory performance · their other clinical groups: schizophrenia (Kali-
sample. A role of executive dysfun~on in nowski et al., 2003), toxic exposure (Diamond
OCD, identified through ROCF perfo~ance, et al., 2003), unilateral neglect (Rapport &
is further described by Savage and Otto ~2003). Webster, 2003), temporal lobe epilepsy (Barr,
In a study by Waber et al. (1994), lon;-tenn 2003), frontal lobe damage (Ruff & Jurica,
survivors of childhood acute leukemia rfalled 2003), brain injury and organic memory im-
fewer organizing-scheme components pn the pairment (Wilson & Watson, 2003), vascular
ROCF but more incidental features iii com- cerebellar lesions (Greve et al., 2003b), stroke
parison to nonnative expectations. The ~thors rehabilitation (Greve et al., 2003a), and pe-
suggest a metacognitive basis for this !weak- diatric closed-head injury (Yeates et al., 2003).
ness, rather than a visuoperceptual de6fit. Lu et al. (2003) suggested that with recent
Shorr et al. (1992) computed measqres of addition of a recognition memory trial (Mey-
copy accuracy, perceptual clustering, 4ncod- ers &: Meyers, 1995b), the ROCF becomes a
ing, and savings ror 50 neuropsychia~c pa- potentially useful instrument for capturing
tients based on their ROCF perfonnan<t. The suspect effort.
authors concluded that perceptual clustering Waber et al. (1989) studied ROCF recall in
in the copy condition, which reflects use of children in the context of a dual-code cogni-
organized strategy in drawing segments. of the tive neuropsychological model. Recall was
figure, was a better predictor of memory compared after copying the figure vs. after
performance than was copy accuracy. Vse of visual inspection for the two experimental
the ROCF in patients with neuropsyclriatric groups, respectively. Fifth-graders who did
disorders is further discussed by Jones' et al. not copy the design prior to recall remem-
(2003). bered its organization better and produced it
Quality of ROCF performance of qormal more configurationally than the other group.
elderly (57-85 years old) was anal~d by However, there was no treatment-group dif-
Mitrushina et al. (1990). The authors lound ference for eighth-graders. The authors ex-
equally efficient recall of the basic strucQue of plain these results from the perspective of
the figure across all ages represented tn, the complementary functioning of the visual sys-
sample. However, a pronounced dec~e in tem, which favors configurational processing,
recall of outer configuration was demon- and the motor system, which relies on se-
strated by older participants. Loss of cJetails quential or part-oriented processing. There-
(coded verbally) with advancing age jlends fore, performance style can be indicative of
partial support to the hypothesis of age- the relative strength of the visual and motor
related compromise in the interactio~ be- codes, which can be interpreted in terms of
tween verbal and visual codes. neuropsychological referents, sequential mo-
Cherrier et al. (1999) compared ~rfor tor programming being associated with the
mance of patients with dementia of Abhei- left cerebral hemisphere and gestalt pattern
mer's type (DAT) and vascular dementia with perception with the right. In the context of
!
this theory, the results of this experiment RELATIONSHIP BETWEEN ROCF

suggested that among preadolescent children PERFORMANCE AND DEMOGRAPHIC
the motor aspect interfered with efficient en- FACTORS
coding of visuospatial information. An expla-
A robust relationship has been found for ROCF
nation based on differential involvement of
copy and recall performance with age in
cerebral hemispheres in information process-
patients and control participants (Ardila &
ing was also offered. Further advancements in
Rosselli, 1989, 2003; Ardila et al., 1989; Ben-
dual-code theory are described by Waber
nett-Levy, 1984; Boone et al., 1993a; Chiulli et
(2003).
al., 1995; Fastenau, 2003; Fastenau et al., 1999;
Further review of the clinical utility of the
Janowsky & Thomas-Thrapp. 1993; King, 1981;
ROCF is provided in Knight (2003), Lezak
Lannoo & Vingerhoets, 1997; Meyers & Mey-
et al. (2004), and Spreen and Strauss (1998).
ers, 1995a; Mitrushina & Satz, 1991a; Mitru-
shina et al., 1995a; Ostrosky-Solis et al., 1998;
Culture-Specific Studies and Normative Powell, 1979; Rapport et al., 1997; Speers &
Data for the ROCF Ribbler, unpublished manuscript; Visser, 1973),
although some negative findings have been
Normative data for the Copy and 10-Minute
reported (Brooks, 1972; Delaney et al., 1992).
Delayed Recall conditions, collected on a sam-
King (1981) concluded:
ple of 300 Hispanic participants stratified by
gender, age, and education, are provided by the strong effect of age . . . indicates a need for
Ponton et al. (1996). Psychometric properties of caution in the interpretation of copy and recall data
the ROCF with children in Spanish-speaking from elderly participants. It may be that the com-
populations are examined by Galindo and plexity of the Rey Figure renders it an unsuitable
Cortes (2003). Normative data based on a sam- task for the discrimination of brain-damaged from
ple of 624 Spanish-speaking children and adults non-brain-damaged elderly participants. Further
living in Bogota, Colombia, stratified by age, data are needed to answer this question. (pp. 642-
education, and gender, are reported by Ardila 643)
and Rosselli (2003).
Guo et al. (2000) investigated the applica- Tombaugh et al. (1992a) used a multiple-
bility and psychometric properties of the trial, intentional learning procedure to explore
ROCF on a sample of 111 healthy retired recall of the ROCF and Taylor figure in
Chinese elderly. A subset of 40 participants healthy participants 20-80 years old. In their
also completed the Taylor figure. Effects of study, participants 60-80 years old scored
demographic variables on the Copy and De- lower than those 20-59 years old on recall of
layed Recall conditions were explored. No both figures, which was attributed by the au-
difference between ROCF and Taylor figure thors to less efficient encoding and retrieval
performance was found. Good test-retest re- strategies used by older people.
liability and construct validity were reported. In another study, Tombaugh et al. (1992b)
Normative data for the Copy and 10-Minute used a similar procedure to explore inten-
Delayed Recall of the ROCF, collected on a tional learning of the Taylor figure in neuro-
sample of 280 normal Italian participants, were logically intact participants 20-79 years old.
reported by Caffarra et al. (2002). Significant The Taylor figure was presented for observa-
effects of age and education on the copying tion for 30 seconds, after which participants
task were reported. Gender affected only the were to reproduce it from memory within
delayed recall performance. The authors re- 2 minutes. The procedure was repeated over
ported inferential cutoffs and equivalent scores four acquisition trials, which were followed by
for the ROCF. a retention trial 15 minutes later, and con-
Normative data for Canadian children and cluded with copying the figure from the model
adults aged ~70+ are provided by Spreen within 4 minutes. An itemized scoring system
and Strauss (1998). was used (Tombaugh, unpublished research).
The results indicated that performance on the memory trials in their sample of healthy
Taylor figure is not affected by gender, de- adults. Similarly, Meyers and Meyers (1995a)
pression, or education and is not related to reported a negligible contribution of educa-
performance on the Vocabulary and Block tion (ranging 0%-5%, with 2% average) to any
Design subtests of the WAIS-R. An effect of of the RCFT scores in their standardization
age, however, was evident, with a greater rate sample.
of age-related decline on recall than on copy Of importance to the clinician is the fre-
trials. The authors concluded that age has an quently reported relationship between ROCF
effect on constructional as well as on memory performance and intellectual level in both
processes. patients and control participants. Hemsley
Hartman and Potter (1998) and Vinger- (1974) documented a significant correlation
hoets et al. (1998) did not find remarkable age between ROCF percent loss on 40-minute
differences in drawing ability or organizational delay and WAIS full-scale IQ (r = - 0.405).
quality of complex figure reproductions. How- He suggested that "it may be necessary to take
ever, robust age differences in recall were IQ into consideration when interpreting
reported. Similarly, Mitrushina et al. (1990) scores on the Rey-Osterrieth test" (p. 1134).
did not find differences on the copy condition Wood et al. (1982) report significant correla-
between 73 young-old (57-70 years) and 80 tions between ROCF 40-minute delayed re-
old-old (71--85 years) participants. However, call and WAIS Block Design scaled scores
pronounced differences between these two age (r=0.628), with minimal associations with
cohorts were reported on the 3-minute de- WAIS Vocabulary (r= 0.258), Digit Span
layed recall condition. Changes in the accu- forward (r=0.079), and Digit Span backward
racy and organizational quality of the copy and (r = 0.221). They argue that ROCF perfor-
recall of ROCF as a function of advancing age mance is "dominated by variation in intellec-
were also reported by Chervinsky et al. (1992). tual skills unrelated to memory" (p. 181). A
Waber and Holmes (1985, 1986) reported majority of other investigators have found
maturational changes in children which affect significant correlations between IQ variables
the accuracy and organization of ROCF repro- and ROCF copy and recall performance
duction in the copy and recall conditions. (Bennett-Levy, 1984; Boone et al., 1993a;
According to these authors, improvement in King, 1981; Mitrushina et al., 1989; Powell,
accuracy of reproductions is evident up to the 1979; Visser, 1973), although some negative
age of 9 years, after which accuracy remains findings have been documented (Speers &
relatively stable. In contrast, developmental Ribbler, unpublished manuscript).
changes in planning and organization of the In general, no differences in ROCF per-
reproduction continue into adolescence. formance between men and women have been
Berry et al. (1991) found a significant rela- found (Berry et al., 1991; Boone et al., 1993a;
tionship of ROCF immediate and 30-minute Browers et al., 1984; Fastenau et al., 1999;
delayed recall with age and education. Scores Meyers & Meyers, 1995b). In some studies,
on the copy condition were not related to age, men outperformed women, but the amount of
education, or gender. variance accounted for by gender has been
The relationship between education and minimal (Ardila & Rosselli, 1989; Ardila et al.,
ROCF performance has been equivocal, with 1989; Bennett-Levy, 1984; King, 1981).
some reporting a significant relationship (Ar- Some studies have addressed the relative
dila & Rosselli, 1989, 2003; Ardila et al., 1989; importance of, or interaction between, ROCF
Berry et al., 1991; Vingerhoets et al., 1998) scores and IQ/demographic factors. Boone
and others failing to document an association et al. (1993a) documented with stepwise re-
(Boone et al., 1993a; Delaney et al., 1992; gression analyses a significant association be-
Hartman & Potter, 1998; Schreiber et al., 1999; tween ROCF scores (copy and delay) and age
Speers & Ribbler, unpublished manuscript). as well as IQ in a healthy middle-aged and
Fastenau et al. (1999) reported that education elderly population, whereas gender and edu-
explained 2%-3% of variance on copy and cation were not predictive of ROCF scores.
Similarly, Bennett-Levy (1984) reported that pants who had completed < 12 years of edu-
age was strongly related to ROCF copy per- cation. It may be that education is a predictor
formance in his young to middle-aged sample, of ROCF scores only at low levels but once a
while copy score and age were significant "threshold" is reached, such as completion of
predictors of recall scores; gender and IQ high school, it no longer has an influence.
were not found to contribute to test scores
once the other variables had been considered.
While Boone et al. {1993a) did not find any METHOD FOR EVALUATING THE
interaction between age and IQ in their age- NORMATIVE REPORTS
restricted sample, their findings considered in
conjunction with Bennett-Levy's results may To adequately evaluate the ROCF normative
suggest that IQ is a less important factor in reports, eight key criterion variables were
ROCF performance during young to middle deemed critical. The first six of these relate to
adulthood but emerges as an important vari- subject variables, and the remaining two relate
able only in advanced age. to procedural variables.
While many studies did not find a relation- Minimal requirements for meeting the cri-
ship between ROCF reproduction and hand- terion variables were as follows.
edness, Weinstein et al. (1990) report an
interaction between handedness and academic
Subject Variables
major in college women. The highest quality
of reproduction (in both copy and recall con- Sample Size
ditions) was seen in left-handed math/science
majors, while the poorest quality was dem-
size. Although this criterion is somewhat
onstrated by familial right-handed non-math!
arbitrary, a large number of studies suggest
science majors. The authors attributed this
difference to an increased ability to coordinate
highly influenced by individual differences
the use of left and right hemisphere process-
ing in high-performing groups.
population mean.
Ardila et al. (1989) found interactions be-
tween education and age and between edu-
Sample Composition Description
cation and gender on ROCF copy scores in
their Spanish-speaking sample, with age and Information regarding medical and psychiatric
gender effects observed primarily in partici- exclusion criteria is important. It is unclear if
pants with no formal education. However, in a geographic recruitment region, socioeconomic
larger sample of participants limited to age 55 status, occupation, ethnicity, handedness, and
and older and classified into 0-5, ~12, and recruitment procedures are relevant. Until
> 12 years of education, while main effects for this is determined, it is best that this infor-
age, education, and gender on ROCF scores mation be provided.
were also documented, no interaction be-
Age Group Intervals
tween age, education, and gender was found
(Ardila & Rosselli, 1989). nus criterion refers to grouping the data into
Rosselli and Ardila (1991) reported effects limited age intervals. This requirement is rel-
of age, educational level, and gender on the evant since a strong effect of age on ROCF
scores for both copy and immediate repro- performance has been demonstrated in the
duction of the figure. literature.
The discrepancy in the Ardila et al. (1989)
and Boone et al. (1993a) reports regarding the Reporting of Educational Levels
importance of education for ROCF perfor- Given the probable relationship between
mance is probably related to differences in the education, especially low levels, and ROCF
educational levels sampled. The Boone et al. performance, information regarding educa-
(1993a) population included only four partici- tion should be provided for each subgroup.
Reporting of Intellectual levels publications provided data for males and fe-
Given the relationship between ROCF per- males separately.
formance and IQ, information regarding in- Several data sets had total sample sizes
tellectual level should be reported for each >50; however, a few studies did not have
subgroup, and preferably normative data adequate medical and psychiatric exclusion
should be presented by IQ levels. criteria. Information on geographic recruit-
ment area was present in the majority of data
Reporting of Gender Composition sets. Ethnicity, handedness, and occupation
Given the possible relationship between gen- were less commonly reported.
der and ROCF performance in favor of males, In regard to procedural variables, all studies
information regarding gender composition reported either mean raw scores or mean per-
should be reported for each subgroup. cent retention and almost all reports provided
SDs. Several studies presented copy scores only
or recall performance only. Although docu-
Procedural Variables mentation of copy strategy through use of col-
ored pencils is commonly employed, only a few
Description of the Scoring System Used
investigators have reported qualitative infor-
A clear statement of the method used for mation on the strategy features of ROCF copy
scoring the ROCF is important because nu- or delay. Similarly, only a few reports have in-
merous scoring systems are available. In ad- dicated that ROCF performance was timed,
dition, given the rather subjective nature of and only a few of these reported the data on
some of the scoring procedures, information performance time. Number and types of error
regarding interrater reliability is desirable. for copy and recall were described in a few
studies.
Data Reporting
The length of time elapsed before delayed
Means and standard deviations, and prefera- recall varied widely: immediate recall, 3-min-
bly ranges, for copy and recall scores are im- ute delay, 10-minute delay, 20-minute delay,
portant. Percent forgetting or percent 30-minute delay, 40-minute delay, 45-minute
retention could substitute for a recall score delay, and 24-hour delay. In only some studies
mean. However, SDs should be used with were the precise scoring systems described.
caution in evaluating the relative standing of Only a few studies reported interrater reli-
an individual score because ROCF scores are ability data for ROCF scores and internal
not normally distributed. consistency indices.
Among all the studies available in the liter-
ature, we selected for review those based on
well-defined samples or that offered some in-
formation not routinely reported. A majority of
OF THE NORMS
authors provide data based on the Osterrieth-
In terms of subject variables, not all studies Taylor scoring system. Any deviations from this
have provided data according to age intervals, standard reporting are specified in the context
although all data sets do indicate the mean age of each table. Scores based on the BQSS are
of the sample. Only a few studies have pro- not reproduced in the tables for this chapter.
vided data for elderly participants. Several The clinician is urged to pay close attention
studies have reported IQs or IQ estimates. to the sequence of trials (e.g., administration of
However, only one publication provided data the immediate recall prior to the delayed recall
by IQ intervals. Mean educational level was trial) and the length of the delay interval (e.g.,
indicated in many studies, but only a few 3- vs. 30-minute delay) as these factors have a
publications reported data by educational le- notable effect on rate of recall (see above).
vel or in age-by-education cells. Information In this chapter, normative publications and
on gender composition of the samples was control data from clinical studies are reviewed
documented in many studies, and several in ascending chronological order. The text of
study descriptions contains references to the Study strengths

corresponding tables identified by number in 1. Information on IQ, age, gender distri-
Appendix 12. Table A12.1, the locator table, bution, geographic recruitment area, and
summarizes information provided in the stud- handedness is provided.
ies described in this chapter. 1 2. Relatively large sample size.

SUMMARIES OF THE STUDIES 1. While no objective evidence of brain
dysfunction was found on lab tests,
[ROCF.1] Powell, 1979 (Table A12.2)
patients were apparently suspected of
The authors report data on 64 Londoners as having dysfunction-hence, the referral
part of an examination of the relationship for neuropsychological testing. Probably
between ROCF performance and IQ and a sizable percent had some type of subtle
verbal memory performance. The 64 "nor- brain dysfunction not detected by the
mal" participants were part of a sample of 150 standard diagnostic laboratory measures.
right-handed patients referred for neurologi- The significantly lower PIQ corroborates
cal screening but "confirmed as having no that this group is probably not "normal."
brain damage on the usual physical tests such 2. Undifferentiated age range.
as the EMI scan" (p. 336). Twenty-one of the 3. Lack of data regarding education.
64 participants were female, and mean age 4. Lack of data regarding copy scores or
was 41.0 (14.05). Mean Verbal and Perfor- raw recall scores.
mance IQs (VIQ, PIQ) prorated from scores 5. The data were collected in England and
on the Comprehension, Similarities, Vocabu- may have limited value for use in this
lary, Block Design, and Object Assembly sub- country.
tests were 107.70 (16.80) and 83.70 (21.55), 6. No information on scoring system or
respectively. interrater reliability.
Means and SDs for percent retention of the
figure following a 40-minute delay compared [ROCF.2] King, 1981 (Table A12.3)
to original copy scores are reported. Significant The authors obtained ROCF data on 71 Ca-
correlations were noted in the sample as a nadian controls as part of a study on the ef-
whole between ROCF percent retention score fects of lateralized nonfocal brain dysfunction
and Block Design (r=0.51), Object Assembly and age on the ROCF and the relationship
(r=0.50), Digit Span (r=0.28), Comprehen- between ROCF and Wechsler Memory Scale
sion (r=0.39), Similarities (r=0.32), Vocab- Visual Reproduction performance in a sam-
ulary (r = 0.27), PIQ (r = 0.38), VIQ (r = 0.30), ple of 185 participants. Controls consisted of
and age (r= -0.33). Powell concludes that healthy volunteers or patients with "non-
ROCF performance is more associated with neurological or psychiatric conditions." Mean
PIQ than VIQ. He cautions that the ROCF age, years of education, and WAIS FSIQ were
"must be viewed in the light of the patient's 39.6 (21.4), 11.4 (2.9), and 104.5 (18.1), re-
intelligence . . . [and] in relation to the pa- spectively.
tient's age, since older people score lower'' Participants copied the ROCF and drew it
(p. 339). Powell describes how the clinician from memory following a 40-minute delay,
can derive an expected memory score based on during which time verbal tasks were admin-
IQ from a regression equation, which can be istered. Copy and recall scores were obtained
used to determine if the patient's actual using the E. M. Taylor (1959) guidelines. A
memory score is worse than expected. percent recall score was also calculated by
multiplying the ratio of recall to copy score
by 100.
'Nonns for children are available in Baron (2004) and Means and SDs for copy. recall, and per-
Spreen and Strauss (1998). cent recall are reported for three age groups:
< 30 (n = 36), 30-60 (n = 17), and >60 (n = deficits similar to those seen in patients with
18). Significant correlations were o}ltained bilateral and diffuse cortical damage.
between ROCF scores (copy, recall, and per-
cent recall) and age and IQ in the pati~ts and Study strengths
controls as a whole. In addition, for the whole I. A nonclinical sample of participants with
sample, males were found to perform. signifi-
good or poor attentional ability was
cantly better than females on recall and per- studied.
cent recall; however, the amount of .Jmance
2. Specification of scoring system.
accounted for was negligible.
3. Information on gender, age, and educa-
tion is provided.
Study strengths 4. Restricted age range.
I. Data are presented by age groupings. 5. Means and SDs are reported.
2. Information regarding IQ, educatiJ>n, and
geographic recruitment area is prl>vided. Considerations regarding use of the study
3. Specification of scoring system. j I. All participants were male.
4. Means and SDs are reported. l 2. Demographic variables for each group
'
I
Considerations regarding use of the st;dy
are not provided.
3. High intellectual level of the participants.
1. An unspecified number of p~ipants 4. No exclusion criteria.
were medical or psychiatric patiepts. 5. No information on interrater reliability.
2. The data were collected in Cana~. and 6. Small sample size.
their usefulness for an American ~ample
is unclear. i [ROCF.4] Bennett-Levy, 1984 (Table A12.5)
3. No information on interrater reli.bility. The authors collected ROCF data on 107
4. Below average educational level. 1 English volunteers as part of a project to de-
5. No information regarding gende~ distri- velop a quantified technique of scoring copy
bution. .i strategy and assessing its relationship with
6. There is an apparent error in tlte data recall. Forty-five participants were hospital-
presentation table: the recall s;'re for ized patients tested prior to nonemergency
controls < 30 years reads 0.0. ( r cal- surgery; 62 were auto-assembly line workers.
culations suggest it should read .0.) Seventy-six were male and 31 were female.
7. The sample sizes for the individ'al age Exclusion criteria were history of head injury
groupings are small. or epilepsy. Mean age was 29.3 (9.3), range
17-49, and mean estimated IQ based on the
[ROCF.3] Huhtaniemi, Haier, Fedio, and. Schonell Graded Word Reading Test and
Buchsbaum, 1983 (Table A12.4) the New Adult Reading Test was 104.9 (7.6).
Participants were instructed as follows:
Four hundred male college student~ were
screened on a measure of vigilance, the Con- "Copy the figure as accurately as [possi-
tinuous Performance Test (CPT). Two groups ble] ... While the subject copied the figure, the
were identified for further study: (1) t good experimenter made a note of every line drawn in
attention group (upper 5% of CPT se<fe dis- sequence . . . When the subject indicated that
tribution, n = 13) and (2) a poor a~ntion she had finished the copy, the figure and the
group (lower 5% of CPT score distriilution, drawing were removed from sight; 40 minutes later,
n = 12). Participants ranged in age :19-29 the experimenter said to the subject: 'You remem-
years, with a mean age of 22 years. i ber that drawing you copied for me. I would now
Participants copied the ROCF and re- like you to recall as much of it as you possibly can.'
produced the figure from memory 3 olinutes When the subject first indicated that s/he could
later. The standard scoring system was 1used. recall no more, the experimenter said: 'Give your-
The authors concluded that atte.tional self a little more time. I always say to people to give
dysfunction is associated with perfor'loance themselves some more time.' In a number of cases,
I
this procedure resulted in one further detail being to consider strategy and age and reliance
recalled. When the subject indicated that s/he had solely on normative recall scores might lead
finished, this was accepted. (p. 110) to an inaccurate interpretation of memory
impairment.
Bennett-Levy developed precise scoring
criteria for the copy and recall of the figure Study strengths
based on Osterrieth's (1944) and E. M. Tay- 1. Careful specification of a scoring system,
lor's (1959) outline. He noted that his scoring including quantification of drawing
system is "almost certainly more stringent strategy, and information on interrater
than those customarily used by clinical reliability.
psychologists." He opted for strict scoring 2. Specification of a regression equation
criteria, to reduce the presence of ceiling ef- which can provide a predicted recall score
fects on the copy trial; a majority of the pro- to be used for comparison with actual test
tocols would have achieved the maximum scores.
possible score if more lenient criteria had been 3. Relatively large overall sample size.
used. Bennett-Levy developed a more lenient 4. Information on IQ, gender, age, occu-
scoring system in addition to his strict criteria pation, and geographic recruitment area
for the recall trials, reasoning that memory is provided.
scores should not be penalized due to sloppi- 5. Presentation of data by gender.
ness or imprecision of the reproduction when 6. Information on time to complete the
it was clear that the details were in fact accu- drawing.
rately remembered. Interrater reliabilities of 7. Means and SDs are reported.
0.96 and 0.98, respectively, for the "strict" and
"lax" scoring systems were obtained on 25 Considerations regarding use of the study
randomly selected recall protocols. Bennett- 1. A unique scoring system, which is much
Levy also developed scores describing "good more complicated than the Taylor sys-
continuation" (when a straight line is drawn as tem and not in wide usage.
one piece until its final intersect with another 2. Use of hospitalized orthopedic patients
line), "symmetry" (successive construction of for nearly half of the subject sample.
symmetrical units), and "strategy total" (sum 3. Undifferentiated age range.
of good continuation and symmetry scores). 4. Minimal exclusion criteria.
Means and SDs for copy score, strict and lax 5. No information on educational level.
recall, copy time, symmetry, good continua- 6. The data were collected in England, and
tion, and strategy total are reported for males their usefulness for an American sample
and females separately and for the sample as a is unclear.
whole. Multiple regression analyses revealed
that strategy, copy time, and age were the [ROCF.S). Speers and Ribbler, Unpublished
primary determinants of copy scores, while Manuscript (Table A12.6)
strategy, copy score, and age were the princi- The authors report data on 40 (20 male, 20
pal predictors of recall. IQ and gender were female) normal participants tested as part of a
significantly associated with copy and recall study on rates of loss for newly learned infor-
performance but did not provide a unique mation over a 24-hour delay. Participants were
contribution to prediction once the other recruited from staff and visitors at a rehabili-
variables were considered. Bennett-Levy pro- tation hospital in California. Those with ab-
vides a regression equation to allow prediction normal neurological histories were excluded,
of lax recall from strategy total and age. He as were those with signs of language, motor, or
suggests that a large discrepancy (i.e., >2 stan- visual disability. Mean age was 35.00 (10.79),
dard errors) between predicted and observed range 23-70, and mean years of education was
recall scores in favor of predicted score would 16.15 (2.77), range 10-22. IQ estimates ob-
argue for the presence of a significant memory tained through the Quick Test revealed a
impairment. He further cautions that failure mean IQ of 107.93 (8.73), range 87-123.
The Rey figure was modified to include an 5. High mean educational level.
additional detail in the left lower quadrant. 6. No data on interrater reliability.
Participants copied the design and then drew 7. Sample size is relatively small.
it from memory immediately and at delays 30
minutes and 24 hours later. They were not [ROCF.61 Ardila, Rosselli, and Rosas, 1989
informed that they would be requested to (Table A 12.7)
recall the figure again after the immediate The authors obtained ROCF data on 200
recall attempt. Participants were kept occu- "normal," Spanish-speaking, right-handed Col-
pied during the 30-minute delay intetval with ombians as part of their assessment of neuro-
a questionnaire and IQ test items; the 24-hour psychological functioning in illiterates. Inclusion
delay period was unstructured. Time to copy criteria for illiterate participants were as fol-
the figure averaged 3 minutes 26 seconds. lows: (1) total illiteracy, (2) illiteracy due to lack
The authors developed a scoring system in- of opportunity to attend schooL (3) no current
volving a 0-2 rating scale for 50 separate scoring or past neurological or psychiatric history or
units, which resulted in a total possible score of sensory or motor impairment, (4) adequate
100. Approximately 12% of the information performance in activities of daily living. All
could not be retrieved on immediate recall, were of low income, and occupations included
with additional!% losses noted on 30-minute factory or construction workers, maids, cooks,
and 24-hour delayed recall. No mean scores or or homemakers.
SDs are reported; however, rates of forgetting High-education participants were recruited
over the three delay periods are plotted in a to match the illiterates on age and gender.
graph format, and the modal reproduction of Participants aged 1~25 were students and
the figure at copy, immediate, and 30-minute had at least 10 years of education; participants
and 24-hour delays are pictured. The in- over age 25 had at least 17 years of formal
vestigators cite significant negative correlations education. No exclusion criteria are reported.
between amount of information loss and age The E. M. Taylor (1959, as reported in Le-
(r= -0.39) but found no significant relation- zak, 1983) scoring criteria were employed.
ship between IQ (r= -0.12) or education Copy and immediate recall data were obtained,
(r= 0.09) and retention of the figure. although only the means for ROCF copy in
gender-by-age-by-education cells are reported;
Study strengths age intervals were 1~25. 2~. 36-45, 46-55,
1. Presentation of data regarding three dif- and ~ and each cell had 10 participants.
ferent recall intervals (although earlier re- Significant education, gender, and age main
call trials no doubt influenced later recall effects were documented for copy and imme-
trials to some unknown extent and, there- diate recall performance, with better perfor-
fore, the data on 30-minute and 24-hour mance associated with males, younger age, and
delays should not be used as a comparison higher education. In addition, significant inter-
reference for participants who were not actions on copy score were noted for education
administered the earlier recall trials). and age, with lower scores primarily observed in
2. Information on IQ estimates, age, gen- the illiterate, rather than the educated, partici-
der distribution, educational level, and pants over age 55. Similarly, a significant edu-
geographic recruitment area is provided. cation and gender interaction was documented,
3. Adequate exclusion criteria. with men outperforming women only in the il-
literate group. No significant interactions were
Considerations regarding use of the study obsetved for recall performance.
1. Modification of the stimulus figure.
2. A unique scoring system, not in common Study strengths
usage, was employed. 1. Provided data on a Spanish-speaking
3. Lack of mean scores and SDs for recall population.
conditions. 2. Provided data on ROCF performance in
4. Undifferentiated age range. illiterates.
3. Large overall sample size. 2. Organization of the data into relatively

4. Data are partitioned in age-by-education- small age intervals.
by-gender cells. 3. Adequate exclusion criteria.
5. Scoring system is specified. 4. Specification of scoring system.
6. Information regarding gender, occupa- 5. Information on IQ, education, gender, and
tion, handedness, and geographic re- geographic recruitment area is provided.
cruitment area is provided. 6. Means and SDs are reported.
1. Small sample size in individual cells. 1. Very high intelligence and educational
2. ROCF scores for immediate recall are level of the sample.
not presented. 2. No information regarding interrater re-
3. No SDs are reported. liability.
4. No information on interrater reliability.
5. No information on exclusion criteria in [ROCF.BJ Berry, Allen, and Schmitt, 1991
educated participants. (Table A12.9)
6. No IQ data available. The authors collected ROCF data on 107 (55
7. Data were collected in Columbia, and male, 52 female) elderly Caucasian partici-
their usefulness for an American sample pants from Kentucky, aged ~79, as part
is unclear. of their evaluation of the psychometric prop-
erties of the ROCF and Taylor figure. Mean
[ROCF.7] Van Corp, Satz, and Mitrushina, age was 65 (8.6), and participants were re-
1990 (Table A12.8) cruited from newspaper ads, flyers at senior-
The authors collected data on 156 healthy citizen centers, and a volunteer subject pool.
elderly adults, aged 57-85, living indepen- Exclusion criteria were history of cardiac,
dently in a retirement community in southern neurological, or psychiatric disease or use of
California as part of their investigation of psychoactive medication. All participants un-
neuropsychological processes in normal aging. deiWent a physical exam by a physician and
Exclusion criteria were history of neurological electroencephalographic (EEG) evaluation by
or psychiatric disorder or substance abuse. a neurologist. Ten recruits were excluded due
Thirty-nine percent (n = 62) of the sample to the discovery of previously undetected
were male and 61% (n = 94) were female. diseases.
Mean years of education were 14.1 (2.9), and Participants were provided with blank
mean WAIS-R (Satz-Mogel format) FSIQ, sheets of 8 ~" x 11" paper and told to draw the
VIQ, and PIQ were 117.21 (12.59), 118.77 complex figure as best they could. With-
(13.27), and 110.74 (13.07), respectively. out being forewarned, participants were in-
Mean scores and SDs are provided based structed to draw the design from memory
on the Taylor (1959) scoring system for the immediately after and 30 minutes later. The
copy and 3-minute delayed recall of the period between immediate and delayed recall
ROCF for four age intervals: 57-65 (n = 28), was occupied with verbal testing. The scoring
~70 (n=45), 71-75 (n=57), and 76-85 system was based on the E. M. Taylor (1959)
(n = 26). Mean VIQ for each age interval protocol, with the following exceptions: (1)
ranged 114.8-122.9, and mean PIQ ranged distorted but properly placed details or cor-
101.~115.1. No differences in VIQ, PIQ, or rectly reproduced but misplaced details re-
education were found between older and ceived 1.5 points (rather than 1 point, as in the
younger participants (<70 vs. >70). Taylor system) and (2) severely misplaced but
correctly drawn details or severely distorted
Study strengths but correctly placed details received 1.25
1. Relatively large sample size, with points (rather than 1 point, as in the Taylor
some individual age groupings approxi- system). Significant correlations were noted
mating 50. between the revised and Taylor systems (copy,
r=0.82; immediate, r=0.93; delay, r:::::0.96); 6. Data on the Taylor figure as well as com-
scores for the revised system averaged 2 prehensive psychometric information for
points higher. Interrater reliability quotients the ROCF.
for the revised scoring system on 87 of the 7. Means and SDs are reported.
protocols were 0.80, 0.93, and 0.96 for the
copy, immediate, and delayed recall trials, Considerations regarding use of the study
respectively; and the most experienced rater's 1. Undifferentiated age range spanning
scores were used for data analysis. three decades.
In a subgroup of 54 participants, alternate 2. Even though the total sample size is
form equivalence was assessed by adminis- large (n = 107), it appeared that the re-
tering either the ROCF or Taylor figure in a ported means and SDs were based on
morning testing session; in the afternqon, the sample sizes of either 54 or 41 but that
version not given in the morning was ,admin- the information on mean age and edu-
istered. Interrater reliability for 37 1 Taylor cation was derived from the whole sam-
protocols was comparable to that docutnented ple.
with the ROCF (copy, r=0.84; imi 1 ediate, 3. Idiosyncratic scoring system.
r = 0.97; delay, r= 0.93), and the mo expe- 4. High mean educational level.
rienced rater's scorings were used r data 5. No information regarding IQ.
an alySIS.
0
i
'
6. The baseline data on the subset of
Test-retest reliability was evaluattfJ in a 41 participants on whom the 1-year ret-
subset of 41 participants who were ~tested est data were obtained have mean recall
1 year after original testing. The ~.uthors scores 5 points below the sample of
conclude that ROCF copy scores w+re not 54 participants, suggesting that the sub-
reliable, with moderate reliability nQted on set was not randomly chosen and was not
immediate and delayed recall. Howeter, ex- representative of the larger sample.
amination of mean scores suggests thlit there
was not a clinically significant change i~ scores [ROCF.91 Tombaugh and Hubley, 1991
over 1 year (e.g., means differed by 0.3-1.0 (Tables A12.1 0, A12.11)
point). The authors assessed the comparability of
Means and SDs are reported for copy, im- scores obtained on the ROCF and Taylor fig-
mediate recall, and 30-minute delaye~ recall ure in copy and recall conditions. Four studies
for the ROCF and Taylor figure for 51 parti- were reported.
cipants as well as for baseline and 1-year Study 1 used an incidental learning para-
follow-up ROCF scores in the subset of 41 digm, in which participants were not informed
participants. Data on internal cons~tency, that recall would follow the copy condition.
construct validity, and criterion-related valid- Participants were 64 undergraduate students
ity are also reported. Significant correlations enrolled in a third-year psychology course who
between ROCF scores and age and education were randomly assigned to copy the ROCF or
were documented for immediate and delayed Taylor figure. Participants reported no history
recall but not for copy scores; gender was not of head injury producing loss of consciousness,
significantly related to any ROCF scores. neurological dysfunction, or current use of
psychoactive medication. Participants were
Study strengths allowed 5.5 minutes to copy the complex fig-
1. Relatively large sample size. ure, after which they were instructed to re-
2. Informationregardingage,education,gen- produce it from memory. After a 4-minute
der, ethnicity, recruitment proc,dures, delay filled with a nonpictorial task, the second
and geographic area is provided. ~ recall condition was offered. Twenty minutes
3. Information regarding interrater reli- later, following completion of verbal learning
ability is provided. tasks, participants were administered the third
4. Well-specified exclusion criteria. i recall condition. A 2-minute time limit was
5. Specification of scoring system. imposed on all recall trials.
Itemized scoring systems allowing a total of sure. Six learning trials were used, followed by
72 points were used for scoring of both figures 20-minute delayed recall trial and a copy trial.
(Denman, 1984, system for ROCF; Tom- A maximum of 2 minutes was allowed for all
baugh, unpublished research, system for Taylor memory trials and 4 minutes for the copy trial.
figure). Itemized scoring systems were used. The re-
The authors concluded that the degree of sults suggested that the difference in per-
accuracy on the copy condition was compa- formance was most pronounced with the
rable for the two figures. However, all recall 60-second presentation interval. The authors
trials yielded significantly higher scores for the inferred that reported differences in recall
Taylor figure (Table A12.10). efficiency of the two figures reflect lower de-
Study 2 used another sample to replicate gree of learning with the ROCF in compari-
the above results and to explore the effect of son to the Taylor figure.
scoring system on the comparability of scores
for the two figures. Study strengths
Participants were 67 undergraduate stu- 1. The authors systematically explore the
dents enrolled in a third-year psychology effect of different variables on perfor-
course who were randomly assigned to copy mance on two figures.
the ROCF or Taylor figure. The procedure 2. Sample sizes are sufficient for these ho-
was similar to that used in study 1, with the mogeneous samples.
exception of omission of the 4-minute delayed 3. Minimally adequate exclusion criteria.
trial and inclusion of a 1-month delayed trial 4. Means and SDs are reported.
(which was based on the data for 52 partici-
pants). In addition, the time required for copy Considerations regarding use of the study
and reproduction of each figure was recorded. 1. Age range and demographic character-
Reproductions were scored according to the istics of the samples are not reported;
itemized system and the original Osterrieth- age range is probably sufficiently narrow.
Taylor system. 2. Itemized scoring systems are not suffi-
Results were consistent with study 1 in that ciently described, and no information is
both figures were comparable on the copy but reported on interrater reliability.
not on the recall condition, irrespective of 3. High educational level (third year of
scoring system. No forgetting was demon- college).
strated between the immediate and 20-minute
delayed recall trials; however, a substantial [ROCF.10] Berry and Carpenter, 1992
decline in scores, equivalent for both figures, (Table A12.12)
was seen over the 1-month interval. The authors report the rate of recall of the
Study 3 used a modified size of the base ROCF over four different delay periods (15,
rectangle of the ROCF to equate it to the base 30, 45, and 60 minutes) in older persons in
square of the Taylor figure (total area of both Kentucky. A sample of 60 participants was
figures= 64 cm2). Itemized scoring systems divided into four groups of 15, which were
were used. Results indicated that differences equivalent in age, gender, and education. All
in recall scores could not be attributed to participants were volunteers who were in
differences in size of the structural compo- good health with no active illness, no history
nents of the two figures. of neurological or psychiatric illness, and no
Study 4 modified the administration of the current use of psychoactive medication. Mini-
figures to explore the effect of intentional Mental State Exam (MMSE) scores were
learning and difference in time of exposure on >24, with a mean of 28.5 (1.7). The sample
recall efficiency. Seventy-two students en- consisted of 31 males and 29 females. All par-
rolled in a first-year psychology course were ticipants were white, and 10% of the sample
instructed to study figures for a specified in- were left-handed.
terval of time (15, 30, and 60 seconds) in order All participants were administered copy and
to reproduce them from memory after expo- immediate recall trials with no time limits,
which were followed by one of the four delay The ROCF and Taylor figure were adminis-
durations. Timing of the delay started from tered in the same order to all participants,
completion of the copy trial. During tbe delay which is consistent with the order used in
periods, participants were administered other clinical practice. The time interval between
neuropsychological tests of a verbal 'nature. administration of the two figures was approxi-
Each protocol was scored for accutacy by mately 1 month. Three conditions were ad-
two independent raters using the sys4em de- ministered for both figures: copy, immediate
scribed by Beny et al. (1991). Interr~er reli- recall, and 20-rninute delayed recall (delay filled
ability for the three trials was as foll~: copy, with nonvisuospatial tasks). Reproductions
r=0.95; immediate recall, r=0.98; belayed were scored according to the standard criteria.
recall, r=0.99. Scores for data anal~s rep- Interrater reliability based on scoring of 10
resent the average of final scores assigned by samples by two experienced neuropsycholo-
two raters for each protocol. gists was 0.91. Correlations with age (- 0.11
The results revealed no significant affect of to- 0.26) and education ( - 0.01 to 0.20) were
delay period on recall. Scores on im~ediate relatively low. The authors concluded that
and delayed-recall trials were silllihtr. The performance on the copy condition for both
authors inferred that most forgetting! occurs figures was nearly identical; however, partici-
very quickly, as a result of "overl$ading" pants performed significantly better on the
working memory. Taylor figure on both recall conditions.
Study strengths . Study strengths

1. Sample composition is well descrf.bed in 1. Information on interrater reliability is
terms of age, ethnicity, gender, jeduca- provided.
tionallevel, handedness, and geographic 2. Information regarding age, education,
location. ; and geographic area is provided.
2. Adequate exclusion criteria were rused. 3. Information on alternate form is provided.
3. Interrater reliability and scoring 1system 4. Sample size approximates 50.
are reported. ; 5. Minimally adequate exclusion criteria.
I
4. Means and SDs are reported. 6. Means and SDs are reported.
5. Age range is probably sufltciently
narrow. Considerations regarding use of the study
1. The data are not broken down by age.
Considerations regarding use of the study 2. SDs for age and education are not
1. While overall sample is adequate, indi- reported.
vidual sample sizes are small. 3. No information regarding IQ or gender.
2. High educational level.
3. No IQ information is reported. [ROCF.12] Kuehn and Snow, 1992
(Table A12.14)
[ROCF.11] Delaney, Prevey, Cramer, and The study explored the comparability of the
Mattson, 1992 (Table A12.13) ROCF and Taylor figure in a clinical sample.
This study addressed the comparability of the Participants were 38 Canadian patients re-
ROCF and Taylor figure in a nonpatient ferred for neuropsychological assessment for
sample and is based on the control iample various forms of brain damage. Patients un-
data collected as part of a large study carried able to draw a Greek cross or administered
out in various locations of the United: States either figure previously were excluded from
on the effect of anticonvulsant medica~ns on the study. Mean age was 46.7 years.
memory functioning. Participants were free of The procedure consisted of copying each
neurological and psychiatric disorders or cur- figure with a lead pencil, followed by 40-
rent drug history. Ages ranged 22--61 years minute delayed recall (without forewarning).
and education, 6-16 years. ' Approximately 3 hours elapsed between
administration of the two figures, during tory tests, or abnormal findings on EEG or
which time tests involving drawings or visual MRI. The final sample included 34 males and
memory were not administered. Two figures 57 females. Seventy-one participants were Cau-
were presented in a counterbalanced order. casian, 10 were African American, five were
The standard scoring systems were used for Asian, and five were Hispanic. Mean educational
both figures. Percent recall was calculated. level was 14.5 (2.5) years. and mean WAIS-R
The authors concluded that performance FSIQ (Satz-Mogel format) was 115.9 (13.0).
on both figures was equivalent for copy and Participants were instructed to copy the
recall scores. Percent recall scores, however, figure onto a blank paper "as carefully as you
were higher for the Taylor figure, when it was can without tracing." Performance was not
administered first. timed, and participants were allowed to make
erasures. Following a 3-minute verbal fluency
Study strengths task and without forewarning, participants
1. Scoring system is specified. were instructed to draw what they could re-
2. Information on gender, age, education, member of the figure on a second sheet of
IQ, and geographic area is provided. blank paper. The E. M. Taylor (1959) scoring
3. Information on alternate form is pro- system was employed.
vided. Means and SDs are reported for copy
4. Means and SDs are reported. scores and percent retention for three age
groupings (45--59, 60--69, and 70--83) and four
FSIQ levels (90-109, 110-119, 120-129, and
Considerations regarding use of the study 130-139). Interrater reliability between two
1. Data are not broken down by age group.
experienced neuropsychologists was 0.82 for
Age range is not specified.
copy and 0.93 for delay. In regression analy-
2. The two groups, used for counter-
ses, a relatively small but significant percent of
balancing, are not comparable in edu-
the variance in ROCF performance was as-
cation but are comparable in IQ.
sociated with age and FSIQ; gender and ed-
3. Clinical sample; no exclusion criteria.
ucation were not predictive of ROCF scores.
4. No information on interrater reliability.
In addition, ROCF copy score was not asso-
5. Small sample size.
ciated with delay score or percent retention.
6. Data were collected in Canada and may
Significantly poorer ROCF scores did not
be problematic for use in the United
emerge until age 70 and older, and individuals
States.
of average IQ showed a trend toward poorer
performance on ROCF delay relative to par-
[ROCF.13] Boone, Lesser, Hiii-Gutierrez, ticipants falling in the very superior intelli-
Bennan, and D'Eiia, 1993b (Table A12.15) gence range. No interaction effects between
The investigators collected data on 91 fluent age and FSIQ were obseiVed. The number
English-speaking healthy older adults recruited and type of errors committed on copy and
in southern California through newspaper ads, recall are summarized.
flyers, and personal contacts as part of their
investigation of the effects of age, IQ, educa-
tion, and gender on ROCF performance. Ex- Study strengths
clusion criteria were current or past history of 1. Information regarding education, gender,
major psychiatric disorder or alcohol or other geographic recruitment area, ethnicity,
substance abuse, neurological illness, and sig- and recruitment procedures is provided.
nificant medical illness which could affect 2. Rigorous exclusion criteria.
central neiVous system function (e.g., uncon- 3. Data are presented by age and IQ
trolled hypertension or diabetes). In addition, groupings.
potential participants were rejected if they had 4. Scoring system is specified, and infor-
abnormal findings on neurological examination, mation on interrater reliability is pro-
metabolic disturbances detected with labora- vided.
5. Information regarding error number and 2. Relatively large sample size, and indi-
type for copy and delay is provi.ed. vidual cells approximate 50.
6. Large overall sample size, although in- 3. Administration system is specified.
dividual cells all fall short of 50. 4. Exclusion criteria are specified.
7. Means and SDs are reported. 5. Information on education, gender, and
geographic recruitment area is reported.
Consideration regarding use of the study 6. The study assessed strategy used in ap-
1. High intellectual and educational level. proach to drawings.
Other comments
1. For participants older than 74, age-cor- Consideration regarding use of the study
rected FSIQs were based on Ryan et al. 1. High educational level.
(1990) tables. 2. Data were checked by a blind evaluator,
but no information on interrater reli-
[ROCF.14] Chiulli, Haaland, LaRue, and ability is provided.
Garry, 1995 (Table A12.16) 3. No information on IQ.
The study explored rates of decline in ROCF
performance after age 70. Participants were [ROCF.15] Meyers and Meyers, 1995a
153 healthy elderly individuals aged· 70-93, (Table A12.17)
living independently, who participated in the The study explored the effect of different
New Mexico Aging Process Study, wliich ex- administration procedures on the rate of recall
plores nutrition and aging. Persons wtth seri- of the ROCF. Participants were undergradu-
ous medical illnesses or taking preseription ate students from a college in Iowa and had no
medications were excluded. The sample was prior history of head injury, drug abuse,
partitioned into three age groups. learning disability, or psychiatric illness. Par-
The ROCF was administered as part of a ticipants were randomly assigned to one of
brief battery of psychological tests. Standard four groups, each of which received a differ-
administration and scoring procedures were ent combination of trials (30 participants
used. A copy condition was followed by in each group). There was no significant dif-
immediate and 30-minute delayed recaU. If the ference between the groups on age, gender, or
reproduction started with the drawing of the education.
large rectangle, the approach was categorized Reproductions were scored according to
as "configura!." All other approaches were the system developed by Meyers and Meyers
determined to be "nonconfigural." All proto- (1992), which is based on the standard scoring
cols were checked by a second, blind ev~uator. system with addition of 114" rule for mis-
The results revealed a significant main ef- placement and a 118" rule for drawing errors.
fect for age group. Accuracy was greitest in In addition, the authors used a recognition
the copy condition but did not differ between trial (Meyers & Lange, 1994).
the immediate and delayed recall conditions. The authors suggest use of a 3-minute recall
The most pronounced decline in performance instead of immediate recall due to its higher
was demonstrated between the first aild sec- correlation with the 30-minute recall.
ond groups, which did not differ consi~erably
from the third group performance. No gender
effects were evident. The number of partici- Study strengths
pants using the configura! approach ~d not 1. Scoring system is described.
differ significantly for the three age gri>ups. 2. Sample composition and demographic
L characteristics are described, as well as
geographic area.
Study strengths 3. Overall sample size is large (n = 120),
1. Data for an elderly sample ar~ parti- although individual groupings are rela-
tioned into three age groups. tively small.
4. Adequate exclusion criteria. 3. Adequate exclusion criteria.

5. Means and SDs are reported. 4. Test administration and scoring proce-
6. Age grouping is suitably restricted. dures are specified.
Consideration regarding use of the study reported.
1. No information regarding interrater re- 6. Data are partitioned by gender x age x
liability or IQ. education.
[ROCF .16] Ponton, Satz, Herrera, Ortiz, Considerations regarding use of the study
Urrutia, Young, D'Eiia, Furst, and Namerow, 1. No information regarding interrater re-
1996 (Table A12.18) liability or IQ.
The ROCF was administered to Spanish- 2. It is unclear which of the two educa-
speaking volunteers as part of a larger battery tional groups included participants with
in a project designed to provide standardiza- 10 years of education.
tion of the Neuropsychological Screening
Battery for Hispanics (NeSBHIS). Volunteers [ROCF.17] Rapport, Charter, Dutra, Farchione,
were recruited through fliers and advertise- and Kingsley, 1997 (Table A12.19)
ments in community centers of the greater The study addressed interrater and internal
Los Angeles area over a period of 2 years. consistency reliabilities of the standard (as
Exclusion criteria were a history of neurolog- described in Lezak, 1995) and Denman scor-
ical or psychiatric disorder, drug or alcohol ing systems for the ROCF. Participants were
abuse, and head trauma. Data for a sample of 318 veterans (312 males, 6 females), aged 18-
300 participants with a median educational 84 years, who were referred to a Veterans
level of 10 years were analyzed. Participants Administration hospital assessment service.
ranged in age 16-75 years, with a mean of38.4 The majority of participants were inpatients.
(13.5) years. Education ranged 1-20 years, Mean age was 55.01 (4.31) years and mean
with a mean of 10.7 (5.1) years. The male-to- education, 12.62 (2.77) years.
female ratio was 40%/60%. The average du- Three independent raters scored copy and
ration of residence in the United States was immediate recall reproductions using stan-
16.4 (14.4) years. Seventy percent of the dard and Denman criteria. Interrater reli-
sample were monolingual Spanish-speaking, abilities are presented for the entire sample
and 30% were bilingual. The proportion of the and for three referral sources separately:
sample respective to their country of origin neurology, psychiatry, and rehabilitation
closely approximates the 1992 U.S. Census medicine. The authors concluded that internal
distribution. Correlations between Marin and consistency and interrater reliabilities for both
Marin (1991) acculturation scale scores and scoring systems were high. Coefficient !X reli-
neuropsychological variables are provided. abilities were also high, indicating psycho-
Participants were instructed to copy the metrically sound inter-item congruity for both
complex figure with no time limit. Reprod- scoring systems.
uctions were scored according to Taylor's Age was modestly related to performance
(1959) criteria. The authors provided norma- on the copy condition and strongly related to
tive data for the copy and 10-minute delayed recall. Education was modestly associated
recall conditions. with copy and weakly associated with recall
performance.
Study strengths
1. Large overall sample, with acceptable Study strengths
sample size for most of the cells. 1. Information on gender, age, education,
2. The sample composition is well described and recruitment procedures is provided.
in terms of age, education, gender, ac- 2. A large sample size.
culturation information, geographic area, 3. Data on psychometric properties of the
and recruitment procedures. ROCF are provided.
4. Two scoring systems are compared. two age groups according to the extended
5. Means and SDs are reported. , 36-point scoring system are presented in
Table A12.20.
Considerations regarding use of the sJudy The authors found that lower performance
1. Participants were V.A. inpatients from for the older group, on the Copy condition,
different wards, including neurology. was the result of minor inaccuracies in drawing
Selection criteria and participants' diag- and, on the Recall condition, the result of
noses are not specified. The dati on test omission of elements. No decline in organiza-
scores are of limited use with the, general tional quality with age was evident. Small age
population due to likely health con- differences were seen on the copy condition,
founds of the sample. with robust differences evident in recall. The
2. The sample was not partitioned Jnto age authors discussed the advantages and dis-
groups. advantages of the BQSS and the extended 36-
3. No information on IQ. point scoring system. Table A12.20 provides
4. Mostly male population. data according to the latter scoring system.
[ROCF.18] Hartman and Potter, 1998
Study strengths
(Table A12.20)
The authors explored the contributio•s of vi- in terms of age, gender, vocabulary test
suospatial ability, organization, and memory to scores, and recruitment procedures.
age differences on the ROCF in ad$lthood. 2. Rigorous exclusion criteria.
Participants were 30 undergraduaf:e and 3. Two scoring systems are compared.
graduate students aged 18-32, with a tnean of 4. Means and SDs for the test scores are
22.3 years, and older adults recruited through reported.
fliers and advertisements in local ne~papers 5. Information on scoring system and in-
and senior-citizen newsletters. Participants terrater reliability is provided.
were screened for history of neurological ill-
ness, head trauma or loss of consciousness,
significant psychiatric illness, untreated hy-
1. The samples are relatively small.
pertension, current use of psychoactive med-
2. Educational levels for the samples are
ication, excessive current use of alcohol, and
high.
dementia. All participants lived independently
3. SDs or ranges for education are not
in the community and reported thems~lves in
provided.
good or excellent health. All older: adults
scored >24 on the MMSE. The t)vo age
groups were selected from a larger satn,ple in [ROCF.19] Ostrosky-Solis, Jaime, and Ardila,
order to match them on Shipley Qartford 1998 (Table A12.21)
Vocabulary Test scores (36.2 vs. 35.5)J The authors investigated an effect of normal
The ROCF was administered actording aging on memory abilities. The sample in-
to Rey's (1941) original instructions, using cluded 105 participants (44 male, 61 female)
different-colored pens handed to p~ipants aged 20-89 years, with a minimum of 6 years
at equal intervals. Copy and immedia~ recall of formal education. The sample was parti-
without forewarning were used. Sco~g was tioned into seven age groups, with 15 partici-
done by two investigators using BQSS ;nd the pants in each group. All volunteers were of
extended 36-point system. BQSS infraclass average socioeconomic status, lived in Mexico
correlations for a subsample of 22 p~tocols City, and were native Spanish speakers. Ex-
ranged 0.79-1.00, with the exception ~f qual- clusion criteria were presence of dementia
itative items (perseveration, confab'itlation, according to the DSM-IV criteria, a score
and neatness), which were low, 0.$--<>.65. < 24 on the MMSE, and a history of neuro-
Intraclass reliability coefficients for tht latter logical or psychiatric conditions, per self-
system ranged 0.79-0.99. Mean scores !for the report questionnaire.
The ROCF was administered according with Recognition and Matching trials. Testing
to Taylor's (1959) instructions. Copy, Imme- and scoring were performed by trained per-
diate Recall, and 20-minute Delayed Recall sonnel. Scores were generated using Os-
conditions were administered. The standard terrieth's (1944) criteria. The data for
scoring procedure was used. conversion of the raw scores into scaled scores
are presented in overlapping age groups using
Study strengths the midpoint interval technique introduced by
1. The sample composition is well de- lvnik et al. (1992a). These tables should be
scribed in terms of age, gender, incentive used in the context of the detailed procedures
for participation, and geographic area. for their application, which are explained by
2. Minimally adequate exclusion criteria. the authors. Therefore, they are not re-
3. Test administration procedures are produced in this book. Interested readers are
specified. referred to the original article.
4. Means and SDs for the test scores are The authors concluded that age and edu-
reported. cation effects were evident on all trials but
5. Information on scoring system is education explained minimal variance on the
provided. copy and memory trials. Gender had a mini-
mal effect on performance.
1. Overall sample is large, but individual
[ROCF.21] Schreiber, Javorsky, Robinson,
cells are small.
and Stern, 1999 (Table A12.22)
2. Recruitment procedures are not re-
ported. The BQSS and the 36-point scoring system
3. Specific information on education is not were compared on samples of adults with
provided, other than "the participants ADHD and matched controls. The control
had a minimum of six years of formal group included 18 participants (9 male, 9 fe-
education." male) aged 18-51, with a mean age of 29.5
4. The data were obtained on Mexican (11.5) years and mean education of 15.1 (1.7)
participants, which may limit their use- years. Exclusion criteria were history of neu-
fulness for clinical interpretation in the rological disorder, major medical illness, psy-
United States. chiatric illness, developmental disorder,
5. No information on IQ is reported. learning disability, ADHD, or significant vi-
6. No information on interrater reliability is sual or auditory impairments.
provided. The ROCF was administered according to
the procedures described in the BQSS manual
[ROCF.20] Fastenau, Denburg, and Hufford, 1999 (R. A. Stem et al., 1999), switching different-
This normative study included 211 healthy colored pens. The Copy, Immediate, and 20-
adults aged 30-85 years, with a mean of 62.9 30 minute Delayed Recall conditions were
(14.2) years. Education ranged 12-25 years, used. The test was administered and scored by
with a mean of 14.9 (2.6) years; 55% were trained personnel using the BQSS and the 36-
women, and over 95% were Caucasian. Parti- point scoring system. The interrater reliability
cipants were recruited using a stratified sam- of these scorers was reported in the BQSS
pling procedure at three different sites as part manual. Table A12.22 provides a score for the
of other studies and financially compensated. copy condition obtained using the 36-point
Exclusion criteria were history of cerebrovas- scoring system. The authors discussed the
cu1ar insult, head injury with loss of con- superiority of the BQSS in discriminating
sciousness exceeding 5 minutes, and chronic between the two groups.
substance abuse, per structured interview.
The Extended Complex Figure Test was Study strengths
administered, which supplements the original 1. The sample composition is well described
Copy, Immediate Recall, and Delayed Recall in terms of age, education, and gender.
2. Rigorous exclusion criteria. 0.69-0.92 for different organizational elements

3. Test administration and scorinJ proce- of the figure. Table Al2.23 provides scores
dures are specified. for the Copy and Immediate Recall conditions
4. Means and SDs for the test sa>res are based on the Meyers and Meyers scoring system.
reported. The authors concluded that organization
during the Copy condition was a strong pre-
Considerations regarding use of the *'dy dictor of subsequent recall.
1. The sample is small, with a vnde age
range. Study strengths
2. Data for the recall conditions are not 1. Relatively large sample.
reported. 2. The sample composition is well described
3. Educational level for the sample is high. in terms of age, education, gender, esti-
4. No information on IQ is reported. mated intelligence level, geographic area,
and recruitment procedures.
[ROCF.22] Deckersbach, Savage, Henil\ 3. Rigorous exclusion criteria.
Mataix-Cols, Otto, Wilhelm, Rauch, Ba.r, 4. Test administration procedures are
and Jenike, 2000 (Table A12.23) specified.
The psychometric properties of ~ scoring 5. Means and SDs for the test scores are
systems measuring organizational apptoach to reported.
the ROCF and influences of copy org;uzation 6. Information on scoring systems and in-
and accuracy on immediate recall were stud- terrater reliability is provided.
ied on individuals diagnosed with oCD and
normal controls. Control participants were Considerations regarding use of the study
recruited through bulletin board noti~s at the 1. The data are provided for a wide age
Massachusetts General Hospital. Th~ control range.
group consisted of 55 healthy adults (38% 2. Educational level for the sample is high.
male) 19-64 years of age, with a meaJl age of
35.13 (12.6) years, and education ranpg 12-- [ROCF.23] Miller, 2003; Personal
20 years, with a mean of 16.7 (2.3) yeap. Beck Communication (Table A12.24)
Depression Inventory scores ranged 0--15, The investigation used participants from the
with a mean of 2.3 (3.2). All particip~ts were Multi-Center AIDS Cohort Study (MACS).
Caucasian and right-handed. Estimtted in- The data were collected from 729 seronega-
telligence level was above averagttI Their tive homosexual and bisexual males for the
health status was determined bas~ on a purpose of establishing normative data for
structured clinical interview. Exclusi«in crite- neuropsychological test performance based on
ria were history of Axis I psychiatric disorder, a large sample. Mean age for the sample was
significant head injury, seizure, neu*>logical 40.4 (7.4) years, and mean education was 16.2
condition, or current medical conditiQil. (2.4) years; 91.2% were Caucasian, 2.5%
Copy and Immediate Recall condiions of Hispanic, 5.6% black, 0. 7% other. All partici-
the ROCF were administered. The ~ pants were native English speakers.
tration procedure used switching ·colored The Copy, Immediate Recall, and 20-
pencils every 15 seconds. The prot~ls were minute Delayed Recall conditions were ad-
scored according to Meyers and Meyers' ministered according to standard instructions.
(1995b) system. In addition, the organt?.ational The data are partitioned by three age groups
approach used during the Copy condition was (25-34, 35-44, 45-59) x three educational
assessed according to the Shorr et ~ (1992) levels (< 16, 16, >16 years).
and Savage et al. (1999) scoring meth¥s. The
interrater reliability for the Savagq et al. Study strengths
method, established on a subsample o( 15 ran- 1. The overall sample size is large, and
domly selected drawings, was mod~rate to most individual cells have more than
high, with Cohen's " coefficients :ranging 50 participants.
2. Normative data are stratified by age x After data editing for consistency and
education. for outlying scores, the following data were
3. Information on age, education, ethnicity, included in the analyses: nine studies, which
and native language is reported. generated 19 data points based on a total of
4. Means and SDs for the test scores are 1,340 participants for the Copy condition;
reported. seven studies, which generated 12 data points
based on a total of 1,086 participants for Im-
Considerations regarding use of the study mediate Recall; seven studies, which gener-
1. All-male sample. ated 11 data points based on a total of 1,056
2. No information on IQ is reported. participants for the Long-Delayed Recall.
3. No information on exclusion criteria. Quadratic regressions of the test scores on
age yielded R2 of 0.899 for the Copy condi-
tion, 0.822 for Immediate Recall, and 0.862
for Long-Delayed Recall, indicating that
82%-90% of the variance in test scores for the
THE ROCF DATA
three conditions is accounted for by the
(See Appendix 12m)
models. Based on these models, we estimated
Data collected from the studies reviewed in scores for the three conditions for age inter-
this chapter were combined in regression vals between 22 and 79 years. If predicted
analyses in order to describe the relationship scores are needed for age ranges outside the
between age and test performance and to reported age boundaries, with proper caution
predict test scores for different age groups. (see Chapter 3), they can be calculated using
Effects of other demographic variables were the regression equations included in the
explored in follow-up analyses. The general tables, which underlie calculations of the pre-
procedures for data selection and analysis dicted scores.
are described in Chapter 3. Detailed results of It should be noted in the context of across-
the meta-analyses and predicted test scores condition comparisons that mean age for the
across adult age groups are provided in Copy condition is considerably higher than
Appendix 12m. mean ages for Immediate Recall and Long-
Only those data based on the standard Delayed Recall because data for two large
36-point scoring system, including Meyers studies based on the older samples were re-
and Meyers' (1995b) approach, were used in ported only for the Copy condition.
the analyses. Data generated using other The scores for the Copy condition of the
methods were not included. Data provided in ROCF for healthy young to middle-aged
Meyers and Meyers' (1995b) manual are not samples are not expected to be normally dis-
included. tributed. It should be noted that the majiority
Separate analyses were performed on the of the studies contributing to the aggregate
Copy, Immediate Recall, and Long-Delayed sample for the Copy condition in our analyses,
Recall conditions. Data for 3-minute delayed reported data for older age groups, with the
recall were not analyzed as only few studies mean age of 62.73 (19.27). The mean Copy
reported data for this condition. The long- score for the aggregate sample is 32.20 (1.79),
delay interval varies widely in the data re- reftecting age-related decline from the opti-
viewed. According to the literature, varying mal performance expected in younger sam-
the delay interval between 15 and 60 minutes ples. Thus, the distribution of scores in our
has minimal effect on the rate of recall (Berry & sample is more normal than expected in
Carpenter, 1992). Therefore, 20-, 30-, 40-, younger samples due to variability in both
and 45-minute delayed recall trials were directions from the mean, avoiding scores
combined in one run of analysis. In all cases, being skewed due to ceiling effect.
the long-delayed recall was preceded by an The pattern of SDs differs across the three
immediate or a 3-minute delayed recall but conditions. For the Copy condition, linear
not both. regression of SDs on age yielded R2 of 0.685;
and for Immediate Recall, quadratic regres- of residuals, which does not affect the
sion of SDs on age yielded R2 of 0.694, indi- estimates of regression coefficients and
cating increase in variability with advancing accuracy of prediction but does infiu-
age, consistent with the literature. Predicted ence the results of significance tests.
SDs, based on these models, are reported. 2. Data for only a narrow range of higher
Regressions of SDs for Long-Delayed Re- levels of education are available for the
call on age suggest that age does not account analyses (12.2-16.2 years). Mean edu-
for a significant amount of variability in SDs cation of 14.33 (0.98) for the Copy
(R2 = 0.482). Though some increase ;in vari- condition is high. We were unable to
ability with advancing age is expeded, this fully explore the effect of education on
trend was not significantly evident in !the col- the test scores because lower educa-
lected data. Therefore, we suggest that the tional levels are not represented in the
mean SD for the aggregate sample be used data. Though reports on the relationship
across all age groups. between education and test scores are
Predicted scores and SDs for 12 age ranges equivocal, a number of studies suggest
across three conditions are summarized in that higher levels of education are asso-
Table A12m.4. ciated with better test performance.
Examination of the effects of demQgraphic Therefore, the predicted values might
variables on the ROCF scores indicated that overestimate expected scores for in-
education did not contribute to the te~ scores dividuals with lower educational levels.
in the data available for analyses. Th. effects 3. Although the effect of intellectual level on
of intelligence level, gender, and hanf}edness ROCF performance has been reported in
on ROCF performance were not expl~d due several studies, we could not include
to a scarcity of data available for revi~. measures of intellectual level in our
analyses due to great variability in the
Strengths of the analyses . type of measures used to assess functional
1. Total sample size of 1,340 for the Copy level among the different studies.
condition, 1,086 for Immediate· Recall,
and 1,056 for Long-Delayed Re~all.
2. R2 of 0.899 for the Copy condition, 0.822
CONCLUSIONS
for Immediate Recall, and 0.~62 for
Long-Delayed Recall, indicating a good A great number of studies exploring the psy-
model fit. chometric properties of the ROCF and its
3. Postestimation tests for parameter speci- clinical utility attest to its popularity among
fications did not indicate problelDS with clinicians and investigators alike. However,
normality or homoscedasticity, With the tremendous variability in administration and
exception of the marginally signifieant test scoring of the ROCF obscures comparability
for normalit:f for Long-Delayed ~- of the results of these studies. To improve
4. It should be noted that the pledicted consistency across different studies, the pro-
values match closely the normative data cedures for administration and scoring need
provided in the Meyers and Meyers' to be highlighted in detail by clinicians and
(1995b) manual for all three conditions, investigators.
with respect to both the extrem' values It should be noted that the distribution of
and the direction/rate of ag~related scores for the ROCF Copy condition deviates
changes. considerably from the normal distribution. A
majority of participants are capable of copying
Limitations of the analyses 1 the figure without major distortions. There-
1. Postestimation test for normality: for the fore, a label of "superior" performance given
Long-Delayed Recall was m~ginally to a subject achieving a high ROCF score is
significant. The Kdensity plot pemon- meaningless. On the other hand, the test is
strated a positive skew in the disttibution highly sensitive to deficits in visuospatial
information processing, and achieving a low aggregate sample. The scope of the research
performance score falling in the outlying literature should be expanded to include lower
range has clinical significance. levels of education and intellectual functioning.
In addition to the numerical expression of a A large number of studies on the learnin!¥
subject's performance, the value of qualitative processing strategies in children and on the
interpretation and the delineation of subject's clinical sensitivity of the test to different neu-
strategy/type of errors was emphasized in rological conditions in adults are available in
several studies reviewed above. In this con- the literature, but only a few studies are dedi-
text, the two avenues of research on the cated to the cognitive/processing strategies is-
ROCF, namely, studies on clinical utility and sues related to older age groups.
on the cognitive processes involved in figure The psychometric properties of different
drawing, are mutually enriching. scoring systems need to be further assessed.
Recommendations for future research on Data on interrater reliability, internal consis-
the ROCF include careful analysis of the ef- tency, and test-retest reliability are scarce.
fects of demographic factors on performance. From the review of existing studies, it appears
The well-documented effects of age and in- that different scoring systems are differentially
telligence (and possibly education) need to be applicable to specific clinical and research
considered in subject selection and data pre- situations.
sentation format. Although education did not Additional information on the current use
have an effect on ROCF performance in the of the ROCF and suggestions for future in-
meta-analyses described in this chapter, this is vestigations, submitted by clinicians, are sum-
due to a narrow range of education in the marized by Knight et al. (2003).
13
Hooper Visual Organization Test
BRIEF HISTORY OF THE TEST In addition to using T-score tables, deter-

mination of impaired vs. normal performance
The Hooper Visual Organization Test (HVOT)
can be made using the cutoff criteria. The
consists of 30 line drawings of familiar objects cutoff scores recommended by the authors
which have been fragmented into pieces. The
vary depending on test administration setting.
task requires the examinee to mentally reinte- In a clinical diagnostic setting, a cutoff score
grate and name the objects, which are arranged
of ~24 is suggested in determining whether
in order of increasing difficulty. The response further assessment is needed. On the other
format can be oral or written, depending on
hand, if the test is used as part of a screening
whether the individual administration or the battery administered to all patients admitted
booklet format is used. The score is the num-
to a facility with a low incidence of organic
ber of correctly identified items, with half-
brain pathology, a cutoff of 20 is recom-
points available for some of the items. Wetzel mended to minimize the rate of false-positive
and Murphy (1991) suggest a discontinuation
errors. Boyd (1981) argued:
rule of five consecutive errors, based on a rat-
ing change of only 1% using this strategy. no single cutoff score can be recommended for use
The test was first published in 1958 and in all clinical situations. Factors such as the sub-
revised in 1983. The test manual for the reject's age, educational level, intelligence, and
vised edition provides conversion tables to whether the situation requires minimization of false
correct raw scores for age and educational positives or false negatives, must all be weighed in
level. Corrected or uncorrected raw scores interpreting test results. (p. 19)
can be converted to T scores according to the
tables provided in the manual, with higher T While the cutoff score suggested by Hooper
scores representing a greater likelihood of was judged by Boyd (1981) to be optimal
neurological dysfunction. The standardization for evaluating chronically ill institutionalized
data reported in the manual are based on patients, it appeared to be too low for less
Mason and Ganzler's (1964) all-male sample of incapacitated patient populations. Further-
231 patients, personnel, and volunteer work- more, Nabors et al. (1997) suggested a cutoff
ers from a Veterans Administration hospital. score of ~ 15 for determination of cognitive
The sample was stratified into nine age co- impairment in medically ill elderly as this
horts:25-29,30-34,35-39,40-44,45-49,50- score provided the best correct classification
54, 55-59, 60-64, and 65-69 years. in their sample of urban medical inpatients at
272
HOOPER VISUAL ORGANIZATION TEST 273
a post-acute geriatric rehabilitation unit (81% 1982a,b; Rathbun & Smith, 1982; Woodward,
sensitivity, 79% specificity). 1982).
Hooper also developed a qualitative system "nte above issue is directly related to as-
of response analysis involving four categories: sumptions as to which cognitive functions are
isolate, perseverative, bizarre, and neologistic measured by the HVOT. Two components of
responses. Lezak et al. (2004) underscores the information processing involved in HVOT re-
benefits of qualitative analysis of errors, sponses are mental reintegration and naming
pointing to the localizing significance of frag- of the objects for each test item. If visual
mentation tendencies. Nadler et al. (1996) perception and synthesis are the primary mech-
concur that qualitative analysis of errors im- anisms involved in item analysis, then non-
proves the differentiation between the effects dominant hemisphere contribution prevails. If
of right vs. left hemisphere dysfunction on test performance also imposes considerable
HVOT performance. naming demands, then both dominant and
Merten and Beal (2000) found item rank- nondominant hemispheres contribute sub-
ing for the HVOT to deviate from empiri- stantially to test performance. Studies explor-
cally based item difficulty in their sample of ing the relative contribution of these cognitive
German-speaking neurological patients and processes to HVOT performance are largely
rules for a number of items to be arbitrary. equivocal.
The authors proposed a revised version based Lezak (1995), Lezak et al. (2004), and Spreen
on empirical item analysis, which retains the and Strauss (1991, 1998) suggest caution in
original items but has a modified set of in- interpreting HVOT failures as a manifestation
structions, order of items, and scoring and ad- of visuospatial deficit due to the contribution of
ministration rules. Merten (2002) developed a the naming component. Schultheis et al. (2000)
short form consisting of 15 items, which was developed the Multiple-Choice Hooper Visual
validated on another sample of German- Organization Test (MC-HVOT), which consists
speaking neurological patients. of the 30 original stimuli presented with four
response choices, in order to remove the nam-
ing demands on test performance. The authors
Construct Validity
found that performance of anomie patients was
The HVOT was developed as a screening in- significantly facilitated by the multiple-choice
strument for organic brain dysfunction. format. Furthermore, patients with both right
However, the issue of the test's sensitivity to and left hemisphere involvement benefited
general vs. lateralized dysfunction remains from diminished naming demands.
controversial. The test authors suggest that In contrast, Ricker and Axelrod (1995) found
the HVOT "is sensitive to general impair- that perceptual organization accounted for
ments, not specific visuopractic functions" 44% of HVOT performance variance, whereas
(Hooper, 1983, p. 6). This view is supported confrontation naming ability was not signifi-
by Boyd (1981, 1982a), Wang (1977), and cantly related to test performance. Similarly, in
Wetzel and Murphy (1991). However, the a study designed to replicate and extend the
HVOTs sensitivity to lateralized dysfunction above research, Paolo et al. (1996c) observed
has been demonstrated in several studies. the HVOT to be a measure of perceptual or-
Lewis et al. (1997) report that HVOT perfor- ganization, whereas performance on the test
mance is vulnerable to acute lesions in the right was not significantly impacted by poor naming
anterior quadrant of the brain. In contrast, Fitz ability. Paul et al.'s (2001) results are consistent
et al. (1992), Rathbun and Smith (1982), and with these findings. Greve et al. (2000) found a
Woodward (1982) demonstrate HVOT sensi- small but significant effect of naming on HVOT
tivity to localized dysfunction of the nondomi- performance, which, however, was interpreted
nant parietal lobe. In fact, a heated debate over by the authors as having little or no practical
general vs. specific sensitivity of the HVOT is impact.
reflected in a series of articles published in Such discrepant findings are likely to be
response to Boyd's (1981) article (Boyd, related to composition of study samples, with
samples comprised of aphasic patients dem- in elementlll)'-school children. An effect of IQ

onstrating the largest effect of naming diffi- is also reported by Gerson (1974).
culty on HVOT performance. Education and gender were unrelated to
Merten and Beal (2000) indicated that the HVOT scores in a study by Wentworth-Robr
HVOT measures visuoperceptual and visuos- et al. (1974). In contrast, Verma et al. (1993)
patial-{)rganizational dysfunction, Seidel (1994) found significant effect of education on HVOT
found it to be a measure of general visual- scores. Based on the analysis of HVOT per-
perceptual-constructional abilities in a pediat- formance of 434 normal children aged ~13,
ric population, and Johnstone and Wilhelm Kirk (1992b) reported that boys attained adult
(1997) concluded that HVOT measures global performance by age 12, whereas girls partici-
visuospatial intelligence and shares 12%-23% pating in this study did not reach the adult
of variance with WAIS-R PIQ subtests. level. Based on these data. Kirk documented
an effect of age and gender on HVOT per-
formance.
An interaction between age and education
Lopez et al. (2003) examined the psychomet- in a sample of cognitively intact elderly was
ric properties of the test on a sample of reported by Richardson and Marottoli (1996).
281 cognitively impaired and intact patients Nabors et al. (1997) found HVOT scores to be
and reported acceptable estimates of internal significantly related to age and education in a
consistency (oc = 0.882) and interrater reli- total sample, which combined cognitively
ability (0.977-0.992). Similarly, an internal intact and impaired elderly urban medical
consistency estimate of >0.88 was reported patients, whereas performance was not signif-
by Merten and Beal (2000) on a sample of icantly related to these demographic variables
320 German-speaking neurological patients. for the cognitively intact group considered
Additional data on the reliability and valid- separately.
ity of the HVOT are provided by Gerson For further information regarding the HVOT,
(1974), Franzen (2000), Franzen et al. (1989), see Lezak et aL (2004) and Spreen and Strauss
Lezak et al. (2004), and Spreen and Strauss (1998).
(1998). Item analysis for use of the HVOT
with Indian participants was performed by
Verma et al. (1993). METHOD FOR EVALUATING THE
NORMATIVE REPORTS
To adequately evaluate the HVOT normative
RELATIONSHIP BETWEEN HVOT reports, seven key criterion variables were
PERFORMANCE AND DEMOGRAPHIC deemed critical. The first six of these relate to
FACTORS subject variables, and the remaining refers to
Age and intelligence level are consistently a procedural issue.
related to HVOT performance. Tamkin and Minimal criteria for meeting the criterion
Jacobsen (1984) report an effect of age and IQ variables were as follows.
on HVOT performance in their sample of
211 male, veteran, psychiatric inpatients. Sim- Subject Variables
ilarly, Wentworth-Rohr et al. (1974) found a
positive relationship between HVOT scores Sample Size
and intelligence level as well as a negative age/ Fifty cases are considered a desirable sample
HVOT relationship beginning in the late 30s. size. Although this criterion is somewhat ar-
Age-related changes in HVOT performance bitrlll)', a large number of studies suggest that
are also documented by Farver and Farver data based on small sample sizes are highly
(1982) and by Tamkin and Hyer (1984). Hil- influenced by individual differences and do
gert and Treloar (1985) documented an effect not provide a reliable estimate of the popu-
of age and IQ level but no gender differences lation mean.
Sample Composition Description several used only selected HVOT items. Only
Information regarding medical and psychiatric studies that report data for the full HVOT for
exclusion criteria is important. It is unclear if nonnal samples are reviewed in this chapter.
geographic recruitment region, socioeconomic In all articles reviewed below, the score re-
status, occupation, ethnicity, or recruitment presents the total number of correct responses
procedures are relevant. Until this is deter- (out of 30).
mined, it is best that this information be In this chapter, nonnative publications and
provided. control data from clinical studies are reviewed
in ascending chronological order. The text of
Age Group Intervals study descriptions contains references to the
This criterion refers to grouping of the data corresponding tables identified by number in
into limited age intervals. This requirement is Appendix 13. Table A13.1, the locator table,
relevant for this test since a strong effect of summarizes information provided in the
age on HVOT performance has been dem- studies described in this chapter. 1
onstrated in the literature.
Reporting of Educational Levels

Given the possible association between edu-
cation and HVOT performance, information [HVOT.1] Rao, Leo, Bernardin, and Unverzagt,
regarding education should be provided for 1991a (Table A13.2)
each subgroup. The authors described the performance of a
control group in their study on cognitive dys-
Reporting of Intellectual Levels function in multiple sclerosis. The control group
Given the relationship between HVOT per- included 100 participants (75 females, 25 males),
formance and IQ, information regarding in- who were paid for their participation. The
tellectual level should be provided for each mean age of the sample was 46.0 (11.6), mean
subgroup, and preferably nonnative data education was 13.3 (2.0), and estimated pre-
should be presented by IQ levels. morbid intelligence (based on demographic
variables) was 106.5 (6.9). All except for one
Reporting of Gender Composition participant were Caucasian. Participants were
Given the possible association between gender recruited from newspaper advertisements. Ex-
and HVOT performance, information regard- clusion criteria were history of substance abuse,
ing gender composition should be reported psychiatric disturbance, head injury or any
for each subgroup. other nervous system disorder, or use of pre-
scription medications. In addition to detailed
Procedural Variables medical and psychosocial history participants
underwent a neurological examination, MRI,
Data Reporting and neuropsychological testing.
Means and standard deviations for the to- The HVOT was administered as part of a
tal number of correct responses should be larger battery. For a description of the ad-
reported. ministration procedure, the authors referred
readers to an earlier article.
SUMMARY OF THE STATUS Study strengths

OF THE NORMS 1. Large sample size.
There are only few studies available in the scribed in tenns of age, education,
literature that provide performance levels for
the HVOT. Several studies have reported data 'Nonnative data for children 5-11 years old are provided
by Seidel (1994) and for those 5-13 years old by Kirk
for psychiatric or neurological samples. Among (1992b). See also Baron (2004) and Spreen and Strauss
the studies providing data for nonnal samples, (1998).
gender, ethnicity, IQ estimate, geo- Consideration regarding use of the study

graphic area, clinical setting, and recruit- 1. Small sample size.
ment procedures.
3. Rigorous exclusion criteria. [HVOT.3] Richardson and Marottoli, 1996
4. Means and SDs for the test scores are (Table A13.4)
reported.
The authors report data for 101 autonomously
living, mostly Caucasian, elderly participants
Consideration regarding use of the study
who comprise a subsample of a cohort of par-
ticipants in Project Safety, a study on driving
group.
performance conducted in New Haven, Con-
necticut. Individuals with a history of neuro-
[HVOT.2] Libon, Glosser, Malamut,
logical disease, excessive use of alcohol, or risk
Kaplan, Goldberg, Swenson, and Sands, 1994 for dementia (based on MMSE score) were
(Table A13.3)
excluded. The sample consisted of 53 males
The HVOT was administered to a sample of and 48 females, with a mean age of 81.47
37 right-handed participants aged 64-94 years (3.30), mean education of 11.02 (3.68) years,
as part of a study examining the relationship and mean MMSE score of 26.97 (2.55). Eth-
between age and cognitive functions in normal nic composition was 90.1% white and 9.9%
aging. Participants were recruited from a local black.
community center and from the Active Life The HVOT was administered and scored
Program, an exercise and fitness program at according to the standard instructions pro-
the Philadelphia Geriatric Center. All parti- vided in the test manual.
cipants scored ?.27 on the Mini-Mental State The data were divided into two age groups
Exam (MMSE) and ~10 on the Geriatric De- of younger-old (76-80) and older-old (81-91)
pression Scale (GDS). All participants passed by two education groups. The results indicated
a physical examination and a graded exercise that the mean performance for participants
cardiac function test. Exclusion criteria were with < 12 years of education was stable across
history of stroke, head injury, seizure disorder, younger-old and older-old age groups and
or major psychiatric problems including sub- considerably lower than for their more edu-
stance abuse or psychoactive medications, per cated counterparts; however, performance for
clinical interviews. the younger-old age group with >12 years of
The sample was divided into the young-old education was superior to that of the older-old
(64-74 years) and old-old (75-94 years) groups. group with comparable education.
There were no between-group differences in
education or MMSE or GDS score.
The HVOT was administered as part of a Study strengths
larger battery. The number of correct re- 1. Data for a relatively large sample of el-
sponses was recorded. derly participants are presented.
terms of gender, education, geographic
Study strengths area, and ethnicity.
1. The sample composition is well de- 3. Adequate exclusion criteria.
scribed in terms of age, education, gen- 4. The data are classified into age-by-
der, handedness, MMSE and GDS education groupings.
scores, geographic area, setting, and re- 5. Means and SDs are reported.
cruitment procedures.
2. Rigorous exclusion criteria. Considerations regarding use of the study
3. Means and SDs for the test scores are 1. No information on intelligence level is
reported. provided.
4. The sample is divided into two age 2. Sample sizes for each age-by-education
groups. cell are relatively small.
[HVOT.4] Walsh, Lichtenberg, and Rowe, 1997 rehabilitation program in a midwestern urban
(Table A13.5) university hospital. Seventy-four patients were
The authors compared HVOT performance identified as cognitively intact. This sample
for three groups of geriatric rehabilitation in- had a mean age of 76.9 (5.9) and mean edu-
patients: cognitively intact, mildly impaired, cation of 10.8 (3.0); 74% were women, 51%
and severely impaired. Patients were referred were African American, and 49% were Euro-
for routine cognitive evaluations from two sites: pean American. All participants were func-
a geriatric rehabilitation service of an urban tionally independent across all cognitive
university rehabilitation hospital and the phys- domains and activities of daily living; had no
ical medicine and rehabilitation unit at a sub- history of neurological disease, psychiatric ill-
urban rehabilitation hospital. The cognitively ness, or substance abuse; and had normal re-
intact group consisted of 32 participants (10 sults of neurological examination.
male, 22 female) who scored 2:123 on the The HVOT was administered as part of a
Dementia Rating Scale or in the unimpaired larger battery.
range on all subtests of the Neurobehavioral
Cognitive Status Examination. Participants had Study strengths
no evidence of closed head injury, stroke, or 1. Adequate sample size.
other neurological conditions which could af- 2. The sample composition is well described
fect cognition, as determined by medical chart in terms of age, education, gender, eth-
review, patient interview, and/or negative ra- nicity, clinical setting, and recruitment
diological findings. procedures.
The HVOT was administered according to 3. Adequate exclusion criteria.
standard instructions. 4. Test administration procedures are spec-
ified.
1. The sample composition is well de- reported.
der, and clinical setting.
group.
specified.
reported. CONCLUSIONS
Considerations regarding use of the study The HVOT has been used clinically as a
1. The sample is relatively small. measure of visual perception and organiza-
2. No information on IQ is reported. tion. However, the effect of naming impair-
ment on HVOT performance remains unclear.
[HVOT.Sl Lichtenberg, Ross, Youngblade, The clinical utility of this test would be en-
and Vangel, 1998 (Table A13.6) hanced with the availability of normative data
The authors compared two groups of geriatric for a large sample of neurologically intact
urban medical inpatients: cognitively intact participants of both genders across a wide age
and impaired. All patients were recruited from span, partitioned by age group and intelli-
consecutive admissions to a geriatric medical gence level. 2
2 Meta-analyses were not perfonned on the HVOT due to
lack of sufficient data.

14
Visual Form Discrimination Test
BRIEF HISTORY OF THE TEST in that each contains two large geometric
shapes and a small peripheral figure. How-
Many of the most commonly administered ever, only one of the smaller designs shown on
tests in neuropsychological practice require the adjoining page below is an exact match for
intact visual perception, and accurate inter- the larger stimulus design above. The other
pretation of visually mediated tests often rests three designs are considered "distracters" and
upon the assumption that visual perceptual are variants of the larger stimulus design. One
skills are intact (Lezak et al., 2004). For ex- of the three distracter designs is created by
ample, in the absence of careful assessment of moving or rotating the peripheral figure, the
visual perceptual abilities, low performance second by distorting one of the major figures,
scores on visual memory tests may be mistak- and the third by rotating one of the major
enly attributed to memory impairment when figures. The subject is requested to point to or
in fact the deficits may be primarily related to "say the number" of the design below that
visual perceptual ability rather than memory. exactly matches the larger stimulus design.
The Visual Form Discrimination Test (VFDT) The VFDT consists of two practice items and
was developed by Arthur L. Benton and col- 16 test items. There is no time limit, and the
leagues (Benton et al., 1983b) as a screening scoring system awards 2 points for each correct
test for visual perceptual deficits. (Please see answer and 1 point for an error that involves
Appendix 1 for ordering information.) only the peripheral figure. Errors involving the
The VFDT is a multiple-choice, matching- major figures receive no points. Scores range
to-sample task. The test is presented using a 0-32. Unimpaired individuals usually can com-
spiral-bound booklet (Benton et al., 1983b). plete the test in less than 5 minutes, and the
The subject views an 81h x 11" inch page in test rarely takes longer than 10 minutes to
the booklet displaying a sample design con- complete regardless of the level of impairment.
taining three geometric elements. Directly be- Because the VFDT is a nonmotoric task, it is
low the stimulus page, the adjoining 81h" x 11" especially useful when assessing senior adults,
inch page presents four smaller three-element patients with severe arthritis or hemiparesis,
designs (numbered 1, 2, 3, or 4). The subject, and/or the medically ill.
therefore, can concurrently view the main The validity of the VFDT to assess visual
stimuli and the four smaller design groupings perceptual impairments with various neuro-
below. The designs on both pages are similar logical conditions has been well established.
278
VISUAL FORM DISCRIMINATION TEST 279
For instance, the VFDT has been used to ex- person would obtain a minimum of four cor-
amine visual perceptual impairments in post- rect answers simply by guessing. A5 such, the
head injury patients (Iverson et al., 1997b, VFDT can additionally be used as a measure
2000; Malina et al., 2001; Millis et al., 2001; of symptom validity and motivation when
Wilde et al., 2000), aphasic patients (Varney, performance scores are especially low (i.e.,
1981), and patients with vascular dementia significantly below chance). However, because
(Mast et al., 2000), Alzheimer's disease (Iver- the task tends to be very easy for most clinical
son et al., 1997a; Kaskie & Storandt, 1995), or populations, performances above chance can
Parkinson's disease (Tang & Liu, 1993). Pa- still be noncredible. Larrabee (2003) found
tients with right hemisphere lesions show the that a raw score of <26 on the VFDT correctly
highest rates of test failure (Benton, 1983a), identified 48% of his definite malingering
although aphasic alexics have been observed neurocognitive dysfunction group (i.e., sub-
to show a 36% failure rate (Varney, 1981) and jects who performed significantly worse than
recovery in letter recognition is accompanied chance on the Portland Digit Recognition
by improvement in visual form discrimination. Test) and 93% of moderate to severe closed
Test-retest reliability has been examined by head injury patients. He reported that no
Campo and Morales (2003) and found to be closed head injury patients scored <24 on the
quite stable over brief intervals (e.g., ~1<? days). VFDT, whereas 32% of his definite malin-
However they did find a practice effect for pe- gering neurocognitive dysfunction subjects
ripheral errors and cautioned that further test- scored <24. He concluded that the VFDT
retest research is necessary to investigate the demonstrates good sensitivity and specificity
stability of the VFDT across longer periods of for discriminating malingerers from head-
time. Internal consistency has ranged from 0.66 injured patients.
(Malina et al., 2001) to 0.75 (Iverson et al., 2000). Since the standard administration of the
In a preliminary study, Caplan and Caffery VFDT usually takes less than 5 minutes to
(1996) examined the test as a motor-free mea- complete with normals and rarely takes more
sure of short-term visual recognition memory. than 10 minutes regardless of level of im-
For the short-term recognition memory for- pairment, the necessity for a shortened VFDT
mat, each stimulus page from the VFDT was version may be questioned. However, very
exposed for 10 seconds and after each exposure brief and accurate screening measures are
the subject was shown the response plate and welcomed when individuals can tolerate only
asked to select the design that he or she rec- minimal testing, due to such factors as low
ognized. This process was repeated for each of frustration tolerance, fatigue, attentional prob-
the remaining stimulus and response plates in lems, or severely impaired cognition. Two
the booklet. After the recognition memory part short form versions of the test have been
of the test was completed, subjects were reex- created by splitting the test in half. The "First-
amined on failed items, using the standard half" short form consists of the first eight
VFDT "matching-to-sample" administration items of the standard test, and the "Front-
procedures. Caplan and Caffery (1996) found back" version consists of the first four (1-4)
that both the standard VFDT administration and last 4 (13-16) items chosen because the
and the 10-second exposure short-term mem- correct answer was presented an equal num-
ory administration effectively discriminated ber of times in the four quadrants. Scores from
between brain-diseased patients and controls; the short forms are multiplied by 2. Iverson
however, use of the combined administration et al. (1997b) evaluated the concurrent valid-
allowed parceling of the effect of perceptual ity and clinical utility of the Front-back short
deficits on multiple-choice recognition mem- form in a sample of patients with closed head
ory performance. injury and observed that the mean difference
Because there are 16 VFDT response cards between the short and full forms was < 1 point
and a one-in-four chance of guessing the and the correlation between the two versions
correct answer on each card, one would ex- was 0.86. In a subsequent publication, Iverson
pect, based on probability theory, that a and colleagues (2000) obtained full and short
form data on patients with heterogeneous significant decline in mean scores compared
neurological and psychiatric diagnoses who to younger age groups. In addition, Campo
were referred for neuropsychological evalua- and Morales (2003) noted a significant influ-
tion. Internal consistencies for the two short ence of age and education on VFDT perfor-
forms were 0.62 (First-half) and 0.63 (Front- mance in their large (n = 397) normative
back). Correlations between the full and short sample, although they also did not observe a
forms were 0.85 (First-half) and 0.86 (Front- gender effect. However, the extent to which
back). For both shortened versions, the ma- their findings, on data collected in Spain,
jority of scores were within 2 points of the full generalize to U.S. populations is unclear.
version (70.7% for First-half, 71.6% for Front-
back). The authors also documented that the
mean difference score between both the short
forms and the full form was less than 1 point METHOD FOR EVALUATING THE
and concluded that use of the short form re- NORMATIVE REPORTS
sults in a very minor loss of accuracy. Clinical To adequately evaluate the VFDT normative
decision rules were provided for improving the reports, five key criterion variables were
overall accuracy of the short forms. The au- deemed critical. The first four of these relate
thors concluded that both short forms had to subject variables and the last to procedural
adequate concurrent validity for both clinical variables.
and research applications, but they expressed Minimal requirements for meeting the cri-
preference for the First-half version because terion variables were as follows.
no research had been conducted on the Front-
back version without administration of the mid-
dle eight items. Unfortunately, no normative Subject Variables
data for either of the short forms have been
published, limiting the clinical utility of these Sample Size
forms. Fifty cases have generally been recommended
(Hayes, 1963; Guilford, 1965) as providing a
reliable estimate of the population mean. For
the purpose of review, a minimum of 50 subjects
per age group interval was deemed adequate.
VFDT PERFORMANCE AND
DEMOGRAPHIC FACTORS Sample Composition Description
The influence of various demographic factors Information regarding medical and psychiatric
on VFDT performance is sketchy, largely be- exclusion criteria is important. It is unclear
cause the studies have been few and sample if geographic recruibnent region, socioeconomic
sizes often have been small. Benton et al.'s status, occupation, ethnicity, or recruibnent pro-
(1983b) original normative sample consisted cedures are relevant. Until this is determined, it
of 85 healthy subjects aged 19-74 years. He is best that this information be provided.
found no effect of age, education, or gender
on VFDT performance. Similarly, Larrabee Age Group Intervals
(2003) studied VFDT performance in samples This criterion refers to grouping of the data
of severe closed head injury patients, moder- into limited age intervals. This requirement is
ate closed head injury patients, and mixed relevant for this test since an effect of age on
neurological and psychiatric patients and ob- VFDT performance has been demonstrated in
served no significant relationship between age some studies.
and education and VFDT performance. The
lack of an age effect has also been observed by Reporting of Education Level
Axelrod and Ridder (as reported by Benton Given that there has been some association
et al., 1983b, 1994b); however, Valdois et al. between educational level and VFDT scores,
(1989) found that older subjects did show a information regarding educational level
should be reported. It is unknown whether and range of performance scores for two
IQ is relevant, so until this is determined, it is age groups of men and women on the VFDT
best that information on IQ be provided. (males 19--54 years old, n = 28; females 19--54
years old, n = 30; males 55-74 years old,
Procedural Variables n = 15; females 55-74 years old, n = 12). The
normative sample consisted of 85 "patients
Data Reporting without history or evidence of brain disease or
Means and standard deviations should be healthy subjects" (p. 58). No further infor-
reported. mation regarding the population from which
With these requirements for reporting in this sample was drawn was provided.
mind, the normative studies for the VFDT
were examined. Study strengths
1. Data are stratified by age and gender.
SUMMARY OF THE STATUS OF
THE NORMS Considerations regarding use of the study
1. Sample size per age/gender category is
Only two studies report normative data for the small.
VFDT. Benton et al. (1983a) report data for a 2. Inadequate exclusion criteria; sample
group of individuals aged 1~74 years who are included patients with unspecified medi-
Uving in the United States, and Campo and cal conditions excluding history or evi-
Morales (2003) report data for a group of dence of brain disease.
adults aged 18--59 years who are Uving in the 3. Age group intervals are too broad.
south and southwest of Spain. The Benton et al. 4. No information is reported regarding the
(1983b) study suffers from small sample size educational level of the various gender/
and no performance SDs reported (although age groups. However, for the total sam-
they were subsequently provided by Caplan ple, 72 subjects had >12 years of edu-
and Schultheis, 1998). As such, data from this cation and 13 subjects had <12 years of
report should be considered "provisional," and schoollng.
further normative research with U.S. popula- 5. No SDs are reported.
tions across the age span is encouraged. The
Campo and Morales (2003) normative data are
Other comments
adequate for patients aged 18--59 years who
1. In examining the distribution of scores of
were hom and educated in Spain.
their normative sample, Benton et al.
In this chapter, normative publlcations are
(1983b) found that the majority of sub-
reviewed in ascending chronological order.
jects had near perfect scores (31-32) and
The text of study descriptions contains refer-
that scores of ~26 were obtained by 95%
ences to the corresponding tables identified
of the group. They concluded that a
by number in Appendix 14. Table A14.1, the
score of 24 or 25 should be considered
locator table, summarizes information pro-
"borderline or mildly defective perfor-
vided in the studies described in this chapter.
mance," a score of 23 reflects "moder-
ately defective performance," and scores
SUMMARIES OF THE STUDIES of <23 indicate "severely defective per-
formance."
This section presents critiques of the two 2. More recently, Caplan and Schultheis
normative studies for the VFDT. (1998) provided the SDs for Benton
et al.'s (1983b) original normative sam-
[VFDT.1] Benton, Hamsher, Varney and Spreen, ple. They also generated an interpretive
1983b (Table A14.2) table with T scores, percentile equiv-
This is the report of the original normative alents, and cllnical performance de-
study. The report provides the mean, median, scriptors based on Heaton et al.'s (1991)
interpretive schemes. They noted that to Benton et al. (1983b, 1994b). The three
because of ceiling effects on the VFDT, error types were categorized according to
only a perfect score would produce an Kaskie and Storandt (1995) (i.e., major dis-
"above-average" rating according to the tortion, major rotation, peripheral error), and
Heaton et al. criteria, while twQ "full- normative data for error types were reported.
credit" errors plus a single 1-poi~t error
(5 points) would be sufficient to yield a Study strengths
"mildly impaired" score (i.e., score= 27). 1. Sample size per age/education category
Using the Benton et al. description, a is adequate (n ~50).
perfect score produces an "'above- 2. The sample composition is well described
normal" performance rating, while three in terms of gender and geographic area.
"full-credit" errors plus a single ~-point 3. Information is reported regarding the
error (7 points) produces a "milfly de- educational level of the various age
fective" score (i.e., score= 25). A lli!Ore of groups.
26 would fall within the Benton et al. 4. Means and SDs for correct responses
normal range, whereas using the ~eaton and errors are reported.
et al. scheme, a score of 26 w~ld be
considered "mildly-moderately imjjaired." Considerations regarding use of the study
Further, a "moderately defectivet score 1. Age group intervals are broad.
of 23, as suggested by Benton :et al., 2. Data were collected in Spain, which
would produce "moderate impmtment" raises questions regarding their appro-
on the Heaton et al. scheme. Capfr1 and priateness for clinical use in the United
Schultheis (1998) concluded tijlt the States.
Heaton et al. scale differs from fhat of
Benton et al. by increasing the : cutoff Other comments
value for impairment. Campo and Morales (2003) found total scores
of ~26 were obtained by 94% of their sample.
This result is practically identical to that re-
[VFDT.2] Campo and Morales, 2003 ported by Benton et al. (1983b), who found
(Table A14.3) that 95% of their sample obtained scores >26.
This study presents normative findings by age
and education for a group of 397 healthy,
unpaid volunteers, recruited by wqrd of
CONCLUSIONS
mouth, and living in the south and southwest
of Spain. The sample was composed bf 191 Indeed, the normative database for the VFDT
males and 206 females. The data are rei>orted has remained meager. The first normative
by two age groups (18--39, n =222; ~9. study was reported by Benton et al. (1983b).
n = 175) and two educational levels (6-8 Mean and median performance data for
years, n = 115; ~9 years, n = 282). T~ be a 85 adults, aged 16--75 years were reported;
participant in this normative study, s..bjects however, no information regarding SDs was
had to speak Spanish as their prim~ lan- provided, which significantly limited the clin-
guage, to be independently functionirt, and ical utility of the test for several years. Campo
to have no history of neuropathologic4J con- and Morales (2003) provided the only other
ditions, psychiatric hospitalization, ab.ormal normative report on a group of 397 healthy
psychomotor development, drug or ~cohol adults aged 18-59, living in the south and
abuse, or psychotropic drug use that: could southwest of Spain. Two additional studies
affect attention/concentration or pf>duce (Lichtenberg et al., 1998; Nabors et al., 1996)
drowsiness. Individuals with chronic medical are found in the literature, which are of-
conditions such as diabetes, hypertensio+, mild ten cited as providing normative data for the
hearing loss, however, were not excludetJ. The VFDT. However, close inspection of the pro-
VFDT was administered and scored accprding cedural details for these studies quickly
.
I
reveals that the data were not collected from Because of its design, the VFDT seems
a group of normal healthy individuals but especially suited for use with senior adults.
from a group of elderly medical rehabilitation Most of the normative data have been col-
inpatients. lected on senior adults in the age range of 60-
Although there has been limited research 74 years. Additional studies with senior adults
using short forms of the VFDT, no normative more advanced in age would be important
data have been reported for these forms. since dementia is most common in old age.
15
Judgment of: Line Orientation
BRIEF HISTORY OF THE TEST: with the test. Following the practice trials, all
30 test items are administered; there is no
The Judgment of Line Orientation Tes~ (JLO) discontinuation rule. Each correct response is
is designed to evaluate visual-spatial skills given one credit, and the total number of
tion and angles of lines in space. The ,W

by assessing the ability to judge the <fienta-
thought to be one of the "purest" tests o visual

is
correct answers out of 30 is recorded.
The test was developed and first published
by Benton et al. (1978). This study assessed
perception as it requires minimal mot<t skills the ability of individuals with left or right
and verbal mediation (Lezak, 1995: l Lezak hemisphere lesions to discriminate line angles.
et al., 2004). The test consists of five ptactice Prior to publication of the JLO, these authors
items followed by 30 test items. Two fCfiils of and others used a tachistoscope to examine
the JLO have been developed (Forms II and line orientation discrimination in healthy
V), which consist of the same 30 items pre- and clinical populations (Benton et al., 1975;
sented in slightly different sequence. In both Fontenot & Benton, 1972; Warrington & Ra-
forms, the test items are organized in th.. order bin, 1970). The JLO was created out of the
of the least to the most difficult. For eac. item, desire to have a more convenient, brief clini-
there is a stimulus card that appears on dte top cal test that could be used with patients at
page of the booklet and a response caid that bedside, if necessary. Further details regard-
I
appears on the bottom page of the b9oklet. ing test development, administration, scoring,
The task is to match a pair of full or partial and norms on the initial sample were later
angled lines appearing on the stimulus 4ard to published in a clinical manual, which also
two of the 11 numbered lines appearingpn the contains several other orientation and visual-
reference card. The 11 lines on the reference spatial tests (Benton et al., 1983a).
card form a semicircle and are separated at 18- Studies have consistently shown far greater
degree intervals. A correct response is ~en to right than left hemisphere association with
an item only when both lines on the s~ulus performance on the JW. In the original
card are correctly matched to the lines ~m the sample of patients with unilateral right or left
reference card. All five practice items E"e ad- hemisphere lesions, Benton et al. (1978)
ministered, and at least two spontaneo s cor- found that right hemisphere brain-damaged
rect responses to the sample items m t be patients performed strikingly worse than
provided by the subject in order to p~d patients with left hemisphere damage. While
284
JUDGMENT OF LINE ORIENTATION 285
left hemisphere-damaged patients performed and incorrectly identifying a vertical line as

poorer than controls, only a minority (10%) an oblique or horizontal line. Flinton et al.
were classified as moderately to severely im- (1998), however, recognized that their sample
paired. Similar findings were reported by York of Alzheimer's patients was slightly older, was
and Cermak (1995) and Trahan (1998), who more educated, and obtained better overall
found poorer performance on the JLO for JLO scores than the sample used by Ska et al.
patients with right hemisphere cerebrovascu- (1990). Patients with Lewy body dementia
lar accidents (CVAs) relative to those with left with psychosis have been reported to commit
CVAs and normal controls, with the poorest more visual perceptual errors than Alzhei-
performance occurring in right CVA patients mer's patients and patients with Lewy body
with left-sided visual neglect. Mehta and dementia with predominantly parkinsonian
Newcombe (1996) suggested that in JLO per- features (Simard et al., 2003).
formance the right hemisphere may be re- As a model for studying diseases such as
sponsible for "metric measurement" but that Alzheimer's, the relationship between cholin-
the left hemisphere keeps tracks and updates ergic system dysfunction and visual-spatial
decisions in line orientation judgment. Con- skills has been examined. Studies employing
sistent with the other studies, Hannay et al. the JLO have reported mixed findings. Meador
(1987) reported increased blood flow to et al. (1993) administered scopolamine (an
temporo-occipital regions bilaterally for par- anticholinergic) vs. placebo to healthy middle-
ticipants performing a line orientation task aged individuals and found that the scopo-
similar to the JLO, but the greatest blood flow lamine group performed significantly worse on
increase occurred in the right hemisphere. the JLO relative to the placebo group. How-
However, Gur and colleagues (2000) reported ever, in a similar study, Obonsawin et al. (1998)
that right parietal temporal activation may be using scopolamine vs. saline in normal middle-
present only in males. aged individuals, found no significant differ-
Impaired performance on the JLO has ences in JLO performance for the two groups.
been reported in patients with various forms Relatively small sample sizes were used in both
of dementia. A number of studies have shown studies (n = 12 per group), and even when
that patients with Alzheimer's disease per- differences were statistically significant, mean
form worse than normal controls on this task score differences were rather small.
(Eslinger & Benton, 1983; Eslinger et al., The few studies conducted on patients with
1985; Ska et al., 1990). Similarly, studies of mental illness have shown that JLO perfor-
Parkinson's disease patients have shown mance does not appear to be altered in patients
lower performance relative to normal controls with schizophrenia. Fleming and colleagues
(Flinton et al., 1998; Montgomery et al., 1993; (1997) found no differences between schizo-
Montse et al., 2001; Ska et al., 1990). Error phrenics and controls on the JLO, suggesting
types, such as vertical and horizontal errors that basic visual-spatial skills are intact in this
or inter- and intraquadrant errors, on the patient group. Similarly, Sweeney et al. (1991)
JLO have also been examined in dementia reported no change in performance on the
patients and found by some authors to dis- JLO from an acute baseline episode to a 1-year
criminate between controls and Alzheimer's follow-up period in patients with schizophrenia;
patients (Ska et al., 1990) and Parkinson's pa- mean scores at both points were within the
tients (Montse et al., 2001), although Flinton normal range. In contrast, JLO scores have
et al. (1998) were unable to replicate these been observed to be lower in depression (Coello
findings in Alzheimer's patients. Specifically, et al., 1990), and JLO performance has been
Flinton et al. (1998) found that only two out reported to be predictive of aggression in fo-
of 10 JLO error types used by Ska et al. (1990) rensic patients (Foster et al., 1993).
occurred with more frequency in Alzheimer's Studies assessing the influence of various
patients relative to controls. The error types other factors on JLO performance have yielded
included misjudging one oblique line with an- interesting results. Rahman and Wilson (2003)
other that is separated by only one spacing found that heterosexual men outperformed
homosexual men on the JLO, but no differ- similar mean scores, that the distribution of
ences were reported between hetero- and ho- the scores across both forms is virtually iden-
mosexual women. The authors suggest that tical, and that when the original JLO cutoff
underlying differences in brain strUctures, scores are used, the shortened versions actu-
particularly those involving the parietal cortex, ally have a high accuracy in classification of
may explain the differences in performance for patient groups. Similarly, Mount et al. (2002)
the two male groups. Interestingly, factors such observed that odd-even short forms are
as near-sightedness do not appear to afftct JLO highly correlated with the total test (r=0.00-
performance (Kempen et al., 1994). finally, 0.93) and that all patient scores derived from
studies have shown that the JLO h~ some short forms were within 2 points of full
limited utility for the detection of maliQgering. scores. Winegarden et al. (1998), evaluated
That is, while scores may be lower 4>r ma- five shortened versions of the JLO, including
lingerers, classification rates for malingerers vs. a 10-item test (items 1-10), two 20-item ver-
nonmalingerers are quite low (lversoD, 2001; sions (items 1-20 and 11-30), and odd-even
Meyers et al., 1999). versions and found that using items 11-30 of
Form V had the highest internal consistency
and best overall correlation with the full ver-
sion. Another shortened version of the JLO
In their manual, Benton et al. (1983a)· report was created by Qualls et al. (2000) in which
the split-halfreliabilityofboth Forms If and V the Latin square randomization strategy was
to be high. The reliability for Form Hj based used to place each of the JLO items into one
on 40 participants, was 0.94, and ~at for of two shortened forms. The shortened ver-
Form V, based on a sample of 124 partici- sions in Qualls et al.'s (2000) study showed
pants, was 0.89. Additionally, in a salilple of adequate internal consistency and validity rel-
37 participants who were administered both ative to the long form, but classification rates
versions of the test (with test probes se;arated (normal vs. impaired) were lower when using
by 6 hours to 21 days), test-retest retability the original JLO cutoff scores.
was 0.91 and the increase in mean scortts from
one test administration to the next was negli-
gible (23.1-23.5), suggesting the absence of a
RELATIONSHIP BETWEEN JLO
practice effect. Validity studies have shown
that the JLO discriminates patients with right
FACTORS
hemisphere lesions from those with left; hemi-
sphere lesions (Benton et al., 1983a, 1994b; Effects of age and gender have consistently
see above discussion). Adequate test relability been reported in the JLO literature, while the
has also been reported for the sh~ened impact of education on JLO performance has
(odd-even) forms of the JLO (Woodard et al., been more equivocal. Benton et al. (1978,
1998). For further information on tlie psy- 1983a) reported a moderate decline in JLO
chometric properties of the JLO, see F)anzen scores with age. Similar age effects were re-
(2000). ' ported by Eslinger and Benton (1983) in
older individuals (55+ years) but not by Ska
et al. (1990). Subsequent studies have shown
Alternate Brief Forms of the JLO
age effects in normal healthy individuals, with
More recently, researchers have qreated poorer performance with advanced age (Basso
shortened versions of the JLO (Qualls~ et al., et al., 2000; Rahman and Wilson, 2003; Wood-
2000; Woodward et al., 1996). One ~ethod ward et al., 1996). Only one study, using a
for shortening the JLO is to split the t~t into shortened version (only odd items were pre-
even- and odd-numbered items. This J1)ethod sented), found no relationship between age
has created two valid and reliable forms (Wood- and JLO performance (Salthouse et al., 1997).
ward et al., 1996, 1998). The authots also This may be due to the restricted number of
found that these two shorter forms produce test items presented to participants.
Benton et al.'s (1978, 1983a) original study critical. The first five of these relate to subject
revealed significant gender differences, with variables, and the remaining two refer to
males outperforming females. In a study with procedural issues.
relatively small sample sizes, Desmond et al.
(1994) found gender differences in JLO perfor-
Subject Variables
mance in healthy controls and stroke patients,
with males outperforming females in both Sample Size
groups. Montse et al. (2001) also found that in
both Parkinson's and healthy control groups, fe-
size. Although this criterion is somewhat ar-
males scored lower than their male counterparts
bitrary, a large number of studies suggest that
on the JLO. Finally, Woodward et al. (1996),
data based on small sample sizes are highly
using a short form of the JLO, also obs_erved
influenced by individual differences and do
better performance in male vs. female patients.
Woodward and colleagues (1996) observed
lation mean.
a decline in JLO performance as a function of
lowered education. However, Benton et al. Sample Composition Description
(1983a) found a more complex relationship
between JLO performance and education.
Specifically, males under the age of 65 years
displayed no differences in JLO scores as a
status, occupation, ethnicity, or recruitment pro-
result of education level (12+ vs. < 12 years);
cedures are relevant. Until this is determined, it
however, females consistently showed a de-
is best that this information be provided.
cline in scores as a result of the interaction of
age and education status. . . Age Group Interval
There is virtually no information regarding
This criterion refers to grouping of the data
JLO performance in different ethnic groups.
into limited age intervals. This requirement is
To date, only one study (Rey et al., 1999) has
relevant for this test since a strong effect of
examined the performance of Hispanics on
age on JLO performance has been demon-
the JLO. These authors report median scores
strated in the literature.
and cutoffs for "defective performance" for a
group of South American, Central American, Reporting of Educational levels
and Cuban Spanish-speaking individuals. They
found that the distribution for Hispanic
cation and JLO scores, information regarding
participants (median JLO score= 24, ~utoff
educational level should be reported for each
score = 17) was virtually identical to published
subgroup.
norms on English-speaking individuals (me-
dian JLO score= 25, cutoff score= 15). While Reporting of Gender Composition
the findings of this study suggest that use of
Given the strong association between gender
the Spanish adaptation of this test is valid,
and JLO performance in favor of males: .in-
further research is needed.
formation regarding gender composition
For further information regarding the JLO,
should be reported for each subgroup, and
see Benton et al. (1983a, pp. 48--54), Lezak
preferably normative data should be pre-
(1995, pp. 400-401), and Lezak et al. (2004,
sented by gender.
pp. 390-391).
METHOD FOR EVALUATING THE Description of Test Version
NORMATIVE REPORTS
Full and shortened versions of the JLO exist.
To adequately evaluate the JLO normative re- Description of the test version allows selection
ports, seven criterion variables were deemed of the most appropriate norms.
Data Reporting used middle-aged (30s and 40s) or older (55+

Group means and standard deviations for the years) samples.
number of correct responses (out of 30) Among all of the clinical studies available in
should be presented. the literature, we selected for review those
that had well-defined samples, presented
means and SDs for the JLO (with the excep-
tion of the original study by Benton et al.,
1983a), and provided descriptive statistics for
OF THE NORMS
demographic factors such as
mean age and
Information presented in the studies report- education for the sample. Additionally, given
ing data for the JLO differs considerably that shortened versions of the JLO save on
across studies. Some of these differences will administration time and may be desirable for
be summarized below. the clinician, one such study with adequate
Only two studies were designed to provide normative data is presented.
normative information on the JLO. Benton In this chapter, normative publications and
et al. (1983a) report the original findings of the control data from clinical studies are reviewed
Benton et al. (1978) study, in which the JLO in ascending chronological order. The text of
was developed; and Woodward et al. (1998) study descriptions contains references to the
report normative data for two shortened, par- corresponding tables identified by number in
allel forms of this test. Data for "normal con- Appendix 15. Table A15.1, the locator table,
trol'' groups from clinical comparison $tudies summarizes information provided in the
are also included in this chapter. studies described in this chapter. 1
Studies that did not include normal control
groups or did not report means and SDs for
the JLO are not reviewed in this chapter
(Benton et al., 1981; Iverson, 2001; Montgo-
mery et al., 1993; Ng et al., 2000; Qualls et al., UL0.1] Benton, Hamsher, Varney, and
2000; Raskin et al., 1990; Seidenberg et al., Spreen, 1983a (Table A15.2)
1995; Sweeney et al., 1991; Trahan, 1998; The normative data in this manual are based on
Winegarden et al., 1998; Woodard et al., the original study in which the JLO was con-
1996). Other studies were not included be- structed and developed (Benton et al., 1978).
cause JLO scores for normal controls were The original study used 144 controls, seven of
abnormally low (Desmond et al., 1994), au- whom were left-handed (Benton et al., 1983a),
thors used individuals referred for neuropsy- but the later manual reports data for 137 con-
chological evaluation as their subject pool trol participants. It is unclear if it is the seven
rather than healthy normal controls (Vander- left-handed individuals whose data are not in-
ploeg et al., 1997), or they used modified ver- cluded in this manual. Patients who evidenced
sions of the JLO for which the administration no brain disease were recruited from the gen-
procedures were not clearly described (Gur eral medical services of the UDiversity of Iowa
et al., 2000; Levin et al., 1991). to serve as the control group. Normative data
The majority of the studies used the 30-item are reported for three age groups (16--49, 50-
JLO (Form V or H), but three studies provide 64, 65-75) by gender. The average educational
norms on 15-item short forms. Most studies level of the participants was approximately 12
report the number of correct responses as the years. Mean JLO scores, without SDs, are re-
outcome data; however, one study reports num- ported for each age-by-gender group.
ber of errors and one study reports T scores.
While many of the studies present statistical Study strengths
differences between males and females, very 1. The sample composition is well described
few stratify the data by gender. Few studies in terms of age, education, gender,
stratify mean scores by age group, and no study
reports data by educational level. Most studies 'Nonns for adolescents are available in Baron (2004).
geographic area, and recruitment proce- 2. Large overall sample size.

dures. 3. Data are partitioned by age group, and
2. Data are partitioned by age groups and in particular, old and very old groups are
gender, and correction scores for age included.
and gender are provided. 4. Adequate exclusion criteria.
3. Test administration procedures are 5. Test administration procedures are
specified. specified.
4. A distribution of corrected scores is pro- 6. Means and SDs for the test scores are
vided along with percentiles and classi- reported.
fication labels based on the performance
of normal participants relative to those Considerations regarding use of the study
with brain lesions. 1. The data are not partitioned by gender,
and there are two to three times more
Considerations regarding use of the study females than males in each age group.
1. No SDs for JW scores are reported. 2. Raw scores are not reported. Means and
2. Control participants are medical patients SDs are reported in T scores.
(with no evidence of brain disease) and 3. It is unclear whether this elderly sample
thus may represent a biased sample. is all community-dwelling.
3. No other exclusion criteria are provided.
4. Overall sample is adequate, but individ- UL0.3] Eslinger, Damasio, Benton, and Van
ual cells are relatively small. Allen, 1985 (Table A15.4)
JW scores were collected on 53 normal vo-
Other comments lunteers (25 males, 28 females) aged 60--88,
1. Specific corrections (point additions) to recruited through senior-citizen and commu-
the achieved scores need to be made, nity organizations in the Iowa City area. Mean
depending on age and gender. age was 73.1, and mean education was 12.0.
All were independent, community-dwelling
UL0.2] Eslinger and Benton, 1983 individuals who were screened for neuro-
(Table A15.3) logical disorder (including head injury and
The control sample included 178 volunteers (35 alcoholism), psychiatric illness requiring hos-
males, 143 females) aged 65-94 years. Partici- pitalization, and any disabling medical or physi-
pants were recruited from senior-citizen orga- cal condition. Subjects considered themselves
nizations and retirement communities in and to be in generally good physical and mental
around Iowa City. Participants with a history of health.
neurological disease and psychiatric hospital-
izations were excluded from the study. The Study strengths
average educational level for males and females 1. Minimally adequate sample size.
was 12.6 and 13.1, respectively. The sample was 2. Information on age, gender, education,
stratified by three age groups (65-74, 75-84, geographic area, and recruitment
85-94) but not by gender. Standard JW ad- strategies.
ministration procedures were used. 3. Means and SDs for the test scores are
The results revealed a steady decline in reported.
performance with advancing age, with an ap-
proximately 0.5 SD decline per decade after Consideration regarding use of the study
the age of 65 years. 1. Data are not stratified by gender or age,
although age grouping is fairly narrow.
Study strengths
1. The sample composition is well de- UL0.4] Rao, Mittenberg, Bernardin, Haughton,
scribed in terms of age, education, gen- and Leo, 1989 (Table A15.5)
der, geographic area, and recruitment This study examined the effects of focal peri-
procedures. ventricular white-matter changes on cognitive
290 PERCEPTUf\l ORGANIZATION: VISUOSPATIAL AND TACTILE
functioning in healthy adults. The authors controls. Error types differed for Alzheimer's
selected 40 participants (10 males, 30 f$nales) patients vs. normal controls.
who had normal brain imaging to s~e as
controls. Their ages ranged 25-60 yeaJS, with Srudy stnm~~
a mean of 42.8 (8.1), mean education of 14.0 1. The sample composition is well de-
(2.3), and mean Verbal IQ of 108.9 (1L9). All scribed in terms of age, education, gen-
participants were recruited from n~paper der, and recruitment procedures.
advertisement in the Milwaukee, Wiseonsin, 2. The data are stratified by age, and in
area. Additional exclusion criteria were a his- particular, older age groups are included.
tory of hypertension, cardiac or cerebrovas- 3. Adequate exclusion criteria.
cular disease, neurological illness, head injury, 4. Test administration procedures are
substance abuse, or psychiatric illness. Parti- specified.
cipants underwent physical and neuniogical 5. Means, SDs, and ranges for the test
exams. Standard JLO administration rroce- scores are reported.
dures were used. 1
Considerations regarding use of the srudy
Srudy~nm~~ , 1. Overall sample is adequate, but the
1. The sample composition is well de!jcribed sample size for the oldest group is rela-
in terms of age, education, WAIS-t;\ Ver- tively small.
bal IQ, gender, MMSE scores, geotaphic 2. The data are not stratified by gender.
area, and recruitment procedures.· There are two or, in some age groups,
2. Adequate sample size. · three times as many females as males.
3. Exclusion criteria are provided. 3. Data were collected in Canada, which
4. Test administration procedures are de- may limit their usefulness for clinical
scribed. interpretation in the United States.
reported. UL0.6] Rao, Leo, Bernardin, and Unverzagt,
1991a (Table A15.7)
Considerations regarding use of study The study examined the pattern of cognitive
1. Relatively small sample size for a wide deficits in patients with multiple sclerosis us-
age range. ing a brief neuropsychological battery. The
2. The data are not stratified by age or authors recruited 100 healthy adults (75 males,
gender. 25 females) with newspaper advertisements
from the Milwaukee, Wisconsin, area. The av-
ULO.S] Ska, Poissant, and Joanette, 1990, erage age of the participants was 46.0 (11.6),
(Table A15.6) average education was 13.3 (2.0), and average
In this study, 95 nonhospitalized norrtal el- Verbal IQ was 106.5 (6.9). Exclusion criteria
derly controls (19 males, 76 females~ were were history of substance abuse, psychiatric
recruited from the community. Exclusion l
illness, head injury, or other neurological dis-
criteria were a history of alcoholism. drug orders. All controls were given neurological
abuse, or neurological or psychiatric illness. evaluations and MRI. Only one participant
The sample was divided into three age gi-oups: was nonwhite. Standard JW administration
55-64,65-74, and 75--84. Average educational procedures were used. All participants were
levels were 10.13 (3.3.8), for the fir$t age paid.
group, 9.46 (3.40) for the second group, and
8.06 (2. 77) for the third group. Standari JW Srudy stnm~~
administration procedures were used. . 1. The sample composition is well described
The authors found a decline in JLO per- in terms of age, education, WAIS-R Ver-
formance with age, but the various types of bal IQ, gender, geographic area, andre-
JW error evaluated in this study did opt dif- cruitment procedures.
fer for different age groups in the ~rmal 2. Large sample size.
3. Adequate exclusion criteria. sample) are actually retest scores and

4. Test administration procedures are may slightly inflate the mean value re-
specified. ported due to a practice effect.
reported. ULO.BJ Kempen, Kritchevsky, and Feldman,
1994 (Table A15.9)
Consideration regarding use of study The authors compared JLO performance of
1. The data are not partitioned by age or 13 individuals (three males, 10 females) with
gender. normal vision to those who were visually im-
paired. Participants ranged in age from 55 to
UL0.7] Meador, Moore, Nichols, Abney, 74, with an average age of 65.2 (5.9), and ob-
Taylor, Zamrini, and Loring, 1993 tained an average of 16.5 (2.9) years of edu-
(Table A15.8) cation. Participants were recruited at the
Twelve healthy individuals (eight males, four University of California San Diego School of
females) served in this study. Participants were Medicine during a routine ophthalmic exam
used in a counterbalanced, repeated-measures by one of the authors. The only exclusion cri-
design, in which they were administered saline terion was poor visual acuity (Snellen distance
or scopolamine (an anticholinergic) at differ- acuity of $20/50). Standard JLO administra-
ent times. Participants ranged in age from 20 tion procedures were used. Interestingly, the
to 42, with a mean age of 31 years, and had authors found that poor vision did not affect
completed 12-20 years of education, with an JLO performance.
average of 16 years. Test sessions were sepa-
rated by at least 72 hours. Their JLO scores Study strengths
when on the saline solution were reported in 1. The sample composition is well de-
this chapter. Participants were staff employees scribed in terms of age, education, gen-
of the Medical College of Georgia and were der, geographic area, and recruitment
paid for their participation. They were procedures.
screened for use of psychoactive drugs and 2. Test administration procedures are
history of neurological, psychiatric, or "major" specified.
medical disease. 3. Means and SDs for the test scores are
reported.
Study strengths
1. The sample composition is well de- Considerations regarding use of study
scribed in terms of age, education, gen- 1. Sample size is small.
der, geographic area, and recruitment 2. Inadequate exclusion criteria used.
procedures. 3. The data are not partitioned by age or
2. Exclusion criteria are provided. gender.
3. Test administration procedures are 4. Educational level for the sample is high.
specified.
4. Means and SDs for the test scores are UL0.9] York and Cermak, 1995 (Table A15.10)
reported. The authors compared JLO scores between
15 control participants (six males, nine fe-
Considerations regarding use of study males) and patients with CVA. The average
1. The sample size is small. age of participants was 61.89 (8.67) and av-
2. The data are not partitioned by age or erage education was 15.07 (3.41) years. Nor-
gender. mal controls were selected from the
3. Educational level for the sample is high. orthopedic floor of two rehabilitation hospitals
4. Participants were staff of the medical cen- in Portland, Maine. Control participants did
ter and may represent a biased sample. not have a history of CVA or other neurolog-
5. Despite counterbalancing methods, JLO ical deficits. Standard JLO administration
scores for six participants (50% of the procedures were used.
Study strengths use "corrected" or "final" scores for tests that might
1. The sample composition is well described present their own age- or education-adjustments.
in terms of age, education, gender, geo- Select the appropriate column in the table for that
graphic area, and recruitment procedures. test. The corresponding row in the left-most column
2. Test administration procedures are in each table provides the MOANS Age-Corrected
specified. Scaled Score ... for your subject's raw score; the
3. Means and SDs for the test scores are corresponding row in the right-most column in-
reported. dicates the percentile range for that same score.
Further, linear regressions should be
Considerations regarding use of the study applied to the normalized, age-corrected
1. Sample size is small. MOANS scaled scores (A-MSS) derived from
2. The data are not partitioned by age or the tables, to adjust patient scores for educa-
gender. tion. Age- and education-corrected scores for
3. Educational level for the sample is high. the JLO (A&E-MSS) can be calculated as
4. Exclusion criteria are very limited (e.g., follows:
no mention of psychiatric disorders) and
not well defined. A&E-MSS1w=K+(W1 • A-MSS1w)
5. Controls are hospital patients and may - (W2 • Education)
represent a biased sample.
where the following indices are specified for
UL0.10] lvnik, Malec, Smith, Tangalos, the JLO:
and Petersen, 1996 (Table A15.11) K 1.54
The study provides age-specific norms for the WI 1.10
JLO test obtained in Mayo's Older Americans w2 o.23
Normative Studies (MOANS) project, which
Education should enter the formula as years
obtains normative data for elderly individuals
of formal schooling.
on different neuropsychological tests. The
The tables of scaled scores per age group
total sample consisted of 746 cognitively nor-
provided by the authors should be used in the
mal volunteers residing in Minnesota, over
context of the detailed procedures for their
age 55, 216 of whom participated in JLO
application, which are explained in Ivnik et al.
testing. Mean MAYO FSIQ (which differs
(1996). Therefore, they are not reproduced in
somewhat from the standard WAIS-R FSIQ)
this book. Interested readers are referred to
for the whole sample was 106.2 ( ± 14.0) and
the original article. Table A15.11 summarizes
mean Mayo General Memory Index on the
sample sizes for the different demographic
WMS-R was 106.2 ( ± 14.2). For a description
groups.
of their samples, the authors refer to their
earlier publications. Standard JLO adminis-
Study strengths
tration procedures were used.
Age categorization used the midpoint in-
IQ, gender, ethnicity, handedness, and
terval technique. The raw score distribution
geographic area is reported.
for each test at each midpoint age was "nor-
2. The data were stratified by age group
malized" by assigning standard scores with a
based on the midpoint interval technique.
mean of 10 and SD of 3, based on actual
3. The innovative scoring system was well
percentile ranks. The authors provided tables
described. The authors developed new
of age-corrected norms for each age group
indices of performance.
(see below). The procedure for clinical appli-
cation of these data is described in the original
article (Ivnik et al., 1996) as follows:
1. Sample size for most age groups is small
first select the table that corresponds to that person's {ranging 2-45), with the exception of the
age. Enter the table with the test's raw score; do not 80-84 group (n = 69).
JUDGMENT OF liNE ORIENTATION 293
2. 'nle measures proposed by the authors UL0.12] Finton, Lucas, Graff-Radford, and
are quite complicated and might be dif- Uitti, 1998 (Table A15.13)
ficult to use in clinical practice. JW scores for 24 normal elderly control
3. Participants with prior history of neuro- participants (13 males, 11 females) were ob-
logical, psychiatric, or chronic medical tained. The average age of participants was
illnesses were included. 70.4 (6.0), average education was 15.7 (2.8),
and average Dementia Rating Scale score was
Other comments 138.1 (3.5). Control participants had no com-
1. 'nte theoretical assumptions underlying plaints of cognitive deficits and did not display
this normative project have been pre- disorders affecting cognition on physical or
sented in Ivnik et al. (1992a,b). neurological examination. Standard JLO ad-
2. 'nte authors cautioned that the validity ministration procedures were used.
of the MAYO indices depends heavily on
the match of demographic features of Study strengths
the individual to the normative sample 1. 'nte sample composition is well described
presented in this article. in terms of age, education, and gender.
3. Correlation of the JLO with age was 2. Adequate exclusion criteria.
-0.25, whereas correlations with educa- 3. Test administration procedures are
tion and gender were 0.21 and- 0.24, specified.
respectively. 4. Means and SDs for the test scores are
reported.
UL0.11 1 Fleming, Goldberg, Binks, Randolph,
Gold and Weinberger, 1997 (Table A15.12) Considerations regarding use of study
'nte control group in this study was comprised 1. Sample size is small.
of 27 (16 males, 11 females) paid volunteers 2. 'nte data are not partitioned by age or
recruited from the Washington, D.C., com- gender.
munity. 'nte average age of participants was 3. Educational level for the sample is high.
26.1 (7.4), and average education was 15.4 4. Recruitment procedures are not reported.
(3.2) years. Participants were screened by
psychiatrists for substance use and major UL0.13] Obonsawin, Robertson, Crawford,
psychiatric illness, neurological disease, and Perera, Walker, Blackmore, Parker, and
other medical diagnoses. Standard JLO ad- Besson, 1998 (Table A15.14)
ministration procedures were used. 'nle study was designed to assess the effects
of scopolamine (an anticholinergic) on neu-
Study strengths ropsychological test performance. Twelve
1. 'nte sample composition is well described healthy participants (three males, nine fe-
in terms of age, education, gender, males) who were injected with a saline solu-
IQ, geographic area, and recruitment tion served as the normal controls. Participant
procedures. ranged in age from 20 to 36, with an average
age of 40.83 (12.55), and obtained an average
3. Test administration procedures are score of 35.09 (7.23) on the National Adult
specified. Reading Test. 'nle authors do not mention
4. Means and SDs for the test scores are precisely from what region of the United
reported.
Kingdom participants were drawn. All parti-
cipants had a medical examination. In addi-
Considerations regarding use of the study tion, abstinence from alcohol for at least 24
1. Sample size is small. hours and from coffee and tea for at least 12
2. 'nte data are not partitioned by age or hours was required. Standard JW adminis-
gender. tration procedures were used. Participants
3. Educational level for the sample is high. were paid.
Study strengths erage age of the male participants was 22.04

1. The sample composition is well de- (3.53) and the average age of the female par-
scribed in terms of age, gender, and es- ticipants was 22.62 (7.24). The average edu-
timated IQ. cation for males was 13.92 (1.01) and the
2. Test administration procedur~ are average education for females was 13.85
specified. (1.05). Males had an average IQ of 112.67
3. Means and SDs for the test scores are (3.78) and females had an average 1Qof112.08
provided. (4.20). No statistical differences were found
between males and females on age, education,
Considerations regarding use of study and IQ. All participants were fluent in English.
1. Sample size is small. There were 25 Caucasian and one African-
2. Educational level is not reported. American females and 21 Caucasian, three
3. Exclusion criteria are not clearly described. African-American, and two Hispanic males. All
4. Recruitment procedures are not ref>rted. participants were assessed to be right-handed
5. The data are not partitioned by rge or by the Edinburgh Handedness Inventory, and
gender. all denied a history of learning disability,
neurological illness, psychiatric disease, or
UL0.14] Meyers, Galinsky, and Volbrech4 head trauma during an interview. Standard
1999 (Table A15.15) · JLO administration procedures were used.
The authors obtained JW scores on $0 (14
male, 16 female) normally functio~g in- Study strengths
dividuals. The average age of particip~ was 1. Adequate sample size.
36.70 (20.50), average education was: 13.67 2. The sample composition is well de-
(3.47), and average IQ was 113.97 (.3.51). scribed in terms of age, education, gen-
Exclusion criteria were history of neurobgical der, and ethnicity.
disease, closed head injury, motor ~hicle 3. Data are partitioned by gender.
accidents, learning disabilities, loss o( con- 4. Adequate exclusion criteria.
sciousness, or "other" conditions. StaJtdard 5. Test administration procedures are
JLO administration procedures were used. specified.
6. Means are reported, and we were able
Study strengths to calculate SDs from the confidence
1. The sample composition is weD de- intervals provided.
scribed in terms of age, educatio~ gen-
der, and IQ. Consideration regarding use of the study
2. Adequate exclusion criteria. 1. Recruitment procedures are not
3. Test administration procedures- are reported.
specified.
4. Means and SDs for the test scores are UL0.16] Bell, Hennann, Woodard, Jones,
reported. Rutecki, Sheth, Dow, and Seidenherg, 2001
(Table A15.17)
Considerations regarding use of the study The study examined object-naming ability and
1. Sample size is relatively small. depth of semantic knowledge in healthy con-
2. Data are not partitioned by age or g~der. trols and patients with early-onset temporal
3. There is a wide spread in the ~e of lobe epilepsy (TLE), using the JLO as part of
participants (SD is nearly 21 years). a larger neuropsychological battery. The con-
4. Recruitment procedures are not reported. trol group included 29 friends, relatives, and
spouses of TLE patients (72% female), aged
UL0.1 5] Basso, Hanington, Matson, 16--60 years, with a mean age of 34.4 (12.5)
and Lowery, 2000 (Table A15.16) I years; FSIQ (as measured with the WAIS-111
The authors sampled 52 healthy undergradu- seven-subtest short form) of 69-110, with a
ate students (26 males, 26 females). ~ av- mean FSIQ of97.7 (6.4); and mean education
l'
of 13.0 (1.7) years. Exclusion criteria were 3. Test administration procedures are
current substance abuse, psychotropic med- specified.
ication use, medical or psychiatric condition 4. Mean error scores and SDs are reported.
that could affect cognitive functioning, past
episode of loss of consciousness >5 minutes, Considerations regarding use of the study
an identified developmental learning disor- 1. The data are not stratified by age or
der, and repetition of a grade in school. Stan- gender.
dard JLO administration procedures were 2. Exclusion criteria are not clearly de-
used. scribed, except to say that participants
had no history of neurological or psy-
Study strengths chiatric illness.
1. The sample composition is well de- 3. There is a wide spread in age (39-85)
scribed in terms of age, education, gen- and education (1-23).
der, FSIQ, and recruitment procedures. 4. The data were obtained in Spain, which
2. Adequate exclusion criteria. may limit their usefulness for clinical
3. Means and SDs for the test scores are interpretation in the United States.
reported.
UL0.18] Rahman and Wilson, 2003
Considerations regarding use of the study (Table A15.19)
1. Sample size is small. This study examined various aspects of cogni-
2. Sample includes a wide age range. tive functioning in heterosexual and homosex-
3. The data are not partitioned by age or ual males and females. The authors recruited
gender. 240 healthy participants. Data are reported for
four groups: heterosexual males, homosexual
UL0.17] Montse, Pere, Carme, Francese, males, heterosexual females, and homosexual
and Eduardo, 2001 (Table A15.18) females. Participants were aged 18-40 years
The authors examined the various types of and were recruited from King's College Lon-
error made by patients with Parkinson's dis- don sources, the local community, and social
ease and normal controls on the JLO. A networks. The authors admit that the partici-
sample of 76 (38 males, 38 females) healthy pants "may have been oversampled from uni-
individuals was selected as a control. Partici- versity sources." All participants were right-
pants ranged in age from 39 to 85, with an handed. Standard JLO administration proce-
average of 63.84 (9.93), and ranged in edu- dures were used. Data are stratified by gender,
cational level from 1 to 20 years, with an av- and information regarding age, education, and
erage of 9.64 (4.17). Participants were free performance on Raven's Matrices is provided.
of neurological and psychiatric illness. Con-
trols were spouses or friends of the Par- Study strengths
kinson's patients and were recruited from 1. Relatively large sample size.
Barcelona, Spain. They were administered 2. The sample composition is well de-
Form H of the JLO. A subset of the sample scribed in terms of age, education, gen-
was retested. der, Raven's Matrices score, geographic
Standard JLO administration procedures area, and recruitment criteria.
were used. The number of errors on the JLO 3. Data are partitioned by gender.
was used as an outcome variable. 4. Test administration procedures are
specified.
1. Adequate sample size. reported.
scribed in terms of age, education, gen- Considerations regarding use of the study
der, geographic area, and recruitment 1. Data are not partitioned by age.
procedures. 2. Educational level for the sample is high.
296 PERCEPTUA,L ORGANIZATION: VISUOSPATIAl AND TACTILE
3. No exclusion criteria are provided. UL0.20] Woodard, Benedict, Salthouse,

4. By the authors' own admission, partic- loth, Zgalijardic, and Hancock, 1998
ipants were recruited mostly· from (Short form) (Table A15.21)
university sources, possibly affecting the Using the odd and even items ofJLO Form V,
generalizability of the data. ' two different short versions of the test were
created, Form E and Form 0. These two
Short Forms versions were administered to 82 (22% male)
"healthy, nondepressed, community dwelling,
UL0.19] Salthouse, loth, Hancock, and geriatric individuals" from western New York
Woodard, 1997 (Short form) (Table A15.2P) State. Mean age, education, and MMSE scores
Using the shortened format of the JLO~ these are reported. Participants ranged in age from
authors examined the relationship bqtween 55 to 84, with an average age of 65.8 (6.7);
age and JLO performance (as well asj other educational level ranged 9-20 years, with an
neuropsychological tests) in a group of~althy average education of 14.0 (2.3) years; and
adults (n = 124; approximately 47% mal~, 53% premorbid IQ ranged 86-121, with an average
female) aged 18-78 years. Participant!4 were IQ of 110.3 (6.6). The sample was primarily
included if reported to be in "reasonabli good Caucasian (97.6%). Additional raw score to
health," to not be a current student~d to standard score conversion tables are provided
have at least 11 years of education. N other in the original article but are not reproduced
exclusion criteria are reported. Rec ent in this chapter.
procedures are not provided, nor is th loca-
tion from which participants were rec~ted. Study strengths
Participants were administered a batttry of 1. Sample size is relatively large.
neuropsychological tests in their home4. The 2. The sample composition is well de-
data were stratified into three age groupings: scribed in terms of age, education, gen-
18-39 years [mean age= 29.0 (4.8); irnean der, and geographic region.
education= 15.5 (1.7)), 40-59 years Fean 3. Test administration procedures are
age= 49.1 (5.1); mean education= 15.2 J2.5)], specified.
and 60-78 [mean age=69.2 (5.11); me$1 ed- 4. Means and SDs are reported for each
ucation= 15.3 (2.6)]. Test procedures f9r ad- short form.
ministering only the odd-numbered itejns of
the JLO were used (Woodard et al., 1996). Considerations regarding use of the study
1. Recruitment procedures are not re-
Study strengths ported.
1. Sample size is relatively large. 2. Exclusion criteria are not well identified.
2. The sample composition is weD de- 3. Data are not partitioned by age or gender.
scribed in terms of age, education, and
gender.
3. Data are partitioned by three age CONCLUSIONS
groups. , As the literature above suggests, the JLO is a
4. Test administration procedures! are clinically useful test for assessing visual-spatial
specified. · functioning in various patient populations.
5. Means and SDs are reported for; each Older age has been consistently associated
short form. with lower JLO performance; however, only
three of the studies stratified data by age, and
Considerations regarding use of the study of these, two included data only for older age
1. Recruitment procedures are : not groups (i.e., 2;::55 years). Further, most of the
reported. normative studies have been based on small
2. Exclusion criteria are not well identi- sample sizes, and of the four studies which did
fied. ' have sample sizes >50, three did not stratify
by age; if they had, the resulting cells would substantially impacts JLO performance is less
have had small sample sizes. Consistent gen- clear, but of concern, the educational levels of
der effects have been reported, with males the samples, when reported, have typically
outperforming females by approximately 2 been high (i.e., 2::13 years in 10 studies) and
points. However, only three studies stratified data have not been stratified by educational
data by gender. Finally, whether education level. 2
2 Meta-analyses for the JW were conducted using data

reported in this chapter. Although the R2 and significance
level for the resulting regression were minimally accept-
able, we felt that the solution was not empirically sup-
ported by the data as the majority of the data points
aggregated at the younger and older age ranges, with a
paucity of data in-between. Therefore, the results of meta-
analyses are not presented in this chapter.
16
Design Fluency Tests
BRIEF HISTORY OF THE TESTS arranged in a symmetrical, circular pattern for

parts I, II, and III, with the two latter parts
The first nonverbal fluency test was developed containing interference objects such as dia-
by Jones-Gotrnan and Milner (1977) and was mond shapes or preconnected dots. In parts
designed to be a nonverbal version of the IV and V, the five dots are arranged in a
Thurstone and Thurstone (1949) verbal fluency nonsymmetrical fashion and there are no in-
test. This original design fluency test consisted terference objects. For each part of the RFFT,
of two trials: a 5-minute unstructured task, in the examinee is instructed to make a unique
which the patient was instructed to rapidly design in each square by connecting two or
draw novel, nonnameable designs (excluding more dots with a straight line. The scores used
scribbles), and a 4-minute trial, in which the for interpretation include total number of
patient was required to rapidly draw novel novel designs produced in the five !-minute
designs containing exactly four lines. How- trials, number of designs repeated (perse-
ever, as rightly pointed out by Hanks et al. verations), and the ratio of perseverations to
(1996), a number of factors, such as the effect unique designs produced. However, addi-
of design complexity on production rate and tional indices of production strategy (Ross
questions regarding what constitutes a persev- et al., 2003; Ruff, 1988) are also available.
erative error vs. unique design, make clinical While this version of design fluency has been
use of this test problematic. gaining widespread use, criticisms have been
Regard et al. (1982) subsequently created raised that, although the RFFT structured
the Five-Point Test, a more structured format format may increase scoring reliability, the
than the Jones-Gotrnan/Milner nonverbal unstructured Jones-Gotrnan/Milner version
fluency test. This test was modified (e.g., dot may better measure the initiation and orga-
arrangement was changed and time limit re- nization found to be deficient in patients with
duced) and used as part I of the subsequent frontal executive dysfunction (suggested by
Ruff Figural Fluency Test (RFFT). The the fact that the free condition on the Jones-
RFFT, developed by Ruff and colleagues Gotrnan/Milner version is more sensitive than
(Evans et al., 1985; Ruff et al., 1987), consists the fixed condition; Harter et al., 1999).
of five parts, with each part containing 35 Less commonly used adaptations of design
items of five dots drawn within a 3 em square, fluency tasks have also been developed: (1)
arranged in a 5" x 7" array drawn in black and providing a sheet with five lines and instructing
white on letter-sized paper. The dots are participants to produce three different shapes
298
DESIGN FLUENCY TESTS 299
(outcome measure is time to complete three patients with right temporal or left frontal le-
designs; Kivircik et al., 2003); (2) providing sions. Subsequently, Jones-Gotman (1991a, b)
four flat pieces of plastic, two straight and two reported that poor performance on design
curved, and instructing participants to create fluency was unique to patients with right fron-
as many designs (representational or nonrep- tal lesions. In a comparison of patients with
resentational, with pieces touching at least primarily right vs. left frontotemporal lobar
once) as possible within 3 minutes (Griffiths, degeneration (FTD) (as demonstrated by
1991), (3) instructing participants to draw as asymmetrical SPECT perfusion patterns), us-
many designs as possible in 2 minutes using ing a difference score consisting of verbal
three lines (two vertical, one semicircular) fluency (FAS) minus design fluency, patients
(Allen et al., 1996), and (4) providing partici- with right FTD displayed higher verbal flu-
pants with three shapes (two straight lines and ency and lower design fluency than patients
an arc) and instructing them to make as many with left FTD (Boone et al., 1999).
different figures as possible using all three Using the Five-Point Test, Lee et al. (1997)
shapes within 2 minutes (Hanks et al., 1996). found that patients with frontal lobe dysfunc-
A design fluency task has also been included tion committed a greater percentage of per-
in the recently published Delis-Kaplan Exec- severative errors than non-frontal neurological
utive Function system (Delis et al., 2001). and psychiatric patients. Furthermore, the
This task involves three !-minute trials: a Five-Point Test more effectively classified
"basic" trial, in which the participant is pre- right frontal lobe patients relative to patients
sented with 35 boxed five-dot arrays and in- with damage to other brain regions. However,
structed to make designs consisting of four Tucha et al. (1999) failed to corroborate this
straight lines which connect to each other; a observation; they reported that right and left
"filter" task, in which the subject is presented frontal lobe--lesioned patients performed
with 35 10-dot arrays (five solid dots and five comparably to each other and to controls.
clear dots) and instructed to draw designs Ruff et al. (1986b) found that patients with
using the empty dots only; and a "switch" trial, severe head injuries produced fewer designs
in which participants are again presented with on the RFFT than those with mild head in-
the identical 35 10-dot arrays but told to draw juries. In a subsequent study, Ruff et al.
designs by switching back and forth between (1994) assessed RFFT and verbal fluency
solid and clear dots. The number of correct performance in patients with right frontal, left
designs, percentage of errors (errors divided frontal, right posterior, and left posterior lobe-
by number of designs), inappropriate designs damaged groups and found that, relative to
(e.g., designs with five lines), and persevera- the other groups, the right frontal group by far
tive errors are tabulated. The above adapta- performed worse on the RFFT but not on the
tions will not be covered in this chapter due to verbal fluency test. Suchy et al. (2003) ob-
the lack of empirical research using these served that the RFFT was better than F AS
techniques. (phonemic verbal fluency) at discriminating
While relatively few clinical studies have between frontal and temporal foci seizures,
been conducted with nonverbal fluency tasks, although F AS was slightly better in classifying
these tests in general have been shown to be right vs. left seizure foci.
sensitive to frontal lobe impairment (Butler Design fluency performance is also sup-
et al., 1993; Taylor et al., 1986), with some pressed in conditions associated with nonlat-
evidence suggesting more sensitivity to right eralized brain dysfunction, such as Parkinson's
frontal damage, although functional imaging disease and Alzheimer's disease (Fama et al.,
studies have revealed activation of both frontal 1998), attention-deficit hyperactivity disorder
lobes in design generation (Elfgren & Risberg, (ADHD) (Rapport et al., 2001), obsessive-
1998). Jones-Gotman and Milner (1977) found compulsive disorder (OCD) (Mataix-Cols
that patients with primarily right frontal or et al., 1999), suicidal behavior (Bartfai et al.,
right frontal-central cortical lesions were most 1990), autism (Turner, 1999), transient global
impaired on design generation, followed by amnesia (Stillhard et al., 1990), amyotrophic
300 PERCEPTUA.L ORGANIZATION: VISUOSPATIAL AND TACTILE
lateral sclerosis (Abrahams et al., 2000), neg- have been reported (Basso et al., 1999; Ruff,
ative symptoms in schizophrenia (Stolar et al., 1996). An average increase of eight designs
1994), head injury (Harter et al., 1~), ex- has been reported on retesting at approxi-
posure to alcohol in utero (Connor et al. mately 1-month (Ross et al., 2003) and 6-
2001), schizoaffective disorder (Beatty; et al., month (Ruff, 1988) intervals, while an average
1993), borderline personality disorde~ (Judd gain of six or seven designs has been docu-
& Ruff, 1993), "high hostile" males (~am mented at 12 months (Basso et al., 1999).
son & Harrison, 2002), and bilateral ahterior However, a much greater practice effect was
cingulate cortex lesions secondary to kingu- reported at 3 weeks, equivalent to an average
lotomy for chronic pain (Cohen et al., 11999). of 17 more unique designs produced on re-
Normal design fluency has been repo~ed in testing (Demakis, 1999). Berning et al. (1998)
dyslexia (Griffiths, 1991) and in inditiduals reported interrater reliabilities of 0.93 (novel
with post-concussion-type symptoms (Chan, designs), 0.74 (perseverations), and 0.66 (error
2001), and improved performance h~ been ratio); Ross et al. (2003) observed intraclass
reported following glucose ingestion !(Allen correlations of 0.82-0.96 for eight different
et al., 1996) and hypnosis (Gruzelier ~ War- RFFT scores; and Sands (1998) documented a
ren, 1993), in participants high in n~ for coefficient of 0.99 for unique design and
achievement (Allen et al., 1992), ud af- perseverative errors.
ter pallidotomy for treatment of adrcmced Factor analyses have revealed that total
Parkinson's disease (Lacritz et al., 2~ unique designs loaded on initiation measures,
RFFI' scores are mostly related to res whereas error ratios loaded on planning fac-
from the Jones-Cotman/Milner versio with tors (Ruff, 1996).
correlations ranging from 0.25 for thf free
condition to 0.38 for the fixed conditimi (De- Design Fluency Test Oones-Gotman/
makis & Harrison, 1997). Further tletails Milner Version)
regarding the RFFT testing materials, ~inis Test-retest correlations at 1 month were re-
tration procedures, and scoring can be o~ned ported at 0.69 for novel designs, while values
in the test manual and kit (see Appendil1 for of 0.77, 0.91, and 0.51 were documented for
ordering infonnation) and in Lezak et al. (2004). unique scoring parameters of complexity, con-
Specific administration and scoring instrQctions crete responses, and drawing quality, respec-
for the Jones-Cotman/Milner design fluency tively (Harter et al., 1999); on average, two
task are contained in Spreen and Strauss (1998, additional designs were produced at 1-month
pp. 199-201). re-retesting.
Chan (2001) documented interrater reli-
abilities of 0.74-0.98. Similarly, Varney et al.
Psychometric Properties of the (1996) reported 90% agreement between two
Design Fluency Tests independent raters on scoring the free con-
dition. Harter et al. (1999) observed interrater
The RFFT reliabilities of 1.00 and 0.98 for novel designs,
The RFFT has demonstrated adequate reli- number of perseverations, and number of
ability and construct validity. Test-rete~ reli- scribbles as well as generally high correlations
ability coefficients at 6 months for ~que for additional scores reflecting complexity
designs produced on the individual I trials (0.97-0.98), number of concrete responses
range 0.58-0.69 (Ruff, 1996). The htghest (0.81-0.90), and number of variations (0.77),
r value is reported for total designs proiuced although ratings of drawing quality were lower
for all five trials, ranging from 0.85 \when (0.46-0.59). Carter et al. (1998) reported in-
testing intervals are separated by 4-7 iveeks terrater agreements of0.77 and 0.66 for novel
(Ross et al., 2003) to 0.88 when testing pnter- output, 0.82 and 0.70 for perseverative errors,
vals are separated by approximately 3 weeks and 0.50 and 0.47 for nameable errors
(Demakis, 1999). Correlation coefficieats of respectively, for the free and fixed conditions.
0.71-0.76 at 6- to 12-month testing in~rvals In addition, an interrater agreement of 0.42
was noted for incorrect number of lines on the Regarding the Jones-Gotman/Milner ver-
fixed condition. However, the ratings reported sion, Varney et al. (1996) observed no signifi-
by Woodard et al. (1992), based on ratings of cant relationship between number of designs
graduate students in clinical neuropsychology, generated and age, educational level, or IQ
were somewhat lower, with coefficients of and no differences in performance between
agreement of 0.64 and 0. 71 for novel designs, men and women. Similarly, Daigneault et al.
0.64 and 0.24 for nameable responses, and (1992) found no relationship between young
0.57 and 0.41 for perseverative responses in and middle-aged adults in design fluency;
the free and fixed conditions, respectively, and however, Mittenberg et al. (1989) reported a
0.68 for wrong number of lines in the fixed significant association between age and design
condition. The data from Carter and col- fluency performance across a larger age range.
leagues were based on the stricter scoring Demakis and Harrison (1997) and Mataix-Cols
criteria of Jones-Gotman, as summarized in et al. (1999) found no effect of gender on de-
Spreen and Strauss (1998). sign generation; however, Harter et al. (1999)
did demonstrate more novel designs produced
by college males compared to college females,
although they note that their sample of men
RELATIONSHIP BETWEEN DESIGN
was small. These authors also failed to docu-
FLUENCY PERFORMANCE AND
ment any effects of age or education on
DEMOGRAPHIC FACTORS
test performance. Interestingly, they observed
Relatively sparse literature exists on design worse performance in left-banders, in terms of
fluency measures. Only a few studies exam- perseverations and scribbled responses, who
ining the influence of demographic factors on also performed more poorly on additional
RFFr performance could be located. In the measures of design quality. A minimally sig-
original normative study, Ruff et al. (1987) nificant relationship with Performance IQ was
found that age and education significantly observed. Design fluency data have been ob-
affect RFFr performance. They documented tained on participants in the United States as
that unique design production decreased as a well as Spain (Mataix-Cols et al., 1999) and
function of age, particularly in those with less England (Abrahams et al., 2000).
than a college education. Similar age and
education effects were noted for persevera-
tive errors. Interestingly, the authors reported
no overall gender differences for design pro-
NORMATIVE REPORTS
duction but revealed a trend toward men
committing fewer perseverative errors than To adequately evaluate the design fluency
women in two subgroups: 40--55 years of age normative reports, five criterion variables
with :512 years of education and 55-70 years were deemed critical. The first four of these
of age with 16 years of education. Overall, the are related to subject variables, and the last
authors concluded that these findings were one refers to procedural issues.
not robust enough to suggest significant
gender differences. Demakis and Harrison
(1997) and Ross et al. (2003) also failed to Subject Variables
demonstrate gender differences in their
samples of college students. Ruff et al. (1987) Sample Size
found no relationship between motor speed Fifty cases are considered a desirable sample
or verbal fluency and RFFr performance but size. Although this criterion is somewhat ar-
did note a significant correlation between bitrary, a large number of studies suggest that
Performance IQ (WAIS-R PIQ) and RFFr data based on small sample sizes are highly
total design production. All normative studies influenced by individual differences and do
appear to have used participants in the Uni- not provide a reliable estimate of the popu-
ted States. lation mean.
Sample Composition Description SUMMARY OF THE STATUS

Information regarding medical and psychiatric OF THE NORMS
exclusion criteria is important. It is unelear if The RFFT
intellectual level, gender, geographic recruit-
ment region, socioeconomic status, occupa- Only one study was designed to provide nor-
tion, ethnicity, handedness, or recrultment mative information on the RFFr: Ruff et al.
procedures are relevant. Until determined, it (1987) presented data for 358 healthy in-
is best that this information be provided. dividuals ranging widely in age and educa-
tional level.
Age Group Intervals Information on age is provided in all stud-
This criterion refers to grouping of the data ies, although only Ruff et al. (1987) stratify on
into limited age intervals. A significant age age. Similarly, all studies report information
effect has been reported for the RFFT, al- on educational level, although again only Ruff
though a relationship between age and the et al. (1987) stratify on education. All studies
Jones-Gotman/Milner version has not been had at least 50 participants with the exception
documented. Given the probable relationship of Demakis (1999). Only Ruff et al. (1986b)
between age and design fluency perforrpance, did not provide information on gender. Ex-
presentation of data by age groupings Iis re- clusion criteria are specified in all but three
commended for these tasks. studies (Berning et al., 1998; Demakis, 1999;
Ruff et al., 1986b). With the exception of one
Reporting of Educational Levels : study (Berning et al., 1998), scores for the five
I
Educational level has been linked to JtFFr RFFr parts are combined. Information re-
scores but not to the Jones-Gotmanlifilner garding geographic area is indicated in all but
version; given the probable relationshtp be- one study (Demakis & Harrison, 1997), and
tween design fluency and education, 'infor- recruitment strategies are specified in only
mation regarding educational level should be three studies (Berning et al., 1998; Demakis,
reported for each subgroup. 1999; Ross et al., 2003).
Reporting of IQ Information
Jones-Gotrnan/Milner Version
IQ level (especially PIQ) has been assoeiated
with RFFr performance, although no such Only five studies report nonnative data for the
link has been documented for the tones- free and fixed conditions separately (Abra-
Cotman/Milner version; given the evidepce of hams et al., 2000; Carter et al., 1998; Demakis
a relationship between IQ and design fthency & Harrison, 1997; Mataix-Cols et al., 1999;
performance, information regarding Iq level Woodard et al., 1992), while three studies
should be reported. f collapse the data (Daigneault et al., 1992;
Harter et al., 1999; Rapport et al., 2001), three
studies provide information for only the free
Procedural Variable trial (Boone et al., 1991, 2001; Varney et al.,
1996), and one study documents data for only
Data Reporting the fixed trial (Beatty et al., 1993). Five studies
Test version should be indicated. F~ the provide data on samples of < 50 (Abrahams
RFFr, group means and standard de~tions et al., 2000; Beatty et al., 1993; Boone et al.,
for the total number of unique designs lcross 1991, 2001; Mataix-Cols, 1999). Data regard-
five parts of the test should be presented, and ing age are provided in every study, and a full
preferably data on perseverations shm4d be age spectrum is represented (i.e., young adult
included. For the Jones-Gotman/Milnet ver- to elderly); however, only one study stratified
sion, group means and SDs should be, pro- by age (Daigneault et al., 1992), although four
vided for the 5-minute unstructured task and/ studies recruited college students (Carter
or the 4-minute structured task. et al., 1998; Demakis & Harrison, 1997;
Harter et al., 1999; Mataix-Cols et al., 1999), Study strengths

which likely represents a relatively restricted 1. Adequate sample size.
age range. All studies provide infonnation on 2. 1be sample composition is well de-
educational level, with most studies reporting scribed in tenns of age and education.
on individuals with > 13 years of education (or 3. Test administration procedures are
in college at the time of testing). Gender specified.
distribution is reported in every study. Five 4. Means and SDs for the test scores are
studies provide data on IQ level and reveal reported.
average to high average intellectual levels.
Seven investigations reported geographic Considerations regarding use of the study
area, with six studies indicating actual re- 1. Exclusion criteria are not described.
cruitment strategies. Ethnic makeup is not 2. Recruitment procedures and gender are
specifically reported for any study. Exclusion not reported.
criteria were judged to be at least minimally 3. 1be data are not partitioned by age.
adequate in all studies.
In this chapter, nonnative publications and [RFFT.2] Ruff, Light, and Evans, 1987
control data from clinical studies are reviewed (RFFT Version) (Table A16.3)
in ascending chronological order. Studies us- 1bese authors collected data on 358 (161
ing the RFFr are reviewed first, followed by male, 197 female) individuals aged 16-70 years
those using the Jones-Gotman/Milner version. with education of 7-22 years. Participants
1be text of study descriptions contains refer- were recruited from various parts of the
ences to the corresponding tables identified United States, with approximately 65% from
by number in Appendix 16. Table A16.1, the California, 30% from Michigan, and 5% from
locator table, summarizes infonnation pro- the eastern seaboard. Participants were ex-
vided in the studies described in this chapter.1 cluded if they had a history of psychiatric
hospitalization, chronic polydrug abuse, or
neurological disorder. 1be data are stratified
by four age groups (16-24, 25--39, 40--54, 55-
70 years) and three educational levels (:::;12,
RFFT Version 13-15, >16 years) but not by gender since no
gender effects were found. Standard test pro-
[RFFT.1J Ruff, Evans, and Marshall, 1986b cedures were used. Based on the data from
(RFFT Version) (Table A16.2) this study, a professional manual was pub-
1be authors compared the perfonnance of lished by Psychological Assessment Resources
moderately and severely head-injured patients (Ruff & Allen, 1996) in which tables con-
to that of 50 nonnal control volunteers. 1be verting raw scores into T scores and percen-
gender composition of the control group is not tiles are provided for the different age groups.
provided. Participants were an average of 28.2
(8.8) years of age, with an average of 13.2 (1.7) Study strengths
years of education. From the general descrip- 1. Large sample size.
tion provided by the authors, it can be assumed 2. The sample composition is well de-
that the control participants were recruited scribed in tenns of age, education, gen-
from the San Diego, California, community. der, and geographic area.
No exclusion criteria are provided. Standard 3. 1be data are partitioned by age and
test procedures were used. Means and SDs for education.
total unique designs drawn, number of per- 4. Adequate exclusion criteria.
severative errors, and error ratios are reported. 5. Test administration procedures are
specified.
1Norms for children and adolescents are available in 6. Means and SDs for the test scores are
Baron (2004) and Spreen and Strauss (1998). reported.
Consideration regarding use of the study [RFFT.4] Fama, Sullivan, Shear, Cahn-Weiner,
1. Recruitment procedures are not reported. Yesavage, Tinklenberg, and Piefferbaum, 1998
(RFFT Version) (Table A 16.5)
Other comments This study examined the verbal and nonverbal
1. Frequency count of educational levels is fluency performance of patients with Parkin-
reported. son's disease. A total of 51 normal controls were
2. The authors removed 20 outlier persev- included. Control participants ranged in age
erative scores (i.e., scores 2:2 SD) to from 52 to 85 years, with an average of 66.7
approximate a normal distribution in (7.4) years. Participants had an average of 16.4
scores for the normative table. (2.3) years of education and an estimated pre-
3. While no overall gender effects were morbid IQ of 115.6 (5.9). A subset of controls
found, in two of the subgroups (44-55 was recruited from the Palo Alto, California,
years old with ~ 12 years of edpcation community via "word-of-mouth." Additional
and 55-70 years old with 2:16 ~ars of data were selected from an archival database
education) women committed more er- collected in the scope of previous studies (ref-
rors than men. erenced in the article). Participants who were
recruited from Palo Alto were paid, but no in-
[RFFT.3] Demakis and Harrison, 1997
formation on the incentive for participation is
(RFFT Version) (Table A 16.4)
provided for those selected from the archival
This study examined the relationship hftween database. Participants were excluded from the
three fluency tests (one verbal and rnf> non- study if they had a significant history of psy-
verbal). The authors recruited 134 (61 n\ale, 73 chiatric disorder, neurological illness, past or
female) college students. Participants ~ere an present alcohol or substance abuse, or other
average ofl9.1 (2.0) years of age, butth~e is no "serious medical conditions," based on an in-
information regarding educationalleve~ None terview and medical examination. Standard test
of the participants had learning disabilitfes and procedures were used. Means and SDs for total
all had been screened for current or p~t neu- number of unique designs drawn are reported.
rological or psychiatric disease. Stand*d test
procedures were used. Means and SDs fur total Study strengths
number of novel designs are reported. . 1. Large sample size.
Study strengths scribed in terms of age, education, and
1. Large sample size. estimated IQ.
2. The sample composition is well de- 3. Adequate exclusion criteria.
scribed in terms of age and gend.r. 4. Test administration procedures are specified.
4. Test administration procedure$ are reported.
specified.
5. Means and SDs for the test scoies are Considerations regarding use of the study
reported. i 1. Broad age ranges are used, and the data
are not partitioned by age.
Considerations regarding use of the study 2. Recruitment procedures are not re-
1. Educational level of the sample is not ported for half of the sample, but refer-
reported. ' ence is made to another study.
2. Recruitment procedures are npt re- 3. Educational and estimated IQ levels for
ported. the sample are high.
Other comments Other comments

1. The authors found no gender iliffer- 1. Demographic information is reported
ences; thus, the data were not! parti- for 51 participants, but the RFFT norms
tioned by gender. ' are based on 50 participants.
[RFFT.S] Berning, Weed, and Aloia, 1998 Study strengths

(RFFT Version) (Table A16.6) 1. The sample composition is well de-
This study examined the interrater reliability scribed in terms of age, gender, educa-
of the RFFT. Participants were 124 under- tion, geographic area, and recruitment
graduate (34 male, 90 female) students re- procedures.
cruited from introductory psychology courses 2. Test administration procedures are
at the University of Mississippi. Participants specified.
ranged in age from 18 to 31 years, with a 3. Means and SDs for the test scores are
median of 20 years. They were given course reported.
credit for their participation. No exclusion
criteria are mentioned. Standard test proce- Considerations regarding use of the study
dures were used. 1. Sample size is small.
The number of unique designs and per- 2. No exclusion criteria are reported.
severative errors as well as error ratios are
presented for each of the five parts and for the Other comments
whole test (i.e., total scores). 1. The study provides data for RFFT
practice effects.
Study strengths
1. Large sample size. [RFFT.7] Ross, Foard, Hiott, and Vincent,
2. The sample composition is well de- 2003 (RFFT Version) (Table A 16.8)
scribed in terms of age, gender, recruit- RFFT data were collected on 90 college stu-
ment procedures, and geographic area. dents (55% female; 89.8% right-handed) re-
3. Test administration procedures are cruited from introductory psychology courses
specified. at an urban, Midwestern university: 44% were
4. Means and SDs for the test scores are Caucasian, 30% were African American, 9%
reported. were Hispanic, and 6% were designated
"other." Ages ranged 18-69, with a mean of
Considerations regarding use of the study 23.9 (7.3), and mean estimated IQ North
1. Exclusion criteria are not described. American Adult Reading Test (NAART) was
2. Educational levels are not reported. 108.1 (9.2). Exclusion criteria were history of
3. While adequate interrater reliability was neurological disorder, learning disability, or
found, the RFFT results were scored by psychiatric conditions involving medication
the participants themselves. usage. Standard test procedures were used.
Participants received course credit for their
participation. Forty-eight participants were
[RFFT.6] Demakis, 1999 (RFFT Version) retested an average of 35.2 days (range 28--51)
(Table A 16.7) later.
Response consistency and test-retest reliability Means and SDs are reported for numbers
were examined in malingering simulators and of unique designs and perseverations, error
controls for various neuropsychological tests. ratios, as well as five production strategy
Normal controls consisted of 21 undergraduate scores (rotational strategies, enumerative
psychology students (67% female) from a small strategies, total strategic clusters, mean cluster
Midwestern liberal arts college. Participants size, and percent designs in strategies).
were an average of 22.5 (7.99) years of age and
had an average of 13.6 (1.46) years of educa- Study strengths
tion. Standard test procedures were used. 1. Large sample size.
There was an approximately 3-week interval 2. The sample composition is well de-
between the initial testing and retesting. Par- scribed in terms of gender, age, esti-
ticipants were paid $10. No exclusion criteria mated IQ, geographic area, recruitment
are provided. procedures, and ethnicity.
306 PERCEPTUAl ORGANIZATION: VISUOSPATIAl AND TACTILE
3. Means and SDs are reported for stan- a study examining neuropsychological function
dard scores as well as for five additional in older individuals. Exclusion criteria were
strategy scores. dementia, current or past neurological illness
4. Adequate exclusion criteria. or injury, substance abuse, drug use, and his-
tory of psychiatric disorder. Average age was
Consideration regarding the use of the study 69.4 (10.6) years, with a range of 51.5-89.6
1. No stratification by age, although the years, and average education was 14.6 (3.4)
sample would generally represent a years.
narrow age range (college students). Protocols were independently scored by
two advanced clinical psychology graduate
Other comments students who had extensive training in clinical
1. Data provided on practice effects. neuropsychology; however, neither rater had
experience with the Design Fluency task prior
Jones-Gotman/Milner Version to this study. Each rater based his or her
ratings on reading of Jones-Gotman and Mil-
[Design Fluency.1] Boone, Ananth, Philpott, ner's (1977, and Jones-Gotman, personal
Kaur, and Djenderedjian, 1991 Oones-Gotman/ communication, 1984) scoring criteria. Coef-
Milner Version) (Table A16.9) ficients of agreement were 0.64 and 0. 71 for
Design fluency performance was collected on novel designs, 0.64 and 0.24 for nameable
16 controls (nine women, seven men) re- responses, and 0.57 and 0.41 for perseverative
cruited in southern California through news- responses in the free and fixed conditions, re-
paper advertisements and from siblings of spectively, and 0.68 for wrong number of lines
patients with OCD as a part of a study on the in the fixed condition.
cognitive characteristics of OCD. Exclusion Means and SDs for total, novel, nameable,
criteria were history of alcohol or drug abuse, and perseverative responses are provided for
head injury, seizure disorder, cerebral vascu- the free and fixed conditions separately, as
lar disease or stroke, psychiatric disorder, or well as for wrong number of lines in the fixed
any renal, hepatic, or pulmonary disease. condition.
Mean age was 35.8 (13. 7) years, mean edu-
cation was 15.2 (2.8) years, and mean WAIS-R Study strengths
(Satz-Mogel) FSIQ was 109.1 (10.9). 1. Adequate exclusion criteria.
Means and SDs for number of novel de- 2. Adequate sample size.
signs generated are reported. 3. The sample composition is well de-
scribed in terms of gender and educa-
Study strengths tional level.
1. The sample composition is well described 4. Means and SDs for various scores in the
in terms ofage, education, gender, IQ, geo- free and fixed conditions reported.
graphic area, and recruitment strategies.
3. Means and SDs for the test scores are 1. Information provided regarding age, but
reported. data are not stratified by age.
2. No information regarding recruitment
Considerations regarding use of the study strategies or IQ level (although mean
1. Small sample size. scaled score [10.0] for the Vocabulary
2. Data not stratified. subtest of the WAIS-R is reported).
3. Data only reported for the free trial.
[Design Fluency.3] Daigneault, Braun, and
[Design Fluency.2] Woodard, Axelrod, and Henry, Whitaker, 1992 (Jones-Gotman!Milner Version)
1992 (Jones-Gotmarv'Milner Version) (Table A16.1 0) (Table A16.11)
Design fluency scores were gathered on Design fluency data were obtained on 128
80 volunteers (35 men, 45 women) as part of French-speaking participants in Canada as
part of a study investigating the effects of aging Considerations regarding use of the study
on prefrontal lobe skills. Participants were re- 1. Data were obtained on French-speaking
cruited through ads, trade union collaboration, participants in Canada, and thus it is un-
and the help of a large sports center. Exclusion clear whether the are appropriate for
criteria were consumption of more than 24 clinical interpretation among English-
beers, five bottles of wine, or 15 oz of spirits speakingparticipants in the United States.
per week; consumption of cocaine, LSD, or 2. No information regarding IQ (although
psychostimulants; or any neurological or psy- mean scores on the Vocabulary subtest
chiatric consultation, psychoactive medication, of the French-language WAIS analog are
head trauma with hospitalization, or major reported).
surgery (e.g., cardiac). Participants were di- 3. Data are not provided for fixed and free
vided into two age groupings: 20-35 years conditions separately.
(mean=27.71, SD=4.05 years, n=70) and
45-65 years (mean= 56.62, SD = 5.29 years, [Design Fluency.4] Beatty, Jocic, Monson, and
n =58). The younger group contained 38 men Staton, 1993 Oones-Gotman/Milner Version)
and 32 women; they were primarily specialized (Table A16.12)
blue-collar workers, although some specialized Data on the 4-minute fixed trial were obtained
white-collar and unskilled blue-collar profes- on 20 controls (9 male, 11 female) as part of a
sions were represented. The older group con- study on memory and executive function in
tained 30 men and 28 women; slightly more schizophrenia and schizoaffective disorder.
than half were specialized blue-collar workers, Mean age was 34.7 (7.7) years, and mean
but some unskilled blue-collar professions, education was 13.6 (1.4) years. Nineteen of
specialized white-collar occupations, and pro- the participants were working or were stu-
fessional occupations were represented. The dents. Exclusion criteria were history of cen-
mean educational level of the younger group tral nervous system disease or injury, major
was 12.36 (2.09) years, and that of the older medical illness, major psychiatric disorder, or
group was 12.11 (3.63) years. current alcohol or drug abuse (one apparently
Means and SDs for total number of different had a past history of substance abuse).
correct drawings is reported. In addition, The authors indicate that an arc, a straight
means and SDs are provided for perseverative line, and a circle each counted as one line.
errors, defined as: (1) identical drawings; (2) Participants were told that the lines did not
none of the parts of the designs differed, in need to touch and that the drawings did not
angle or dimension, by>100%, or (3) the only have to represent a real object. Means and SDs
difference between two designs was a rotation, are reported for number of designs generated
a mirror representation, or a change in the and number of rule violations (i.e., drawings
global dimension. Perseverations were scored containing fewer or more than four lines).
by two assistants, and the average ratings are
reported. No significant differences in perfor- Study strengths
mance between the two groups were detected. 1. Procedure well specified.
2. Means and SDs reported for number of
Study strengths designs and number of rule violations.
1. Good exclusion criteria. 3. The sample composition is well de-
2. Large overall sample size, and each of the scribed in terms of age, education, and
two age groupings has >50 participants. gender.
3. The sample composition is well de- 4. Good exclusion criteria.
scribed in terms of educational level,
gender, occupation, geographic area, and Considerations regarding use of the study
recruitment procedures. 1. Small sample size.
4. Means and SDs for total number of 2. No information regarding IQ or re-
correct designs and perseverations are cruitment strategy.
provided. 3. Data available only for the fixed trial.
[Design Fluency.S] Varney, Roberts, Struchen, screened for current or past neurological or
Hanson, Franzen, and Connell, 1996 psychiatric disease. Means and SDs for the
(Jones-Gotman/Mi/ner Version) (Table At6.13) free and fixed conditions and total score are
Data on the 5-minute free condition were reported.
collected on 87 volunteers (28 males~ 59 fe-
males) with no history of neurolo3ical or Study strengths
psychiatric illness, loss of consciousnes• due to 1. Relatively large sample size.
head trauma, or severe febrile illnes!#. Mean 2. n.e sample composition is well de-
age was 27.7 (13.1) years, with a rang4 of 18- scribed in terms of age and gender.
77, and mean education was 14.4 (2) years, 3. Adequate exclusion criteria.
with a range of 12-21. · 4. Test administration procedures are
Means and SDs for number of nqvel de- specified.
signs produced are reported. n.e agrtement 5. Means and SDs for the test scores are
rate between two independent ratqrs was reported.
90%. Ninety-five percent of particip~ts pro-
duced one or fewer nameable desigqs, 95% Considerations regarding use of the study
made three or fewer repeated desi~s. and 1. Educational level of the sample is not
scribbling errors were rare. : reported.
No significant relationships betwee~ num- 2. Recruitment procedures are not re-
ber of designs generated and age or edUcation ported.
and no significant differences in performance
between men and women were absented. Other comments
1. n.e authors found no gender differ-
Study strengths ences; thus, the data were not parti-
1. Large overall sample size. tioned by gender.
2. n.e sample composition is w'U de-
[Design Fluency.7] Carter, Shore, Hamadek,
scribed in terms of age, educati~n. and
and Kubu, 1998 (Jones-Gotman/Milner Version)
gender. '
(Table A 16.15)
4. Means and SDs for the test scores are Sixty-six participants (19 males, 47 females),
reported. primarily undergraduates in Ontario, Canada,
were tested. Inclusion criteria were age 18--60,
Considerations regarding use of the study right-handed, English as first or main lan-
1. Data are not stratified. guage, FSIQ > 79, and no significant neuro-
2. Data available only for the free condition. logical, systemic, or psychiatric illness. Mean
3. No information regarding IQ, ethnicity, age was 25.06 (7.83) years, with a range of 19-
geographic area, or recruitment 56; mean education was 15.21 (1.60) years;
procedures. and mean WAIS-R FSIQ estimate was 100.85
(11.07), with a range of 81-124.
[Design Fluency.6] Demakis and Harrisoa, Participants were administered the free and
1997 (Jones-Gotman!Milner Version) fixed conditions per the instructions of Jones-
(Table A 16.14) Cotman (in Spreen & Strauss, 1998). Means
This study examined the relationship between and SDs for total designs, novel output, per-
three fluency tests (one verbal and t\\J> non- severative errors, and nameable errors in the
verbal). The authors recruited 134 (61 male, free and fixed conditions separately are re-
73 female) college students, who a~ raged ported as well as for incorrect number of lines
19.1 (2.0) years of age. n.e order of ll!st ad- in the fixed condition.
ministration was counterbalanced so th•t each
test was administered first, second, ana third Study strengths
the same number of times. None of the par- 1. n.e sample composition is well described
ticipants had learning disabilities or h~ been in terms of age, education, gender,
apparent recruitment strategy, FSIQ, Considerations regarding use of the study

language, and geographic area. 1. Data are not stratified, although it can be
2. Adequate exclusion criteria. assumed that the sample had a relatively
3. Means and SDs using the Jones-Gotman narrow age range (college students).
(in Spreen & Strauss, 1998) scoring cri- 2. No information regarding IQ.
teria are reported. 3. Data are not provided for the free and
fixed conditions separately.
1. Data are not stratified by age, although a Other comments
relatively narrow age range is assumed. 1. Data regarding interrater reliability and
2. High educational level but average IQ. test-retest reliability are provided.
Other comments [Design Fluency.9] Mataix-Cols, Barrios,

1. Information regarding interrater reli- Sanchez-Turet, Vallejo, Junque, 1999
ability is reported. (Jones-Gotman/Milner Version)
(Table A 16.17)
[Design Fluency.&] Harter, Hart, and Harter, 1999 Data were collected on 27 (23 women, 4 men)
Oones-Gotman/Milner Version) (Table A16.16) undergraduates in Barcelona, Spain, as part of
Design fluency data were obtained on 64 a study of the impact of subclinical obsessive-
college students (91% female) apparently en- compulsive symptoms on design fluency per-
rolled at Texas Tech University as part of a formance. Mean age was 19.1 (1.3) years, and
study on expanded scoring criteria for the one subject was left-handed. An exclusion
Jones-Cotman/Milner design fluency version. criterion was history of a psychiatric disorder.
Ages ranged 17-58 years, with an average of Means and SDs for the free and fixed
20 years. Eighty-one percent of participants conditions are reported.
reported no history of neurological disorder;
11 reported history of a blow to the head re- Study strengths
sulting in loss of consciousness, although only 1. The sample composition is well de-
one subject alleged continuing difficulties re- scribed in terms of age, educational
lated to the injury (headaches and reading status, gender, geographic area, and re-
comprehension problems), and three reported cruitment strategy (ethnicity assumed
recent signs of possible neurological dysfunc- Spanish).
tion, including seizures, dizziness, fainting, 2. Means and SDs for the test scores are
and memory difficulties. Eighty-six percent of reported.
the sample were right-handed. 3. Minimally adequate exclusion criteria.
Means and SDs for number of novel de-
signs, scribbles, and perseverations as well as Considerations regarding use of the study
additional scores reflecting complexity, num- 1. Small sample size.
ber of concrete responses, and drawing quality 2. No information regarding IQ.
are provided both for the sample as a whole 3. Data were collected in Spain, which may
and with the 12 students with possible neu- limit their usefulness for clinical com-
rological dysfunction excluded. parison in the United States.
Study strengths [Design Fluency.1 0] Abrahams, Leigh, Harvey,

1. Large sample size. Vythelingum, Grise, and Goldstein, 2000
2. The sample composition is well de~ (Jones-Gotman/Milner Version) (Table A16.18)
scribed in terms of age, education, gen- Design Fluency performance was measured in
der, geographic area, handedness, and 25 controls (16 males, 9 females) as part of a
recruitment procedure. study of cognition in ALS. Mean age was 55.8
3. Means and SDs are reported for several (11.6) years, mean education was 14.1 (3.1)
scores. years, and mean NART FSIQ was 114.6 (9.9).
Controls were recruited in London lfrom a 2. Adequate exclusion criteria.

local volunteer organization, a local e<ltcation 3. Means and SDs for the test scores are
center, and friends of ALS patients . .t\11 were reported.
right-handed. Exclusion criteria were: neuro-
logical disorder or significant head inj~ry. Considerations regarding use of the study
Means and SDs for number of designs 1. Data are not stratified by age.
produced in the free and fixed conditipns are 2. Small sample size.
reported, as well as % perseverative responses 3. All male sample.
and % unacceptable responses. · 4. Data are provided only for the free
condition.
Study strengths j
1. The sample composition is Wfll de-
scribed in terms of age, education,
CONCLUSIONS
FSIQ, handedness, gender, georaphic
area, and recruitment strategies. • The relatively few studies conducted on design
2. Minimally adequate recruitment strategy. fluency tests have generally shown these tasks
3. Means and SDs for the test scotes are to be sensitive to frontal lobe impairment. The
reported. · two different tests selected for review in this
chapter, the RFFT and the Jones-Gotman/
Considerations regarding use of the stody Milner design fluency task, have unique ad-
1. Small sample size. · vantages and disadvantages. The RFFT is a
2. Data were collected in England,: which structured test, which is relatively easy to ad-
may limit their usefulness for ~linical minister and score. This test demonstrates
comparison in the United States.~ adequate inter-rater and test-retest reliability.
3. Data are not stratified. However, criticisms have been raised that due
to its structured format the RFFT may not
[Design Fluency. 11] Boone, Swerdloff, ¥iller, measure initiation and organization skills in
Geschwind, Razanl, Lee, Gaw Gonzalo, tladdal, patients with frontal dysfunction as well as the
Rankin, Lu, and Paul, 2001 ()ones-Gotma,J!Mil- Jones-Gotman/Milner. On the other hand, the
ner Version) (Table A 16.19) complexity of designs on production rate and
Design Fluency performance was asseSsed in questions regarding what constitutes a per-
22 male controls as part of a study op neu- severative error vs. a unique design make
ropsychological function in adult Klinefelter's clinical use of the Jones-Gotman/Milner ver-
syndrome. Participants were recruitecl from sion somewhat problematic.
newspaper and radio ads and flyers in the A review of the literature for design fluency
southern California area and paid foj- their tests found relatively sparse normative studies.
participation. Exclusion criteria were history For the RFFT, no studies reported gender
of learning disability, major psychiatric{ disor- effects but effects of age and education were
der, substance abuse, or neurological! disor- documented. In general, unique design pro-
der. All participants were fluent in E',tglish. duction decreases and perseverative error
Mean age was 34.32 (14.81) years, mE:ian ed- rates increase with advanced age, particularly
ucation was 13.36 (2.15) years, mean WAIS-R in individuals with less than a college educa-
(Satz-Mogel) VIQ was 106.46 (17.01), and tion. Additionally, one study (Ruff et al., 1987)
mean PIQ was 107.46 (16.58). found that total design production increased
Means and SDs for number of no~el de- with higher Performance IQ scores. While
signs in the free condition are reporte4. normative data and conversion tables have
been created for a wide age range in the
Study strengths RFFT manual (Ruff, 1996), it is clear that
1. Information regarding age, education, additional normative studies using larger
gender, IQ, geographic area, l~age, sample sizes are needed. Significant practice
and recruitment procedures is reported. effects have been reported for the RFFT,
particularly when the test-retest interval is duction with advancing age (Mittenberg et al.,
5 weeks or less. 1989). Similarly, some studies have reported
For the Jones-Gotman!Milner version, no differences in design production between
mixed results have been reported regarding males and females (Mataix-Cols et al., 1999;
the influence of age, education, and gender. Varney et al., 1996), while others have re-
Most studies report no effects of education ported significantly greater production of no-
(Harter et al., 1999; Mittenberg et al., 1989; vel designs by males relative to females
Varney et al., 1996). Some studies have found (Harter et al., 1999). A minimal relationship
no age effects (Daigneault et al., 1992; Harter between Performance IQ and design pro-
et al., 1999; Varney et al., 1996), while others duction has also been reported (Harter et al.,
have reported decreased design fluency pro- 1999). 2
'Meta-analyses were not performed as sufficient amounts

of homogeneous data were not available for either version
of the test.
17
Tactual Performance Test
BRIEF HISTORY OF THE TESJ scores regarding lateralization and localiza-

tion, and test performance in specific patient
The Tactual Performance Test (TPT) ij based populations; the interested reader is referred
on the Sequin-Goddard Formboard. ~It was to this publication.
incorporated by Halstead (1947) in his triginal The TPT involves several different abilities,
neuropsychological battery and subse~uently including tactile perceptual skills, propriospa-
employed by Reitan (1979) in his expa4sion of tial ability, tactile/spatial memory, and visual
the Halstead Battery. Participants are blind- constructional skills. Deficits in TPT perfor-
folded, and 10 blocks of differing shapds and a mance have been associated with frontal le-
matching 10-hole formboard are pl~f!ed in sions (Halstead, 1947; Shure & Halstead,
front of them. They are instructed td insert 1958), posterior dysfunction (Reitan, 1964),
the blocks in the board as quickly as they can and right hemisphere disturbance (Reitan,
with their dominant hand only. Follo*g this 1964; Schreiber et al., 1976). Some reports
trial, they are required to place the blocks have indicated that increased right-hand time
with their nondominant hand and th$1. both is associated with left hemisphere damage and
hands together. The blocks and formboard are increased left-hand time is tied to right hemi-
removed, followed by the blindfold, ~d the sphere dysfunction (Dodrill, 1978a; Reitan,
participants are asked to draw the fonnboard 1964), but no consistent lateralization findings
and as many of the block shapes as ~y can have been documented for the memory and
remember in their relative location. Specific localization scores (Heilbronner et al., 1991;
instructions are provided in Reitan's •(1979) Thompson & Parsons, 1985). Of interest,
Manual for Administration of Neuropsycho- normal right-handed males show a left-hand
logical Test Batteries for Adults and Clfildren, advantage on time and memory scores when
Reitan and Wolfson's (1985) The Htf.stead- the left hand is tested first, suggesting an en-
Reitan Neuropsychological Test Batteiy, and hanced role of the right hemisphere in TPT
Swiercinky's (1978) Manual for the· Adult performance (Jenkins & Parsons, 1989).
Neuropsychological Evaluation. Th~pson TPT performance has typically been re-
and Parsons (1985) summarize the literature ported in terms of time to complete the task
through 1983 on the status of the 1PT in with the preferred hand, nonpreferred hand,
terms of test construction, effect of $Ubject both hands, and total time, as well as number
variables on performance, interpretation of of blocks correctly recalled and located.
312
TACTUAl PERFORMANCE TEST 313
Reitan (1979) recommends Halstead's (1947) Heaton, 1991) and that this percentage in-
cutoff of 15.6 minutes for total time, five or creases with age (Goldstein & Braun, 1974;
fewer blocks recalled, and four or fewer Price et al., 1980; Thompson et al., 1987).
blocks located. However, given the significant Criticisms have emerged regarding the lack
association of TPT scores with age, IQ, and of standardization of some aspects of TPT
possibly education and gender, a single cutoff administration, which could substantially in-
would not appear to be appropriate, particu- fluence the obtained test scores. Snow (1987)
larly in older participants, as documented notes that Reitan recommends discontinua-
by the following reports. Price et al. (1980) tion of a time trial at 15 minutes if the patient
found that using Halstead's (1947) cutoffs for is discouraged and not close to finishing the
total time, memory, and localization, 89.8%, task but continuing if the patient is near
12.2%, and 77.6%, respectively, of a healthy, completion, while other clinicians (e.g., Rus-
elderly sample with a mean age of 71.9 were sell et al., 1970) routinely stop at 10 minutes.
misclassified as brain-damaged. Similarly, Snow (1987) points out that differing amounts
Ernst (1987) documented in his sample aged of exposure time to the blocks could influence
65-75 years misclassification rates of 77%, memory and localization scores. Snow (1987)
36%, and 89% for total, memory, and locali- also observes that no precise scoring criteria
zation scores, respectively. Bak and Greene have been developed for the memory and lo-
(1980) reported that 40% of their healthy calization scores; for example, some clinicians
sample aged 50-62 were misclassified on total allow a four- or five-point star, while others
time. The mean scores of this group fell be- give credit for a six-point star. Chavez et al.
low the cutoffs for memory and localization, (1982) note that some clinicians follow a
and the mean scores of an older group (67-86) consistent order in placing the blocks before
did not surpass the cutoff for any of the TPT blindfolded participants while other examin-
measures. Cauthen (1978a), studying a ers randomly arrange the blocks. They also
broader age range, documented 22%, 15%, report that the order of block presentation did
and 53% misidentification rates for total time, not affect the time trials with their normal
memory, and localization in his sample aged college student population but that a stan-
20-60 years; only 2% of 20-year-olds were dardized block presentation format was asso-
misclassified on total time, while 63% of those ciated with higher memory scores, and a trend
50-60 years old were miscategorized. Dodrill was noted toward higher localization scores.
(1987) documented 21.7%, 5%, and 39.2% Kupke (1983) observed better total time
misclassification rates for total time, memory, scores in his college sample using a portable
and localization, respectively, in a young version of the TPT rather than the standard
control sample. Chavez et al. (1982) noted equipment.
that the mean localization score (4.87) of their In addition to the above concerns regarding
male college students fell below the Reitan TPT administration and interpretation, Lezak
cutoff of 5. Bornstein and colleagues (1987b) et al. (2004) point out that the TPT is very
emphasized that cutoff scores may be useful frustrating and stressful for patients due to its
but only if considered in the context of other length and difficulty and question its practical
neuropsychological information obtained in a utility. Also, because moderately to severely im-
test battery and if age, education, and other paired individuals fail the test so completely,
appropriate adjustments are made. little information can be gleaned regarding
Reitan and Wolfson (1985) and others extent of impairment in these participants
(Golden et al., 1981b; Jarvis & Barth, 1984) (Russell, 1985). In an attempt to shorten and
have suggested that nondominant hand per- simplify the test, Lezak et al. (2004) have re-
formance should be 30% faster than dominant commended substituting the six-block child's
hand performance, although several investiga- TPT (TPT-6) for the 10-block adult version.
tors have documented that a sizable percentage Russell (1985) reported that TPT-6 scores
of normals show superior dominant hand were highly correlated with the 10-block TPT
performance (Cauthen, 1978a; Thompson & and that the TPT-6 successfully discriminated
controls from brain-damaged participants. colleagues have also addressed additional psy-
Also, test administration was shortened by chometric properties of the TPT, reporting
two-thirds. Unfortunately, although tllese re- internal consistency reliabilities of 0.6588 for
ports regarding the utility of substituling the the preferred hand, 0.8715 for the non-
six-block TPT for the 10-block version in preferred hand, 0.8199 for both hands, 0.8953
adults are promising, the only "normative" for total time, 0.69 for memory, and 0. 79 for
data on the six-block TPT in adults are Clark localization (Charter et al., 2000; Charter,
and Klonoffs (1988) 79 coronary byp.ss sur- 2001c). Reliabilities for the difference scores
gery patients and Russsell's (1985) 1.9 "con- across trials have been poor (i.e., 0.0892
trols" referred for suspected but unconfirmed for preferred hand minus nonpreferred hand,
brain dysfunction. 0.0838 for both hands minus preferred
In spite of the above-mentioned liabilities, hand, and 0.2065 for both hands minus non-
the TPT enjoys some popularity in clinical preferred hand), although the blocks per min-
assessment. It was found to be one of the ute reliabilities have been somewhat higher
HRB tests which are most useful in the as- (i.e., 0.3122 for nonpreferred hand minus
sessment of brain impairment (total time and preferred hand, 0.6387 for both hands minus
location scores; Mutchnick et al., 1991) and to preferred hand, and 0.5058 for both hands
be associated with daily living skills in geriatric minus nonpreferred hand) (Charter, 2001b).
patients (Searight et al., 1989), driving ability The test blocks are not of comparable dif-
in head-injured patients (Brooke et al., 1992), ficulty; for example, on the Memory trial, the
and failure to return to premorbid educational rhombus is the most difficult to recall for both
or vocational levels in brain tumor survivors patients and controls while the circle is the
(Hochberg & Slotnick, 1980). It was the easiest (Charter & Dutra, 2001a,b). Three
most sensitive measure in the identifi(!Jltion of blocks from the Memory trial (diamond, oval,
brain dysfunction in blind patients (:Bigler & and rhombus) and one block from the locali-
Tucker, 1981). In addition, TPT performance zation trial (oval) obtained discrimination in-
is lower in verbal (Davis et al., 1989) and dexes <0.3, suggesting that these items do not
nonverbal (Hamadek & Rourke, 1994) learn- discriminate well between good and poor
ing-disabled individuals, patients with obses- performers. In addition, on the block place-
siv~ompulsive disorder (Insel et al., 1983), ment trials, blocks at the top of the board are
inmates referred for psychiatric treatment more difficult to correctly position than blocks
(Young & Justice, 1998), and alcoholics on the bottom (Charter, 2000b).
(Gurling et al., 1991; Hesselbrock et al., 1985; Clark and Klonoff (1988) examined the in-
Hochla et al., 1982; Loberg, 1980), especially ternal consistency, test-retest reliability, and
left-hand performance (Jenkins & ~arsons, construct validity of the TPT-6 and concluded
1981). Of interest, TPT scores were Dot im- that it was reliable and had construct validity
pacted by chronic marijuana use in medical characteristics comparable to those of the 10-
students (Rochford et al., 1977). Further dis- block TPT.
cussion of TPT interpretation is proviiled by
Bradford (1992).
RELATIONSHIP BETWEEN TPT
Psychometric Properties of the TPT PERFORMANCE AND DEMOGRAPHIC
FACTORS
Initial investigations of interrater reliability
revealed moderate agreement for IJ)emory Age has been consistently related to TPT per-
scores (71 %-76.3%) and poor correspohdence formance in normal individuals (Bak & Greene,
for localization scores (56. 7%-63.8%; ~rtin & 1980; Cauthen, 1978a; Elias et al., 1993;
Greene, 1978); however, more recentlj' Char- Fromm-Auch & Yeudall, 1983; Heaton et al.,
ter et al. (1998) reported interscorer reli- 1986, 1991; Moore et al., 1984; Reed & Reitan,
abilities of 0.9810 for the memory trial and 1962, 1963b; Reitan, 1955d; Yeudall et al.,
0.9773 for the localization trial. Ch~r and 1987) as well as brain-damaged (Fitzhugh et al.,
1964; Prigatano & Parsons, 1976; Reed & for by gender was minimal (Heaton et al.,
Reitan, 1962; Vega & Parsons, 1967), medical 1991). Ernst (1987) documented better male
(Alekoumbides et al., 1987; Reed & Reitan, performance on memory, localization, and all
1963b), and psychiatric (Alekoumbides et al., time scores except dominant hand, while Elias
1987; Ernst et al., 1987; Prigatano & Parsons, et al. (1993) also found a male superiority in
1976) patients, with older age associated with memory scores. Conversely, other reports
poorer scores. To our knowledge, no pub- have suggested that women score better than
lished study has failed to document a corre- men on memory (Fabian et al., 1981), locali-
lation between TPT and age. zation (Chavez et al., 1982), or both scores
Higher IQ also appears to be associated with (Gordon & O'Dell, 1983), although Gordon
better TPT performance in various groups, and O'Dell (1983) comment that the gender
including psychiatric (Warner et al., 1987) and differences they observed were not of "prac-
mentally retarded individuals (Matthews, tical significance."
1974). IQ levels have also been associated with In the two studies addressing the relationship
TPT scores in normals (Cauthen, 1978a; Wiens between lateral preference and test perfor-
& Matarrazzo, 1977), with PIQ showing more mance, no differences in test scores between
of a correlation than VIQ (Wiens & Matarrazzo, right- and left-banders were documented
1977; Yendall et al., 1987). (Gregory & Paul, 1980; 'n.ompson et al., 1987).
The data on the relationship of education Arnold et al. (1994) documented a signifi-
and gender with TPT indices have been more cant effect of acculturation (i.e., higher per-
equivocal; taken as a whole, the available lit- formance associated with higher acculturation
erature does not demonstrate a convincing level) on TPT dominant hand, nondominant
relationship between these variables and TPT hand, and total time, with no effect of accul-
performance. Ernst (1987), Finlayson et al. turation on localization and memory scores.
(1977), and Yendall et al. (1987), failed to de-
tect a relationship between educational level
and TPT scores in normals, while Heaton et al.
(1986, 1991) observed significant differences
NORMATIVE REPORTS
across educational levels only for the memory
score. No association has been detected be- Our review of the literature located nine TPT
tween education and TPT performance in normative reports for adults published since
brain-damaged patients (Finlayson et al., 1977; 1965 (Cauthen, 1978a; Dodrill, 1987; Ernst,
Prigatano & Parsons, 1976; Vega & Parsons, 1987; Fromm-Auch & Yendall, 1983; Harley
1967; Warner et al., 1987), although Vega and et al., 1980; Heaton et al., 1991; Pauker, 1980;
Parsons observed an effect of education on Schear, 1986; Yendall et al., 1987), as well as
memory and timed scores in a medical patient the original Halstead (1947) and Reitan (1955b,
sample and Alekoumbides et al. (1987) ap- 1959) normative data and three interpretive
parently detected a relationship between ed- guides for the HRB (Golden et al., 1981b;
ucation and TPT scores in a medical and Reitan & Wolfson, 1985; Russell et al., 1970).
psychiatric sample. Hundreds of other studies have also reported
Most publications suggest that there is no control subject data, and we have included
difference between males and females on TPT discussion of 13 of those investigations which
performance in normal participants (Dodrill, involved some unique feature, such as large
1979; Elias et al., 1990; Filskov & Catanese, sample size (~100), retest data, elderly pop-
1986; Fromm-Auch & Yendall, 1983; King ulation, non-English-speaking sample, and use
et al., 1978; Moore et al., 1984; Pauker, 1980; of a shorter version of the test (Alekoumbides
Thompson et al., 1987; Yendall et al., 1987) et al., 1987; Anthony et al., 1980; Bak &
and in patients (Dodrill, 1979). However, Greene, 1980; Bornstein et al., 1987a; Clark &
Heaton et al. (1986) reported that males out- Klonoff, 1988; El-Sheikh et al., 1987; Heaton
performed females on total time, although the et al., 1986; Klove & Lochen, cited in Klove,
amount of variance in test scores accounted 1974; Matarazzo et al., 1974; Moore et al.,
1984; Russell, 1985; Thompson & Heaton, Sample Composition Description

1991; Wiens & Matarazzo, 1977). (We also As discussed previously, information regarding
found a study in which military personnel medical and psychiatric exclusion criteria is
were administered the TPT during various important; it is unclear if geographic recruit-
field maneuvers of differing intensity, to ment region, socioeconomic status or occupa-
evaluate the effect of environmental stress on tion, ethnicity, handedness, and recruitment
TPT performance [Arima, 1965]. However, procedures are relevant. Until determined, it
given the questionable relevance of this inis best that this information be provided.
formation for the typical neuropsychological
testing session, we decided not to review this Age Group Intervals
publication.) Given the association between age and TPT
Russell and Starkey (1993) developed the performance, information regarding the age of
Halstead-Russell Neuropsychological Evalua- the normative sample is critical and normative
tion System (HRNES), which includes 22 data should be presented by age intervals.
tests. In the context of this system, individual
performance is compared to that of 576 brain- Reporting of JQ Levels
damaged participants and 200 participants Given the relationship between TPT perfor-
who were initially suspected of having brain mance and IQ, data should be presented by
damage but had negative neurological find- IQ intervals, or at least information regarding
ings. Data were partitioned into seven age intellectual levels should be provided.
groups and three educational!IQ levels. This
study will not be reviewed in this chapter Reporting of Educational Levels
because the "normal" group consisted of the Given the possible, although minor, association
V.A. patients who presented with symptoms between educational level and TPT scores, it is
requiring neuropsychological evaluation. For preferable that information regarding highest
further discussion of the HRNES, see Lezak educational level completed be reported.
et al. (2004).
Of note, few relevant manuscripts have Reporting of Gender Distribution
emerged since the 1980s, perhaps due either Given the possible, although minor, associa-
to the publication of Heaton et al.'s (1991) tion between gender and TPT scores, it is
comprehensive normative tables or to the in- preferable that information regarding gender
creasing use in research and clinical practice be reported.
of flexible neuropsychological test protocols
which include newer tasks rather than tradi-
tional fixed neuropsychological batteries.
To adequately evaluate the TPT normative Data Reporting
reports, seven key criterion variables were
Means and standard deviations, and prefera-
deemed critical. The first six of these relate to
bly ranges, for time in seconds or minutes on
subject variables, and the remaining refers to
the TPT for the dominant and nondominant
procedural issues.
hands separately and together and the total
across all three trials should be reported, as
terion variables were as follows.
well as means and SDs for memory and lo-
calization scores.
Subject Variables
Sample Size SUMMARY OF THE STATUS

OF THE NORMS
As discussed in previous chapters, a minimum
of 50 participants per grouping interval is All but 10 data sets had total sample sizes lar-
optimal. ger than 100 (Bak & Greene, 1980; Bornstein
TACTUAL PERFORMANCE TEST 317
et al., 1987a; Clark & Klonoff, 1988; El-Sheikh Matarazzo, 1977; Yendall et al., 1987), and
et al., 1987; Halstead, 1947; Klove & Lochen, information regarding ethnicity was presented
cited in Klove, 1974; Reitan, 1955b, 1959; in four data sets (Alekoumbides et al., 1987;
Russell, 1985; Wiens & Matarazzo, 1977). Only Dodrill, 1987; Halstead, 1947; Russell, 1985).
three publications consistently had at least 50 Exclusion criteria were judged to be adequate
participants in individual subject groupings in only 11 publications (Anthony et al., 1980;
(Ernst, 1987; Heaton et al., 1986; Schear, 1984), Bak & Greene, 1980; Bornstein et al., 1987a;
although some reports had some subgroups Dodrill, 1987; Fromm-Auch & Yendall, 1983;
which met this criterion (Fromm-Auch & Heaton et al., 1991; Moore et al., 1984;
Yendall, 1983; Harley et al., 1980; Pauker, 1980; Pauker, 1980; Thompson & Heaton, 1991;
Yendall et al., 1987). Wiens & Matarazzo, 1977; Yendall et al.,
Approximately half of the studies summa- 1987). Geographic recruitment areas were
rized in this chapter present TPT data ac- specified in all but two publications (Bornstein
cording to circumscribed age ranges (Bak & et al., 1987a; Dodrill, 1987). Fourteen data
Greene, 1980; Cauthen, 1978a; Ernst, 1987; sets were obtained in the United States
Fromm-Auch & Yendall, 1983; Harley et al., (Alekoumbides et al., 1987; Anthony et al.,
1980; Heaton et al., 1986, 1991; Moore et al., 1980; Bak & Greene, 1980; Halstead, 1947;
1984; Pauker, 1980; Schear, 1984; Wiens & Harley et al., 1980; Heaton et al., 1986, 1991;
Matarazzo, 1977; Yendall et al., 1987). Infor- Klove & Lochen, cited in Klove, 1974; Reitan,
mation on IQ levels is reported in all but se- 1955b, 1959; Russell, 1985; Schear, 1986;
ven studies (Barrett et al., 2001; El-Sheikh Thompson & Heaton, 1991; Wiens & Matar-
et al., 1987; Ernst, 1987; Heaton et al., 1986; azzo, 1977), six in Canada (Cauthen, 1978a;
Klove & Lochen, cited in Klove, 1974; Russell, Clark & Klonoff, 1988; Fromm-Auch &
1985; Shear, 1984), and two reports presented Yendall, 1983; Moore et al., 1984; Pauker,
TPT data for age-by-IQ groupings (Cauthen, 1980; Yendall et al., 1987), one in Norway
1978a; Pauker, 1980). Similarly, educational (Klove & Lochen, cited in Klove,1974), one in
level was also indicated in all but four studies Egypt (El-Sheikh et al., 1987), and one in
(Bornstein et al., 1987a; Cauthen, 1978a; Australia (Ernst, 1987).
Moore et al., 1984; Pauker, 1980), and Heaton Total mean time in seconds or minutes and
et al. (1986, 1991) organized data by educa- SDs to complete the task across the three time
tional levels. Information on gender compo- trials as well as means and SDs for Memory
sition of the samples was available in all but and Localization were reported in 19 datasets
four reports (Anthony et al., 1980; Harley (Alekoumbides et al., 1987; Anthony et al.,
et al., 1980; Klove & Lochen, cited in Klove, 1980; Bak & Greene, 1980; Bornstein et al.,
1974; Thompson & Heaton, 1991); five data 1987a; Cauthen, 1978a; Dodrill, 1987; El-
sets included only male (Clark & Klonoff, Sheikh et al., 1987; Ernst, 1987; Fromm-
1988; Schear, 1984; Wiens & Matarazzo, Auch & Yendall, 1983; Harley et al., 1980;
1977) or nearly all male (Alekoumbides et al., Heaton et al., 1986, 1991; Klove & Lochen,
1987; Russell, 1985) populations. Ernst (1987) cited in Klove, 1974; Pauker, 1980; Russell,
and Heaton et al. (1991) presented data sep- 1985; Schear, 1984; Thompson & Heaton,
arately for male and females. 1991; Wiens & Matarazzo, 1977; Yeudall
Information on other subject variables was et al., 1987). Twelve reports provided data for
provided less frequently; data on handedness the preferred and nonpreferred hands sepa-
was indicated in eight studies (Bak & Greene, rately and together (Alekoumbides et al.,1987;
1980; Cauthen, 1978a; Clark & Klonoff, 1988; Bak & Greene, 1980; Cauthen, 1978a; El-
Dodrill, 1987; Fromm-Auch & Yendall, 1983; Sheikh et al., 1987; Ernst, 1987; Fromm-Auch
Russell, 1985; Schear, 1984; Yendall et al., & Yendall, 1983; Heaton et al., 1991; Russell,
1987), occupation or socioeconomic status was 1985; Schear, 1984; Thompson & Heaton,
described in five reports (Alekoumbides et al., 1991; Wiens & Matarazzo, 1977; Yendall
1987; Dodrill, 1987; Halstead, 1947; Wiens & et al., 1987). Three studies reported score
ranges (Fromm-Auch & Yendall, 1983; Hal- history of brain injury. The eight participants
stead, 1947; Harley et al., 1980). who carried diagnoses of mild psychoneurosis
Several publications reported supplemen- were male soldiers aged 22--38 (mean= 29.6);
tary TPT scores such as T-score equivalents some had had combat experience but none had
(Harley et al., 1980); T scores corrected for a history of head injury. The last six participants
age, education, and gender (Heaton et al., were aged 27-39 and included a depressed
1991); IQ-equivalent scores (Dodrill, 1987); military prisoner facing execution, a severely
test-retest data (Bomstein et al., 1987a; depressed female with suicidal and homicidal
El-Sheikh et al., 1987; Matarazzo et al., 1974); impulses tested prior to lobotomy, and a sui-
and data for a six-block version (Clark & cidaVhomicidal female and a suicidal male tes-
Klonoff, 1988; Russell, 1985). In addition, ted pre- and post-lobotomy. Educational level
Harley et al. (1980) provide the perce~tage of ranged from 7-18 years, and the following oc-
patients correctly placing blocks, and Harley cupations were represented: artist, entertainer,
et al. (1980), Ernst (1987), and Schear (1984) farmer, housewife, semiskilled and unskilled
report means and SDs for number of blocks laborers, professional, secretary, teacher, tech-
correctly placed by each hand, both hands, nician, trade, and student. Ethnic background
and total. Alekoumbides et al. (1987) present included American, Balkan, English, French,
means and SDs for number of blocks placed German, Irish, Polish, Scandinavian, and Scot-
per minute, and Harley et al. (1980) report tish. IQ levels ranged from 7(}..140.
mean and SD for time to place each block for Mean total time for the three trials and
the three trials. mean scores for memory and localization are
The text of study descriptions contains ref- reported for the total group and each control
erences to the corresponding tables identified subgroup, as well as the individual scores for
by number in Appendix 17. Table A17.1, the each subject. The TPT criterion scores used in
locator table, summarizes information pro- calculating the Impairment Index were >15.6
vided in the studies described in this chapter. 1 minutes for total time, fewer than six blocks
recalled, and fewer than five blocks located.
Study strengths
1. Information provided regarding IQ, edu-
Given that the TPT has typically been used cation, occupation, age, gender, ethnicity,
within the context of the HRB, the aaistead and geographic recruitment area.
(1947) and Reitan (1955b, 1959) data and in-
terpretation formats will be reported first, Considerations regarding use of the study
followed by a summary of the other inter- 1. Small sample size including use of two
pretation formats. Then, the normative pub- participants twice.
lications and control groups from clinical 2. Inclusion of participants with psychiatric
comparison studies will be reviewed in as- diagnoses and post-lobotomy.
cending chronological order. 3. No reporting of SDs.
4. Undifferentiated age range.
Original Studies [TPT.2] Reitan, 1955b, 1959 (Table A17.3)
[TPT.1] Halstead, 1947 (Table A17.2) The author obtained TPT scores on 50
participants in Indiana who had apparently
The author obtained data on 28 control partic-
been referred for neuropsychological testing
ipants in Chicago, half of whom had psyChiatric
and "who had received neurological exami-
diagnoses. The 14 participants without psychi-
nations before testing and showed no signs
atric diagnoses were nine male and five· female
or symptoms of cerebral damage or dysfunc-
civilians aged 15-50 (mean= 25.9), without
tion. . . . None . . . had positive anamnes-
'Norms for children are available in Baron (2004) and tic findings" (p. 29); but participants hospitalized
Spreen and Strauss (1998). with paraplegia and neurosis were included.
The sample included 35 men and 15 women, data regarding the normative sample on which
and mean age and educational level were these guidelines were developed.
32.36 (10.78) and 11.58 (2.85), respectively.
Mean WAIS VIQ, PIQ, and FSIQ were Considerations regarding use of the study
110.82 (14.46), 112.18 (14.23), and 112.64 The authors argue that these norms were
(14.28), respectively. meant as "general guidelines" and that "exact
percentile ranks corresponding with each
Study strengths possible score are hardly necessary because
1. Information regarding IQ, education, the other methods of inference are used to
gender, age, and geographic recruitment supplement normative data in clinical inter-
area is provided. pretation of results of individual participants"
2. Adequate sample size. (p. 97). However, we maintain that more
3. Means and SDs are reported. precise scores as well as separate normative
data for different age, IQ, and educational
levels are necessary to avoid false-positive er-
rors in diagnosis.
It is not clear how cutoffs were developed
2. Insufficient medical and psychiatric ex-
(not reproduced here); they do not match the
clusion criteria; the sample included
cutoffs recommended by Halstead (1947).
participants hospitalized with spinal cord
Golden et al. (1981b, pp. 2(}..21) recom-
injuries and psychiatric disorders.
mend that if nondominant hand performance
is 40% better than dominant hand perfor-
Interpretive Guides mance, a dominant hemisphere lesion should
be inferred, and if the nonpreferred hand score
[TPT.3] Reitan and Wolfson, 1985 is <25% better than the preferred hand score,
The authors suggest that nondominant hand a nondominant hemisphere lesion should be
performance should be about one-third faster hypothesized.
than dominant hand time and that perfor- BU88eU et al. (1970, pp. 42-44), in con-
mance on the third trial (both hands) should structing their neuropsychological key ap-
be one-third faster than nondominant hand proach, devised six rating equivalents of TPT
performance. They provide general guidelines raw scores based primarily on "rules of
for TPT score interpretation in the form of thumb" recommended by P.M. Rennick and
"severity ranges:" perfectly normal (or better apparently derived from a small group of pa-
than average), normal, mildly impaired, and tients. Russell (1984) subsequently modified
seriously impaired. They list the test total the ratings as reflected in Table 17.1.
completion time in minutes and number of Russell et al. (1970) suggest that left
blocks recalled and located that correspond to hemisphere damage is indicated if the right-
each severity range. No other information is hand score is worse than the left-hand score
provided, such as score means and SDs or any by 2 rating points and vice versa.
Table 17.1. Rating Equivalents of the TPT Raw Scores, According to Russell et al. (1970)
Rating Equivalents
0 2 3 4 5 6
Dominant ~3.9 4-7.9 8-9.9 10-12.9 13-15.9 16-19 20,X

Nondominant ~2.9 3-5.4 5.5-6.7 6.8-9.7 9.8-13.9 14-18 19-20,X
Both ~1.7 1.8-3.3 3.4-3.9 4-5.7 5.8-9.5 9.6-18 19-20,X
Total ~8.9 9-16.6 16.7-20.5 20.6-28.5 28.6-39.7 40-56 57-60
Memory 9-10 6-8 4-5 2-3 1 0 6=TPT total
Localization 7-10 5-6 3-4 1-2 0 M=O 6=TPT total
320 PERCEPTUAl ORGANIZATION: VISUOSPATIAl AND TACTILE
Normative Studies and Control Groups provided, as well as means and SDs for mem-
from Clinical Comparison Studies ory and localization scores. A random sub-
sample of 29 applicants was readministered the
[TPT.4] Klove and Lochen (cited in Klove, TPT 14-24 weeks following the original ad-
1974) (Table A 17.4) ministration (Matarazzo et al., 1974) as part of
The authors obtained TPT data on 22 Ameri- an examination of the Halstead Impairment
can controls from Wisconsin and 22 Norwegian Index. Means and SDs for TPT total time in
controls as part of a validation study on the minutes and memory and localization are re-
ability of the HRB to detect brain damage. ported for both the original testing and the
Mean age, educational level, and IQ for the retest. One of the 29 participants obtained a
American participants were 31.6, 11.1, and score outside Halstead's (1947) suggested cut-
109.3, respectively, and for the Norwegian off for total time, while nearly a third (nine of
participants, 32.1, 12.2, and 111.9, respectively. 29) of the participants scored below Reitan's
The TPT data are presented in terms of cutoff on localization. Significant correlations
mean for total time in minutes and memory were observed between the time scores and
and localization scores for each group. PIQ in both control groups (Wiens & Matar-
azzo, 1977), while significant correlations were
Study strengths documented between time scores and FSIQ in
1. This publication is unique in providing only one control group; VIQ was not associated
TPT data on a Norwegian population. with any of the TPT scores, and only the time
2. Information regarding educational level, scores were related to IQ measures.
IQ, age, and geographic recruitment area
is reported. Study strengths
1. Information on test-retest performance.
Considerations regarding use of the study 2. Adequate sample size for the small age
1. Small sample size. range.
2. Undifferentiated age ranges. 3. Adequate medical exclusion criteria.
3. No SDs reported. 4. Information provided regarding educa-
4. No exclusion criteria are specified, and tional level, IQ, gender, occupation, and
no information regarding gender distri- geographic recruitment area.
bution of the sample is provided. 5. Means and SDs are reported.
5. Individual times for each hand sepa-
rately and combined are not reported. Considerations regarding use of the study
6. No information on gender. 1. High IQ level.
7. Relatively low educational level of the U.S. 2. High educational level.
sample, although IQ is average. 3. All-male sample.
[TPT.5] Wiens and Matarazzo, 1977 [TPT.6] Cauthen, 1978a (Table A 17.6)
(Table A17.5) The author obtained TPT data on 117 partici-
The authors collected TPT data on 48 male pants recruited from hospital volunteers and
applicants to a patrolman program in Portland, service clubs in Canada. Those with "evidence
Oregon, as part of an investigation of the WAIS of organic dysfunction" associated with head
and MMPI correlates of the HRB. All partici- injuries, illnesses, and symptoms of organic
pants passed a medical exam and were judged dysfunction were excluded (although the au-
to be neurologically normal. Participants were thor indicates that five participants "were
divided into two equal groups, which were judged to have performed in a manner con-
comparable in age (23.6 vs. 24.8), education sistent with central nervous system dysfunc-
(13.7 vs. 14.0), and WAIS FSIQ (117.5 vs. tion" on the TPT and "apparently were
118.3). Mean time in minutes and SDs to suffering from such dysfunction"). The sample
complete the TPT with the preferred hand, included 35 male and 82 female Caucasian,
nonpreferred hand, both hands, and total are urban residents. All but three participants
were right-handed, and given that there were 3. Minimal exclusion criteria.
no apparent differences in performance be- 4. No data on individuals with less than
tween right- and left-banders, the three left- average IQ.
banders were included in the sample. 5. IQ groupings are somewhat odd, and no
Mean time in minutes and SD are reported information is provided regarding how
for preferred hand, nonpreferred hand, both they were derived.
hands, and total, as are means and SDs for 6. The mean total time for those 30-39
memory and localization. The TPT data are years old with average IQ is in error.
presented in four age (20-29, 30-39, 40-49, 7. Several unusual variations occurred in
and 50--60) by three WAIS IQ (91-111, the data, which are probably due to the
112-122, and 123-139) groupings, with indi- small individual cell sizes. For example,
vidual cell sizes ranging 5-18. Significant dif- the 30-39 group with average to high
ferences across age groups were documented average IQ had mean scores on pre-
on all TPT scores, and significant differences ferred hand time which were higher
across IQ groups were documented for time for than the 40-49 group with the same IQ
both hands, total time, and memory and local- level. Also of concern, the 50--60 group
ization. Using Halstead's (1947) cutoffs of 15.7 in the highest IQ range scored lower
minutes for total time, :::;5 for memory, and :54 than those of lower IQ levels on non-
for localization, 22% of the sample were mis- preferred hand time and memory and
classified as impaired for total time, 15% for localization scores.
memory, and 53% for localization. The authors 8. Data were collected in Canada, raising
conclude that "the inclusion of over half the questions regarding their generalizability
normals in the dysfunctional range of perfor- for clinical interpretation in the United
mance on location indicates that the cutoff States.
point requires adjustment" (p. 458). They also
observed that the percent exceeding the cutoff [TPT.7] Harley, Leuthold, Matthews,
for total time increased dramatically across age and Bergs, 1980 (Table A 17. 7)
groups (e.g., 2% for the 20-29 group, up to 63% The authors collected TPT data on 193 V.A.
for the 50--60 group). The authors provide re- hospitalized patients in Wisconsin aged 55-79.
vised cutoff scores for each age decade for total Exclusion criteria were FSIQ <80, active
time, memory, and localization; when all three psychosis, unequivocal neurological disease or
cutoffs were employed, 29%-37% of partici- brain damage, and serious visual or auditory
pants in each age group fell below the cutoffs acuity problems. Patients with a diagnosis of
for at least one score. Participants decreased chronic brain syndrome were included. Pa-
their nondominant hand performance by an tient diagnoses were as follows: chronic brain
average of 1.3 (1.9) minutes, but 19% failed to syndrome unrelated to alcoholism (28%),
show faster nondominant hand performance. psychosis (55%), alcoholism (37%), neurosis
(9%), and personality disorder (4%). Mean
Study strengths educational level was 8.8 years. The sample
1. Large overall sample size. was divided into five age groupings: 55-59
2. Presentation of data in age-by-IQ (n =56), 60-64 (n = 45), 65-69 (n = 35),
groupings. 70-74 (n=37), and 75-79 (n=20). Mean
3. Information regarding handedness, eth- educational level and percent of sample
nicity, gender, and geographic recruit- included in each diagnostic classification are
ment area. reported for each age grouping. The authors
4. Means and SDs are reported. also provide test data on a subgroup of 160
participants equated for percent diagnosed
Considerations regarding use of the study with alcoholism across five age groupings. The
1. Small individual cell sizes. "alcohol-equated sample" was developed "to
2. No information regarding educational minimize the influence that cognitive or motor/
level. sensory differences uniquely attributable to
alcohol abuse might have upon gro~p test aged 19-71 and included 152 men and 211
performance levels" (p. 2). This su~ample women. Exclusion criteria consisted of signifi-
remained heterogeneous regarding represen- cant physical disability, sensory deficit, current
tation of the other diagnostic categori$. medical illness, use of medication that might
T-score equivalents for raw scores and affect test performance, history of actual or
mean time in minutes per block are ntported suspected brain disorder, and alcoholism.
for dominant hand, nondominant han6, both MMPI profiles "could not suggest severe
hands, and total time per block for thf three disturbance" or include more than three clin-
trials combined by age groupings. In atldition, ical scales with T scores ~70 or an F -scale
total number of blocks correctly placed; score >80.
number of blocks correctly placed py the The TPT was administered according to
dominant hand, nondominant hand, ~ both Reitan's guidelines. Means and SDs for total
hands; time per block for three triaJt
com- time in seconds and memory and localization
bined; and percentage of patients c~rrectly scores are reported for the sample as a whole
placing blocks are provided by age grqupings and by three age groupings (19-34, 35-52,
for the total and alcohol-equated samptes. We 53-71) by four WAIS IQ levels (89-102,
reproduce only the mean, SD, and ra;.ge for 103-112, 113-122, 123-143). Individual cell
the four time scores. · sample sizes ranged from 4-60. Age-by-IQ
categories were determined "in a compromise
Study strengths , between what would be desirable and what
1. Large sample size in many individual the obtained sample characteristics and size
cells, approximately 50. · dictated" (p. 1). No differences in TPT per-
2. Reporting of data on IQ, eduqational formance between men and women were
level, and geographic recruitment area. documented.
3. Data are presented in age groupiPgs.
4. Means and SDs are reported. · Study strengths
1. Large sample size, although individual
Considerations regarding use of the sttJdy cell sizes are substantially below 50.
1. The presence of substantial ne...ologic 2. Presentation of the data in age-by-IQ
(chronic brain syndrome), substance groupings.
abuse, and major psychiatric diso~ers in 3. Adequate medical and psychiatric ex-
the sample. : clusion criteria.
2. Low educational level, although IQ lev- 4. Information regarding gender, recruit-
els are average. · ment procedures, language, and geo-
3. No information regarding gem~r, al- graphic recruitment area.
though given that data were obtafned in 5. Means and SDs are reported.
a V.A. setting, the sample is likel,r all, or
nearly all, male. Considerations regarding use of the study
1. No information regarding education.
Other comments 2. Individual times for each hand sepa-
The scores for the two oldest age groips are rately and together are not reported.
identical in the whole sample and the :4cohol- 3. Participants were recruited in Canada,
equated group because these two gro~~s did raising questions regarding usefulness
not have overrepresentation of alcohopcs, so for clinical interpretation in the United
they did not need to be adjusted. . States.
4. The age-by-IQ cell representing partici-
pants aged 53-71 with IQ of 89-102
[TPT.8] Pauker, 1980 (Table A17.8) contained only four participants; Pauker
'
The author obtained TPT scores on 363loronto comments that this category "should not
citizens fluent in English recruited thro$gh an- be considered to be of any more than
nouncements and notices. Participantf' were interest value" (p. 2).
5. At least one subject 53-71 years old in investigation of the effect of age on perfor-
the 123-143 IQ range scored particularly mance on the HRB and the Wechsler Mem-
poorly on total time and localization, ory Scale. Participants were equally divided
causing the means to be artificially low into two age groupings: 50--62 and 67-86
and the SDs to be excessively large for years. Participants were fluent in English and
this age-by-IQ cell. denied history of CNS disorders, uncorrected
6. IQ levels below the average range are sensory deficits, or illnesses or "incapacities"
not represented. which might affect test results; participants in
poor health were excluded. Mean ages of the
[TPT.9] Anthony, Heaton, and Lehman, 1980 two groups were 55.6 (4.44) and 74.9 (6.04)
(Table A 17.9) years, respectively. Participants in the first
The authors amassed TPT data on 100 normal group were hom between 1916 and 1929, and
volunteers from Colorado as part of a cross- participants in the second group were born
validation of two computerized interpretive between 1892 and 1912. Nine individuals in
programs for the HRB. Participants had no the first group and 10 in the second group
history of medical or psychiatric problems, were female. Four WAIS subtests were ad-
head trauma, brain disease, or substance abuse. ministered (Information, Arithmetic, Block
In addition, for 85% of the controls, normal Design, Digit Symbol); mean scores on these
EEGs and neurological exams were obtained; measures suggested that IQ levels were within
in the remaining 15% of participants, it appears the high average range or higher. Mean edu-
that this information was not available. Mean cational levels for the two groups were 13.7
age was 38.88 (15.80) years, and mean educa- (1.91) and 14.9 (2.99) years, respectively.
tion was 13.33 (2.56) years. Mean WAIS FSIQ, Mean and SD times in seconds for the right
VIQ, and PIQ were 113.54 (10.83), 113.24 hand, left hand, both hands, and total are re-
(11.59), and 112.26 (10.88), respectively. ported, as are means and SDs for memory and
TPT data are presented in terms of mean localization. The groups differed significantly
and SD time in minutes divided by number of on all the time measures.
blocks placed and mean and SD for memory
and localization scores. Participants incor- Study strengths
rectly identified as brain-damaged (according 1. Data on a very elderly age group not
to Russell et al., 1970) were older, less edu- found in other published reports.
cated, and less intelligent than participants 2. Adequate exclusion criteria.
correctly classified as non-brain-damaged. 3. Information regarding education, IQ,
gender, handedness, fluency in English,
Study strengths and geographic recruitment area.
1. Large sample size. 4. Means and SDs are reported.
3. Information regarding education, IQ, Considerations regarding use of the study
age, and geographic recruitment area. 1. High IQ level.
4. Means and SDs are reported. 2. High educational level.
3. The older age grouping spans nearly two
Considerations regarding use of the study decades and may be too broad for opti-
1. The large undifferentiated age grouping. mal clinical interpretation.
2. The IQ range is high average. 4. Small sample sizes.
3. No information regarding gender.
4. Individual times for each hand sepa- [TPT.11] Fromm-Auch and Yeudall, 1983
rately and together are not reported. (Table A 17.11 )
The authors obtained TPT data on 193 Ca-
[TPT.1 0] Bak and Greene, 1980 (Table A 1 7.1 0) nadian participants (111 male, 82 female) re-
The authors gathered TPT data on 30 right- cruited through posted advertisements and
handed Texan participants as part of an personal contacts. Participants are described
as "nonpsychiatric" and "nonneurological." scores, causing the means to be artifi-

Eighty-three percent of the sample were cially low and the SDs to be excessively
right-handed, and mean age was 25.4 (8.2) large for this age grouping.
years, with a range of 15-64 years. Mean ed-
ucation was 14.8 (3.0) years, with a range of [TPT.12] Moore, Richards, and Hood, 1984
8-26 years, and included technical and uni- (Table A17.12)
versity training. Mean WAIS FSIQ, VIQ, and This data set of 284 participants was actually
PIQ were 119.1 (8.8, range 98-142), 119.8 provided by Pauker (1980) but is again in-
(9.9, range 95-143), and 115.6 (9.8, range cluded because these authors divide the data
89-146), respectively. No subject obtained an into smaller age ranges than those provided by
FSIQ which was lower than the average range. Pauker. Specifically, the following six age
Participants were classified into five age groupings were employed: 19-27 (mean=
groupings: 15-17 (n = 32), 18-23 (n = 74), 23.1; 24 men, 32 women), 2fh'36 (mean=
24-32 (n =56), 33-40 (n = 18), and 41-64 31.8; 36 men, 28 women), 37-45 (mean= 40.8;
(n = 10) (total sample for this test= 190). 31 men, 28 women), 46--55 (mean= 50.8; 26
Mean time in minutes, SD, and range are men, 34 women), 56-65 (mean= 61.2; 8 men,
reported for preferred hand, nonpreferred 12 women), and 66-76 (mean=69.5; 8 men,
hand, both hands combined, and total time for 17 women). Mean FSIQ for the six groups
each age grouping. Similarly, mean correct were as follows: 115, 112, 111, 116, 115, and
blocks, SD, and range are summarized on lo- 115. Participants were recruited through
calization and memory trials. The authors note newspaper advertisements and paid $10. Ex-
that none of their participants required more clusion criteria were nonHuency in English,
than 15 minutes to place the blocks with history of central nervous system disorder,
preferred, nonpreferred, and both hands. No current treatment for emotional disorder, and
gender differences were documented, and major physical illness or disability.
male and female data were collapsed. Standard test administration procedures
were followed. Means for number recalled,
Study strengths number correctly located, average completion
1. Large overall sample, with some indi- time across the three placement trials,
vidual cells approximating 50. and "location proportion" (not defined) are
2. Presentation of the data by age grouping. reported.
3. Information regarding mean IQ and
educational levels, handedness, gender, Study strengths
recruitment procedures, and geographic l. Large sample size, although the two
recruitment area. older age groupings have fewer than 50.
4. Some psychiatric and neurological ex- 2. Presentation of the data in age group-
clusion criteria. ings.
5. Means and SDs are reported. 3. Adequate medical and psychiatric ex-
clusion criteria.
Considerations regarding use of the study 4. Information regarding gender, recruit-
1. The high intellectual and educational ment procedures, and language.
level of the sample. 5. Means reported.
2. An age grouping of 41-64 with 10 par-
ticipants would not appear to be partic- Considerations regarding use of the study
ularly useful. 1. No information regarding education.
3. Participants were recruited in Canada, 2. Individual times for each hand sepa-
raising questions regarding the useful- rately and together are not reported.
ness of the data for clinical interpreta- 3. Participants were recruited in Canada,
tion in the United States. raising questions regarding usefuJness
4. At least one subject in the 18-23 group for clinical interpretation in the United
scored particularly poorly on the time States.
4. IQ levels below the average range not of the TPT in adults. Participants had been
represented. admitted to a neurology ward of the Miami
5. SDs not reported. V.A. Medical Center for a suspected neuro-
logical condition but showed no evidence of
[TPT.13] Schear, 1984 (Tables A17.13, A17.14) brain damage upon neurological evaluation,
The author reports norms for the TPT from a including brain CT scans. Mean age was 43.5
Kansas neuropsychiatric sample consisting of (13.6) years, and mean educational level was
556 right-handed males with no "peripheral 14.8 (6.4) years. All but two participants were
injuries or defects" which could adversely af- male. Exclusion criteria were severe psychiat-
fect test performance. The sample reflected a ric disturbance (e.g., psychosis, severe de-
large number of diagnostic categories: 35% pression), left-handedness, and inability to use
had evidence of various signs of brain damage one or both hands.
(organic brain syndromes, alcohol encepha- The TPT was administered according to the
lopathy, epilepsy, etc.) and 49% exhibited Rennick procedures (Russell et al., 1970);
psychiatric disturbance (nonorganic psychotic specifically, if a time trial was not completed
disorders, schizophrenia, alcoholism, etc.). after 10 minutes, the trial was discontinued
A maximum of 10 minutes was allowed for and the score was prorated b~ed on the
each of the three TPT timed trials. Means, number of blocks remaining to obtain a com-
SDs, and ranges are provided for years of bined time score. The 19 control participants
education and time in minutes for right hand, were tested as part of a group of 80 partici-
left hand, both hands, and total for five age pants administered both the six-block and 10-
decades: ~29 (n = 111), 30-39 (n = 112), block TPT. Forty participants were given the
40-49 (n = 111), 50-59 (n = 155), and 60--69 10-block version first, and the remaining
(n = 67). Means, SDs, and ranges are also sample was administered the six-block version
reported for memory and localization and for first during the course of a comprehensive
number of blocks placed with the right hand, neuropsychological evaluation. The interval
left hand, both hands, and total. between administration of the two versions
ranged from 1 hour to 2 days.
Study strengths Mean time in minutes and SD are reported
l. Large sample size, with individual cells for the dominant hand, nondominant hand,
exceeding 50. both hands, and total, as are means and SDs for
2. Presentation of the data by age decades. memory and localization for both versions.
3. Information regarding education, gender, Correlations of the time scores between the
handedness, and geographic recruitment two tests ranged from 0.62 (nondominant
area. hand) to 0.82 (total time), suggesting that these
4. Data regarding mean number of blocks scores "are measuring approximately the same
placed. attributes and the TPT 6 could be substituted
5. Means and SDs are reported. for the TPT 10" (p. 73). The correlation for the
memory scores was 0. 71, but the correlation for
Considerations regarding use of the study localization was only 0.55. Despite the overall
l. Insufficient exclusion criteria; partici- strong association between the two versions,
pants were diagnosed with organic, psy- the TPT-10 was found to be much more diffi-
chiatric, and/or medical illnesses. cult than the TPT-6; the time to complete the
2. All male sample. TPT-6 was approximately one-third that of the
3. No information on IQ. full TPT. The localization score had poor reli-
ability due to its marked variability.
[TPT.14] Russell, 1985 (Table A17.15)
The author obtained data on the six-block Study strengths
children's version of the TPT in a sample of 19 l. Data on the 6-block version in adults and
Caucasian "control" participants as part of his its relationship to the 10-block test are
examination of the use of a shortened version reported.
2. Information on educational level, geo- scores. Significant differences in performance

graphic recruitment area, gender, eth- were found between males and females; men
nicity, handedness, and age is given. significantly outperformed women on total
3. Means and SDs are reported. time.
Considerations regarding use of the study Study strengths

1. Small sample size. 1. Large size of overall sample and indi-
2. Undifferentiated age range. vidual cells.
3. Insufficient exclusion criteria. 2. Information regarding education, gender,
4. The mean score for the TPT-10 for both age, and geographic recruitment area.
hands appears to be in error since it is 3. Data grouped by age and educational
twice that of either hand alone. level.
5. High mean educational level. 4. Minimally adequate exclusion criteria.
6. No data on IQ.
7. Mostly male sample. Considerations regarding use of the study
1. No reporting of TPT score SDs.
[TPT.15] Heaton, Grant, and Matthews, 1986 2. Mean individual WAIS subtest scaled
(Table A17.16) scores reported but not overall IQ scores.
The authors obtained TPT data on 553 normal 3. Individual times for each hand sepa-
controls in Colorado, California, and Wiscon- rately and together are not reported.
sin as part of an investigation into the effects of
age, education, and gender on HRB perfor- [TPT.16] Alekoumbides, Charter, Adkins,
mance. Nearly two-thirds of the sample were and Seacat, 1987 (Table A17.17)
male (356 males, 197 females). Exclusion cri- The authors report TPT data on 135 medical
teria were history of neurological illness, sig- and psychiatric medical inpatients and outpa-
nificant head trauma, and substance abuse. tients without cerebral lesions or histories of
Participants were aged 15-81, with a mean of alcoholism or cerebral contusion from V.A.
39.3 (17.5). Mean education was 13.3 (3.4) hospitals in southern California as part of their
years, with a range of 0-20. The sample was development of standardized scores corrected
divided into three age categories (<40, 40-59, for age and education for the HRB. Among the
and ;?:60) with sizes of 319, 134, and 100, re- 41 psychiatric patients, nine were diagnosed as
spectively, and classified into three education psychotic and 32 as neurotic. In addition to
categories (<12, 12-15, and ;?:16 years) with psychiatry services, patients were drawn from
sizes of 132, 249, and 172, respectively. medicine (n =57), neurology (n = 22), spinal
Testing was conducted by trained techni- cord injury (n = 9), and surgery (n = 6) units.
cians, and all participants were judged to have Mean age was 46.85 (17.17) years, ranging
expended their best effort on the task. TPT 19-82, and mean education was ll.43 (3.20)
mean total time in minutes per block and years, ranging 1-20. Frequency distributions
mean memory/localization scores are reported for age and years of education are provided.
for the six subgroups, as well as percent clas- Mean WAIS FSIQ, VIQ, and PIQ were within
sified as normal using Russell et al.'s (1970) the average range, i.e., 105.89 (13.47), 107.03
criteria. Approximately 20%-30% test score (14.38), and 103.31 (13.02), respectively;
variance was accounted for by age, but only means and SDs for individual age-corrected
approximately 5%-10% of test score variance subtest scores are also reported. All partici-
was associated with educational level. Signifi- pants except one were male; the majority were
cant group differences on all TPT scores were Caucasian (93%), with 7% African American.
found across the three age groups; significant The mean score on a measure of occupational
group differences across educational levels attainment was 11.29.
were documented only for the memory score, No differences were found in test perfor-
and a significant age-by-education interaction mance between the two psychiatric groups
was documented for memory and localization and the nonpsychiatric group, and the data
were collapsed. Mean and SD for time in for time, Memory, and Localization scores for
minutes to correctly place all the blocks are both testing sessions are provided, as are raw
presented for the preferred hand, non- score change and SD, median raw score
preferred hand, both hands, and total time, as change, and mean percent of change. Signifi-
well as memory and localization scores. In ad- cant improvements were noted in time,
dition, mean and SD blocks correctly placed memory, and localization scores. Correlations
per minute are summarized for the preferred between such demographic variables as age,
hand, nonpreferred hand, both hands, and to- education, mean percent of change, and mean
tal time. The latter scores were included change were generally small, with education
because the first set of scores would not dis- accounting for up to 17% of variance and age
criminate between participants who success- accounting for up to 8% of variance.
fully placed all the blocks except one in the
allotted time vs. participants who correctly Study strengths
placed no blocks. Both age and educational I. Information on short-term (3-week)
level had significant associations with TPT retest data.
scores in the expected direction, and regres- 2. Information on IQ level, gender, and age.
sion equation information to allow correction 3. Minimally adequate exclusion criteria.
of raw scores for age and education is included. 4. Means and SDs are reported.

l. Large sample size. l. Undifferentiated age range.
2. Information regarding IQ, age, educa- 2. No information regarding education.
tion, ethnicity, gender, occupational at- 3. TPT time score is not defined.
tainment, and geographic recruitment 4. Individual times for each hand sepa-
area. rately and together are not reported.
3. Regression equation data for computa- 5. Small sample size.
tion of age- and education-corrected
scores. [TPT.18] EI-Sheikh, EI-Nagdy, Townes,
4. Means and SDs are reported. and Kennedy, 1987 (Table A17.19)
The authors reported TPT data on 32 under-
Considerations regarding use of the study graduate and graduate Egyptians at the
I. Data were collected on medical and American University at Cairo as a part of their
psychiatric patients. cross-cultural investigation of the Luna-
2. Undifferentiated age range (mitigated by Nebraska and Halstead-Reitan batteries. The
the regression equation information). average age was 20.6 years (range 17-24). No
3. Nearly all male sample. subject had a history of known brain damage.
Participants were described as "Arabic and
[TPT.17] Bornstein, Baker, and Douglass, 1987a English speaking." TPT instructions were
(Table A17.18) translated into Egyptian colloquial Arabic by
The authors collected TPT test-retest data on the first author and checked by two indepen-
23 volunteers (14 women, 9 men) aged 17-52, dent judges fluent in both Arabic and English.
with a mean of 32.3 (10.3) years, as part of an In case of disagreement between these two
examination of the short-term retest reliability judges, a third judge was consulted.
of the HRB. Exclusion criteria consisted of a The TPT was administered in English to
positive history of neurological or psychiatric 23 participants and in Arabic to nine partici-
illness. Mean VIQ was 105.8 (10.8), with a pants and readministered 2 weeks later.
range of 88-128, and mean PIQ was 105.0 Means and SDs for time in minutes are
(10.5), with a range of ~121. reported for the preferred hand, non preferred
Participants were administered the HRB in hand, both hands, and total, as are means and
standard order both on initial testing and SDs for memory and localization. No differ-
again 3 weeks later. Means, SDs, and ranges ences in performance were found between
administration in English and Arabic. Signifi- scores were converted to WAIS-R equivalents
cant practice effects were documented for time by subtracting 7 points from the VIQ, PIQ,
for both hands, total time, and localization. and FSIQ. Mean FSIQ, VIQ, and PIQ scores
were 100.00 (14.35), 100.92 (14.73), and 98.25
Study strengths (13.39), respectively. IQ scores ranged from
1. Data obtained on an Arabic sample. 60-138 and reflected a normal distribution.
2. Information on test-retest scores. Mean time in minutes and SDs are re-
3. Information regarding educational level, ported for TPf total time, as are means and
age, and geographic recruitment area. SDs for memory and localization. In addition,
4. Means and SDs are reported. IQ-equivalent scores for various levels of in-
telligence are presented. Using Halstead's
Considerations regarding use of the study (1947) cutoffs of ~15.6 minutes for total time
1. Small sample size. and 2:6 and 2:5 for memory and localization,
2. Minimal exclusion criteria. respectively, 21. 7%, 5.0%, and 39.2% of a
3. No information regarding intellectual subgroup of the sample were misclassified as
level. brain-damaged.
4. Undifferentiated age range, although it
can be assumed it is fairly restricted. Study strengths
[TPT.19] Dodrill, 1987 (Table A17.20)
2. Comprehensive exclusion criteria (al-
The author collected TPf data on 120 par- though the appropriateness of including
ticipants in Washington during the years mentally retarded individuals could be
1975-1976 (n = 81) and 1986-1987 (n = 39). questioned).
Half of the sample was female, and 10% were 3. Information regarding education, IQ,
minorities (six African American, three Native age, occupation, gender ratio, handed-
American, two Asian American, one un- ness, ethnicity, recruitment procedures,
known). Eighteen were left-handed. There and geographic area.
were 45 students, 11 homemakers, and one 4. IQ-equivalent scores provided.
retiree; 37 were employed, and 26 were un- 5. Data for different IQ levels provided.
employed. Participants were recruited from 6. Means and SDs are reported.
various sources, including schools, churches,
employment agencies, and community service
agencies. They were either paid for their
participation or offered an interpretation of
2. Individual times for each hand sepa-
their abilities. Exclusion criteria were history
rately and together are not reported.
of "neurologically relevant disease (such as
3. On the IQ-equivalent scores, the two
meningitis or encephalitis)," alcoholism, birth
highest IQ groups have lower scores on
complications "of likely neurological signifi-
localization than the 100-120 IQ group.
cance," oxygen deprivation, peripheral ner-
vous system injury, psychotic or psychotic-like
disorders, or head injury associated with un- [TPT.20] Yeudall, Reddon, Gill, and
consciousness, skull fracture, persisting neu- Stefanyk, 1987 (Table A17.21)
rological signs, or diagnosis of concussion or The authors obtained TPf data on 225
contusion. One-third of potential participants Canadian participants recruited from posted
failed to meet the above medical and psychi- advertisements in workplaces and personal so-
atric criteria, resulting in a final sample of 120. licitations. Participants included meat packers,
Mean age was 27.73 (11.04) years, and mean postal workers, transit employees, hospital lab
education was 12.28 (2.18) years. Participants technicians, secretaries, ward aides, student
tested in the 1970s were administered interns, student nurses, and summer students.
the WAIS, and participants assessed in the In addition, high school teachers identified for
1980s were administered the WAIS-R; WAIS participation average students in grades 10-12.
Exclusion criteria were evidence of "forensic Considerations regarding use of the study
involvement," head injwy, neurological insult, 1. 11te sample was atypical in terms of its
prenatal or birth complications, psychiat- high average intellectual level and high
ric problems, or substance abuse. Participants level of education.
were classified into four age groupings devised 2. 11te data were obtained on Canadian
to ensure group homogeneity: 15-20, 21-25, participants, which may limit their use-
26-30, and 31-40 years. Information regarding fulness for clinical interpretation in the
percent right-banders, mean years of educa- United States due to possible subtle
tion, and mean WAIS/WAIS-R VIQ and PIQ cultural differences.
are reported for each age grouping for males 3. Examination of the data reveals odd,
and females separately and combined. For the unpredicted variability, with those 21-25
sample as a whole, 88% were right-handed and years old performing worse on the time
had completed an average of 14.55 (2.78) years scores than those 26--30 years old.
of schooling. 11te mean FSIQ, VIQ, and PIQ
were 112.25 (10.25), 112.60 (10.86), and 108.13 [TPT.21] Ernst, 1987 (Table A17.22)
(10.63), respectively. 11te author obtained TPI' data on 110 primarily
TPI' data were gathered by experienced Caucasian (99%) residents of Brisbane, Aus-
testing technicians who "motivated the par- tralia, aged 65-75. Fifty-nine were female and
ticipants to achieve maximum performance" 51 were male, and mean educational level was
partially through the promise of detailed ex- 10.3 years; men and women did not differ in
planations of their test performance. years of education. Participants were recruited
Means and SDs for time in seconds to ex- primarily through random selection from the
ecute the task with the preferred and non- Queensland State electoral roll (n = 97), with
preferred hands separately and together are the remainder (n = 13) solicited through
presented, as are means and SDs for Memory senior-citizen centers. Exclusion criteria were
and Localization scores, each age grouping, history of significant head trauma or neurologic
and each age-by-gender grouping. disease. Nearly one-half of the sample were
No significant relationships were found diagnosed with at least one chronic disease
between TPI' scores and gender or educa- (hypertension = 33, heart disease= 9, thyroid
tional level. Significant age effects were noted dysfunction= 7, asthma= 5, emphysema= 2,
for the time, memory, and localization scores; diabetes= 1), for which they were receiving
and significant correlations were documented treatment and which was described as well-
between all scores and PIQ, particularly in controlled. Sixty-six participants were receiving
males. An association between TPI' memory medications, primarily for the diseases listed
score and VIQ was also noted. Because no above.
significant differences were found between Test administration was according to the
men and women, only the combined sample Reitan instructions. All participants were ad-
data are reproduced below. ministered the Trailmaking Test first, followed
by either the TPI' or Booklet Category Test.
Study strengths Mean times in minutes and SDs are re-
1. Large sample size, with individual cells ported for the preferred hand, nonpreferred
approximating 50. hand, both hands, and total. In addition, mean
2. Data grouped by age. number of blocks and SDs are presented for
3. Data availability for a 15-20 age group. each time measure and for memory and lo-
4. Adequate medical and psychiatric ex- calization. Using cutoffs of 15.7 minutes for
clusion criteria. total time and five blocks for memory and four
5. Information regarding handedness, edu- blocks for localization, 77%, 36%, and 89%,
cation, IQ, gender, occupation, geo- respectively, of the sample were classified
graphic recruitment area, and recruitment as impaired. Men outperformed women on
procedures. memory, localization, and all time scores
6. Means and SDs are reported. except dominant hand. No differences in TPI'
scores emerged between participants with and nitroglycerin= 47%, antiarrhythmics = 7%,
without chronic disease, although participants digitalis= 7%). In addition, six participants
taking medications scored better on memory were prescribed antianxiety medication,
and localization. Educational level did not two were receiving sleeping medication, and
appear to be related to TPr performance. A one was on an antidepressant. Also, several
third of the sample (34.5%) failed to show were receiving treatment for chronic medical
superior nondominant hand performance on illnesses such as diabetes (n = 2), ulcers
the second trial, and this subgroup was not (n = 3), gout (n = 2), allergies (n = 3), and ar-
overrepresented by men or women, older age thritis (n = 2). Mean age was 55.5 (8.0) years,
(<70% ), chronic illness, or medication usage with a range of35-68 years, and mean WAIS-R
and did not show poorer scores on the TPr FSIQ was 105.9 (12.2), with a range of77-137.
measures. Participants administered the TPr The majority of the sample had completed at
prior to the Booklet Category Test obtained a least 9 years of education (84%), and nearly
higher mean number of blocks placed for the one-third had completed some post-high
preferred hand, but no other test order effects school work (29.1%).
were documented. The TPr-6 was administered according to
the instructions for the 10-block TPr to each
Study strengths subject 3 weeks before surgery and 3, 12, and
1. Large sample size in a restricted age 24 months postsurgery. Mean time in minutes
range. and SDs for time to complete the task with
2. Presentation of the data by gender. each hand both hands, and total are reported,
3. Information regarding education, geo- as are means and SDs for localization and
graphic recruitment area, recruitment memory for each of the four testing sessions.
procedures, and ethnicity. No significant correlations were noted be-
4. Information regarding test administra- tween measures of presurgical cardiac status
tion order effects. and TPr scores, and only one significant dif-
5. Means and SDs are reported. ference in test performance was noted across
the four largest cardiac medication groups;
Considerations regarding use of the study however, this may have been a chance result
1. Approximately half of the participants given the multiple comparisons. The authors
had at least one chronic illness, and over argue that "these findings suggest that the
half were taking prescribed medications. sample is an appropriate normative group in
2. No information regarding IQ. that the stress of a disease state and upcoming
3. Low mean educational level. surgical intervention was present but the dis-
4. Data were collected in Australia and may ease per se or specific medications were not
be unsuitable for clinical use in the directly related to test performance" (p. 177).
United States. The authors note that the TPr-6 had good
test-retest reliability, with little practice effect
[TPT.221 Clark and Klonoff, 1988 across the four testing sessions. Construct
(Table A17.23) validity appeared to be consistent with that of
The authors obtained data on the children's the 10-block TPr.
version of the TPr-6 in a sample of 79 male,
right-handed, coronary bypass surgery partic- Study strengths
ipants in Canada as part of their evaluation of 1. Data for the six-block version.
the reliability and construct validity of the 2. Test-retest data (although it is techni-
shortened TPr in adults. Exclusion criteria cally not a true test-retest study since
were postsurgical complications and stroke or the participants underwent an interven-
other neurological conditions. Apparently all ing surgical procedure).
of the sample were prescribed coronary med- 3. Information regarding education, gen-
ications (p blockers= 75%, calcium channel der, age, IQ, geographic recruitment
blockers= 65%, long-acting nitrates =59%, area, and handedness.
4. Large sample size. 4. Means and SDs are reported for total
5. Means and SDs are reported. time, memory, and localization trials.
1. All patients were medically ill and re- 1. Data on hands separately not reported.
ceiving various medications. 2. No information regarding ethnicity or
2. Undifferentiated age range. geographic area (although it can be as-
3. Low overall educational level, although sumed it was Maine, given the academic
IQ is average. affiliations of the authors).
4. Data collected in Canada, raising con- 3. High IQ and educational level.
cerns regarding their usefulness for
clinical interpretation in the United [TPT.24] Thompson and Heaton, 1991
States. (Table A 17.25)
5. All-male sample. The authors report TPT data on 489 partici-
pants apparently recruited from California,
[TPT.23] Elias, Podraza, Pierce, and
Colorado, Ohio, and/or Michigan as part of
Robbins, 1990 (Table A 17.24)
their examination of the relationship between
The authors recruited 183 community-dwelling patterns of TPT performance and other neu-
participants (76 men, 107 women) from church ropsychological test scores. Exclusion criteria
groups, businesses, professional organizations, were history of head trauma, neurological ill-
and community service organizations for older ness, substance abuse, "serious" psychiatric
persons as part of a study on the impact of illness, and peripheral injury which could in-
hypertension on cognition. Exclusion criteria terfere with test performance. Mean age and
were no major chronic or acute disease, years of education were 39.43 (17.76) and
including hypertension, treatment for a neu- 13.19 (3.46), respectively. Mean WAIS FSIQ
rological disorder, brain trauma, mental ill- was 113.09 (12.07).
ness, or any cardiovascular or cerebrovascular The TPT was administered by trained
disease. Skilled clerical, supervisory, blue- technicians according to standard procedures.
collar, and professional-executive occupations Mean time in minutes and SDs for dominant
were represented. hand, nondominant hand, both hands, and
Participants were divided into three age total are provided, as are numbers of blocks
groupings: 20-31 (41 men, 47 women), 37-49 correctly recalled and located. A reversal in
(23 men, 38 women), and 55--67 (12 men, 22 the expected pattern of improvement from
women). Mean educational levels for the dominant to nondominant hand performance
three groups were 15.4, 15.7, and 14.9, re- occurred in 30% of participants and was as-
spectively (range 12-20 for each age group); sociated with relatively poorer scores on some
and mean WAIS VIQ and PIQ were 119 and WAIS Performance subtests.
116 for the youngest group, 122 and 122 for
the middle-aged group, and 124 and 121 for
Study strengths
the oldest group.
Means and SDs for TPT total time (in
2. Information regarding education, IQ,
minutes), Memory (number correct), and
age, and geographic recruitment area.
Localization (number correct) are reported.
Study strengths
1. Large overall sample size (with individ-
ual subgroup sizes of 88, 61, and 34). Considerations regarding use of the study
2. Adequate exclusion criteria 1. Undifferentiated age range.
3. Information regarding gender, educa- 2. No information regarding gender distri-
tionallevel, IQ, and recruitment strategies bution.
is provided; and data are stratified by age. 3. High mean educational and IQ levels.
[TPT.25] Heaton, Grant, and Matthews, 1991 (dominant hand) to 14% (memory) of score
The authors provide normative data on the variance; and gender accounted for at most
TPT from 486 (378 in the base sample and 1% of score variance. These demographic
108 in the validation sample) urban and rural variables in combination were associated with
participants recruited in several U.S. states 7% (dominant hand) to 34% (total) of score
(California, Washington, Colorado, Texas, variance. For the sample as a whole, minutes
Oklahoma, Wisconsin, Illinois, Mifhigan, per block for the dominant hand, nondomi-
New York, Virginia, and Massachusetts) and nant hand, both hands, and total were 0.7
Canada. Data were collected over a !5-year (0.8), 0.6 (0.6), 0.4 (0.6), and 0.5 (0.3), re-
period through multicenter collali>rative spectively. Numbers of blocks recalled and
efforts; the authors trained the test :ufminis- correctly located were 7.6 (1.6) and 4.3 (2.5),
trators and supervised data collectiop. Ex- respectively.
clusion criteria were history of l~ng In 2004, the authors published revised
disability, neurologic disease, illnesses \affect- norms, based on a sample of over 1,000
ing brain function, significant head ttauma, normal adults. In addition to age, education,
significant psychiatric disturbance ; (e.g., and gender stratification, the data are parti-
schizophrenia), and alcohol or substance tioned by race/ethnicity (African American
abuse. Mean age for the total sample was 42.0 and Caucasian).
(16.8) years, and mean educationalle.Jel was
13.6 (3.5) years. Sixty-five percent ~f the Study strengths
sample were male. Mean WAIS FSIQ, VIQ, 1. Large sample size.
and PIQ were 113.8 (12.3), 113.9 (13.$), and 2. T scores corrected for age, educa-
111.9 (11.6), respectively. . tion, and gender. The 2004 edition pres-
Participants were generally paid for their ents the data for two race/ethnicity
participation and judged to have prpvided groups.
their best efforts on the tasks. The nrr was 3. Comprehensive exclusion criteria.
administered according to Reitan and Wolf- 4. Information regarding IQ and geo-
son's (1985) instructions, with the exdeption graphic recruitment area.
that time trials were discontinued at 10 min-
utes unless the subject was progressing well, Consideration regarding use of the study
near to finishing the task, or judged to have 1. Above average mean intellectual level
the potential for becoming distressed if forced (which is probably less of an issue given
to discontinue before completion. ~inutes that these are WAIS, rather than WAIS-R,
per block (number of minutes divided by IQ data).
number of blocks correctly placed) was the
performance parameter employed for the Other comments
preferred hand, nonpreferred hand, an~ both I. The interested reader is referred to the
hands and for total time; numbers of blocks Fastenau and Adams (1996) critique of the
recalled and correctly located were ~d for Heaton et al. (1991) norms, and Heaton
memory and localization scores. A T-score et al.'s (1996) response to this critique.
system with demographic correction. was
developed on 378 participants and .cross- [TPT.26] Elias, Robbins, Walter,
validated on 108 participants. Extensive T- and Schultz, 1993
score tables corrected for age, educatiop, and (Table A17.26)
gender are provided; and the int~ested TPT data on 427 participants, including those
reader is referred directly to the handbook for from the 1990 study and reflecting the same
these data. exclusion criteria, are provided for men and
Age accounted for 7% (dominant harid) to women separately for six age groupings: 15-
29% (localization) of the variance in TPT 24 (37 men, 24 women), 25-34 (40 men,
scores; education was associated with 3% 56 women), 35-44 (36 men, 56 women),
45-54 (25 men, 46 women), 55-64 (25 men, Mean scores for localization, memo:ry,
35 women), and 65+ (24 men, 23 women). dominant hand, nondominant hand, and both
Participants with <12 or >19 years of edu- hands are reported.
cation were excluded because participants
outside this range were disproportionately Study strengths
distributed across the age and gender 1. Huge sample size.
groupings. Mean WAIS Vocabulary and In- 2. Information regarding age, occupational
formation subtest scores ranged from 13.9- status, and recruitment strategy with
14.7 and 13.2-13.7, respectively, across the some limited data regarding ethnicity,
age groups. educational level, and gender (it is as-
Means and SDs for total time, Memo:ry, and sumed all were male).
Localization trials are provided.
Study strengths 1. No exclusion criteria.
1. Large overall sample size, although most 2. No stratification of data by age or edu-
individual age x gender cells were <50. cation.
2. Adequate exclusion criteria. 3. No SDs reported for the test scores.
3. Information regarding educational level
(although only ranges provided) and
WAIS subtests (Information, Vocabu-
lary) is provided, and data are stratified CONCLUSIONS
by age and gender. Review of the validity studies and norma-
4. Means and SDs reported. tive data for the TPT reveals considerable
controversy regarding the utility of this test.
Considerations regarding use of the study Although some authors found the TPT to be
1. No information regarding ethnicity or sensitive to different aspects of brain dys-
geographic area (although it can be as- function, many pointed to notable drawbacks
sumed it was Maine, given the academic limiting its practical utility. One of the major
affiliations of the authors). liabilities of this test is that it is long and
2. No data reported for hands separately. difficult and requires the examiner to
blindfold the examinee. This often creates
considerable psychological discomfort for the
[TPT.27] Barrett, Morris, examinee (see Lezak et al., 2004, pp. 469-
Akhtar, and Michalek, 2001 471). Use of the TPT-6 can alleviate some
(Table A17.27) of the problems associated with the full 10-
TPT data were obtained on 1,052 Air Force block version. However, data should be
veteran controls who setved in Southeast collected on the validity/reliability of the
Asia from 1962 to 1971 in a study examining short version and on its comparability to the
the effects of Agent Orange on cognition. full version if it is to be used in clinical
Subjects averaged 43.9 (7.6) years of age, and practice.
5.3% were African American (with the rest The clinical usefulness of the TPT would be
"nonblack"); 37% were officers, 16.6% were improved by adjusting cutoffs relative to a
enlisted Byers, and 46.0% were enlisted subject's age, intellectual level, and possibly
ground crew; most of the officers were education and gender, although the effect of
college-educated, and most enlisted person- the latter two demographic variables on TPT
nel were high school-educated. No exclusion performance needs to be further explored.
criteria are listed aside from low exposure Taking demographic factors into account in
to dioxin and epilepsy. Participation was assigning participants to impaired vs. non-
voluntary. impaired groups would improve the specificity
334 PERCEPTUAl ORGAN IZATION: VISUOSPATIAL AND TACTILE
of the TPT. This would also reduce excessively The above suggestions address major criti-
high rates of misclassification of "normal" cisms voic d by investigators regarding the
participants in the impaired range reported in validity of the TPT. With further improve-
the studies reviewed. ments in administration and refinement
Another aspect of the test which would in interpretation guidelines, the utiJity of
benefit from further research attention is this widely used test might be considerably
standardization of test administration. .
1mprove d.2
' Meta-analyses were not pe rform ed for the TPT as th e

data available for review are hete rogeneous in terms of
measures reported, country wh re data were collected,
and sample composition, with s ve ral studies reporting
data for patients with medical or neurological condi-
tions.
v
VERBAL AND VISUAL LEARNING
AND MEMORY
18
Wechsler Memory Scale
(WMS-R, WMS-111, and WMS-IIIA)
BRIEF HISTORY OF THE TEST Information, which included six relatively sim-
ple general and personal information questions;
The Wechsler Memory Scale (WMS) made its
(2) Orientation, which asked five questions re-
first appearance in 1945 and was one of the
lating to place and time; (3) Mental Control,
first standardized memory batteries. Revisions
which required the patient to count backward
of the WMS test battery were published in
from 20, recite the alphabet, and count by 3s,
1987 (i.e., WMS-R) and 1997 (i.e., WMS-III),
with bonus points awarded for fast, perfect
with an abbreviated administration and scor-:-
performance; (4) Logical Memory, which tested
ing version appearing in 2002 (i.e., WMS-
only immediate auditory memory for two sep-
IIIA). (Please see Appendix 1 for ordering
arate orally presented stories; (5) Digits For-
information.) The WMS-III and WMS-IIIA
ward and Digits Backward, which assessed
are the versions generally currently used in
attention span and immediate auditory memo-
standard practice; however, as of this writing
ry; (6) Visual Reproduction, which tested im-
(January 2004), the WMS-R is still used in
mediate visual memory for geometric designs
some clinical practices and research settings
after a 10-second exposure; and (7) Associate
and, indeed, is still available for purchase from
Learning, which tested recall for an orally pre-
The Psychological Corporation.
sented list of five semantically related (easy) and
five unrelated (hard) word pairs over three
WMS (1945) trials.
Although Wechsler (1945) originally intro-
Wechsler's (1945) original test battery was duced two forms of the WMS, only form I was
developed with the intention of providing "a normed by Wechsler for clinical use. Re-
rapid, simple and practical memory examina- searchers and clinicians, therefore, relied on
tion" (p. 87). The original WMS required about form I of the WMS to provide evidence for
30 minutes to administer and consisted of the integrity of memory function.
seven subtests: (1) Personal and Current As the popularity of the original WMS
increased around the world, numerous norma-
The authors gratefully acknowledge the conbibutions of tive studies emerged to enhance its clinical
Paul Satz and David Schretlen to an early version of this
chapter, as it originally appeared in D'Elia, Satz &: utility. Of the 27 normative reports that were
Schretlen (1989). published for the original WMS form I, 15
337
338 VERBAl AND VISUAl lEARNING AND MEMORY
presented data for groups residing in the United and screening device for use as part of a
States (Abikoff et al., 1987; Bak & Greene, 1980; general neuropsychological examination, or
Bigler et al., 1981a,b; Haaland et al.• 1983; any other clinical examination requiring the
Heaton et al., 1991; Hulicka, 1966; Ivnilc et al., assessment of memory functions" (Wechsler,
1991; Mitrushina & Satz, 1991b; Russell, 1975; 1987, p. 1). Several changes were made in this
Russell & Starkey, 1993; Ryan et al., 1987; first major standardized revision of the WMS,
Schaie & Strother, 1968a; Trahan et al., 1988; including modifications of test items and ad-
Van Gorp et al., 1989; Wechsler, 1945). Nine ministration procedures. New subtests were
studies reported data for individuals living added to assess spatial and figural memory,
elsewhere, including three studies of Clfladian and delayed recall testing on most subtests
individuals (Klonoff & Kennedy, 1965~ 1966; was incorporated as a standard procedure.
Cauthen, 1977), four of Australian groups Scoring accuracy was greatly improved by the
(desRosiers & Ivison,1986; Ivinskis et al~ 1971; provision of detailed scoring procedures.
Ivison, 1977, 1986), and one study each for Overall, the significant improvements in
British (Kear-Colwell & Heller, 197$) and scoring criteria and administration procedures
Turkish (Gilleard & Gilleard, 1989) populations. for the WMS-R permitted a more valid as-
Russell (1975) was the first to add a delayed sessment of memory than was possible with
recall condition to the test, and almost Without the original WMS. Although the WMS-R is
exception, WMS studies published subsequent still used in some laboratories and in some
to Russell's article included a delayed recall research applications, it has largely gone the
condition on one or more subtests, usually way of the original WMS, and the WMS-III
Logical Memory and Visual Reprod11ction. and WMS-IIIA should now be considered the
However, some investigators (Bak & Greene, versions of standard practice.
1980; Haaland et al., 1983; Ivnik, et al.; 1991; A brief overview of the WMS-R follows.
Trahan et al., 1988) followed Russell's proce- The full battery includes information and
dure of interposing 30 minutes betwe~n the orientation questions, eight short-term mem-
immediate and delayed recall conditions. while ory tasks, and four delayed recall trials, all of
others used delays of 45 minutes (Mitrushina & which take about 45 minutes to 1 hour to
Satz, 1991a,b; Van Gorp et al., 1990), 1 hour administer. Four of the six information and all
(Cauthen, 1977), and even 24 hours (Abi<off et five of the orientation questions from the
al., 1987). Despite all the normative research, original WMS were retained, and three new
data for various WMS age groups remained questions were added; thus, there are a total
scant or nonexistent. For subjects uncler 50 of 14 scorable information and orientation
years of age, only Abikoff et al. (1987) prC>vided questions (as opposed to 11 in the original
WMS normative data for delayed recall of the WMS). The presentation of the Mental Con-
Logical Memory subtest. Trahan et al. (1988), trol subtest was left unchanged, except that
Ryan et al. (1987), and Ivnik et al. (1991) pro- bonus points are no longer assigned for fast,
vided normative data for the delayed rocall of perfect performance.
the Visual Reproduction subtest of the WMS. The Digit Span subtest uses a different se-
Interestingly, no studies provided reliable quence of numbers from the original WMS
WMS delayed recall normative data tor the and begins with series that are shorter by one
Associate Learning subtest for U.S. pOpula- digit. Paragraph 1 of the WMS-R Logical
tions. Delayed recall data for any subteshiof the Memory subtest is similar to the one used in
original WMS were scant for subjects under 20 the original WMS, with only some slight
years of age. Finally, in 1987, the long-awaited modifications. Immediate and 30-minute delay
standardized revision of the test was pub¥shed. recall are assessed. A detailed scoring method
awards full credit for either a verbatim or a
"gist" response. The WMS-R Verbal Paired
WMS-R (1987)
Associates subtest uses four of the six easy and
The WMS-R was introduced in 1987 and de- all four of the hard pairs found on the Asso-
scribed by the test publisher as a "diagnostic ciate Learning subtest of the original WMS.
WECHSLER MEMORY SCALE 339
The revised edition provides up to six trials to pattern. Part two requires the subject to tap
correctly learn all the pairs, and equal credit is out the pattern in the reverse order.
awarded to all pairs, regardless of difficulty. The original WMS Memory Quotient
Delayed recall for the pairs is assessed after 30 (MQ) was replaced with five composite scores
minutes. intended to differentiate separate memory
Regarding the Visual Reproduction subtest, mechanisms. However, it should be noted that
two of the three original WMS stimulus cards the scores bottom out at 50, so the test may
were retained. Two additional cards were ad- generate an overestimation of memory func-
ded: one containing one design, the other tioning in individuals with severe memory
containing two designs. Although this subtest impairments.
was intended to assess nonverbal learning and The WMS-R is purported to provide norms
memory, all stimulus cards are, unfortunately, for individuals aged 16 years 0 months to 74
easy to verbally encode. Immediate and 30- years 11 months; however, close inspection of
minute delay recall for the designs are as- the technical manual reveals that the norms
sessed. A detailed set of scoring criteria for the for the age groups 18-19, 25-34, and 45-54
designs was developed. were interpolated on the basis of the scores of
Three additional subtests were added to the the adjacent sampled age groups.
WMS-R, which purport to assess aspects of Although the WMS-R was clearly an im-
nonverbal (visual) memory. The Visual Paired provement over the original WMS, several
Associates subtest was developed as an analog reviews suggested there was still room for
to the Verbal Paired Associates (word pairs) improvement (D'Elia et al., 1989; Loring,
subtest. This test presents six colors, one at a 1989; Chelune et al., 1989; Elwood, 1991).
time, in association with a different design. (Please refer to Lezak (1995, pp. 502--505)
Immediately following presentation of the six and Spreen and Strauss (1998, pp. 391-415)
pairs, the design is presented alone and the for information regarding neuropsychological
subject is to remember the name of the color findings as well as further presentation of the
paired with the design. Although this subtest merits and limitations of the WMS-R.) In
was developed with the intention to "mini- 1997, the new standardization of the WMS
mize the role of verbal mediation in memo- appeared (WMS-III).
rizing and responding to the figure-color
pairs," it appears that four of the six designs
WMS-111 (1997)
can be readily verbally encoded. A 30-minute
delay recall of the figurEH!olor pairs is also The WMS-III is an individually administered
assessed. battery of 11 subtests of learning, memory,
The Figural Memory subtest was developed and working memory. Six of the subtests are
as a measure of nonverbal (visual) recognition included in the core battery, and five subtests
memory. The subject is shown a set of shaded are considered optional I supplemental. In
geometric designs which are difficult to ver- comparison with the WMS-R, administration
bally encode. After the designs are removed, time has been reduced. Administering the
multiple-choice recognition memory for the WMS-III core battery (i.e., six subtests) takes
design is assessed. approximately 30-35 minutes, and adminis-
The Visual Memory Span subtest is a vari- tering the five supplementary scales adds an-
ant of the Corsi Cube Test, which is itself a other 30 minutes to the process. Most of the
variant of the Knox Cube Imitation Test. (See memory subtests include a delayed recogni-
Lezak et al., 2004, pp. 354-356 for discussion tion procedure so that a differentiation be-
of Knox Cube and Corsi Block-Tapping tests.) tween retrieval vs. encoding deficits can be
For this WMS-R subtest, eight like-colored made. The WMS-III was co-normed with the
squares are printed on a card in random or- WAIS-III, and their joint factor index scores
der. As in Corsi's task, every time the examiner permit ability/memory comparisons.
taps the squares in a prearranged sequence, The WMS-III includes six subtests from the
the subject attempts to copy the tapping WMS-R, although most have been altered and
340 VERBAL AND VISUAL LEARNING AND MEMORY
improved (Information and Orientation, Digit regarding the speed (slow, medium, fast) of
Span, Mental Control, Verbal Paired Associ- reading the stories to the subject nor does it
ates, Logical Memory, and Visual Reproduc- comment on the intonation, pauses, or inflec-
tion). The scoring sensitivities of these snbtests tion of the presentation. It is known that the
have been generally improved by extEnding speed of presentation of similar prose passages
the floor and raising the ceiling. The WMS-R can affect delayed recall performance (Shum
Figural Fluency and Visual-Paired A$ociate et al., 1997). To ensure a standardized ad-
subtests have been deleted from the wMS-III ministration and improve the reliability and
battery. Four new subtests have been a4ded to validity of the subtest (especially when the test
the test battery (Faces, Family Picturest Word is being administered by trainees, interns,
Lists and Letter-Number Sequencing) etc.), a cassette-taped version of the two stories
A brief description of the WMS-III spbtests using a medium speed (Shum et al., 1997)
follows. With the exception of minor cftanges should be considered for future versions of this
to the wording of one question, the Informa- subtest.
tion and Orientation subtest is essentially un- The Faces subtest is new to the WMS-III
changed from the WMS-R version. Rerftrding but will be familiar to anyone who has used the
the Logical Memory subtest, in story 4, poor Warrington Faces Test (Warrington, 1984) or
old Anna is still living in the same neighbor- the Denman Neuropsychology Memory Scale
hood and experiencing the same problems. (Denman, 1984). The WMS-III Faces subtest
One very minor change to the description of uses a recognition paradigm to assess visual
her plight was made. Regarding story iB, the immediate and visual delayed memory by
macho truck driver from the WMS-R was presenting a series of 24 faces to the subject
evidently fired, and a new story is offered and then showing them a second series of 48
about a man living in San Francis~ who faces and requesting that they identify only the
prefers watching old movies. The new !lory B 24 faces that they had been previously shown.
was reportedly developed so that it wo~d "be Approximately 30 minutes later, the subject is
less likely to evoke an emotional reactio• from again shown the series of 48 faces and again
some examinees" {p. 13). requested to identify the 24 faces initially
Regarding the immediate recall condition, presented.
similar to the WMS-R, the examiner sepa- All eight Verbal Paired Associates word
rately requests recall for story A and story B pairs from the WMS-R have been replaced
following each presentation. However, unlike with novel, unrelated word pairs. The ceiling
the WMS-R, the WMS-III examiner p~sents has been raised by an expanded word-pair list.
story B a second time, to increase the.likeli- The WMS-R provided for eight learning trials;
hood of learning story B details. Im~diate however, only four learning trials of the word-
recall for story B is then reassessed. Folbwing pair list are administered for the WMS-111.
an approximately 30-minute delay, recall for Following an approximately 30-minute delay,
stories A and B is again captured. cued recall is elicited, with the examiner of-
Scoring procedures for verbatim or near- fering the first word of each pair and the
verbatim recall of the stories have • been subject expected to provide the second, or
tightened; numerous examples are offered in "associated," word. There is also a subsequent
the manual to improve inter-rater reliab~ty. A recognition task, where the examiner reads a
supplemental scoring procedure has also been list of 24 word pairs and the subject must
developed to allow examiners to charaQterize identify the pair as either "new" or from the
the subjects' "gist" or thematic recall ff the previous condition.
stories. Following free recall for both stories The Family Pictures subtest is new to the
after an approximately 30-minute delay, rec- WMS-III, is purported to be the visual analog
ognition memory is tested, using 30 qu~tions to the Logical Memory subtest, and requires
that probe for details about stories A ~d B. recall of the characters, their scene activity,
The WMS-III manual does not present and spatial location. The Family Pictures
any standardized administration gui~lines subtest is a multimodality test of memory in
that the family scenes are visually presented numbers, the first starting with just a single
but can also be verbally encoded. The subject number-letter pair and the final string con-
is initially shown a "family portrait" card of six taining eight elements in four pairs (e.g., 7-N-
family members along with the family dog and 1-Q-4-V,-3--0). Following each string pre-
told that these will be the characters on four sentation, the subject must remember the
subsequent scene cards. The subject is then numbers and letters and then repeat them,
shown the four cards for 10 seconds each and saying the numbers first in ascending order
requested to tiy and recall as much about and then the letters in alphabetical order.
each of the scenes as possible. After all four The WMS-111 Spatial Span subtest stimuli
cards have been exposed, recall for the char- have been modified from the WMS-R two-
acters in the scene, their location, and their dimensional, eight like-colored square format
action is elicited for each card. Approximately to a three-dimensional, 10-block board format
30 minutes later, the recall paradigm is re- patterned after the WAIS-R NI Spatial Span
peated for all four cards. Board (Kaplan et al., 1991). The administra-
The Word Lists subtest is new to the scale tion is otherwise identical to the WMS-R: the
and is considered optional; however, it will be subject is asked to replicate an increasingly
familiar to anyone who has administered the long series of visually presented spatial loca-
Rey Auditory-Verbal Learning Test (Rey, tions that are tapped out on the stimulus block
1964) or the California Verbal Learning Test tops. Every time the examiner taps the blocks
(Delis et al., 1987). The WMS-111 version in a prearranged sequence, the subject at-
presents a list of 12 semantically unrelated tempts to copy the tapping pattern. Part two
words to be learned across four learning trials. of the test requires the subject to tap out the
Following the learning trials, the subject is pattern in the reverse order.
presented with a one-trial distracter list to The Mental Control subtest is considered
learn, followed by recall of the original list. an optional test and consists of eight items.
The subject is told that recall of the first list The ceiling has been raised by expanded
will be later assessed. Approximately 30 min- content. Counting backward by 3s has been
utes later, recall for the first list is obtained, deleted, and six new tasks have been created.
followed by an auditory recognition protocol. As was true for the WMS (1945) but not for
The Visual Reproduction subtest, a former the WMS-R (1987), bonus points are awarded
core WMS-R subtest, is optional for the WMS- for fast, perfect performance. As a result, a
111 and consists of five design cards, some of possible 40 points can be generated from the
which are easy to verbally encode. Cards A , C, eight items.
and D remain from the WMS-R; however, card The Digit Span subtest is optional on the
B has been dropped. Two additional design WMS-111 and is identical to the WAIS-III
cards are offered, one of which is a modifica- version. It is similar to the WMS-R version,
tion of a card resurrected from the original with the exceptions that the Digit Span For-
WMS (1945). Similar to the WMS-R, imme- ward test now hegins with a two-digit se-
diate and 30-minute delay recall for the designs quence and ends with a nine-digit sequence.
are assessed. However, unlike the WMS-R, For Digit Span Backward, an eight-digit se-
following the delay condition, there is a 48- quence has been added. The administration
item recognition task, a direct copy condition, and scoring procedures are otherwise identi-
and a seven-item discrimination condition, al- cal to the WMS-R.
lowing for evaluation of motor vs. memory Only six of the WMS-III subtests are con-
performance. The scoring criteria have been sidered primary (Logical Memory, Verbal
improved to allow for partial credit, rather than Paired Associates, Faces, Family Pictures,
the ali-or-nothing WMS-R approach. Letter-Number Sequencing, and Spatial
The Letter-Number Sequencing subtest is Span), and all must be administered in order
new to the WMS-111 and is a measure of au- to calculate the WMS-111 index scores. As
ditory working memory. The subject is read noted earlier, the primary subtests can be
seven ever-increasing strings of letters and administered in 30-35 minutes.
The Information and Orientation, Word country: northeast, north central, south, and
Lists, Visual Reproduction, Digit Span, and west. Each age group was required to have a
Mental Control subtests are considered op- full-scale IQ score of 100. However, some age
tional and do not contribute to the WMS-III groups did not meet the IQ standardization
index scores but are used to obtain supple- requirement, and others did not meet other
mentary information. demographic criteria. Therefore, a weighted
The structure of the WMS-III is substan- sample was derived to improve the fit to the
tially different from that of the WMS-R in that census data and to more closely approximate
the number of summary indices has been an average (IQ) of 100 for each age group.
increased from five to eight and they reflect The WMS-III standardization sample in-
the significant revisions to the content of the cludes 1,250 individuals aged 16-89 years,
test. The WMS-III provides for the calcula- divided into 13 age group categories (16-17,
tion of eight primary index scores and four 18-19, 20-24, 25-29, 30-34, 35-44, 45-54,
Auditory Process Composite scores. The pri- 55-64, 65-69, 70-74, 75-79, 80-84, 85-89).
mary indexes are considered by the publishers However, although the final WMS-111 stan-
to be core to evaluating memory functioning, dardization sample is reported to have 1,250
whereas the Auditory Process Composites are cases, only 1,032 cases in the standardization
considered supplementary scales to better are actually unique. Because the WAIS-111
isolate and characterize various processes in and WMS-111 were co-normed, one can
memory functioning, such as single-trial readily compare performance on the WMS-
learning vs. learning over trials, retention, and III with IQ. Both tests have summary indices
retrieval. with means of 100 and standard deviation
The method of calculating the WMS-III in- (SD) scores of 15 points.
dex scores is also different from that for the
WMS-R (see pp. 203-211, W AIS-Ill, WMS-lll
WMS-IIIA (2002)
Technical Manual-Updated, [Wechsler,
2002a]). There are three primary indices for The WMS-IIIA (Wechsler, 2002b) was de-
auditorially processed material (Auditory Im- veloped to allow quick estimates of general
mediate, Auditory Delayed, and Auditory memory functioning of individuals aged 16-89
Recognition Delayed), two primary indices for years when extended memory testing was not
visually processed materials (Visual Immedi- indicated and/ or feasible. The WMS-IIIA is
ate, Visual Delayed), and three primary indices basically a "carve-out" and repackaging of the
for multimodality memory functioning (Im- WMS-III as a screening instrument to help
mediate Memory, Working Memory, and determine whether further memory or neuro-
General Memory). psychological evaluation is needed. Speciflcally,
The WMS-111 and WAIS-III were co- the WMS-IIIA allows reliable assessment of
normed, with the WMS-III randomly admin- immediate and delayed auditory and visual
istered to approximately one-half of the memory abilities and provides "clinicians with
WAIS-III standardization participants within a method for evaluating memory problems in
each standardization stratification variable their patients when the clinician does not in-
category. The standardization sample for the tend to parse out specific memory factors or
WAIS-III included 2,450 adults selected to be make statements regarding brain-behavior
representative of the U.S. population of adults relationships" (WMS-IIIA manual, p. 30). The
aged 16-89 years in 1995. From this larger WMS-IIIA consists of test stimuli and norma-
sample, 1,032 adults were tested using the tive data derived directly from the WMS-111
WMS-III protocol. This WMS-111 sample was and its original standardization. A trained ex-
then weighted to match the 1995 census data. aminer can usually administer the WMS-IIIA
The collected sample was stratified for age, in 15-20 minutes.
gender, racelethnicity, educational level, and The WMS-IIIA consists of four subtests:
geographic region. The standardization test Logical Memory I, Family Pictures I, Logical
sites were divided into four regions of the Memory II, and Family Pictures II. The test
WECHSlER MEMORY SCAlE 343
stimuli, administration, and scoring proce- the WMS-IIIA is a short form of the WMS-III,
dures for the WMS-III Logical Memory sub- it should not be considered a parallel form. A
tests are identical to those for the WMS-IIIA major limitation and difference between the
Logical Memory subtests. Therefore, on the two tests is that there is no recognition testing
WMS-IIIA Logical Memory I subtest, the following the delayed recall procedures for the
examinee listens to two different stories, A two subtests used to assess memory function-
and B, and immediately after hearing each ing on the WMS-IliA. As such, clinicians using
story is asked to retell it from memory. Story B the WMS-IIIA are not able to comment on
is then read to the examinee a second time, to possible contributions to the scores from en-
increase the likelihood of learning story B's coding and/or retrieval difficulties. Also, since
details. Immediate recall for story B is then the Logical Memory subtest primarily mea-
reassessed. The Logical Memory I score is sures auditory verbal learning and memory
based on the accuracy of the immediate recall and the Family Pictures subtest is a multi-
of these stories. Without warning, approxi- modality measure of memory (since the pic-
mately 30 minutes later, the subject is again tures can also be verbally encoded), the
asked to recall the two stories. The Logical examiner cannot comment on possible later-
Memory II score reflects the accuracy of the alization of memory findings. To perform such
delayed recall of these stories. Identical to analyses, a more comprehensive assessment of
the WMS-III Family Pictures subtest, the memory using the WMS-III and/or other
WMS-IIIA Family Pictures subtest requires combinations of memory tests, with recogni-
the subject to view four different scenes with tion testing, would be required.
family members for 10 seconds each. Subjects The standardization sample for the WMS-
are then requested to recall as much about IIIA, which was psychometrically derived
each of the scenes as possible. The scoring from the original WMS-111 standardization
protocol is identical to that used for the data, included 1,032 adults selected to be
WMS-111, and points are assigned for correct representative of the U.S. population of adults
recall of the characters, their scene activity, aged 16-89 years in 1995. As noted earlier, the
and spatial location (Family Pictures I). original WMS-111 standardization test sites
Without warning, approximately 30 minutes were divided into four regions of the country:
later, the recall and scoring paradigm is re- northeast, north central, south, and west.
peated for all four cards (Family Pictures II). Participants were originally tested on the
Using the norms tables developed for the WMS-111 protocol. The sample was stratified
WMS-IIIA, the raw scores for each subtest are for age, sex, race/ethnicity, educational level,
converted to scaled scores. Each subtest per- and geographic region. Regarding education,
formance score can therefore be expressed in the sample was stratified according to five
terms of a standard scaled score or percentile. levels (~8. 9-11, 12, 13-15, ;:::16 years). In-
Three additional scores may also be calculated terestingly, for individuals aged 16-19 years,
by summing the appropriate subtest scaled information regarding the educational attain-
scores and converting these sums to standard ment of the parent was used. This initial
scores. Specifically, the Immediate Memory WMS-IIIA standardization sample was then
Composite is composed of Logical Memory I divided into 13 age group categories (16-17,
and Family Pictures I; the Delayed Memory 18-19, 20-24, 25-29, 30-34, 35-44, 45--54,
Composite is composed of Logical Memory II 55-64, 65-69, 70-74, 75-79, 80-84, 85-89).
and Family Pictures II. A Total Memory Each age group was required to have an av-
Composite score is based on the sum of the erage FSIQ of 100, but this was found not to
four subtest scaled scores. The WMS-IIIA be the case. Further, some of the age groups
composites are reported to have psychometric did not meet the 1995 population criteria on
properties approaching those observed for the some of the stratification variables. As such,
WMS-111 index scores, and the WMS-IIIA "cases were randomly selected within stratifi-
overall is reported to have clinical sensitivity cation parameters for duplication" and a
similar to the WMS-111. However, although "weighted sample was derived to improve the
fit to the census data as well as to more•closely preting Index scores (p. 77)" since these scores
approximate an average (IQ) of 100 for each reflect performance relative to an age-peer
age group" (page 8). Therefore, altho"gh the group. Wechsler (1987) notes that an index
final standardization sample is repotted to score of 100 obtained by an older adult in his
have 1,250 cases, only 1,032 are actudny un- or her 70s, although average, would not reflect
ique. In the final WMS-IIIA standar<lization the same level of absolute performance when
sample, there are 100 cases in each ag~group, compared to a young adult in his or her 20s
except for the two oldest age groups,! which obtaining the same index score.
have 75 cases each. Although the samjle was The question of gender differences in per-
also stratified by educational level, as noted formance has been more controversial. On the
earlier, the normative data for the WMS-IIIA Visual Reproduction subtest, while some au-
are reported only by age category. . thors have reported no gender effects (Trahan
In summary, the WMS-III repre~nts a et al., 1988), Ivison (1977) found that women
significant improvement over previou~ incar- performed slightly worse than men. The
nations of Wechsler's battery and should be WMS-R manual reports that males and fe-
considered a welcome addition to the ~euro males do not significantly differ on the WMS-
psychologist's armamentarium. A WAIS-III- R indices. Moderate correlations have been
WMS-III computerized scoring program is generally reported between education and
available from The Psychological Corp~ation. memory functioning Lezak et al. (2004);
The scoring program. in addition to pr~ding Wechsler, (1987). lvnik et al. (1992b), in ex-
normative data by age category, can P.J'ovide amining Visual Reproduction performance,
age/education-corrected norms. found it was a significant variable in older
adults, with better performance noted for
those who were better educated. Years of
education were significantly correlated with
RELATIONSHIP BETWEEN TESl
all WMS-R indexes (Wechsler, 1987). Ri-
PERFORMANCE AND DEMOGRAPJiiC chardson and Marottoli (1996) note that ad-
FACTORS
justment for lower levels of education is
It has long been accepted that memory func- especially important when testing the elderly,
tions decline with age (Botwinick, 198l). Ex- since the "educational attainment of both
amining the normative data provided 'n the White and non White Americans aged 75
manuals, advancing age does appear to have a years and older is less than 12 years (median
negative effect on the memory subtest ~of the for Whites= 11.6 years, median for non
WMS-R, WMS-III, and WMS-IIIA. Whites= 8 years, median for all ethnic= 10.9
The WMS-R VISual Reproduction ~btest years)." As Stem et al. (1992) have empha-
appears to be the most sensitive to deteriora- sized, "utilization of standard non-education
tion with age (Ivnik et al., 1992b; Le~ et al. corrected normative data will frequently result
(2004), whereas immediate recall performance in normal subjects being misclassified as im-
on the Logical Memory subtest rem~ rela- paired or demented, especially if they are
tively stable through middle age and then older and have low education (e.g.,<8 years)."
starts to decline. However, Logical M~mory The effect of IQ on subtest performance
delayed recall performance begins to gr~ually remains unclear. However, an examination of
decline in the 30s-50s, after which the decline correlation tables provided in the WMS-R
accelerates (Wechsler, 1987). Lezak et al. manual suggests modest correlations with IQ
(2004) caution that for older age groups~n the for many of the subtests.
WMS-R standardization, education ru*l age Regarding demographic influences on
are strongly negatively correlated, thereby WMS-111 performance, the WMS-III manual
making it difficult to disentangle the~ indi- does not specifically present information on the
vidual impact on performance. The ~S-R effects of sex and education!IQ on individual
manual, however, notes that "the user ~ould aspects of subtest performance; however, ex-
keep the effects of age in mind when rinter- amination of the normative data tables does
show the expected decline in performance Reporting of IQ and/or Educational levels

starting in the 30s and accelerating in the Since performance on the WMS-R and WMS-
50s-70s. In further analyses of the standardi- III is positively correlated with IQ and edu-
zation data, Haaland et al. (2003) report that cation, the means and SDs for years of
the oldest groups always performed the poor- education and/ or estimated IQ should be
est. They also noted a more significant age reported for each subgroup, and preferably
deterioration on the Visual Reproduction sub- normative data should be presented by level
test than on the Logical Memory subtest for of education and/or IQ.
immediate, recall and recognition conditions.
Reporting of Gender
Given the equivocal evidence of an effect of
METHOD FOR EVALUATING THE gender on performance, information on the
NORMATIVE REPORTS gender distribution of the sample should be
reported.
To adequately evaluate the WMS-R, WMS-111,
and WMS-IIIA normative reports, eight key
criterion variables were deemed crucial. The Procedural Variables
first five of these relate to subject variables,
and the remaining three relate to procedural Inclusion of a Delayed Recall Condition
variables. To assess storage and rates of forgetting over
Minimal requirements for meeting the cri- time, the addition of a delayed recall condition
terion variables were as follows. to assess long-term memory is essential.
Description of Scoring Procedures

Subject Variables
A clear statement of the method used for
Age Group Interval scoring the WMS-R, WMS-III, and WMS-IIIA
Since Wechsler's original intention was to is critical. Several methods are available for
scoring the WMS-R (Crosson et al., 1984b;
tie memory to IQ and because of the well-
Power et al., 1979; Schear, 1986; Schwartz &
documented correlation between memory and
Ivnik, 1980; Wechsler, 1945). It is important to
age, WAIS-R age interval groupings were
keep in mind that application of any one of
adopted as the standard against which to
compare the WMS-R studies, whereas the age these scoring procedures to the same WMS-R
group intervals of the WAI S-Ill were used for protocol will result in a different score. There-
the WMS-III and WMS-IIIA. fore, the scoring method used in the nonnative
report must be identical to that used by the
Sample Size clinician, to insure appropriate comparison.
As Wechsler (1987) noted, 50 cases have
generally been recommended (Guilford, 1965; Data Reporting
Hayes, 1963) as providing a reliable estimate Mean and SD scores should be provided, at
of the population mean. Following Wechsler's minimum.
(1987) lead and for the purpose of review, a With these requirements for reporting in
minimum of 50 subjects per age group inter- mind, the normative studies for the WMS-R,
val was deemed adequate. WMS-111, and WMS-IIIA were examined.

OF THE NORMS
status, occupation, ethnicity, and recruitment Performance on the WMS batteries is posi-
procedures are relevant. Until determined, it tively correlated with IQ (Lezak et al., 2004);
is best that this information be provided. however, reports of WMS-R, WMS-111, and
WMS-IIIA comparison data which control for Cullum (2003) noted, "One-hundred percent
IQ are surprisingly unavailable. Mittenberg retention of one word from Verbal Paired
et al. (1992) reports WMS-R scores for sub- Associates still indicates poor performance on
jects grouped by IQ estimates within each age that test" (p. 525).
group interval; however, these results were Problems with ceiling effects in the stan-
also based on small sample sizes. Almost every dardization sample on some subtests can also
study compensates, to some degree, for this affect scaled score performance. For instance,
limitation by reporting mean IQ and/or edu- Lacritz and Cullum (2003) point out that be-
cational level for subject groups. Still, an cause of ceiling effects in the WMS-111 stan-
empirical basis for judgments about the range dardization data for the Family Pictures and
of normal variance in memory that is attrib- Faces subtests, subjects under age 45 must
utable to variation in IQ cannot be found in retain 97-99% of the learned information just
the literature at this time for the WMS-R, to obtain a scaled score of 10 (i.e., 50th per-
WMS-III, or WMS-IIIA. centile). They also note that because almost
Seven of the eight WMS-R normative everyone in the standardization study obtained
reports reviewed present data from U.S. perfect recognition scores, in general, the
populations (with Marcopulos et al., 1997, distribution scores for the WMS-111 Auditory
presenting data for low-education, rural indi- Delayed Recognition Index are restricted and
viduals), and one study reports normative data negatively skewed. They suggested, therefore,
based on an Australian sample. that this should be considered more of a def-
There are no published data available re- icit index than an ability index.
garding recognition testing on delay for any of
the WMS-R or WMS-IIIA subtests. Without
assessing delay recognition for the material
not freely recalled, one is unsure whether SUMMARIES OF THE STUDIES
the patient primarily has an encoding or a This section presents critiques of the various
retrieval problem. Therefore, the WMS-R and normative studies for the WMS-R, WMS-111,
WMS-IIIA tests should be considered as and WMS-IIIA. The studies are reviewed in
providing a limited assessment of memory chronological order. Studies using the WMS-R
functioning until recognition testing proce- are presented first, followed by those using the
dures are developed and data become avail- WMS-III and the WMS-IIIA. The text of
able. Fortunately, the WMS-111 corrects this study descriptions contains references to the
problem by providing a recognition compo- corresponding tables identified by number in
nent following the delayed recall condition on Appendix 18. Table A18.1, the locator table,
most of the memory subtests. The WMS-111 summarizes information provided in the stud-
and WMS-IIIA are adequately normed for ies described in this chapter.
ages 16-89 years; however, the WMS-111
norms for recognition memory testing appear
to be quite forgiving. [WMS-R.1] Wechsler, 1987
Caution is warranted when interpreting This is the standardization of the WMS-R.
retention rates for the WMS-111 memory The standardization sample was "designed to
subtests. A high retention rate does not nec- represent the normal population of the United
essarily reflect an excellent memory. The States." Each case needed for the standardi-
reason is that the retention rate score does not zation sample was prespecified according to
allow one to make a distinction between the age, sex, race, and geographic region. Educa-
person who initially remembered a high per- tional levels of the standardization sample
centage of the material and later recalled a were "0-11 years, 12 years (high school
high percentage and the person who initially graduation or its equivalent), and 13 years or
remembered only a couple of elements of the more." The mean FSIQ estimate of the sam-
material and later recalled the same couple of ple at each age group was 100, with SD of 15
elements. In other words, as Lacritz and (see Other Comments, below). Although the
manual reports that the WMS-R provides screened over the telephone for neuro-
"norms stratified at nine age levels" (p. 2), close psychological, neurological, psychiatric, and
inspection reveals that normative data were medical disorders. The authors' stated inten-
collected for only six of the nine age groups and tion was to provide "preliminary" WMS-R
that the other "normative data" have been es- norms for healthy elderly individuals over age
timated statistically by interpolation. Specifi- 74 since the WMS-R standardization cuts off at
cally, no data were collected for three of the age 74 years. Educational level is presented for
nine reported age groups (18-19, 25--34, and each age group: 50-70, mean= 14.4, SD = 2. 7
45-54 years). The WMS-R is a proprietary test, years; 75-95, mean= 14.6, SD = 3.0 years.
and the interested reader is referred to the Delayed recall for lngical Memory, Verbal
manual for normative data presentation. Paired Associates, Visual Reproduction, and Vi-
sual Paired Associates was assessed at 30 minutes.
Study strengths
1. Age group intervals mirror WAIS-R in- Study strengths
tervals. 1. Adequate exclusion criteria. Sample
2. Sample size generally meets criteria. composition description is adequate.
3. Information regarding gender, ethnicity, 2. Information regarding education is
and geographic region is provided. provided.
4. Delayed recall procedures are followed 3. Delayed recall is assessed.
for four of the nine subtests. 4. Scoring and method of administration
5. Scoring procedures are well described in were according to Wechsler (1987).
the manual. 5. Data reporting included mean and SD
6. Data reporting include mean and SD scores.
scores. 6. Sample size is sufficiently large to allow
7. Sample composition description is cautious interpretation.
adequate.
Considerations regarding use of the study 1. The age group intervals are broad;
1. Normative data are not available by however, they are probably sufficiently
education and/or IQ level, although narrow to allow cautious interpretation.
information regarding IQ and educa- 2. High educational level.
tional characteristics of the sample was 3. No information regarding gender.
described.
2. Exclusion criteria were not fully detailed.
Other comments
Other comments 1. Forgetting rates (i.e., "savings scores") are
1. IQ scores were generally estimated. The also provided. Savings scores (SS) were
complete WAIS-R was administered calculated for I..ngical Memory, Visual
only to those 35-44 or 55--69 years old. A Reproduction, Verbal Paired Associates,
four-subtest short form (Vocabulary, and Visual Paired Associates using the
Arithmetic, Picture Completion, Block following formula: SS = (delayed recall/
Design) was administered to the other immediate recall) x 100.
four age groups on which normative data 2. The subject's last learning trial was used
were collected. as the measure of immediate recall.
3. All subjects were administered the De-
[WMS-R.2] Cullum, Butten, Troster, and mentia Rating Scale (DRS) (Mattis,
Salmon, 1990 (Tables A18.2, A18.3) 1976) to assess overall cognitive status.
The sample is described as a group of healthy, Ages DRS Total Scores
"above average" educated, community-
dwelling, older adult volunteers recruited via 50-70 141.1 (3.7)
flyers and newspaper advertisements and 75-95 139.8 (4.3)
4. The authors caution against application Considerations regarding use of the study
of the standard WMS-R index scores for I. Normative data are not reported by ed-
individuals over age 74 because this ucational or IQ level because sample size
"would likely result in an underestima- is small relative to this purpose.
tion of older subject's true abilities." 2. No information regarding gender.
Therefore, the WMS-R index scores
"must be used only as approximate esti- Other comments
mates of memory function" in individu- I. The authors note that the interpolated
als over age 74.
WMS-R norms presented in the manual
underestimate performance on the
[WMS-R.3] Mittenberg, Burton, Darrow, and "Attention/Concentration Index at the
Thompson, 1992 (Tables A18.4, A18.5) lower end of the score distribution and
This study provides empirically derived nor- overestimate scores on the VJSual Mem-
mative data for an age group (25--34) ne- ory and Delayed Recall Indexes at the
glected in the WMS-R standardization. The upper end of the distribution."
WMS-R manual reports estimated "norms"
statistically interpolated from adjacent age [WMS-R.4] Lichtenberg and Christensen, 1992
groups in the standardization sample. (Table A18.6)
Subjects were volunteers recruited in
This study, entitled "Extended Normative
Florida from "local businesses, evening/
Data for the Logical Memory Subtest of the
weekend adult education and vocational I
Wechsler Memory Scale-Revised: Responses
technical classes" and screened to exclude
from a Sample of Cognitively Intact Elderly
neurological, psychiatric, and alcohol prob-
Medical Patients," provides clinical compari-
lems. The sample was designed to match 1980
son data for a group of cognitively intact ge-
U.S. Census data stratified on age, gender,
riatric medical patients aged 70-99 years,
ethnicity, and education. Educational level of
seeking treatment in a hospital in Detroit.
the sample was described as 0-11 years
Michigan, one-third of whom were being seen
(n = 12), 12 years (n = 20), and 13+ years
for hip fractures from falling, one-third for
(n = 18). Mean WAIS-R estimated FSIQ was
knee replacements due to arthritis, and one-
101.3 (14.58, range 72-131, median= 100).
third for stabilization and recovery from a
The WAIS-R FSIQ estimate was based on
lengthy illness. The sample was comprised of
administration of the Vocabulary and Block
25 men and 43 women (35 Caucasian, 31
Design subtests only.
African American) screened for the absence of
neurological dysfunction. Information on
Study strengths years of education is provided for each age
I. Age group interval is adequate. group interval:
2. Sample size met minimal criteria.
70-74 group, education= 12.0 (3.7) years
3. Exclusion criteria are well described and
sample composition description is
adequate.
4. Educational level and IQ estimate of the Although the title of this report suggests
sample are described. that normative data are being supplied, the
5. Delayed recall procedures were accord- authors actually provide WMS-R Logical
ing to Wechsler (1987). Memory subtest clinical comparison data for a
6. Scoring procedures were according to sample of cognitively intact geriatric medical
Wechsler (1987). inpatients. Indeed, the author's caution that
7. Data reporting include mean and SD "the data presented here will be best used
scores. when applied to geriatric medical patients
8. Information provided regarding geo- seeking treatment in an urban medical set-
graphic area and recruitment procedures. ting" (p. 746).
Study strengths of 441 are as follows: VIQ = 105.5 (10.0),

1. Age group intervals are generally ade- PIQ= 107.3 (11.4), and FSIQ= 106.6 (10.5).
quate; however, the 80-99 age category "nle MOANS recruits participants from two
is broad. ongoing research projects at the Mayo Clinic
2. Information regarding years of educa- in Rochester, Minnesota. One of the goals of
tion is provided for each age group research project 1 is to obtain age-specific
interval. norms on traditional and experimental cogni-
3. Delayed recall for the Logical Memory tive tests. Potential normal volunteers were
subtest is assessed according to Wechs- captured by sampling all recent medical ex-
ler (1987). aminations performed at Mayo's Department
4. Scoring procedures were according to of Community Internal Medicine (CIM). A
Wechsler (1987). participant was deemed "normal" if he or she
5. Data reporting included mean and SD was able to function independently and lacked
scores. any active neurological or psychiatric condi-
6. Information regarding gender, ethnicity, tions that might compromise cognitive status.
and geographic area is provided. Sample All subjects had a thorough medical screening;
composition description is adequate. however, chronic medical illness was not an
exclusion criterion. For instance, persons with
Consideration regarding use of the study diabetes, cardiac problems, and hypertension
1. Sample size within each age group was were included in the "normal" sample. Indi-
relatively small. viduals deemed normal by chart review were
added to a possible participants list. Partici-
Other comments pants cuJled from research project 1 were
1. Medical history ruled out suspected ce- willing to undergo a 4-hour outpatient clinic
rebral damage, and the Mattis Dementia visit, during which a battery of neuropsycho-
Rating Scale (cutoff score= 129) was logical tests was administered.
administered to ensure intact cognitive Research project 2 is a component of the
functioning. Mayo Clinic's ongoing Alzheimer's Disease
Patient Registry. "ntis project follows up all
70-74 group, DRS= 134.7 (3.6)
newly diagnosed persons with dementia who
75-79 group, DRS= 134.9 (4.2)
present for examination at Mayo's Department
80-99 group, DRS= 133.3 (3.9)
of CIM. Every patient is demographically
matched to a normal control. "nle normal
[WMS-R.SJ lvnik, Malec, Smith, Tangalos, control is obtained by recruiting individuals
Petersen, Kokmen and Kurland, 1992b who present for a general medical examination
This study presents age-corrected (aka, age- to any CIM internist. To be considered "nor-
specific) norms for the WMS-R derived from a mal," the individual must have a Mini-Mental
sample of 441 cognitively normal individuals State Exam (Folstein et al., 1975) score >24,
aged 56-94 who are participants in Mayo's and the internist must satisfy "him or herself
Older American Normative Studies (MOANS). that the patient is indeed cognitively normal"
Age groups were ~74 (n=274) and 75-94 (Ivnik et al., 1990). "nlese individuals also re-
(n = 167). Information regarding the IQ of the ceived a neurological examination. Data from a
sample was provided; however, valid WAIS-R potential recruit were reviewed by a team,
Verbal IQ (VIQ), Performance IQ (PIQ), and which included a geriatrician, neurologists,
FSIQ could be calculated only for the 274 in- and neuropsychologists. "nlese individuals had
dividuals aged ~74 since the standardization also been administered a battery of neu-
data could be obtained from the WAIS-R ropsychological tests, but the results of this
manual. For this group, the actual WAIS-R IQ evaluation were not used in determining the
summary data are as follows: VIQ = 106.1 normality of the potential subject. The criteria
(10.0), PIQ = 108.0 (11.7), and FSIQ = 107.3 for normality were otherwise the same as for
(10.6). "nle MAYO values for the entire sample project 1.
Potential participants were randomly soli- age 74. To accomplish the extension,
cited from projects 1 and 2, and approximately data were collected on normal volun-
34% agreed to participate. A further exclusion teers aged 56-74. Data collected from
criterion was inability to complete any neu- this younger sample were then com-
ropsychological test that was administered, pared to the actual 56-74 WMS-R
regardless of reason. norms available in the test manual to
The MOANS normative sample consists of determine how the two groups differed.
primarily well-educated (high school or Knowing how the younger sample dif-
greater) Caucasian adults living in Rochester, fered allowed Ivnik and colleagues to
Minnesota, and the immediately surrounding, "correct" the norms collected from their
"primarily agricultural" communities of Olm- 74-94 sample so that they would corre-
stead County. spond with what probably would have
been the WMS-R norms for those 74-94
Study strengths years old had they actually been col-
1. Age group intervals are generally ade- lected during the WMS-R standardiza-
quate, with the exception of the table tion. In other words, this research
listing data for an 88-year midpoint. (The provides a statistically derived estimate
age group interval for this midpoint va- of probable WMS-R values for individ-
lue table is actually 83-94 years.) uals older than age 74. The differences
2. Sample size is adequate. noted in performance values between
3. Sample composition description is well the WMS-R standardization sample and
detailed in the current and prior reports the MOANS sample most probably re-
(Ivnik et al., 1990, 1991, 1992a; Malec flect the different sampling procedures
et al., 1992). employed in the two studies as well as
4. Reporting of IQ and educational levels differences in sampling national vs. re-
was detailed. gional populations.
5. Delayed recall conditions were admin- MAYO procedures convert all WMS-R
istered according to the WMS-R manual. subtest raw scores to age-corrected, nor-
6. The scoring procedure is well described malized scaled scores "before any other
according to standardized procedures algebraic or tabular conversions" are un-
detailed in the test manual. dertaken.
7. Adequate exclusion criteria with the ex- Regarding the "normalizing" proce-
ception of chronic medical illness. dure, each raw score was first converted
to a percentile rank "based upon the ac-
Considerations regarding use of the study tual cumulative percent distribution."
1. Data reporting does not include pre- Each percentile rank was then converted
sentation of mean and SD scores. Ra- to a scaled score (M = 10, SD = 3).
ther, tables are provided which convert Within each age group, "the percentile
WMS-R raw scores to percentile ranks range encompassed by each scaled score
and age-specific scaled scores for mid- was set so that the resultant distribution
point ages occurring at 3-year intervals of scaled scores was as normal as possi-
from 61 through 88. The age range ble," resulting in age-corrected scaled
around each midpoint age is ±5 years. scores. In determining age-corrected
MOANS WMS-R summary indices, lin-
Other comments ear transformations of the sums of
1. The sample consists of primarily well- MOANS scaled scores were computed to
educated, urban, Caucasian adults, convert the distribution to the accepted
which should be considered when using WMS-R standard (i.e., M = 100,
these data. SD = 15). The resulting sumrruuy scores
2. One of the purposes of this study was to are referred to as MAYO Verbal Memory
extend the norms on the WMS-R above Index, MAYO Visual Memory Index,
MAYO General Memory Index, MAYO data, they are not reproduced here. The
Attention/Concentration Index, MAYO reader is referred to the original article.
Delayed Recall Index, MAYO Percent
Retention Index. The MAYO Percent
[WMS-R.61 Richardson and Marottoli,
Retention Index is new and computa-
1996 (Table A18.7)
tionally derived.
3. A major assumption of Ivnik and col- This study provides education-corrected nor-
leagues is as follows: mative data for the performance of adults aged
persons above age 74 who agreed to senre as 75-91 years on commonly administered neu-
nonnal volunteers are demographically similar ropsychological tests, including the WMS-R.
to those age 56 through 74, since they were The sample consists of a subset of 101 adults
drawn from the same population via identical (53 males, 48 females) recruited from a larger
sampling procedures. We further assume that longitudinal study (Project Safety, n = 1,103),
the differences which exist between our youn- aged 76-91 years, who were active drivers, free
ger sample and the [WMS-R] national sample from neurological and psychiatric disease, and
would be similar for the persons above age 74 living independently in an urban community in
if we were able to compare these older per- the northeastern United States (e.g., New
sons to a national nonnative sample of like age. Haven, CT).
(p.5) Educational levels were reported as follows:
4. The administration and scoring proce- Age 76-80 years 81-91 years
dures for the MAYO system are identical Education 10.44 (3.86) 11.59 (3.45)
to the WMS-R with the exception of the Sample size 50 51
Verbal and Visual Paired Associates I Range for entire sample: 4th grade through college
subtests. Specifically, for the Verbal and Sample sizes for age education categories
Visual Paired Associates I subtests, if the are as follows:
criterion was not reached by the third
learning trial, additional trials were not Age 76-80 81-91
administered as required by the WMS-R Education <12years ~12years <12years ~12years
manual. As a result, the MAYO Delayed
Sample size 26 24 18 33
Recall Index is not directly comparable
to the WMS-R Delayed Recall Index.
5. Since Ivnik and colleagues are applying Study strengths
the same norming procedure to each 1. Age group intervals are adequate.
test in their extensive neuropsychological 2. Adequate exclusion criteria. The sample
battery, this should enhance the com- description was adequate.
parability of test scores across tests. 3. Education for the sample was reported.
6. MAYO indices differ from WMS-R in- 4. Delayed recall was assessed for the
dices, even for those aged 56-74 years. Logical Memory and Visual Reproduc-
Therefore, if you are evaluating someone tion subtests.
who was previously tested at age 74 or 5. Description of the administration and
younger and is now above age 74, you scoring procedures was according to the
will need to convert the prior exam WMS-R test manual.
scores to MOANS equivalents in order 6. Data reporting included mean and SD
to compare old vs. current performance. scores.
7. MAYO and Wechsler summary scores 7. Information regarding gender and geo-
are not interchangeable. Ivnik et al. graphic area is reported.
(1993) note that the difference can be as
large as ±17 points. Consideration regarding use of the study
Due to the innovative nature and the 1. Sample size is adequate for data reported
voluminous amount of the normative by age category; however, it becomes
quite small when data are preset.ted by were reported by age category as follows: 0-4,
age and education. 5, 7, and 9-10 years. The sample contains pri-
marily females.
Other comments
1. Logical Memory subtest data for the less Study strengths
educated 81-91 group should he used 1. Age group intervals are generally ade-
with caution due to small sample size quate.
and the finding that the norms deviated 2. Adequate exclusion criteria. Sample
significantly from normal when com- composition description is adequate.
pared to other available normative data. 3. Education data were reported and
2. The authors note that since their ~ample grouped by age category.
was not medically screened an~ indi- 4. Delayed recall for the Logical Memory
viduals were not ruled out if ~y had and Visual Reproduction subtests is as-
more common medical diseaseS (e.g., sessed.
hypertension and diabetes), thq "data 5. Data reporting included mean and SD
should be viewed as reflecting the 'nor- scores.
mative' population of urban driver older 6. Information on gender, ethnicity, and
than age 75 rather than the perfotmance geographic area is provided.
of 'normal' individuals."
1. Sample size is generally too small to al-
[WMS-R.7] Marcopulos, Mclain, and Giqliano,
low valid use of the normative data.
1997 (Tables A18.8-A18.11) .
2. Scoring procedures were not specifi-
The goal of this study was to develop norma- cally described but presumably fol-
tive data for rural, community-dwelling adults lowed the protocol in the WMS-R
aged 55 and older with no more than 10 years manual.
of formal education. Participants w~e ad- 3. The sample is primarily female; there-
ministered a battery of nine neuropsJcholo- fore, caution is suggested when applying
gical tests, which included the WMS-~. the data to males.
The data were obtained from a biracial
sample (n = 133; white = 64, African ~meri Other comments
can = 69) of nondemented, healthy adults aged 1. The percent retention as a measure of
55 and older (mean age= 76.48, SD 7.87,f the rate of forgetting (i.e., savings score)
range unknown), who attended schoollfor 10 was calculated using the following
years or less (mean education= 6.65: years, formula:
SD=2.14, range 0-10 years) and we~ edu-
Savings score% = (delayed recaU/
cated and primarily lived in a rural cominunity
immediate recall) x 100
setting (central Virginia). Subjects wEire ex-
cluded if they reported the presenct of a The authors found the savings score to
chronic or severe psychiatric disorder, lustory be relatively impervious to the effects of
of extensive psychotropic drug use, long-term age and suggested that it may be "the
substance abuse, neurological diseaset elec- most sensitive and specific indicator of
troconvulsive therapy, head injury, or 1oss of abnormal memory functioning."
consciousness. Subjects were paid $~ and 2. The authors highlight the fact that most
received a certificate of appreciation for com- of the normative data in the literature
pleting a 2-hour test battery, which was are based on urban, high school-
administered in one session. Data were re- educated, white adults. They caution
ported by age and again by age/educati~. Age that "the use of extant norms with lower
group categories were as follows: 55-64, 65-74, educated, rural-dwelling, older adults
75-84, and 85 +. The upper range of thqoldest can overestimate degree of cognitive
age category is not known. Educatio~ data impairment." Since 26% of the elderly
live in nonurban areas and rural indi- Considerations regarding use of the study
viduals tend to be disadvantaged with 1. Age group interval is too broad.
regard to education, the authors urge the 2. Data reporting did not include mean and
development of more normative data for SD scores; however, raw scores and their
this group. corresponding scaled score equivalents
can be obtained from the tables.
[WMS-R.Bl Shores and Carstairs, 2000 Other comments

One of the goals of the Macquarie University 1. The authors present several normative
Neuropsychological Normative Study (MUNNS) tables regarding various aspects of the
was to develop WMS-R normative data for WMS-R in their excellent article, and
adults aged 18--34 living in Australia. Partici- the interested reader is referred to this
pants were administered the WMS-R as part document for Australian normative in-
of a larger battery of neuropsychological tests formation.
that were also being normed for Australian
populations. The WMS-R was administered [WMS-111.11 Wechsler, 1997
and scored according to the methods de- This is the standardization of the WMS-111.
scribed in the standardization manual. The The normative information was generated
only deviation from standard procedure was from a national sample stratified by age, sex,
that Logical Memory subtests I and II were race/ethnicity, educational level, and geo-
administered with the Australian idiom stories graphic region, representative of the 1995
reported by Ivison (1993). U.S. population. The data are a subset
The data were obtained from a sample of (n = 1,250) of a larger stratified sample col-
399 healthy adults (193 males, 206 females) lected while norming the WAIS-III (n =
living in the Sydney metropolitan area. The 2,450). Subjects were located "primarily
methodology and sample characteristics were through the use of marketing research firms in
described in detail in an earlier report by 28 U.S. cities in the Northeast, North Central,
these researchers (Carstairs and Shores, South, and West regions." Subjects were re-
2000). Data were reported by age/educational cruited by random telephone calls, newspaper
level and again by gender. Education data ads, and flyers posted at senior centers and
were reported by age category as follows: various community organizations. Several in-
<12 (n = 91), 12 (n = 91), and >12 (n = 217) dependent examiners recruited and tested
years. additional subjects. Standardization sampling
sites (marketing researchers and independent
researchers) were located in 47 of the 50
Study strengths states (Hawaii, Utah, and North Carolina were
1. Adequate exclusion criteria and sample evidently not included). All participants
composition description per Carstairs were paid for taking the tests. Participants
and Shores (2000). were excluded if they were color-blind, had
2. Sample size is adequate. uncorrected hearing loss or visual impairment,
3. Education data were reported and were in current treatment for drug or alcohol
grouped by age category. dependence or consumed more than three
4. Delayed recall for the Logical Memory alcoholic beverages on more than two nights a
and Visual Reproduction subtests is as- week, were taking any psychiatric medica-
sessed. tions, were seeing any professional for think-
5. Information on gender and geographic ing or memory problems, had a disability
area is provided. that would affect motor performance, re-
6. The administration and scoring criteria ported a history of unconsciousness for ~5
follow the WMS-R manual. Exceptions minutes, or had a history of any medical or
to the standard administration are ap- psychiatric condition that could potentially
propriately referenced. alter cognitive functioning (see WAIS-III,
WMS-III Technical Manual-Updated 2. The WMS-III battery, similar to the

[Wechsler, 2002a], p. 21 for full details). Per- WMS and WMS-R, is designed so that
formance scores are reported for 11 age individual subtests can be administered,
groupings in the range 16-79 years. Each of scored, and interpreted as needed.
these age group categories contains data on 100 3. For ages 16-64, an equal number of
participants. The two oldest age groui?ings in males and females participated; how-
the range 80-89 years have a sample si2lfl of 75. ever, for ages 65-89, the proportions
This is a proprietary test, and the normative reflected 1995 census data.
data are available in the test manual.: Scaled
scores can be converted to percentile~ by ex- [WMS-IIIA.1] Wechsler, 2002b
amining Table E.1 on page 200 of thai WMS-
lll Administration and Scoring )lanual As noted earlier, the WMS-IIIA is basically a
(Wechsler, 1997). Demographically ajljusted "carve-out" and repackaging of the WMS-III
norms are also available when usihg the as a screening instrument to help determine
computerized scoring program that ii avail- whether further memory or neuropsychologi-
able from the test publisher. ! cal evaluation is needed. As such, the
WMS-IIIA uses Logical Memory and Family
Study strengths Pictures subtest stimuli that are identical to
1. Age group intervals are adequata those used in the WMS-III. The scoring pro-
2. Sample size met criteria. tocols are also identical to the WMS-111. Since
3. Information on education, gender, eth- the norms for this short-form version were
nicity, geographic area, and recn.Ptment carved out of the original WMS-111 stan-
procedures is provided. ' dardization, details of the standardization
4. Delayed recall and recognition iproce- sample will not be repeated here. Please
dures are standard on all apprppriate refer to the descriptions provided above
subtests. for the WMS-111 for details regarding the
5. Scoring procedures are well descdbed in standardization.
the manual. ' This is a proprietary test, and the normative
6. Adequate exclusion criteria. Sample data are available in the test manual.
composition description is adequate.
Study strengths
Considerations regarding use of the st~dy 1. Age group intervals are adequate.
1. Data reporting did not include mean and 2. Sample size met criteria.
SD scores; however, raw scores arld their 3. Information on education, gender, eth-
corresponding scaled score equwalents nicity, geographic area, and recruitment
can be obtained from the tables. ~ procedures is provided.
4. Delayed recall procedures are standard
Other comments on the subtests.
1. Education is positively correlateJl with 5. Scoring procedures are well described in
performance on memory tests (Lezak the manual.
et al., 2004). In 2002, the publishet of the 6. Adequate exclusion criteria. Sample
WMS-III made available age/eduf:ation- composition description is adequate.
corrected norms; however, the age/
education-corrected norms may be ob-
tained only by purchasing a CODJputer- Considerations regarding use of the study
ized scoring program for the WAjS-III- 1. Data reporting did not include mean and
WMS-III (see Appendix 1 for oidering SD scores; however, raw scores and their
information). Data are not available by corresponding scaled score equivalents
age and IQ; however, some comptlrisons can be obtained from the tables.
can be made with the WAIS-III si.lce the 2. No recognition testing following the
two tests were co-normed. delay recall condition.
Other comments interpolation between points on a continuous

1. Although education is positively corre- variable like age" (Bowden & Bell, 1992,
lated with performance on memory tests p. 34), granting permission to a commercial
(Lezak et al., 2004), WMS-IIIA norma- test developer/distributor to interpolate data
tive data are not available by age and across decade increments seems unjustified.
education. Data are also not available by We feel strongly that reliance on such
age and IQ; however, some comparisons estimated normative data for interpretive
can be made with the WAIS-111 since purposes raises professional concerns. High-
the two tests were co-normed. lighting our concern, Mittenberg et al. (1992)
2. For ages 16-64, an equal number of provided empirically derived provisional
males and females participated; how- norms for the 25-34 age group. Differences
ever, for ages 65--89, the proportions between these norms and the WMS-R pub-
reflected 1995 Census data. lished index scores appear to be clinically
significant. (With the more recent develop-
ment of the WMS-111, the publishers noted
that no data were interpolated in the revised
CONCLUSIONS
standardization.)
There are currently three published versions The WMS-R normative data developed by
of Wechsler's memory battery that are com- lvnik et al. (1992b) are also recommended as
mercially available: WMS-R (1987), WMS-111 long as the demographics of lvnik's partici-
(1997), and WMS-IIIA (2002). The strengths pants match closely the demographic charac-
and considerations regarding use of the extant teristics of the subject under study.
comparison data for the various versions of the Despite the fact that educational level was
test battery are summarized below. 1 significantly related to test performance in the
standardization sample, WMS-R normative
data stratified by age and education, as well as
WMS-R (1987)
data for various ethnic and cultural groups,
The WMS-R has remained popular since its are lacking. Although four of the nine subtests
introduction over 15 years ago, and WMS-R of the WMS-R require delayed recall of the
subtests have continued to be included in material presented, a recognition testing for-
some research protocols, which insures that we mat was not provided, making it impossible to
will continue to see data using this test for determine whether a poor recall score is due
some years. However, given the improvements to an encoding problem or a retrieval problem
in assessment achieved with the WMS-111, the or both. The interested reader is referred
WMS-R should be phased out and replaced to Fastenau (1996a), who has developed
with the WMS-111 in clinical settings. multiple-choice recognition stimuli for the
Regarding norms for the WMS-R, for WMS-R Logical Memory and Visual Repro-
Americans of average intelligence in the age duction subtests.
ranges 16-17, 20-24, 35-44, 55--64, 65-69, Over 15 years have passed since the intro-
and 70-74, the data provided with the WMS-R duction of the WMS-R, and considerable re-
should receive the strongest attention. The search has accumulated regarding its utility
data presented for age groups 18-19, 25-34, with various clinical populations. Although the
and 45-54 most probably should be used only WMS-R is a better instrument than the orig-
with appropriate caution since the values were inal WMS, the WMS-111 should be considered
estimated by statistical interpolation. Although the preferred version of the test for the as-
"test standardization of necessity involves sessment of memory functions. Research us-
ing the WMS-R is winding down and should
'Meta-analyses were not performed as this review was not soon cease. Therefore, clinicians and re-
intended to summarize all of the voluminous literature searchers are encouraged to use the WMS-111.
available on this test. Conversely, comprehensive sets of
norms are available in the test manuals for all versions of However, it still may be necessary to occa-
theWMS. sionally use sections of the WMS-R when
evaluating a patient previously assessed on and educated in the United States. However,
this older form of the test. although recognition memory testing has
been added to several of the WMS-III sub-
tests, this is an area that needs further im-
WMS-111 (1997) and WMS-IIIA (200~)
provement in terms of test development and
The larger size of the normative data sample norms. Since education has been consistently
(n = 1,250, aged 16-89) for the w.MS-III found to enhance WMS test performance,
compared to the WMS-R represen~ a sig- future normative studies involving the WMS-
nificant improvement. The WMS-~ was III and WMS-IIIA should report data by age/
normed on only 316 subjects aged ·16-74. education categories in the test manual.
Because the WMS-III was co-normad with Relative to earlier editions, the WMS-111
the WAIS-III, one can now perform analyses definitely reflects the improvements in psy-
of difference scores as well as more s4>phisti- chometrics and advances in scientific and
cated test pattern analyses. The W~S-IIII clinical understanding of learning and mem-
WMS-IIIA standardization norms pro~ded in ory function and dysfunction that have ac-
the manual appear to be adequate ftr ages crued over the years. However, we do look
16-89, especially if applied to individu. born forward to the WMS-IV.
19
List-Learning Tests
REY AUDITORY-VERBAL LEARNING TEST Rey AVLT has gained remarkable popularity
among clinicians. Interpretations of clinical
The Rey Auditory-Verbal Learning Test (Rey
data, however, are obscured by considerable
AVLT) has been extensively used to evaluate
variability in administration of the test. There
memory functioning in normal samples and in
is very little uniformity in various procedural
a variety of clinical samples representing dif-
aspects.
ferent medical and psychiatric conditions. The
test was introduced by the Swiss psychologist
Andre Rey (1941) as a measure to assess the 1. The administration procedure varies
inconsistency in relative performance on the widely. The standard administration in-
recall task vs. the recognition task. The test was cludes five successive presentations of
described in English by Taylor (1959). The the original list of 15 words, followed by
English translation of the Rey AVLT does not free recall on each trial; an interference
correspond exactly with the original French trial, involving presentation and free re-
version: three French words were substituted call of another list of 15 words; post-
in the translation (bell for belt, moon for sun, interference free recall of the words
and nose for mustache). Later, Rey (1964) from the original list; and a recognition
modified the procedure to include five free trial (Lezak et al., 2004). A number of
recall trials and a recognition task. Contem- studies have also utilized delayed recall
porary versions of this test (described in Lezak, and delayed recognition trials (Lezak
1995; Lezak et al., 2004) include an interfer- et al., 2004).
ence trial (first introduced by Taylor, 1959) In contrast to this standard, some
and a postinterference recall trial; the subse- studies use a different number of recall
quent additions are the delayed recall and trials, which varies from three (White,
delayed recognition trials (Lezak, 1995; Lezak 1984) to six (Madison et al., 1986). The
et al., 2004; Spreen & Strauss, 1991, 1998). interference trial is omitted in several
studies (Miceli et al., 1981; Bolla-Wilson
& Bleecker, 1986; Bleecker et al., 1988).
Variability in Administration
Uzzell and Oler (1986) omitted the
of the Rey AVLT
postinterference recall trial and pre-
Due to its usefulness in detection and identi- sented recognition immediately follow-
fication of faulty memory mechanisms, the ing recall of the interference list. Squire
357
and Shimamura (1986) and Shimamura differences in delay intervals should be

et al. (1987) modified their procedure to taken into consideration, even though
present a recognition trial after each of Lezak et al. (2004) report a minimal de-
the five acquisition trials, without using cline in recall over a 30-minute interval.
free recall trials. Shimamura et al. (1987) In addition, the delay interval is 6lled
presented the words in a different order with various activities in different studies,
on each acquisition trial. which may influence performance on the
2. The format of the recognition trial varies delayed recall and recognition trials.
widely: Rey (1964) and Lezak et al. 4. The rate of presentation differs across
(2004) described a story format in which studies. According to Rey (1964) and
all 15 words from the original list are Taylor (1959), each word should be
imbedded. The original story described separated by a !-second interval. Rey
by Rey contained twice as many dis- also suggested recording the number of
tracter nouns as the one being used in words remembered in each 15-second
current studies. According to Rey block to provide an indication of the
(1964), the story was to be read to the rhythm of recall. According to Lezak
subject, who was instructed to stop the et al. (2004), the rate of presentation
examiner when a word was recognized. should be one word per second. Some
The story described by Lezak was to be authors have used a slower rate of pre-
read by the subject with the instructions sentation, e.g., one word every 2 seconds
to circle recognized words. In addition to (White, 1984).
the story format, Lezak described other 5. Time allowed for recall differs between
versions of the recognition trial, consist- studies. According to Rey's (1964) in-
ing of lists of 50 words which include structions, the first presentation of the
words from list A, the interference list, list should be followed by recall within
and 20 words phonemically and/or se- 60 seconds, and on subsequent acquisi-
mantically similar to the words from tion trials, 90 seconds should be allowed
both lists presented to the participants. for recall. The majority of investigators,
Presentation of the lists can be either however, allow unlimited time for recall.
auditory or visual. It should be noted This aspect of administration is not
that other versions of the recognition list usually described by the authors.
have been proposed. For example, Ivnik The degree of encouragement on the
et al. (1987, 1990, 1992c) used a 30-word part of the examiner to elicit maximal
list for recognition. effort from the subject also varies be-
The order of administration of the tween the studies.
recognition trial also varies between 6. The extent of feedback about the sub-
studies. The recognition trial is admin- ject's performance on each trial might
istered after the postinterference recall influence the results. According to Rey
trial, after the delayed recall trial (with (1964) and Taylor (1959}, the feedback
varying delay intervals), or after both- had to be given to the subject each time
postinterference and delayed recall tri- the word was repeated within one trial.
als, which influences performance on the Rey also instructed test administrators to
last recognition trial. provide the subject with the feedback on
3. The interval for the delay varies from 15 the number of words recalled on each
minutes (Miceli et al., 1981) to 60 min- trial and to inform the subject on the
utes (Ivnik et al., 1987). Geffen et al. fifth recall trial that this is the last trial.
(1990, 1994), Seines et al. (1991), and The current standard administration
Miller (personal communication) used a does not follow these guidelines. Ac-
20-minute delay, whereas Ivnik et al. cording to Lezak et al. (2004}, the ex-
(1990, 1992c) and Savage and Gouvier aminer should not volunteer information
(1992) used a 30-minute delay. These about repetition of the same words
LIST-LEARNING TESTS 359
within one trial unless this infonnation is factor-analytic study conducted by Ryan et al.
solicited by the subject because this (1984). Vakil and Blachstein (1993) factor-
might cause distraction. analyzed the Rey AVLT perfonnance of 146
nonnal participants. The basic factors ex-
In view of the variability in administration tracted were acquisition and retention. The
procedures for the Rey AVLT and considering latter factor was further subdivided into stor-
the usefulness of the test and extensive re- age and retrieval. 'ntis information suggests
search database, Schmidt (1996) published a that verbal memory functioning can be sub-
handbook which summarizes the current sta- divided into separate mechanisms, which is
tus of the administration, scoring, and nor- consistent with empirical data accumulated in
mative resources for the Rey AVLT. The the field of experimental psychology. The in-
author outlines the historical development of tegrity of these mechanisms is reflected in the
the test, the role of demographic factors on test different indices derived from Rey AVLT
perfonnance, and statistical properties with performance.
respect to different clinical groups; describes Recall on trial I represents immediate
alternate fonns and test-retest comparisons; memory span for words. In most cases, im-
addresses issues of qualitative and quantitative mediate memory span for words is expected to
interpretation of the results; and suggests ad- be consistent with the immediate memory
ministration instructions, which are derived span for digits. A discrepancy between recall
from the instructions described by Lezak on trial I and digit span in favor of digit span
(1976, 1983) and Spreen and Strauss (1991). can be attributed to infonnation overload
The review includes summaries of nonnative (Geffen et al., 1990). Similarly, recall on trial I
data for basic scores, additional scores, and is expected to be roughly similar to recall on
derived indices for different age groups- the interference trial because both present
children, adolescents, adults, and elderly-in equivalent word lists for the first time. Supe-
the context of reviews of the empirical studies riority in recall on trial I over the interference
from which these data are gathered. In addi- trial may indicate the effects of proactive in-
tion, the author provides metanonns, which terference on the latter. The opposite pattern
are derived by calculating a pooled mean and might suggest difficulties in changing response
standard deviation (SD) for the relevant stud- set (Lezak et al., 2004). Furthermore, compar-
ies. The handbook contains test fonns and ison of recall on trial V vs. the postinterference
stimulus sheets which are reprinted from the trial provides a measure of retention for newly
literature. Peaker and Stewart (1989) also learned infonnation and its loss due to retro-
published a review of the Rey AVLT. active interference. Both proactive and retro-
Additional infonnation on the history and active interference are thought to reflect
procedural aspects of the Rey AVLT is pro- executive and/or memory dysfunction (Bluse-
vided in Lezak et al. (2004) and Spreen and wicz et al., 1996; Gershberg & Shimamura,
Strauss (1998). 1995; Kareken et al., 1996; Paulsen et al.,
1995a; Torres et al., 2001; Yehuda et al., 1995).
Fleming et al. (1995), Kareken et al. (1996),
functioning of Different Memory
and Torres et al. (2001) found increased sus-
Mechanisms, as Assessed by
ceptibility to retroactive interference in
the Rey AVLT
schizophrenic patients, which was attributed
'nte advantage of the Rey AVLT is that it not by Torres et al. (2001) to the frontally medi-
only provides a measure of rote verbal mem- ated central executive dysfunction.
ory but also allows partitioning of memory Change in performance over five acquisi-
processes into their components and identi- tion trials indicates a learning curve, and its
fying faulty memory mechanisms that lead to slope provides a measure of verbal learning.
memory loss. Verbal learning over five acquisition trials is
The Rey AVLT was reported to load on also commonly identified as the recall differ-
a verbal learning and memory factor in a ence between trial V and trial!. Some authors
define learning as the difference between the with ruptured and repaired anterior commu-
highest recall on any trial and recall on trial I nicating artery aneurysm. However, on delayed
(Query & Megran, 1984). recall, these patients demonstrated neither a
Analysis of the recall pattern might shed primacy nor a recency effect, which indicates
light on the use of learning strategies and or- very poor recall from any part of the list. The
ganization of the material. Mungas (1983) serial position effect also reflects the predictive
introduced a measure of consistency of se- role of recall efficiency from the middle seg-
quential organization, reflecting the degree to ment of the list. Ryan et al. (1992) reported
which pairs of words that are recalled con- significantly lower recall of words from the
secutively on one trial are recalled consecu- middle segment in AIDS patients compared to
tively on the next trial. Woodard et al. (1999a) controls in a federal corrections sample.
used analytical decomposition of the learning The rate of forgetting over time can be ex-
curve over five acquisition trials to examine the plored through comparison of recall on the
relative contribution of acquisition and con- postinterference trial and the delayed recall
solidation mechanisms to the multitriallearn- trial. Comparison of performance on recall vs.
ing deficit in patients with Alzheimer's disease. recognition conditions provides additional in-
This method yields measures of gained items sight into faulty encoding, storage, or retrieval
and lost items across consecutive learning mechanisms. Comparison of the number of
trials, which are expressed as gained access words recognized from list A and the inter-
(acquisition) and lost access (consolidation ference list suggests differences in storage of
failures that lead to rapid intertrial forgetting). overlearned vs. once-learned material.
The authors assert that this method is useful in Analysis of errors also provides useful in-
uncovering the cognitive processes underlying formation on the integrity of different memory
learning deficits in persons with memory dis- mechanisms. Rey (1964) instructed that, in
orders and in characterizing potential areas for addition to the number of correct responses,
remediation. the numbers of false responses, repetitions,
Serial position effects provide information on and repetitions where the subject questioned
the functioning of specific memory mecha- recalling the word previously should be re-
nisms. Primacy and recency effects in the recall corded. Spreen and Strauss (1998) and Lezak
pattern provide another index of vulnerability et al. (2004) suggest a notation system to
to proactive and retroactive interference. These identify the quality of errors on recall and
indices are useful in distinguishing amnesias recognition trials.
affecting encoding mechanisms (in which case The quality of intrusion errors has been
the predominance of the recency effect would analyzed in several studies. Intrusions of words
be evident) and those conditions which com- from list A on recall of the interference list or
promise the efficiency of recall but spare basic from the interference list on postinterference
encoding processes (demonstrating intact pri- recall provides evidence of proactive interfer-
macy and recency effects). For example, Tier- ence and weakness in source or context
ney et al. (1994) reported greater recency than memory (Geffen et al., 1990). Extralist intru-
primacy effects in Alzheimer's patients, sions reveal a tendency for semantic or pho-
whereas Parkinson's dementia patients as well netic confusion or confabulatory responses.
as control participants demonstrated intact Intrusions as well as the tendency to repeat
primacy and recency effects. Similarly, Bigler words from the list more than once reflect
et al. (1989) reported only recency effects in a impairment in self-monitoring functions
sample of Alzheimer's patients. In contrast, (Lezak et al., 2004).
Crockett et al. (1992) did not find differences lvnik et al. (1990, 1992c) and Geffen et al.
in the magnitude of primacy/recency effects (1990, 1994) developed additional measures of
in patients with anterior and posterior brain performance, which allow exploration of the
damage and in psychiatric inpatients. Stefanova effect ofdifferent memory mechanisms on mem-
et al. (2002) reported intact primacy/recency ory functioning. Similarly, Vakil and Blachstein
effects during five acquisition trials in patients (1994) developed a measure to assess incidental
liST-LEARNING TESTS 361
learning of temporal order, which provides an presented with the same vs. different forms of
additional index of retention. the Rey AVLT on retest (27 ± 3 days after
The contribution of the Rey AVLT to the original testing) indicated that the practice
accuracy of diagnostic determination in head effect was evident only in the group exposed to
injuries, dementias, amnesias, and frontal lobe the same list. Therefore, the practice effect is
syndrome has been well documented (Arm- largely due to the retention of specific test
strong et al., 1996; Bigler et al., 1989; Blach- material rather than to the metamemoric fac-
stein et al., 1993; Glennerster et al., 1996; tor and can be overcome using an alternate
Guilmette & Rasile, 1995; Heubrock, 1995; form on the retest. Uchiyama et al. (1995)
Janowsky et al., 1989; Lezak et al., 2004; Lucas found a significant practice effect on 1-year
& Sonnenberg, 1996; Mitrushina et al., 1994, longitudinal follow-up in males with an aver-
1995b; Woodard et al., 1999a). Data for a age age of 36.55 (7.19), which was of a com-
sample of recent-onset spinal cord injury pa- parable magnitude for two alternate forms.
tients are presented by Kurylo et al. (2001). Shapiro and Harrison (1990) suggested that
In addition to its usefulness in the identifi- when the test-retest sessions are spaced very
cation of faulty memory mechanisms, several close to each other (5 days), there remains a
Rey AVLT indices can be used as measures of general practice effect due to repeated ad-
motivational level and cooperation in the ministrations in healthy college students but
testing procedures (Ashendorf et al., 2003; not among the older patient population. Data
Bernard, 1990, 1991; Bernard et al., 1993; on repeated administration are also presented
Binder et al., 1993, 2003; Greiffenstein et al., by McCaffrey et al. (2000).
1994; King et al., 1998; Nelson et al., 2003; These results indicate a need for alternate
Sherman et al., 2002; Sullivan et al., 2001, forms of the AVLT which can be used in
2002). Barrash et al. (2004) developed an ex- longitudinal evaluation of changes over time
panded version of the test (AVLTX) aimed at in patients' verbal memory. A number of in-
detection of inadequate effort or malingering vestigators have developed alternate forms
by adding 60-minute delayed recall and rec- and reported their psychometric properties
ognition trials. The authors identified perfor- (Crawford et al., 1989; Geffen et al., 1994;
mance patterns that are highly inconsistent Ryan et al., 1986; Shapiro & Harrison, 1990).
with performance of brain-damaged patients. Lezak (1983, 1995) and Lezak et al. (2004)
provide several alternate forms. Ryan and
Geisser (1986) and Uchiyama et al. (1995)
Practice Effect and Alternate Forms
report high comparability between forms A
ofthe Rey AVLT
and C provided by Lezak.
The effect of repeated administration of the Criteria for generation of alternate word
Rey AVLT was investigated by several au- lists vary between different studies; however,
thors. Lezak (1982) reported a small but sta- the most salient requirements include a match
tistically significant practice effect observed at between the original and alternate lists with
6- and 12-month retest in a group of normal regard to the following characteristics: proba-
participants. Mitrushina and Satz (1991a) re- bility of the occurrence of the word in English
ported improvement in recall on trial I (at- usage, which is assessed using the Thorndike-
tributed to the practice effect) over repeated Large tables (1944); word length (one- or two-
annual probes, which was consistent for all syllable nouns); serial position; the imagery
age groups in a sample of healthy elderly. value of the words, based on Paivio et al. 's
Crawford et al. (1989) hypothesized that two (1968) tables; and control for semantic or
factors can lead to practice effects on memory phonetic associations between words. A good
tests: (1) retention of specific test material and match of the word lists on the first criterion is
(2) a metamemoric factor (i.e., exposure to a especially important since, according to Fuller
similar task may facilitate development of op- et al. (1997), low-frequency words result in
timal strategy). Comparison of the test-retest lower recall and higher recognition than high-
performance for those participants who were frequency words.
Different studies have yielded a wide array CALIFORNIA VERBAL LEARNING

of alternate form reliability coefficients, but TEST-SECOND EDITION
generally coefficients fell within the accept-
The California Verbal Learning Test (CVLT)
able range (>0.60) for up to a 1-month in-
was originally published by Delis et al. in
terval and between 0.30 and 0.89 for a 1-year
1987, and the revised version (CVLT-11},
interval.
which considerably improved test utility and
For further information on the psychomet-
the normative database, was published in
ric properties of the Rey AVLT, see Franzen
2000. The structure and administration of the
(2000), Lezak et al. (2004), Schmidt (1996),
CVLT/CVLT-11 are similar to those of the Rey
and Spreen and Strauss (1998).
AVLT (see above). However, in contrast to the
Rey AVLT, the word lists contain 16 items,
Assessment of Auditory-Verbal Learning which are drawn from four categories.
with the Rey AVLT in Different Languages
and Cultures
Structure of the CVLT-11 and Desctiption
To account for the effect of demographic
of the Normative Data Provided in the
variables on Rey AVLT performance, Tuokko
Test Manual
and Woodward (1996) developed a demo-
graphic correction system (using the Heaton List A, which is used in five acquisition trials,
et al., 1991, procedure), which incorporates includes four items from each of the following
various neuropsychological measures, includ- categories: vegetables, animals, ways of trav-
ing the Rey AVLT. It was developed and eling, and furniture. List B, which is used in
validated on samples of community-dwelling the interference trial, includes items from two
and institutionalized Canadian elderly over of the categories in list A (vegetables and an-
the age of 65 years. In this system, raw scores imals) and from two new categories (instru-
are converted to T scores corrected for age ments and parts of building). Recall of list B is
and education, which can be plotted on a followed by short-delay free recall and cued
profile sheet containing suggested optimal recall of list A. After a 20-minute delay filled
cutoff points for impairment. The set-up of with nonverbal testing, long-delay free recall
the profile sheet allows grouping of the data in and cued recall are assessed, followed by yes/
accordance with the DSM-111-R criteria for no recognition of list A. Ten minutes later, an
dementia. The authors suggest that applica- optional forced-choice recognition trial is ad-
tion of the system considerably improves the ministered.
accuracy of diagnostic interpretations of test Performance on the CVLT-II generates
scores. 28 variables, which reflect various aspects of
Lannoo and Vingerhoets (1997) provided learning and memory. Factor analysis using
data for the Rey AVLT for a large sample of principal components analysis of scores on
healthy Flemish adults partitioned into two 19 key variables from the normative reference
age groups x two education groups x gender. sample extracted six factors: general verbal
Normative data for a Brazilian version of learning, response discrimination, primacy-
the Rey AVLT for Brazilian adults 15-93 years recency effects, organizational strategies, re-
old are reported by Diniz et al. (2000). The call efficiency, and acquisition rate. Factor
Portuguese version of the test is provided. analysis on the mixed clinical sample yielded
Lee et al. (2002) provided data for the roughly similar results in terms of the types of
Chinese Rey AVLT (C-RAVLT), among other factors and the variables loading on each fac-
neuropsychological tests, collected in Hong tor. However, the authors argue against data
Kong on a sample of 475 Cantonese-speaking reduction through the use of factors since it
Chinese 13-46 years old. Data are reported leads to loss of information provided by indi-
in three education x two achievement x two vidual variables.
gender groups for adolescents and in three The U.S. Census-matched normative ref-
education x two gender groups for adults. erence sample consists of 1,087 adults aged
16--89 years. The test manual contains nor- cued recall, yes/no recognition, and an op-
mative data stratified by seven age groups: 16- tional forced-choice recognition trial admin-
19 (n = 150), 20-29 (n = 190), 30-44 istered after a 5-minute delay. Although the
(n = 200), 45-59 (n = 150), 60-69 (n = 145), short form yields fewer performance indices,
70-79 (n = 145), and 80--89 (n = 107). In ad- the authors cite literature supporting its use-
dition, the data were stratified by gender for fulness for diagnostic purposes. Raw scores
those variables that revealed significant gen- for the short form were calibrated to raw
der differences. The normative data represent scores for the standard form using equi-
transformations of raw scores to T or z scores percentile equating (due to the skewness of
with increment values of ± 0.5. The distribu- some variables on the short form) based on a
tion for each variable was hand-smoothed for sample of 278 participants administered both
minor irregularities both within and across age forms in a counterbalanced order.
groups. For the variables with skewed distri-
butions precluding z-score transformations,
Review of the Recent Literature
the data are reported in frequency and cu-
on the CVLT and CVLT-II
mulative frequency. The authors pointed out
that the norms for the CVLT-11 are less The CVLT-11 manual (Delis et al., 2000)
stringent than the norms for the original ver- contains an extensive review of the studies
sion of the CVLT; thus, the same raw score on which used the original version of the CVLT.
a particular variable may yield a higher stan- Studies examining the performance of various
dardized score on the CVLT-II than on the clinical groups, age and gender effects on
original CVLT. CVLT performance in healthy groups, detec-
Scoring software that computes the multi- tion of exaggeration/malingering, and predic-
ple raw and standardized scores for the stan- tion of everyday functioning as well as
dard, alternate, and short forms of the test is critiques of the first edition of the test are
available. summarized in the manual. Demographically
The authors report good psychometric adjusted norms for the CVLT based on a
properties of the test. sample of over 1,000 normal adults aged
20--85 years, stratified by age, education,
gender, and race/ethnicity (African American
Alternate and Short Forms of the CVLT-11
and Caucasian), are presented by Heaton et al.
The alternate form uses words from different (2004). Additional studies addressing the
categories in comparison to the standard form. utility of the CVLT and CVLT-11, which were
The two forms were, however, equated in published after the CVLT-11 manual was
terms of word frequency. Raw scores on the written, are summarized below.
alternate form were calibrated to raw scores
on the standard form using linear equating, Effect of Semantic Organization on Recall
based on a sample of 288 participants who Word lists in the CVLT are comprised of
were administered both forms in a counter- words taken from four semantic categories.
balanced order. Words from the same category are never pre-
The authors developed a short form of the sented consecutively, which allows assessment
test to be used for screening purposes or for of semantic clustering, reflecting the "extent to
testing patients who may feel overwhelmed by which the examinee has actively imposed an
the full-length form. The short form uses only organization on the list of words according to
one list of nine words taken from three cate- shared semantic features" (Delis et al., 2000).
gories, which is administered over four ac- This strategy facilitates encoding and retrieval
quisition trials. Short-delay free recall follows of words. The rationale and computational
a distractor task of counting backward from formulas for semantic clustering indices were
100 for 30 seconds. The subsequent trials revised in the CVLT-11. Whereas the semantic
include long-delay free recall following a clustering index in the original version used
10-minute delay filled with non-verbal tasks, the words recalled during a given trial as the
baseline for calculating expected values organization of information becomes a deter-

(recall-based expectancy), the clustering indi- mining factor in recall efficiency. (It should be
ces in the second edition use the word list as pointed out that decline in speed of mental
the baseline (list-based expectancy)' (Delis processing is another factor compromising
et al., 2000; Stricker et al., 2002). : recall rates in the elderly as slowing down in
The following recent studies addre~ed the the rate of CVLT word list presentation re-
effect of semantic organization on recall in duced recall differences between young and
healthy samples using the CVLT. Shear et al. elderly groups in the study by Weible et al.,
(2000) examined the effect of semanti4 cueing 2002.)
on free recall of the word list in 154 :healthy
young adults. The sample was divided ~to four Anatomical Correlates
groups, each receiving either the stan<lardized Brain mechanisms subserving learning and
administration procedure or a vaqant of memory, as measured by the CVLT, were in-
modified procedure. The results indical:ed that vestigated in several recent studies. Baldo et al.
delayed free recall was comparable, i¥espec- (2002) investigated the role of frontal cortex in
tive of receiving or not receiving stmantic long-term memory using the CVLT-11 in pa-
cueing. Cueing did, however, promptigreater tients with focal frontal lesions and age- and
use of semantic clustering on delayeq recall. education-matched controls. The performance
Prior knowledge about semantic categc:Jies did pattern of frontal patients was characterized
not facilitate learning. The authors co+cluded by overall low recall, an increased tendency
that semantic cueing does not enhuce de- to make intrusion errors, reduced semantic
layed recall in healthy participants!. They clustering, and impaired yes/no recognition
pointed out, however, that cueing migh;: have a due to endorsement of semantically related
greater effect in patients with poor s.mantic words and words from an interference list. The
organizational skills. I authors concluded that these findings support
The role of semantic organization ip recall earlier reports relating false recollections and
efficiency was further investigated by \fegesin source memory deficits to semantic confusion
et al. (2000) on 32 young and 68 elderly par- attributed to frontal dysfunction.
ticipants. The latter sample was su~vided Savage et al. (2001) explored the neural
into young-old and old-old groups based on a basis of spontaneous and directed semantic
median age of 74 years. Two word lists were organization using positron emission tomo-
used: lists A from the standard and *emate graphy (PET) technology. The CVLT proce-
forms of the CVLT. Encoding strat~ was dure was modified to include three encoding
manipulated by presenting each participant conditions that manipulated semantic organi-
with words grouped into semantic categories zation: spontaneous, directed, and unrelated.
(blocked list) and words intermix~ (un- This design allowed the authors to manipulate
blocked list). Item and source memo*r were semantic clustering levels over three encoding
tested 20 minutes after the final reciill trial conditions. The imaging results revealed two
and again 1 week later. The results inPicated distinct activations in the left inferior pre-
that semantic blocking enhanced iteni mem- frontal cortex (inferior frontal gyrus) and left
my for the older groups but not for th• young dorsolateral prefrontal cortex (middle frontal
participants. In addition, use of s$mantic gyrus), corresponding to the levels of semantic
clustering declined with age and was pdsitively clustering observed in the behavioral data.
related to source memory efficiency~ which Blood How in the orbitofrontal cortex was
disproportionately declined in two older strongly correlated with semantic clustering
groups, compared to item memory. : scores during immediate free recall (encod-
The results of these studies suggQSt that ing). The authors concluded:
semantic organization of presented i~forma
tion is not critical for individuals wi~ intact orbitofrontal cortex performs an important, and
semantic organizational capacity. Hewever, previously unappreciated, role in strategic memory
as this capacity declines with age, s~mantic by supporting the early mobilization of effective
LIST -lEARNING TESTS 365
behavioral strategies in novel or ambiguous situa- (n = 150 and 151). Cluster analysis on seven
tions. Once initiated, lateral regions of left pre- indices revealed five distinct performance
frontal cortex control verbal semantic organization. pattern clusters, which included two intact
(p. 219) memory clusters differing in encoding strategy
(semantic vs. serial) and three clusters indica-
Involvement of bilateral frontotemporal ar-
tive of deficits in consolidation, retention, and
eas in verbal memory is documented by John-
retrieval memory processes, respectively. The
son et al. (2001) in their study exploring brain
latter cluster had accompanying problems with
activation on functional magnetic resonance
increased number of intrusions and persever-
imaging (fMRI) during CVLT performance.
ations, pointing to poor memory control.
The results provide correlational evidence of
In contrast, Demery et al. (2002) found two
right frontal and medial temporal lobe (hip-
distinct level of performance clusters (within
pocampal) participation in verbal memory,
normal limits and moderate to severe im-
which mediates the success of word processing
pairment) in their sample of 160 TBI patients.
by the left medial temporal lobe as measured
No pattern of performance clusters was
by performance on the CVLT.
identified in this study.
Vanderploeg et al. (2001) proposed that
Assessment of Learning and Memory impaired learning and memory in TBI pa-
in Traumatic Brain Injury tients is due to a consolidation deficit, rather
Construct and criterion validity of the CVLTin than to deficient encoding or retrieval pro-
assessment of learning and memory deficits cesses. The authors compared CVLT perfor-
associated with traumatic brain injury (TBI) mance of 55 patients with moderate to severe
have been tested in several studies. The mul- TBI, 55 controls matched on age and per-
tifactorial structure of the CVLT was reex- formance on trial 5 and total for trials 1-5
amined by Wiegner and Donders (1999), who (acquisition-matched), and 55 controls mat-
identified a four-factor model in their study ched on demographic characteristics but
with 150 TBI patients, which differed from the not on CVLT performance (demographic-
six- or five-factor model identified by the test matched). The results revealed comparable
authors. The factors were interpreted as at- rates of learning across groups, indicating no
tention span, learning efficiency, delayed re- encoding differences. Rate of forgetting was
call, and inaccurate recall. Cluster analysis, significantly more rapid for the TBI group in
performed on the variables with the highest comparison to the two control groups, con-
factor loadings on each of the four factors, re- sistent with consolidation problems. Lower
vealed four performance subtypes. Two of rates of proactive interference for TBI pa-
them differed in the level of performance tients in comparison to demographic-matched
(average scores across all cluster variables vs. controls provided further evidence of a con-
significantly below average on all variables), solidation problem in TBI patients. Compa-
which was related to injury severity, with the rably low rates of proactive interference
other two subtypes reflecting distinct perfor- for the TBI and acquisition-matched groups
mance patterns. The authors confirmed the point to impaired acquisition in both. All
usefulness of the CVLT in assessing learning groups benefited equally from semantic or
abilities in patients with TBI and underscored recognition retrieval cues, indicating no dif-
its sensitivity to the general severity of brain ferences in retrieval process. These findings
injury. support the authors' assertion of a consolida-
Subsequent studies revisited the issue of tion deficit underlying memory impairment in
level of performance vs. pattern of perfor- TBI. However, the authors point to the limi-
mance in subtyping TBI patients. Curtiss et al. tation of their study in using a primarily male
(2001) used CVLT and the Wechsler Memory sample and address the possibility that other
Scale-Revised (WMS-R) Digit Span to derive subgroups of TBI patients might display en-
seven indices of short- and long-term memory coding and retrieval deficits, as reported in
processes on two samples of TBI patients previous studies.
As to the low rates of proactive intel!ference compared serial position effects produced on
reported by Vanderploeg et al. (2001), Numan trial! in 25 patients with mild dementia [Mini-
et al. (2000) suggested that buildup and re- Mental State Exam (MMSE) = 20], 25 patients
lease from proactive interference in TBI can with very mild dementia (MMSE =25.5), and
be uncovered if appropriate methods of 50 age- and education-matched normal con-
analysis are used. Whereas the commonly trols. Performance of the very mildly de-
used measure of proactive interference (trial! mented group was also compared to that of a
recall on list A minus recall on list B).did not group of 11 patients with transient amnesia
reveal any evidence of proactive interference arising from a series of ECT treatments. Pri-
in either the TBI or control group to their macy and recency effects were defined in this
study, proactive interference was dettteted in study as recall of the first two and the last two
both groups when relative recall of shared and items, respectively. The results indicated sig-
nonshared category items from the ~o lists nificantly lower overall recall as well as a sig-
was taken into account. · nificantly reduced primacy effect, with a
Furthermore, the impact of TBI on tipecifi.c normal recency effect in the demented group
memory components should be vi~ed on compared to the control group. Performance
a continuum. For example, Duchniclc et al. of patients with ECT-induced amnesia was
(2002) examined the discrepancy qetween comparable to that of very mildly demented
long-delay free recall and recognition dis- patients on most standard CVLT measures;
criminability (LDFR/RD), and the discrep- however, they revealed an expected primacy
ancy between free recall and seml¥}tically and recency pattern. The authors concluded
cued recall as indicators of retrieval p110blems that a reduction in the primacy effect is an
in 122 TBI patients. The results support a early feature of memory impairment in Alz-
continuum of retrieval deficit severi\Y: per- heimer's disease. They pointed out, however,
formance improvement with recogni~on but that it is not a necessary feature of all causes of
not semantic cueing points to more~ severe memory impairment.
retrieval deficit, whereas improvem4t with Paul et al. (2002) examined the serial posi-
both recognition and semantic cuei~ indi- tion effect in vascular dementia. The authors
cates less severe retrieval deficit. compared performance of 19 patients with
Thus, there is no consensus on the number mild dementia and 17 with moderate demen-
of factors that represent learning and Jllemory tia. The mildly demented group demonstrated
constructs which are measured by the CVLT intact primacy and recency effects, while nei-
or on the number or the basis (level vs. pattern ther primacy nor recency effects were pro-
of performance) for subtyping TBI wtients. duced by the moderately demented group.
However, all studies point to the clinifal sen- The authors concluded that absence of serial
sitivity of the instrument to leamieg and position effects may occur in more advanced
memory deficits in TBI. The difference$ across dementias regardless of dementia type.
studies in hypothesized mechanisms ~derly
ing memory impairment in TBI might be due Repeated Administration and
to differences in demographic charac~ristics Practice Effects
of the samples, severity of TBI, degree of re- McCaffrey et al. (2001) evaluated CVLT
covery, and time elapsed between TBI and performance of 22 HIV-negative individuals
assessment. across five assessment probes equally spaced
over approximately 1.5 years, in the context of
Assessment of Serial Position Effect ; a National Institute of Mental Health study on
in Dementias the natural progression of HIV infection. Total
Studies using the CVLT to assess dettentias recall (trials 1-5) was analyzed as part of a
focused on a serial position effect. Bayley et al. larger battery. A considerable practice effect
(2000) examined performance pattern 'in Alz- was observed on this measure, with a statisti-
heimer's disease and amnesia associabpd with cally significant improvement at the second
electroconvulsive therapy (ECT). The authors administration and relative stability in recall
for the following three administrations. The (2000b) and Slick et al. (2000) addressed the
authors interpreted the initial administration sensitivity of different cutoff scores to detect
of the test as a period of "adaptation" and suboptimal effort on the CVLT. Baker et al.
learning, with subsequent stabilization in (2000) developed two formulas to detect in-
performance, and cautioned test users to complete effort on the CVLT. The discrimi-
consider this trend when interpreting changes nant function formula, which incorporates
in CVLT performance on the retest. Data for scores on three CVLT measures, yielded an
repeated administration of the CVLT are also acceptable false-positive rate of 7.46%.
presented in McCaffrey et al. (2000).
In a follow-up article based on the same
Use of the CVLT in Other Languages
study, Duff et al. (2001) proposed a dual-
and Cultures
baseline approach to minimize the practice
effect. The test was administered twice within Nolin (1999) used the French translation of
a 2-week period to 18 HIV-positive/sympto- the CVLT in a study on memory functioning
matic, 25 HIV-positive/asymptomatic, and 26 in head-injured individuals. Psychometric
HIV-negative participants and later adminis- properties of the test were examined on 309
tered two more times at approximately intact individuals, 25 male closed head injury
6-month intervals. The short-term stability patients, and 26 bilingual university students.
coefficients for recall trials for the asymp- Reliability indices varied depending on the
tomatic and control groups ranged 0.~. 77 method used. Acceptable translation validity
(with the exception of one coefficient falling and criterion-related validity were demon-
below this range). Significant practice effects strated. Significant intercorrelations between
were found on six of the CVLT measures for the French and English versions were found.
all three groups over the initial short test- Kim and Kang (1999) reported normative
retest interval. Across four assessments, the data for the Korean version of the CVLT (K-
authors noted a cubic trend on most of the CVLT) collected on a sample reflecting Ko-
measures for all three groups, with an initial rean census data (n = 357). The data for 22
increase in scores from time I to time II, a indices stratified by age and gender are re-
subsequent decrease from time II to time III, ported. Consistent with the analyses pre-
followed by an increase from time III to time sented in the test manual (Delis et al., 2000),
IV. This pattern of performance change across factor analysis of 19 K-CVLT indices extracted
repeated administrations supports the dual- six factors accounting for 73.3% of common
baseline method for controlling practice ef- variance. The scores on various K-CVLT var-
fects. In addition, the dual-baseline approach iables were generally lower than the norma-
led to improvement in a number of stability tive data for the English CVLT, and SDs were
coefficients. The authors cautioned, however, larger.
that this approach is appropriate not for all Barker-Collo et al. (2002) examined the
patient groups and not with all neuropsycho- effect of the American content of the CVLT
logical instruments. on the performance of New Zealanders. The
For further information on the psycho- CVLT and a modified version developed to
metric properties of the CVLT/CVLT-11, see reflect New Zealand content (NZ-VLT) were
Elwood (1995), Franzen (2000), Lezak et al. administered to 90 healthy adults. The authors
(2004), and Spreen and Strauss (1998). report significantly better performance on the
NZ-VLT on several measures.
Assessment of Effort with the CVLT
The pattern of CVLT performance associated
Adaptations and Alternate Versions
with malingering and insufficient effort has
of the CVLT
been described in several studies. Demakis
(1999) reported that simulators were less The children's version of the CVLT (CVLT-C,
likely to consistently recall the same word Delis et al., 1994) follows the same approach
across successive learning trials. Sweet et al. to assessment of learning and memory as the
adult version. 'nle word lists consist of 15 forms), drawn from the Battig and Montague
shopping items falling into three semantic (1969) category exemplar collection. 'nle two
categories. Utility of the children's version has most common responses were used as non-
been recently addressed by Goodman et al. target distractors for the recognition tasks. 'nle
(1999), Donders (1999a,b), and Beebe et al. lists were closely matched for mean frequency
(2000). of occurrence of the words in the Battig and
A 9-item dementia version of the CVLTwas Montague study and for mean word frequency
described by Spreen and Strauss (1998), in printed text (Frances & Kucera, 1982). In-
Woodard et al. (1999b), and Davis et al. terform reliability reported by Brandt (1991)
(2002). Several test measures were reported to was high, especially for the free recall trials.
be sensitive to specific memory deficifs char- A revised version of the test (HVLT-R) was
acteristic of Alzheimer's disease and ischemic introduced in 1996 (Benedict et al., 1996,
vascular dementia (Davis et al., 2002); 1998; Frank & Byrne, 2000) and made com-
Stevens (2000) and Stevens et al. (2001) mercially available by Brandt and Benedict in
described development of a pictorial version 2001. The administration procedure in the
of the CVLT, the Connecticut Pictorial revised version was modified to include a
Learning Test (COPLT). Results of two ex- 20-25 minute delayed recall trial, with rec-
periments were reported. The analyses per- ognition testing following the delayed recall.
formed in the first experiment reveaied that In all other respects, the revised test is iden-
the new verbal stimuli to be adapted for pic- tical to the original version.
torial use were relatively equivalent in diffi- The HVLT-R generates 10 performance
culty to the original CVLT. The authors indices-recall on each learning trial, total
proposed use of these stimuli as an alternate recall across three learning trials, delayed re-
form III of the CVLT. 'nle second experiment call, percent retention on delayed recall re-
explored psychometric properties of the pic- spective to the higher of the recall scores
torial format of the test. The authors reported from learning trials 2 and 3, and four indices
good internal consistency and concurrent va- of recognition-number of true positive re-
lidity of the pictorial version. sponses, semantically related false-positive
errors, semantically unrelated false-positive
errors, and the Recognition Discrimination
Index, derived by subtracting the number of
HOPKINS VERBAL LEARNING TEST
false-positive errors from the number of true-
The Hopkins Verbal Learning Test (HVLT) positive responses.
was originally introduced by Brandt in 1991 to The normative data reported in the test
allow brief repeated assessments over time for manual (Brandt & Benedict, 2001) are based
use with demented patients or those who are on a sample of 1,179 community-dwelling in-
too impaired for administration of more dividuals free of neurological or psychiatric
lengthy and complex tests. The HVLT eonsists disorders (300 men, 879 women), aged 16-92
of a 12-item word list, composed of four words years, with a mean age of 59.00 (18.62) years
drawn from each of three semantic categories. and mean education of 13.47 (2.88) years. The
Three free recall trials are followed by a yes/no normative data stratified into eight age ranges
recognition trial using a list of 24 words, which are reported in raw scores. In addition, the
includes 12 targets and 12 distracters. Half of tables reported in the appendix allow raw to
the distractors are drawn from the sa'me se- T-score conversion for total recall, delayed
mantic categories as the targets (related dis- recall, percent retention, and the Recognition
tractors) and half, from other categories Discrimination Index, using the overlapping
(unrelated distractors). The HVLT incl~des six age strata technique.
equivalent forms. The items chosen for the Analysis of interform reliability of the
word lists represent four very high-frequency HVLT-R suggested high equivalency of the six
responses from each of 18 semantic categories forms for the free recall trials. However, the
(three categories for each of the six eq~valent six forms for the recognition trial yielded very
small but statistically significant differences in discriminative validity of the HVLT-R be-
performance. Therefore, the four recognition tween patients with Alzheimer's disease and
indices are listed separately for two homoge- demographically matched controls, a three-
neous groups of forms: forms 1, 2, and 4 and variable discriminant equation correctly clas-
forms 3, 5, and 6. sified 90% of participants (Shapiro et al.,
The structure of the HVLTIHVLT-R also 1999).
allows examination of the qualitative aspects The utility of the HVLT in assessing
of performance, such as extent of category memory deficits associated with head injuries
clustering, similar to the CVLT. was demonstrated by Guskiewicz et al. (2001)
The psychometric properties of the test are and Morey et al. (2003).
generally good. Stability coefficients over a 9- A comparison of the HVLT/HVLT-R and
month period in a sample of healthy elderly the CVLT/CVLT-R suggests that the HVLT/
were moderate yet comparable with similar HVLT-R can adequately assess basic verbal
properties reported for other tests of verbal learning capacity. However, it has limitations in
memory (Rasmusson et al., 1995). Test-retest assessing more complex and qualitative aspects
reliability coefficients for the four primary of verbal learning and memory (Lacritz &
variables of the HVLT-R over a 6-week period Cullum, 1998; Lacritz et al., 2001). On the other
in a normal elderly sample ranged 0.39--0.74 hand, the advantages of the HVLTIHVLT-R
(Benedict et al., 1998). The authors noted that make it highly suitable in situations where more
restricted and non-normal distribution of complex measures cannot be administered.
several variables limited the range of reliabil- Among its advantages, noted across different
ity coefficients. studies, are the following: it is brief, easy to
A considerable practice effect was noted on administer, and repeatable; has no ceiling ef-
repeated administration of the same HVLT-R fects on many variables for older groups; and
form over four testing probes 2 weeks apart, does not require adjustment for education.
whereas no practice effect was seen on re- In this chapter, we reviewed a recent article
peated administration of the alternate forms (Friedman et al., 2002) reporting normative
(Benedict & Zgaljardic, 1998). data for elderly African Americans based on a
For further information on the psychomet- sample of 237 participants stratified by two
ric properties of the HVLT, see Franzen age groups (60-71, 72-84), gender, and three
(2000), Lezak et al. (2004), and Spreen and educational levels (< 12, 12, > 12).
Strauss (1998).
The usefulness of the HVLT/HVLT-R in
distinguishing between demented/amnesic
WHO-UCLA AUDITORY VERBAL
groups and normal elderly has been addressed
LEARNING TEST
in several studies (Brandt, 1991; Brandt &
Benedict, 2001; de Jager et al., 2003; Frank & The WHO-UCLA Auditory Verbal Learning
Byrne, 2000; Hogervorst et al., 2001; Krebs, Test (WHO-UCLA AVLT; Maj et al., 1993)
1994; Kuslansky et al., 2004; Lynch, 2002; was developed by a group of UCLA re-
O'Connor, 2002; Shapiro et al., 1999). The searchers for the World Health Organization
specificity of the Total Recall and Recognition (WHO), for use in a multicultural context. It
Discrimination Index of the HVLT ranged is designed to minimize the cultural bias
98%-100%, whereas sensitivity was somewhat inherent in the Rey AVLT while preserving its
lower, ranging 87%-94%, in screening for original format. The test consists of two lists,
dementia of Alzheimer's type (Brandt, 1991; each containing 15 words from five cross-
Hogervorst et al., 2001). Kuslansky et al. culturally relevant categories: body parts,
(2004) reported a modest sensitivity of 69% animals, tools, household objects, and trans-
with a specificity of 89% for the HVLT Total portation vehicles. Three exemplars for
Recall in differentiating between non- each of the five categories, which are selected
demented elderly and those meeting DSM-IV from the 250-item lexicon of universally fa-
criteria for dementia. In a study on the miliar concepts compiled by Snodgrass and
Vanderwart (1980), are presented in fixed assessment of memory deficits in a sample of

random format. A copy of the WHO~UCLA 138 individuals with the diagnosis of probable
AVLT and administration instructions :can be Alzheimer's disease. The results revealed
found in Appendix 3. Due to "universal fa- modest but statistically significant associations
miliarity" of the words used in this test, its between these tests on many variables, with
intercultural variability is low (Lezak et al., total number of words learned over three tri-
2004), which warranted its use by the WHO in als showing the strongest association. The
a study on the cognitive sequelae of. HIV-1 authors concluded that the results support the
infection across different countries. utility of shorter list-learning tasks in patients
The test includes five acquisition trials, with mild to moderate Alzheimer's disease
interference trial, recall after interfere4ce, 20- and suggested use of the CERAD word list as
to 30-minute delayed recall, and a '\terbally a good alternative in situations where admin-
presented yes/no delayed recognition Qut of a istration of the CVLTis not feasible. However,
list of 30 words, which includes 15 ~ts and they suggest caution in applying interpretive
15 distracters. strategies that are derived on more compre-
Normative data for a Spanish transl~tion of hensive measures to shorter measures.
the WHO-UCLA AVLT (Ponton et al.~ 1996), The psychometric properties of the CERAD
stratified by age and education, are I repro- Word List Learning task in assessment of
duced in this chapter. The construct va$dity of memory functioning in nondemented elderly
this version of the test was assessed by Ponton and in differential diagnosis of dementia across
et al. (2000) using factor analysis of a latge test ethnic samples have been examined in many
battery assessing various cognitive d«?mains. studies (Andel et al., 2003; Calm et al., 1995;
The WHO-UCLA AVLT measures incloded in Chen et al., 2000; Fillenbaum et al., 1998;
the analysis (recall on trial 5, postinterference Galasko et al., 1995; Ganguli et al., 1991, 1996;
recall, and 20-minute delayed recall) lo~ed on Grady et al., 2002; Guruje et al., 1995; Morris
a unique factor, with very low loadings on four et al., 1989, 1993; Stewart et al., 2001;
nonmemory factors extracted in this· study. Unverzagt et al., 1996, 1999; Welsh et al.,
The validity of the Spanish version of the test 1994).
and demographic predictors of test perfor- Normative data for the CERAD battery
mance were further examined by Lopez- based on a sample of 83 healthy African-
Carlos (1999) and Mares (2002). American individuals were reported by Un-
verzagt et al. (1999). The Word List Learning
task is included in a Brazilian-Portuguese
version of the CERAD battery (Bertolucci
CERAD LIST-LEARNING TEST
et al., 2001). Norms on CERAD for use in an
The Consortium to Establish a Registry for Australian setting were reported by Collie
Alzheimer's Disease (CERAD) battery in- et al. (1999) on a sample of 243 healthy aging
cludes a word list learning task, whiqh taps adults. The Word List Learning task discrim-
verbal memory using a list of 10 unrelated inated well between normal and demented
words presented visually across three 1Jials at individuals in a Jamaican sample (Unverzagt
a rate of one every 2 seconds. The ofder of et al., 1999).
presentation changes with each trial. qelayed
recall and recognition are tested after a brief
delay filled with a nonverbal distract'r task
SELECTIVE REMINDING TEST
(Morris et al., 1989). The highest test-retest
reliability coefficients (ranging 0.62--0.~) for The Selective Reminding Test (SRT) was
clinical and control samples were obtained on originally introduced by Buschke (Buschke,
the total number of words recalled over three 1973; Buschke & Fuld, 1974) for assessment
learning trials. of memory and learning. Hannay and Levin
Kaltreider et al. (2000) compared th~ utility (1985) described and examined the psycho-
of the CERAD test and the CVLT in the metric properties of the Verbal Selective
Reminding Test (VSRT), which is based on injury (Dikmen et al., 1995; Paniak et al.,
this paradigm. 1989), multiple sclerosis (Beatty et al.,
The test uses a list of 12 unrelated words. It 1996a,b; Chiaravalloti et al., 2003; DeLuca
differs from other list-learning tests in that et al., 1998; Demaree et al., 2003; Faglioni
only the words that were omitted by the ex- et al., 2000b; Pan et al., 2001), temporal lobe
aminee on the preceding trial are repeated by epilepsy (Drane et al., 1998), end-stage pul-
the examiner, whereas the task for the exam- monary disease (Crews et al., 2001, 2003;
inee is to repeat the entire word list on each Ruchinskas et al., 2000), and brain tumors
trial. The administration procedure includes (Torres et al., 2003) as well as depressed pa-
up to 12 free recall trials. The learning crite- tients receiving ECT (Datto et al., 2001).
rion, according to Larrabee et al. (1988), is Use of versions of the SRT with different
recall of all words for three consecutive trials ethnic and cultural groups has been de-
without reminders. scribed. Xavier et al. (2002) used an SRT with
The test structure allows distinguishing be- a Brazilian sample of nondemented partici-
tween list learning and item learning. It also pants 80-95 years old. Butman (2001) in-
allows separating retrieval from long-term cluded the Buschke SRT as part of a battery
storage (when items are retrieved without fur- for early diagnosis of dementia in Latin
ther reminders) and retrieval from short-term America. Two Spanish versions of the VSRT
storage. The test yields several derived indices were developed and validated on healthy and
of verbal learning, which include Short-Term demented groups (Campo & Morales, 2004;
Retrieval, Long-Term Retrieval, Long-Term Campo et al., 2000, 2003). A Hebrew version
Storage, Random Long-Term Retrieval, and of the SRT with three parallel forms was de-
Consistent Long-Term Retrieval. veloped by Gigi et al. (1999). An adaptation of
Various modifications of the SRT have been the SRT for use in the visual in addition to the
reported. Buschke (1984) added a cued recall verbal modality and its validation on a normal
component to the test. This version, Free and sample were reported by Indian researchers
Cued Selective Reminding Test (FCSRT), has (Rao & Andrade, 1998).
been used with demented and amnesic patients
and with nondemented elderly (Buschke, 1984;
Degenszajn et al., 2001; Grober et al., 1997,
OTHER VERBAL AND NONVERBAL
1998, 2000; Ivnik et al., 1997; Petersen et al.,
LIST-LEARNING TESTS
1999).
The VSRT normative data, psychometric Two parallel forms of a 10-word list-learning
properties, and use with nondemented sam- test are included in the Repeatable Battery for
ples have been described in several studies the Assessment of Neuropsychological Status
(Larrabee & Levin, 1986; Larrabee et al., 1988; (RBANS), published by Randolph (1998).
Ruff et al., 1989b; Trahan & Larrabee, 1993). Artiola i Fortuny et al. (1999) included a 16-
The validity of a six-trial version of the test was word list-learning task in their standardized
examined by Smith et al. (1995), Drane et al. and validated battery of neuropsychological
(1998), and Larrabee et al. (2000). These tests culturally adapted for Spanish-speaking
studies demonstrated comparable psychomet- individuals. Normative data based on 390
ric properties of the 12- and six-trial versions. participants, which were collected in Spain,
A number of studies report utility of the Mexico, and the United States, are stratified
original Buschke SRT, the VSRT, and modi- by geographical area x age x education.
fied versions of the test with various clinical Toglia and Battig (1978) developed the Af-
populations, including those with amnesia fective Auditory Verbal Learning Test, which
or Alzheimer's/Parkinson's dementia (Bartok consists of positively and negatively valenced
et al., 1997; Boeve et al., 2003; Faglioni et al., word lists. Snyder and Harrison (1997) sug-
2000a; Kuzis et al., 1999; Masur et al., 1989, gested that the affective valency of the word
1990; O'Connell & Tuokko, 2002; Sliwinski list yielded different magnitudes of primacy
et al., 1997; Stem et al., 1998, 1999), head vs. recency effects, which may be useful in the
evaluation of individuals suffering from af- more information overload (forward digit span
fective disorders. The effect of positively and >trial I), confusion regarding information
negatively valenced words on physi.,logical source (e.g., misclassify list B words as list A
measures of arousal was described by 'Snyder words), and less efficient retrieval. However,
et al. (1998). rate of learning (learning curve), loss of infor-
Majdan et al. (1996) developed a nooverbal mation over a distractor or an extended delay,
analog of the Rey AVLT, the Aggie Figures recognition ability, and false-positive errors
Learning Test (AFLT). The three forD1f of the may be resistant to aging (Bleecker et al., 1988;
AFLT, described by the authors, wete con- Bolla-Wilson & Bleecker, 1986; Cohen et al.,
structed according to the Rey AVLT! format personal communication; Geffen et al., 1990;
using abstract figures that do not§lenthem- Mitrushina et al., 1991; Wiens et al., 1988).
selves to verbal encoding. The auth em- Petersen et al. (1992) concluded that learning
phasized that the AFLT is superior to e Rey performance declines uniformly with age but
Visual Design Learning Test (see . et al., forgetting remains relatively stable across age
2004), which has only one version an4 is de- when adjusted for the amount of material ini-
signed to assess learning over five acqUisition tially learned. Salthouse et al. (1996) suggested
trials and recognition only. Rey's anal«tg does direct age-related influences on memory in-
not contain an interference list, whiqh pre- dependent of speed of performance, based on
cludes evaluation of retention rates. Ill addi- the existence of the direct path from age to
tion, Rey's stimuli represent simple geo~etrical memory in their structural equation model.
designs that can be easily encoded verllally. Similarly, age-related performance decline
In addition to these nonverbal +alogs, was reported by Norman et al. (2000) for a
Lezak et al. (2004) described the Jtctorial number of CVLT indices. According to the
Verbal Learning Test (PVLT), whic~J~llows literature review by Delis et al. (2000), several
the five-trial learning format of the ReyjAVLT. studies using the original CVLT version
For more information on list-learning tests, (Murphy et al., 1997; Woodruff-Pak & Fink-
refer to Lezak et al. (2004) and Spreen and biner, 1995) reported decline in the rate of
Strauss (1998). acquisition, primacy effect, use of semantic
clustering, and tendency for intrusion errors on
delayed recall trials. Delis et al. (2000) reported
a high inverse relationship between age and
RELATIONSHIP BETWEEN LIST-LEARNING CVLT -II recall, with age explaining 25% of the
TEST PERFORMANCE AND , variance in total recall for five acquisition trials.
DEMOGRAPHIC FACTORS Age was also significantly related to several
performance indices of the HVLT-R (Brandt &
Effect of Age
Benedict, 2001; Friedman et al., 2002; Van-
The effect of demographic variables on Rey derploeg et al., 2000), accounting for 19% of
AVLT performance has been extensiVely ex- the variance in total recall according to Brandt
plored in the literature. Studies cons~tently and Benedict (2001). However, Kuslanskyet al.
demonstrate an effect of age on rec~, and (2004) did not find a relationship between age
some studies report an effect of age on rec- and performance on the HVLT in a large
ognition (Bleecker et al., 1988; Geffen et al., sample of elderly aged ~70 years.
1990; Graf & Uttl, 1995; Mitrushina et al.,
1991; Mitrushina & Satz, 1991a,b; Q,ery &
Effect of Education
Berger, 1980; Query & Megran, 1983; Savage
& Gouvier, 1992; Selnes et al., 1991; Uchiyama The effect of education is less clear. Cohen
et al., 1995; Wiens et al., 1988). Som~ inves- et al. (personal communication), Delaney et al.
tigators suggest that age impacts only specific (1992), Query and Berger (1980), Query and
AVLT scores. For example, total number of Megran (1983), and Uchiyama et al. (1995)
words recalled on the five learning trials may observed a relationship between education
be lower in older participants, who may show and Rey AVLT scores, while Bolla-Wilson and
Bleecker, (1986), Mitrushina et al. (1991), vocabulary storage on Rey AVLT recognition
Petersen et al. (1992), and Wiens et al. (1988) performance. Delis et al. (2000) reported a
did not find a significant association. Bolla- correlation of 0.46 between the CVLT-11 total
Wilson and Bleecker (1986) argued, based on recall over five trials and the WASI Vocabu-
their multiple regression analyses, that edu- lary raw score. Similar association between
cation does not account for AVLT test score CVLT performance and WAIS-R Vocabulary
variance over that associated with IQ. The was reported by Keenan et al. (1996).
other studies reporting a relationship between
education and AVLT performance either did
Effect of Gender
not include an examination of the effects of IQ
on scores or simply reported correlations be- Gender differences in Rey AVLT perfor-
tween education and AVLT scores without mance have been reported in several studies.
controlling for the significant association be- Women have outperformed men on the
tween education and IQ. learning, interference, and delayed recall tri-
An effect of education on CVLT perfor- als (Bleecker et al., 1988; Bolla-Wilson &
mance was reported by Norman et al. (2000). Bleecker, 1986; Cohen et al., personal com-
In respect to the CVLT-II, the correlation munication; Geffen et al., 1990; Vakil &
between total recall over five trials and edu- Blachstein, 1997). Other studies, however, did
cational level was 0.29, according to the test not demonstrate an effect of gender on any of
manual (Delis et al., 2000). the Rey AVLT measures (Savage & Gouvier,
On the HVLT-R, education accounted for 1992; Wiens et al., 1988).
5% of the total variance or less across different As to the CVLT, Norman et al. (2000) re-
variables, according to the test manual (Brandt ported a significant relationship between gen-
& Benedict, 2001). Vanderploeg et al. (2001) der and performance, with female superiority
did not find a relationship between education on several indices. Delis et al. (1987, 2000)
and HVLT-R scores in a predominantly white, reported a significant association between
community-dwelling, elderly sample, whereas gender and performance on 13 out of 22 in-
Friedman et al. (2002) found a significant, dices of the original version of the CVLT, with
moderate effect of education on HVLT-R women outperforming men on all of the list A
performance in a sample of African-American and list B recall variables, semantic clustering,
elderly. This discrepancy might be due to percent middle (serial position) recall, and
greater variability in educational ranges in the number of free recall and cued recall intru-
latter study. Kuslansky et al. (2004) did not sions, whereas men outperformed women on
find a relationship between education and percent recency (serial position) recall. Ac-
performance on the HVLT in a large sample of cording to Delis et al.'s (2000) literature re-
elderly. view, the 1988 study by Kramer et al. supports
gender differences found in the above study.
Women recalled about one word per trial more
Effect of Intelligence Level
than men, although their results did not differ
Bolla-Wilson and Bleecker (1986) and Query in terms of error type, learning slope, and
and Megran (1983) reported an association forgetting, which was 10%-15% from trialS to
between IQ and Rey AVLT measures, al- short-delay recall, with no further forgetting
though there has been some discrepancy re- from short-delay to long-delay recall. Similarly,
garding which specific scores are affected. Wiens et al. (1994) found significant gender
Wiens et al. (1988) reported a relationship differences on 26 CVLT variables in a sample
between FSIQ and age with recall but not of 700 healthy job applicants.
recognition, which suggests the utility of rec- In their summary of the literature review,
ognition in assessment of pathological condi- Delis et al. (2000) concluded: "Differences
tions, free of confounds from demographic between males and females, in favor of fe-
and IQ factors. Similarly, Bleecker et al. males, are to be expected on the CVLT, and
(1988) did not find an effect of age, sex, or these differences seem to stem from the use of
374 l
more efficient, semantically based l~arning
VERBAL AND VISUAL LEARNING AND MEMORY
variables, and the remaining two refer to

strategies by women than by men" (p. ~8). On procedural issues.
the other hand, Randolph et al. (1999) ques- Minimal requirements for meeting the cri-
tioned the interpretation of female su!friority terion variables were as follows.
in the CVLT normative reference sample as
evidence for a female advantage in: verbal
Subject Variables
memory functions. The authors propo!led that
"this advantage was stimulus driv~n and Sample Size
therefore neither strictly langua~e bf;ed. or
memory based. A different selection ~f stim-
ulus items on the CVLT might confi~ably
result in superior performance by m~n (p.
data based on small sample sizes are highly
495). Such an int~rpretation was well j~tified
in reference to the original version of the
CVLT, which was presented in the fo of a
lation mean.
shopping list. However, stability of the r,attern
of gender differences, which is also evftent in Sample Composition Description
the second edition of the CVLT, has~romp Information regarding medical and psychiatric
ted researchers to revisit this issue, gi n that
the CVLT-11 does not use the shop ng list
format and its items are gender-neufral. In
status, occupation, ethnicity, or recruitment
fact, according to Delis et al. (2000), ootal re-
procedures are relevant. Until determined, it
call across five acquisition trials for }"Omen
is best that this information be provided.
was on the average five words higher ~an for
men on both the original CVLT 3.lld the Age Group Intervals
CVLT-11. Similarly, prominent gender:effects
were seen on many CVLT-11 variabl~, with
women outperfoqning men. ! especially relevant for this test since an un-
Examination of the effect of gen~er on
equivocal effect of age on list-learning test
HVLT-R performance yielded conflic.ng re-
performance has been demonstrated in the
sults. Gender accounted for at most l 7% of
literature.
variance on total recall according to ~e test
manual (Brandt & Benedict, 2001). Si~ilarly, Reporting of Educational Levels
Kuslansky et al. (2004) did not find Ia rela-
Given the association between education and
tionship between gender and perform~. ce on
performance on list-learning tests, informa-
the HVLT in a large sample of elderly: How-
tion regarding educational level should ~e
ever, in a study by Vanderploeg et al.\(2000)
reported for each subgroup, and preferably
using multiple linear regression analyst;, gen-
normative data should be presented by edu-
der contributed 8.5% of unique vari!Jnce in
I cational levels.
the prediction of total recall, with wom~n out-
performing men. Similar findings w;re re-
Reporting of Intellectual Levels
ported by Friedman et al. (2002) on a ~ample
of African-American elderly. r
Given the relationship between performance
on list-learning tests and IQ, information re-
garding intellectual level should be reported
for each subgroup, and preferably normative
METHOD FOR EVALUATINGj data should be presented by IQ levels.
THE NORMATIVE REPORTS :
To adequately evaluate the normative ~ports, Reporting of Gender Composition
eight key criterion variables were clee~ed Given the probable relationship between
critical. The first six of these relate to ~ubJeCt gender and performance on list-learning tests
in favor of males, information regarding gen- that particular table. The clinician is urged to
der composition should be reported for each pay close attention to the sequence of trials
subgroup, and preferably normative data (e.g., administration of the recognition trial
should be presented by gender. prior to delayed recall) on which the data set
is based as it has a notable effect on rate of
recall (see above).
Procedural Variables In this chapter, normative publications and
control data from clinical studies are reviewed
Description of Administration Procedures
in ascending chronological order. Data for the
Administration procedures for each list-learning Rey AVLT are presented first, followed by
test differ widely among studies. A detailed review of one study on the HVLT-R and one
description of the procedures allows selection of study on the WHO-UCLA AVLT. The text of
the most appropriate norms or corrections in study descriptions contains references to the
interpretation of the data to account for differ- corresponding tables identified by number in
ences in administration procedures. Appendix 19. Table A19.1, the locator table,
summarizes information provided in the
Data Reporting
studies described in this chapter. 1
The group mean and standard deviation for
the number of words correctly recalled/
recognized on each trial should be presented
at minimum. SUMMARIES OF THE STUDIES
[RAVLT.1] Rey, 1941,1964 (Table A19.2)
The author provided normative data for the
five learning trials on a sample of 132 French-
speaking Swiss participants. The AVLT per-
OF THE NORMS
formance of five groups is reported: manual
Studies reporting normative data for the list- laborers (n = 25), professionals (n = 30), stu-
learning tests vary in their descriptions of dents (n = 47), elderly laborers (n = 15), and
procedural and subject variables and in the elderly professionals (n = 15). Age ranges are
grouping of obtained data into categories. specified only for the last two groups: elderly
Among all the studies available in the lit- laborers ranged 70-90 and elderly profes-
erature, we selected for review those based on sionals ranged 70-88. No other descriptive
well-defined samples or that offer some as- data are provided, such as mean educational or
pects of information that are not routinely intellectual level and exclusion criteria. Mean
reported. Most of the reviewed studies pro- scores and SDs across the five learning trials
vide data for the Rey AVLT. We did not in- are reported. No other data are available.
clude normative reports for the CVLT-11 as
comprehensive normative data are provided Study strengths
in the test manual. One study providing norms 1. Large overall sample size.
for the HVLT-R collected on a large sample of
African-American participants was reviewed. Considerations regarding use of the study
In addition, we reviewed one study reporting 1. Data were collected more than 50 years
normative data for the WHO-UCLA AVLT, ago.
Spanish version, collected on a large sample of 2. No data are available for the interfer-
monolingual and bilingual Spanish-speaking ence trials or delayed recall.
participants. 3. No description of exclusion criteria, IQ
The majority of the studies present data in levels, years of education, or composition
number of words correctly recalled/recog- of the sample by gender.
nized by participants on different test trials.
When reported units deviate from this format, 1Nonnative data for children for Rey AVLT are available
the system used is described in the context of in Baron (2004) and Spreen and Strauss (1998).
4. Participants were French-speaking, and 5. Information on gender, education, IQ,

as discussed earlier, the French test and geographic area is presented.
stimuli differ from the currently used
English words (i.e., moustache, stm, and Considerations regarding use of the study
belt were included in the French stimuli). 1. Data for acquisition trials II-IV are not
provided.
Other comments 2. Demographic characteristics for each
1. Chiulli et al. (1985) reported that use of group are not reported, even though
Rey's norms led to misclassificltion of some of them were used in the analyses.
22% of their control group as iiOpaired 3. The exclusion criteria indicate that "se-
(using criteria of 1 SD below the verely brain damaged" individuals were
appropriate age and education 'lllean). not included in the sample. This implies
They recommend caution in using Rey's that mildly to moderately brain-damaged
norms, and this concurs with o$r own patients were included.
clinical experience. 4. The data were collected on medical pa-
tients in an acute-care hospital.
[RAVLT.2] Query and Megran, 1983 5. Mean IQs nearly fell within the low av-
(Table A19.3) erage range of intellectual ability, and
The study provides norms for 677 male am- mean educational level was below aver-
bulatory inpatients aged 19-81 tre~ted at age. Thus, the data may be relevant
North Dakota Veterans Administration Medi- only for individuals with IQs in the 80s
cal Center for a variety of physical complaints. and 90s.
Mean education was 11.44; mean IQ was 6. Data are available only for males.
93.83. Psychotic, severely brain-damaged pa-
tients and those suffering from major depres- [RAVLT.3] Rosenberg, Ryan, and Prifitera,
sive disorder were not included in the study. 1984 (Table A19.4)
The sample was divided into age groups. Ninety-two male psychiatric and neurological
The standard administration procedtire was inpatients from the V.A. Medical Center in
used; recognition was measured with a story in Chicago were divided into memory-impaired
which the subject was instructed to circle words and non-memory-impaired groups. Those par-
from the learned list. Learning was measured ticipants whose memory quotient (MQ) based
by the difference in number of words recalled on the WMS was 12 or more points lower than
on the highest trial vs. the lowest trial. the WAIS FSIQ and/or those whose MQ was
According to the authors, the data suggest <85 were classified as memory-impaired. The
effects of education and IQ on maintenance groups did not differ significantly with respect
of learning ability for younger men. A more to age or education. Participants were referred
complicated relationship was found for older to the psychology service for routine psy-
men. The authors concluded that there is chological and/ or neuropsychological evalua-
progressive short-term memory loss with ad- tion and selected without regard to diagnostic
vancing age, which results in declines in re- classification. Mean age, education, and FSIQ
call, then recognition, followed by l~arning for the entire sample were 48.05 (14.03), 11.37
ability. Recognition is highly affected by IQ in (2.82), and 93.11 (13.43), respectively.
older men. The test was administered according to
Lezak's (1976) description. The paragraph was
Study strengths used for the recognition task.
1. The data are broken down by age group.
2. The administration procedure is well Study strengths
outlined. ; 1. Information on age, education, IQ, gen-
3. Sample size is large, and most in<lvidual der, recruitment procedures, and geo-
cells exceed 50. graphic area is provided.
4. Means and SDs are reported. 2. Sample sizes are sufficient.
3. Test administration procedures and the 3. No reported exclusion criteria. Partici-

type of recognition task are identified. pants with histories of head trauma and
4. Means and SDs are reported. chronic illnesses (e.g., hypertension)
were included.
Considerations regarding use of the study 4. No information on IQ.
1. Data are not partitioned into age groups. 5. An immediate recognition trial was used,
2. The "normal" control sample is not which facilitates performance on delayed
representative of the overall population. recall and delayed recognition.
It consists of neurological and psychiat-
ric patients. [RAVLT.S] Ryan, Geisser, Randall, and
3. All-male patient sample. Georgemiller, 1986 (Table A19.6)
4. Somewhat low IQ level. The study provides alternate form reliability
and equivalency in a group of diagnostically
[RAVLT.4] Cohen, Andres, and Smolen, Personal heterogeneous inpatients from the V.A. Med-
Communication (Table A19.5) ical Center in Kansas (82 males, three females)
Participants were elderly volunteers (57 wo- referred for psychological and/or neuro-
men, 28 men) from Peoria, Illinois, and the psychological assessment. The sample re-
surrounding communities, aged 60-89 years. presented a wide range of psychiatric and
Mean education was 13.8 years; 30.6% of the neurological diagnoses, including undiagnosed
sample were suffering from hypertensive ill- patients and vocational counseling clients. This
ness, and 18.8% reported a history of head sample, therefore, is not representative of any
trauma. These participants were included in identifiable diagnostic group.
the study since the results of preliminary re- For the original administration of the test,
gression analyses indicated a lack of associa- the authors followed the standard procedure
tion between test performance and these described in Lezak (1983). The recognition
conditions. However, gender was found to be trial consisted of a list of 50 words which in-
related to performance. Therefore, norms are cluded words from list A and list B (interfer-
presented for males and females separately. ence list) and 20 words phonemically and/ or
The administration procedure included ac- semantically similar to those in lists A and B.
quisition trials, postinterference recall with For the alternate form, the authors used list
immediate recognition, and 30-minute de- C provided by Lezak (1983) and Taylor
layed recall followed by delayed recognition. (1959). Interference and recognition stimuli
The authors concluded that gender, age, were constructed by the authors following
and education are related to performance, Lezak's model.
with women performing better than men. The original and alternate forms were pre-
sented in a counterbalanced order, with a
Study strengths mean test-retest interval of 140 minutes.
1. The data are partitioned into age groups. Alternate form reliability coefficients ran-
2. The data are presented for males and ged 0.60-0.77. Differences between means
females separately. were less than 1 point on each of the five ac-
3. A comprehensive set ofscores is presented. quisition trials, postinterference trial, and
4. Information regarding education, gen- recognition trials. The forms were judged to
der, and geographic area is provided. be equivalent measures.
5. Test administration procedures are gen-
erally specified. Study strengths
6. Means and SDs are reported. 1. Administration procedure is well identi-
fied.
Considerations regarding use of the study 2. Information on age, education, IQ, gen-
1. Sample sizes are small. der, ethnicity, recruitment procedures,
2. The format of the recognition procedure and geographic area is provided.
was not specified. 3. Sample size is large.
4. Information on alternate form is pro- addition, means and SDs for number of words
vided. learned (trial V minus trial I) are reported
5. Means and SDs are reported. separately for men and women aged 40-65
and 66-89 (not reproduced in this book). Al-
Considerations regarding use of the study though males and females were comparable in
1. The sample is clinically heterogeneous, verbal intelligence, women outperformed men
consists of inpatients, and cannot be con- on the acquisition trials, especially among the
sidered "normal;" no exclusion crteria. older participants. Older participants scored
2. Undifferentiated age group. · lower than younger participants, except for
3. Predominantly male sample. the recognition trials. Perseverations, confab-
4. Low IQ level of the sample. ulations, and intrusions were not associated
with age or gender.
[RAVLT.6] Bleecker, Bolla-Wilson, Agnew, and A stepwise regression analysis showed that
Meyers, 1988 (Table A19.7) age and gender accounted for a significant
The authors report AVLT data on a large portion of the variance on each acquisition
sample (n = 196) of Maryland part4!ipants trial. Vocabulary accounted for a significant
aged 40-89 drawn from the Johns !iopkins portion of the variance only on trials IV and V.
Teaching Nursing Home Study of !Normal Performance on the recognition trial was not
Aging as part of an investigation of tqe con- affected by age, gender, or vocabulary. Overall
tribution of age, gender, vocabulary ! range, performance was higher for women in com-
and depression to AVLT performancf.l. The parison to men, with an increase in this ten-
sample includes participants descri$ed in dency with age.
Bolla-Wilson and Bleecker (1986). 1
Participants with histories of head ~uma Study strengths
with loss of consciousness, stroke, seizures, 1. Large sample size, but individual cells
uncontrolled hypertension, congestive! heart are small.
failure, abnormal thyroid function, electrocon- 2. Presentation of data by age decades and
vulsive therapy, sleep disorders, coma, psychi- by gender.
atric disorders, or alcohol or drug abu9e were 3. Stringent medical and psychiatric exclu-
excluded. All participants had Mini-Mental sion criteria.
State Examination scores >23. W AIS-R Vo- 4. Use of a screening instrument for de-
cabulary raw scores were used to e$timate pressive symptomatology.
verbal intelligence. The presence of depressive 5. Information regarding educational level,
symptoms was assessed with the Be<fl< De- estimated verbal intelligence, and geo-
pression Inventory. Means and SDs ate pre- graphic recruitment area is provided.
sented for males (n = 87) and females (n := 109) 6. Specification of an administration format.
separately by five age decades: 40-49, So-59, 7. Means and SDs are reported.
60--69, 70-79, and 80-89. ·
Test administration generally followed the Considerations regarding use of the study
instructions recommended by Lezak :(1983) 1. Very high educational levels for some
and involved presentation of words at a rate of age groups.
one per second for five acquisition trials, each 2. Test administration did not include an
followed by a free recall period. After t~e fifth interference or delayed recall trial. In
trial, participants were presented with; a rec- addition, the recognition trial followed
ognition trial containing 50 words incl.uding trial V, and thus, the normative data de-
the 15 test words and phonemically ~d se- rived from this recognition trial may not
mantically similar distractor words. No inter- be representative of recognition perfor-
ference or delayed recall trials were em~loyed. mance associated with the traditional
Means and SDs for the five acquisitiOn tri- administration format involving an in-
als and recognition trial are reported fqx: each terference trial prior to the recognition
decade for males and females separately. In trial.
[RAVLT.7] Wiens, McMinn, and Crossen, 1988 [RAVLT.8] Crawford, Stewart, and Moore,
(Tables A19.8-A19.10) 1989 (Tables A19.11, A19.12)
The study reports normative data for 222 job The study compared test-retest performance
applicants, currently employed in a variety of with the same form of the Rey AVLT and with
occupations (white collar and blue collar), a parallel version of the AVLT. The parallel
who had previously passed basic academic version was developed by the authors based
skills tests and physical examinations and were on the criteria identified in the article.
free from physical illness or limitations. The Sixty participants, free of neurological, psy-
applicants represented an occupational cross- chiatric, and sensory disability, were recruited
section of the community. Participants were from nonmedical health-service personnel and
free from alcohol and other substance abuse the fire service. Participants were divided into
and ranged in age 19-51 years, with a mean pairs matched for gender, age (±3 years), and
age of 29.1 (6.0). Sample composition was years of education (±1 year) to form two
87% male, 13% female; 94.6% Caucasian, groups. There were no significant differences
5.4% racial minority. between groups in mean estimated IQ based
Standard administration was used with a !- on National Adult Reading Test (NART)
second interval between each word during scores (106 and 108, respectively).
presentation. On the recognition trial, partic- One group was administered the original
ipants were to circle words from the learned AVLT and the other, the parallel version.
list in a paragraph. Participants were retested following a delay of
The data are stratified by WAIS-R FSIQ, 27 days (±3 days), with half of each group
age, and education. receiving the same version and the other half,
The authors noted better recall at higher IQ the alternative version.
levels and an inverse trend between age and all The standard administration procedure was
acquisition and delayed recall trials. A proactive used. On the recognition condition, partici-
interference effect was observed for all groups; pants were asked to inform the examiner of
recall for the second word list was inferior to the the stimulus words that had been contained in
initial recall of the first word list. Performance the previously presented word lists. A recog-
on the recognition trial was unrelated to age nition score was obtained by subtracting the
and IQ, which points to its importance in number of false-positive identifications from
studying pathological memory loss. the number of words correctly identified.
The authors concluded that the parallel
version can be used as an equivalent form of
Study strengths
the AVLT. A significant practice effect was
1. The data are stratified by age, education,
seen for participants who were administered
and IQ levels.
the same versions.
2. Administration procedures are well
outlined.
Study strengths
3. Exclusion criteria appear to be adequate.
1. Information on an alternate form and
4. Sample sizes vary between the groups,
practice effects.
but for the majority of the groups they
2. Administration procedure was described.
are adequate.
5. Information regarding recruitment pro-
4. Information on IQ and geographic re-
cedures, gender, ethnicity, and geo-
cruitment area.
graphic area is provided.
Considerations regarding use of the study 1. Demographic characteristics of the
1. Age groups are restricted to younger sample are not described.
range. 2. Relatively small sample sizes.
2. Some sample sizes are very small. 3. Data are not presented by age groupings.
4. Data were collected in the United 3. Adequate exclusion criteria.

Kingdom, and it is unclear to what extent 4. Means and SDs are reported.
they are appropriate for clinical use in 5. Administration procedure is specified.
the United States.
[RAVLT.9] Nielsen, Knudsen, and Daugbjerg, 1. No information regarding educational
1989 (Table A19.13) · level.
The authors gathered AVLT normative data 2. Some information regarding occupa-
on 101 Danish participants aged 2<h54. The tional status is reported, but not all of the
majority (n = 87) had undergone minor sur- categories are defined.
gery (primari1y arm and leg) and were tested 3. It is unknown if the stimulus words were
several weeks postsurgery "when they were translated.
free of post-surgical inconveniences." Four- 4. No interference trial was administered;
teen recruits were hospital laundry wt>rkers. thus, the delayed recall information may
Exclusion criteria included history of head not be comparable to that obtained if
trauma, alcohol abuse, prolonged expoiure to any interference trial had been admin-
organic solvents or other toxic agents, pres- istered prior to the delayed trial.
ence of somatic or psychiatric disease fwhich 5. Individual cell sizes are rather small.
might adversely affect neuropsychological 6. Data were collected on a Danish sample,
functioning," or use of medications "which which might limit their clinical useful-
could affect intellectual capacity." Fifty-three ness in the United States.
participants were male and 48 were female;
91% were right-handed. The sample was [RAVLT.10] Roth, Davidoff, Thomas, Doljanac,
classified into three age groupings: 20-29 Dijkers, Berent, Morris, and Yarkony, 1989
(n = 35), 30-39 (n = 27), and 40--54 (n = 39); (Table A19.14)
only two participants were older than 50. The authors obtained AVLT data on 61 paid
Mean prorated Verbal IQ (based on a, trans- control participants as part of their examina-
lation of the WAIS-R) was 98.61 (12.21), with tion of neuropsychological deficits in acute
a range of 78--140; means and SDs for seven spinal cord-injured patients. Mean age was
subtests are reported. Information on occu- 27.5 (standard error= 1.0), and mean educa-
pational status indicated that skilled workers tion was 12.8 years (standard error= 0.2).
were somewhat overrepresented compared to Forty-five participants were male and 16 were
the general population, while self-employed female; 39 participants were recruited in De-
individuals were underrepresented. troit and 22 in Ann Arbor. Exclusion criteria
Test administration appeared to be: based were history of closed head injury and recent
on the Lezak (1983) instructions. Specfically, high-frequency alcohol or substance abuse.
"a list of 15 discrete nouns was read aloud at a The AVLT appeared to have been admin-
rate of one per second for five consecutive istered according to the Lezak (1983) in-
learning trials, each followed by a free recall structions. Means and standard errors are
test. Delayed recall was tested 15 min after reported for the five acquisition trials, list B,
completion of the fifth learning trial" (p. 39). postinterference recall, and recognition.
It is not reported if the words were tr~lated.
Means, SDs, and ranges are repo~d for Study strengths
the five acquisition trials, total score across the 1. Information regarding age, gender, ed-
five trials, and delayed recall. ucation, and geographic recruitment
area is provided.
Study strengths 2. Minimally adequate exclusion criteria.
1. Data are presented by age groupings. 3. Adequate sample size.
2. Information regarding gender, IQ, geo- 4. Means and standard errors are reported.
graphic recruitment area, age, ~ded 5. Test instructions are not specified but
ness, and occupation is presented. appear to be standard.
Considerations regarding use of the study the recognition score by taking into ac-
1. Undifferentiated age range, although the count misassignments from list A to list
small standard error suggests that the B and vice versa; and total number of
age spread is not large. false-positive misidentifications.
2. Recognition procedure is not specified. 3. Serial position effect was measured for
list A averaged over the five acquisition
[RAVLT.11l Geffen, Moar, O'Hanlon, Clark, trials. Serial positions of the words were
and Geffen, 1990 (Tables A19 .15-A19.17) collapsed into five groupings (words 1-3,
The article provides norms for 153 adults re- 4-6, 7-9, 10--12, 13-15).
siding in Australia in age groups spanning 4. Functional indices, such as proactive and
seven decades (16--86 years; M = 44.5, SD = retroactive interference effects, forget-
20.2). Age groups included approximately ting, retrieval efficiency, and item infor-
equal numbers of males and females and were mation overload, were computed as
matched for intelligence, education, and oc- ratios of pairs of trials. All of the above
cupation. All participants were physically indices were presented for males and
healthy and free of neurological symptoms by females separately.
self-report. Participants' occupations (current
or preretirement) ranged from unskilled to Based on participants' performance on dif-
professional. Average education was 11.2 (2.2) ferent trials, the authors made inferences re-
years, with a range of 7-22 years. Estimates of garding the memory mechanisms involved.
FSIQ were derived from error scores on the The authors noted significant associations
National Adult Reading Test (NART, Nelson, between age, gender, and Rey AVLT perfor-
1982). Average estimated IQ was 111.5 (7.3), mance, with females consistently outperforming
with a range of94-127. All participants spoke males. Performance of younger participants was
English as the first language. superior to that of older participants.
Standard administration procedures were
used for recall trials. Twenty-minute delayed Study strengths
recall followed by a recognition condition was 1. Demographic characteristics are well
used. The delay was filled with the Digit Span, described in terms of age, occupation,
WMS Logical Memory, and NART. The rec- education, IQ, fluency in English, and
ognition condition consisted of a list of 50 geographic area.
words which included words from lists A and B 2. Normative data are presented by age
as well as 20 phonemically or semantically group and separately for males and fe-
similar distractor words. Participants were males.
to identify as many of the previously learned 3. The administration procedure is well
words as possible, as well as the specific list of described. Additional indices were de-
origin (list A or B). A specially designed com- veloped to explore different memory
puter program was used for scoring. mechanisms.
The following variables were included in 4. Large overall sample size.
the analyses. 5. Adequate exclusion criteria.
1. Recall trials: recall of list A on trials 1-V;
total recall over five trials; total number Considerations regarding use of the study
of repeated words; extralist intrusions; 1. The 20-minute delay interval was filled
recall of list B; postinterference recall with verbal memory tasks, which might
(retention); and 20-minute delayed recall. serve as a source of interference.
2. Recognition trial: number of words rec- 2. The sample had relatively high estimated
ognized from lists A and B; a nonpara- IQ values, which might limit generaliz-
metric signal detection measure of ability of the data.
recognition performance, p(A) = 0.5( 1 + 3. Estimated IQs of the youngest age group
hit rate- false-positive), which corrects were up to 10 points lower than those of
the other age groups; however, this may 3. Short-Term Percent Retention repre-
have been an artifact of the NART since sents trial VI recall as a proportion of
the teenagers probably had not y~t been trial V recall.
exposed to all the vocabulary items on 4. Long-Term Percent Retention reflects
the NART and, as a result, the tost may the Delayed Recall as a proportion of
have underestimated their IQs. · trial V recall.
4. Sample sizes per group are small.
5. Data are obtained in Australia, which The data are broken down into seven age
may limit usefulness for clinical inter- groups. In addition, the data for the summary
pretation in the United States. scores are presented using overlapping inter-
vals at specified age midpoints. The authors
[RAVLT.12] lvnik, Malec, Tangalos, Petersen, provide justification for this approach, assert-
ing its advantage over more conventional
Kokmen, and Kurland, 1990
(Tables A19.18-A19.20) group assignment. The latter tables should be
used in the context of the detailed procedures
The study provides age-specific norms for the for their application, which are explained in
Rey AVLT, derived from a sample pf 394 Ivnik et al. (1990) and not reproduced here.
cognitively intact volunteers aged :55--97, Therefore, only raw data tables are reproduced
living in Olmsted County, Mi$esota. in this book.
Participants received general medical exami-
nations performed by their primary-care Study strengths
physicians prior to enrollment in the study. 1. Demographic characteristics of the
Participants were excluded if they had. an ac- sample are well described in terms of
tive psychiatric or central nervous system geographic area, age, handedness, gen-
condition, complaints of cognitive difficulty der, education, IQ, and marital status.
during history taking and systems review, 2. The data are partitioned into age groups.
findings on physical examination suggesting 3. The scoring system is well described.
disorders with potential to affect cognition, 4. The sample sizes for each group are
certain types and dosages of psychoactive large.
medication, or prior history of disorders 5. Stringent exclusion criteria are used.
causing residual cognitive deficit; ahronic 6. Means and SDs are reported.
medical illnesses were excluded only when 7. A new technique using overlapping in-
they were reported by the physician to com- tervals at specified age midpoints is de-
promise cognition. These normative d'ta are scribed.
believed to be a reasonably unbiased :repre-
sentation of the elderly living in th. geo- Considerations regarding use of the study
graphic region. . 1. The procedure for the recognition trial is
The standard administration procedJ¥e was not specified.
used, including 30-minute delayed recall and 2. The technique proposed by the authors
recognition trials. The standard scoring pro- is quite complicated and should be used
cedure was used. Scores were also provided in the context of the detailed procedures
for the number of errors on the recopition for its application described in the orig-
trial. In addition, four summary measur~s (two inal article.
"learning" and two "memory" scores) were
computed: 1 [RAVLT.13] Miller, Seines, McArthur, Satz,
Becker, Cohen, Sheridan, Machado, VanGorp,
1. Total Learning (TL) represents a total of and Visscher, 1990 (Table A19.21)
five acquisition trials. · The article described the results obtained on
2. Learning Over Trials [LOT= TL- (5 x homosexual/bisexual males recruited in the
trial I)] is an estimate of an individual's Multi-Center AIDS Cohort Study (MACS), an
improvement over trials. · epidemiological project designed to assess the
natural history of HIV-1 infection. The article illness, change in psychotropic medication
explores the effect of HIV serostatus and during the study, and being present on the
symptom status on cognitive and motor func- ward for less than 2 weeks.
tioning. The administration procedures out- 2. Twenty-five participants were under-
lined in Lezak (1983) were employed. graduate students, with a mean age of 19
years.
Study strengths
1. Sample sizes are large. Four alternate forms of the AVLT {forms
2. Demographic characteristics of each AB, CD, EF, and GH) were used {one trial for
sample are described in terms of gender, the acquisition and interference lists, respec-
race, age, education, and geographic lo- tively). List pair AB represented the standard
cation. form used for the Rey AVLT; form CD was
3. Means and SDs are reported. the alternate list pair provided by Lezak. The
4. Administration procedures are specified. remaining two list pairs were generated based
on their match to the above lists, according to
Considerations regarding use of the study the criteria developed by the authors. The
1. The study recruited participants aged order of administration was counterbalanced
21-72. Norms are presented for all ages across dates and participants. Only one form
combined. They would be most accurate was used per session. The interval between
for young and middle-aged adults. sessions varied 2-13 days (mean= 5, SD =
2. The data are presented only for trial V 3.6). The standard administration procedure
and the total for trials 1-V. was used for each form.
3. High educational level of the sample. Means and SDs are reported for the five
4. All-male sample. acquisition trials, list B, and postinterference
5. Exclusion criteria are not specified. recall.
The authors concluded that all four forms
yielded comparable mean recall scores. Al-
Other comments
ternate form reliability coefficients for each
1. In addition to other demographic vari-
comparison varied 0.67-0.90.
ables, the paper reports race composi-
The results suggested that the use of alter-
tion (white, black, Hispanic, other), CES
nate forms may eliminate direct practice
Depression Scale scores, and CD4 cells/
effects; however, a general practice effect re-
mm 3 count.
mained due to repeated administrations when
the tests are spaced as much as 5 days apart.
[RAVLT.14] Shapiro and Harrison, 1990 This effect persisted for a number of days in
(Tables A19.22-A19.24) healthy college students but not among the
The authors developed criteria for word older patient population.
selection to generate new lists to be used as
alternate forms of the AVLT. Four AVLT Study strengths
forms were compared on two samples of 1. Established reliability indices for alter-
participants: nate forms.
1. Seventeen elderly participants were re- 3. Test administration procedures are
cruited from the patient population at the specified.
V.A. Medical Center at Salem, Virginia,
with a mean age of 66 years. They were Considerations regarding use of the study
rehabilitating primarily from stroke or 1. Sample sizes are small.
limb surgery but also had associated 2. Demographic characteristics of the par-
medical illness. Seven of these participants ticipants are scarcely described (no in-
carried a clinical diagnosis of dementia. formation on gender, educational level
Exclusion criteria were as follows: acute of the older group, IQ, or age ranges).
3. No exclusion criteria are reported for the Considerations regarding use of the study
younger group. 1. The sample represents highly function-
4. Participants in the older groUp had ing, well-educated, elderly individuals,
either neurological or medical illl)ess. which might limit the generalizability of
results.
[RAVLT.15] Mitrushina, Satz, Chervinsky and 2. Sample sizes for the youngest and oldest
D'Eiia, 1991 (Tables A19.25-A19.27) groups are relatively small.
The study explored the effect of age ~m dif- [RAVLT.16] Mitrushina and Satz, 1991a
ferent memory mechanisms in a sample) of 156 (Tables A19.28, A19.29)
healthy elderly participants (62 males,: 94 fe-
The study examined the magnitude of the
males) aged 57-85. The sample was.' parti-
practice effect in repeated administration of
tioned into four age groups, which did not
neuropsychological measures in a sample of
differ significantly in level of educa4on or
122 healthy elderly participants (49 males, 73
FSIQ. Participants with a history of reuro-
females) aged 57-85 recruited in southern
logical or psychiatric illness (per self-feport)
California. This study represents a longitudi-
were excluded. MMSE scores for all 'artici-
nal follow-up of the sample described in Mi-
pants were >24. All participants were;native
trushina et al. (1991). Participants with a
English speakers and active in the comrfunity.
history of neurological or psychiatric illness
The standard administration procedt.te was
(per self-report) were excluded. Mini-Mental
a
used: recognition trial, consisting of para-
State Exam scores for all participants
graph, followed a 10-minute interval :rtler the
were >24. All participants were native English
last recall trial, during which nonverb4! tests
speakers and active in the community. Mean
were administered. :
age was 70.4 (5.0) years, mean education was
The authors reported recall on five aJ:quisi-
14.1 (2.7) years, and mean WAIS-R FSIQ was
tion trials, postinterference recall, r~tion,
118.2 (13.0). The sample was partitioned into
number of false-positive misidentifi<htions,
four age groups, which did not differ signifi-
forgetting rates (loss of information from!trial V
cantly in level of education or gender.
to trial VI), and acquisition rates (trial v:minus
Standard administration procedures were
trial I score). :
followed. The longitudinal data over three
In addition, a serial position effect was ex-
annual probes for trial I, trial V, and post-
plored by comparing performance o~ three
interference recall are presented.
segments of the list: beginning (wo~ 1-5),
The authors concluded that recall on trial I
middle (6-10), and end (11-15).
improved over repeated probes for all age
The authors concluded that a si~ficant
groups, which might be attributed to a prac-
effect of age is evident in recall on fl.re ac-
tice effect, whereas cross-sectional compari-
quisition trials, whereas rates of acqu~ition,
sons revealed decreased scores on this trial
forgetting, and recognition are not affe~ed by
with age. The effect of longitudinal testing on
age. In addition, primacy/recency eff~ were
verbal learning and forgetting (trial V and
equally strong for all groups.
postinterference recall trial) was negated by
the effect of immediate rehearsal of to-be-
Study strengths , remembered information over five trials.
1. The data are divided into age groups.
2. Sample composition is described in Study strengths
terms of gender, native languag*', age, 1. The data are divided into age groups.
IQ, education, and geographic ar€fa. 2. Sample composition is described in
3. Administration procedure is specined. terms of age, gender, education, IQ,
4. Adequate exclusion criteria. ! English fluency, and geographic area.
5. Data on a comprehensive set of tile Rey 3. Administration procedure is specified.
AVLT scores are provided. · 4. Adequate exclusion criteria.
6. Means and SDs are reported. 5. Means and SDs are reported.
Considerations regarding use of the study [RAVLT.18] Delaney, Prevey, Cramer,

1. The sample represents highly function- and Mattson, 1992 (Table A19.31)
ing, well-educated, elderly individuals, The authors collected data on 42 control
which might limit the generalizability of participants as part of an investigation on
results. partial complex and generalized seizures and
2. Sample sizes for the youngest and oldest memory as well as anticonvulsant efficacy.
groups are relatively small. Participants were recruited in Connecticut,
3. No data for trials II-IV. California, Florida, Virginia, Massachusetts,
New York, and Minnesota. Exclusion criteria
[RAVLT.17] Seines, Jacobson, Machado, were history of neurological or psychiatric
Becker, Wesch, Miller, Visscher and
disorder or "current drug history that could
McArthur, 1991 (Table A19.30) affect performance." Mean age was 45.8 years
This article summarizes results from the (range 22-67), and mean education was 12.8
MACS described above (see Miller et al., years (range ~16).
1990). Data for 733 seronegative homosexual The administration format detailed in Lezak
and bisexual males are presented for the (1983) was employed, with the exception that
purpose of establishing norms for neu- participants were forewarned that a 20-minute
ropsychological test performance based on a delayed recall would occur. During the 20-
large sample. minute delay period, additional testing in-
The standard administration procedure was volving motor, attentional, and verbal fluency
used. In addition, recall after a 20-minute de- tasks was conducted. An alternate list (C;
lay was assessed, followed by a delayed rec- Lezak, 1983) was administered 1 month later.
ognition trial. (Recognition was not tested after Means and SDs for trials I, III, and V;
administration of the postinterference trial.) postinterference recall; 20-minute delayed
The total score for five acquisition trials and recall; and recognition are reported for both
the score on trial V as well as postinterference test forms. The two forms correlated highly
recall, delayed recall, and delayed recognition (acquisition trials, r=0.61-0.86; delayed re-
data for three age groups and three education call trials, r=0.51-0.72), providing support
groups are reported. for their comparability. Significant correla-
The authors concluded that age and edu- tions were documented between test scores
cation are important determinants of perfor- and age (r =-0.22 to -0.55) and education
mance on the Rey AVLT. (r=0.14 to 0.54).

1. Data are stratified by age and education. 1. Information on alternate forms is provided.
2. The demographic composition of the 2. Adequate exclusion criteria.
sample is well described in terms of age, 3. Information regarding age, education, and
gender, ethnicity, education, and geo- geographic recruitment area is provided.
graphic area. 4. Most test administration procedures are
3. The administration procedure is well specified.
outlined. 5. Means and SDs are reported.
4. Sample sizes are large.
5. Means and SDs are reported. Considerations regarding use of the study
1. The exact recognition procedure was not
Considerations regarding use of the study specified.
1. The generalizability of the results is 2. Participants were forewarned regarding
limited due to high educational level of delayed recall.
the majority of the sample. 3. Undifferentiated age range.
2. Data for trials I-IV are not reported. 4. No information regarding IQ or gender.
3. All-male sample. 5. Somewhat small sample.
4. No exclusion criteria are reported. 6. No data for trials II and IV.
[RAVLT.19] lvnik, Malec, Smith, Tangalos, remembered across triaJs 1-V) cor-
Petersen, Kokmen, and Kurland, 1992c rected for immediate word span (triaJ
(Table A19.32) I): LOT=TL-(5xtria1 I score).
The study provides age-specific norms for the This index represents the ability to
Rey AVLT obtained in Mayo's Older ~men improve upon triaJ I performance
cans Normative Studies (MOANS). Informa- during each of the subsequent four
tion provided in this article updates p~vious learning trials.
normative data reported by Ivnik et aJ. ~1990). 2. Recall of list B represents an index of
The present study extends the normatif base proactive interference on later word span.
from 394 to 530 participants, refines scoring 3. The Mayo Auditory-Verba] Delayed
procedures, and develops a uniform rnkthod- Recall Index (MAVDRI) includes the
ology for producing normative information on following measures:
many different tests. There were no clanges a. Recall after interference represents
in the administration procedure compted to memory status after a brief delay.
the earlier publications. b. Recall after 30-minute delay repre-
The sample consisted of cognitively normaJ sents memory status after an ex-
volunteers aged 5~97. Age categorization tended delay.
used the midpoint intervaJ technique.IMean 4. Recognition efficiency.
MAYO VIQ, PIQ, and FSIQ (which Idiffer 5. The MAYO Auditory-Verba] Percent
somewhat from standard WAIS-R IQs) for the Retention Index (MAVPRI) reflects the
whole sample were 104.8 (10.4), 106.6 (11.5), common clinicaJ practice of evaJuating
and 105.8 (10.8), respectively. The sa~ple is information recalled after a delay as a
aJmost exclusively Caucasian and living in an function of the amount of data originally
economically stable region. learned and includes the following
The standard administration procedute was measures:
used for recall triaJs. In addition, 30-minute a. Short-Term Percent Retention
delayed recall followed by a recogniti~ triaJ (STPR) expresses recall after inter-
was administered. The recognition tri~ preference as a proportion of triaJ V re-
sented 30 words italicized in Lezak's! 1976 call: STPR = 100 x (tria] VI recall!triaJ
paragraph as a two--column list, which ~ sub- V recall).
ject reads. Participants indicated recof'ition b. Long-Term Percent Retention (LTPR)
of a learned word by crossing it off the 'st. expresses delayed recall as a proportion
The AVLT component scores wer. con- of triaJ V recall: LTPR= 100 x (de-
verted to age-corrected and normaJized scaJed layed recall score/triaJ Vrecall).
scores, with a mean of 10 and an SD of3, ~hich
are derived from the cumulative fre<fency Conversion of raw and computed scores
distributions of each raw score withiQ each into scaJed scores aJlows greater comparability
midpoint age group. ScaJed scores are grpuped of these indices to each other and to perfor-
and summed within groups to derive· three mance on other tests.
summary indices: MAVLEI, MAVDRI, and These data should be used in the context of
MAVPRI. ' the detailed procedures for their application,
The following measures derived frorp Rey which are explained in lvnik et aJ. (1992c).
AVLT performance were summarized in the Therefore, they are not reproduced in this
tables provided by the authors: book. Interested readers are referred to the
originaJ article.
1. The MAYO Auditory-Verba] Le~ing
Efficiency Index (MAVLEI) conststs of Study strengths
the following measures: : 1. Demographic characteristics of the
a. Recall on triaJ I. · sample are well described in terms of
b. Learning Over TriaJs (LOT) ~aJcu geographic area, age, gender, education,
lated as TotaJ Learning (sum of~ords IQ, handedness, and ethnicity.
2. The data are broken down by age group and medical questionnaires. The number of
based on the midpoint interval technique. correctly recalled words from list A as well as
3. The administration procedure is well the number of commission errors or words
described. incorrectly identified as being on the list were
4. The innovative scoring system is well recorded. Immediate and delayed recognition
described. The authors developed new trials were used, which consisted of under-
indices of performance to explore dif- lining the ]earned words in the same para-
ferent memory mechanisms. graph for both recognition trials.
5. The sample sizes for each group are large. The authors concluded that there is no ef-
fect of gender for any of the trials. An effect of
Considerations regarding use of the study age was evident for all trials, with the excep-
1. Participants with prior history of neuro- tion of trials I and II and list B.
logical, psychiatric, or chronic medical
illnesses are included. Study strengths
2. The technique proposed by the authors 1. Administration procedures are well de-
is quite complicated and should be used scribed.
in the context of the detailed procedures 2. The data are partitioned by age and
for its application described in the orig- gender.
inal article. 3. Adequate excJusion criteria.
4. Information regarding education, re-
Other comments cruitment sources, and geographic area
The theoretical assumptions underlying this is provided.
normative project have been presented in Iv- 5. Means and SDs are reported.
nik et al. (1992a,b).
The authors cautioned that validity of the Considerations regarding use of the study
MAYO AVLT indices depends heavily on the 1. An immediate recognition trial was used,
match of demographic features of the indi- which facilitates performance on delayed
vidual to the normative sample presented in recall and recognition.
the article. 2. No information on mean educational
level.
3. Sample sizes for each group are small.
[RAVLT.20] Savage and Gouvier, 1992 4. Normative data for older groups (60--69
(Tables A19.33, A19.34) and 70-76) are not reported for the de-
The study explores the effect of age and gen- layed recall trials.
der on Rey AVLT performance. Participants
were 134 undergraduate students, senior citi- [RAVLT.21] Crossen and Wiens, 1994
zens from community programs, and others (Table A19.35)
(66 males, 68 females), forming seven age The study compares performance on the Rey
groups ranging from the late teens through the AVLT and the CVLT. Participants were 60
seventh decade. Only those participants who individuals (52 men, eight women) who had
completed 12th grade were included in the applied for jobs in the civil service involving
study (with the exception of those 16-19 years public safety and had passed medical screen-
old). All participants completed an extensive ing examinations. Mean age was 29.9 (6.2)
medical history questionnaire. Participants years, mean education was 14.7 (1.6) years,
with a history of head injury, alcoholism, and mean WAIS-R FSIQ was 106.3, with a
mental illness, cardiovascular disease, or other range of 88-133.
conditions associated with impaired memory All participants were administered both
functioning were excJuded. forms in a counterbalanced order, with an
The standard administration procedure was interval of about 2-4 hours between admin-
used, including recall after a 30-minute delay, istrations. The Rey AVLT was administered
during which participants 6lled out personal according to standard procedures. On the
recognition trial, participants were to ·.read a memory tests. A list of 50 words was used for
paragraph and circle the learned wor<Js. The the delayed recognition trial. Participants
CVLT was administered according to Ute in- were asked to identify words from lists A and
structions in the manual (Delis et al., l987). B and to specify the list of origin.
Data for acquisition trials I-V and the total The authors reported recall on five acquisi-
for five acquisition trials, list B, postinterkrence tion trials, the total number of words recalled
recall, and recognition are reported. over five trials, the interference trial, post-
The authors concluded that the CYI4T yiel- interference recall, delayed recall trial, num-
ded higher scores than the Rey AVLT.: There ber of words recognized from lists A and B,
were significant differences in performajlce on and a nonparametric measure of recognition
all parameters of the AVLT and correspf>nding performance, p(A) = 0.5( 1 + HR - FP), which
variables of the CVLT, which amounted fo one- corrects the recognition score (hit rate, HR) by
half to one full word difference on evet trial. taking into account false-positives (FP).
The results suggested no order effect and! a min- The authors concluded that the forms are
imal practice effect for different list-l~ng equivalent. In test-retest conditions, the
tests administered in the same test batteiy. highest reliability was demonstrated by the
total number of words learned over the five
Study strengths acquisition trials (r=0.77) and performance
1. Information regarding age, gend~r, IQ, on the postinterference trial (r= 0.70).
and educational level is provided. I
2. Administration procedure is outli.;ed. Study strengths
3. Sample size is sufficient. ! 1. Demographic characteristics of the sam-
4. Adequate exclusion criteria. ple are described in terms of age, gender,
5. Means and SDs are reported. , education, IQ, and geographic area.
I 2. Administration and scoring system are
Considerations regarding use of the study described.
1. Undifferentiated age range. f 3. Sample size is adequate.
2. High educational level. : 4. A comparison of two alternate forms is
3. Mostly male sample. provided.
[RAVLT.22] Geffen, Butterworth, and Geffen,
1994 (Tables A19.36-A19.39) Considerations regarding use of the study
The study explored the equivalence between 1. The 20-minute delay interval was filled
the original form of the Rey AVLT (form 1) with verbal cognitive and verbal memory
and a new form (form 4), as well as the test- tests, which might cause interference
retest reliability of both forms. Participants with delayed testing.
were 51 volunteers (25 males, 26 females) 2. The sample has high estimated IQ val-
living in Australia with no self-reported history ues, which might limit the generaliz-
of head injury or neurological abnormality. ability of data.
Mean age was 31.3 (12.7) years; ed~ation 3. Data were collected in Australia, which
ranged 6-20 years, with a mean of 12.2 (2.4); may limit their usefulness for clinical
and estimated IQ (based on NART) J.1mged interpretation in the United States.
100-128, with a mean of 115.6 (6.26). ; 4. Minimal exclusion criteria.
A new form of AVLT (form 4) was generated 5. Undifferentiated age range.
by the authors based on criteria develo:£1ed by
them. All participants were tested onj both [RAVLT.23] Torres, Flashman, O'Leary,
form 1 and form 4 in a counterbalanced order, and Andreasen, 2001 (Table A19.40)
with an interval of6-14 days between se.ions. The authors reported data for 160 healthy
The standard administration procedu~ was adult controls in a study on the effects of in-
used, with the exception of a 20-minuteldelay terference on word list recall in schizophrenia.
interval filled with verbal cognitive and terbal The sample included 84 males and 76 females
LIST-lEARNING TESTS 389
from the community, with mean age of 29.0 seronegative homosexual and bisexual males
(10.2) and 30.5 (11.7) years and mean edu- for the purpose of establishing norms for
cation of 14.4 (2.1) and 14.8 (1.9) years, re- neuropsychological test performance based on
spectively. WAIS-R VIQ and FSIQ as well as a large sample. Table A19.41 overlaps with the
parental socioeconomic status are reported. data reported earlier by Seines et al. (1991),
Participants were screened for psychiatric, which were reanalyzed to provide norms
neurological, and substance abuse history. stratified by age x education.
The Rey AVLT was administered as part of Mean age for the sample was 38.2 (7.4)
a comprehensive battery, which included years and mean education was 16.3 (2.4)
measures of various aspects of frontaVexecu- years; 91.5% were Caucasian, 2.6% Hispanic,
tive functions and memory. The standard ad- 4.9% black, and 1% other. All participants
ministration procedure was used (Lezak, were native English speakers.
1995). Recall on acquisition trials I and V, The standard administration procedure was
interference list, and recall after interference used. In addition, recall after a 20-minute
are reported for males and females separately. delay was assessed, followed by a delayed
In addition, the authors analyzed patterns recognition trial. (Recognition was not tested
of learning and proactive interference for after the postinterference trial.)
patients and controls equalized on trial I Scores for acquisition trials I and V, totals
performance as well as relationships between for five acquisition trials, interference trial, 20-
measures of proactive/retroactive interference minute delayed recall, delayed recognition,
and executive or memory functions. These and number of false-positive errors for three
data are not replicated in this book. age groups x three educational levels are re-
The authors concluded that increased ported.
susceptibility to retroactive interference is
related to frontally mediated central executive Study strengths
functions. 1. The overall sample size is large, and
most of the individual cells have more
Study strengths than 50 participants.
1. Large sample size. 2. Normative data are stratified by
2. The sample composition is well de- age x education.
scribed in terms of age, education, gen- 3. Information on age, education, ethnicity,
der, VIQ, and FSIQ. and native language is reported.
4. Test administration procedures are reported.
specified.
reported. 1. All-male sample.
Considerations regarding use of the study 3. No information on exclusion criteria.
1. Recruitment procedures are not re- 4. Data for trials II-IV are not reported.
ported.
2. The data are not partitioned by age [HVLT-R.1] Friedman, Schinka, Mortimer, and
group. Borenstein Graves, 2002 (Table A19.42)
3. FSIQ for the sample falls within the high Data were obtained from the Hillsborough
average range. Elder African American Ufe Study (HEALS).
4. Data for trials II-IV are not reported. The authors examined the influence of de-
mographic characteristics on HVLT-R per-
[RAVLT.24] Miller, 2003; Penonal formance in a community-dwelling sample of
Communication (Table A19.41) 237 African-American older adults (108 men,
The investigation used participants from 129 women), aged 60-84 years and living in
the MACS. Data were collected from 920 Tampa, Florida. Mean age and education are
provided for each demographic group sepa- 3. No information on IQ is reported.

rately. Information about the particip~ts was 4. The administration procedure was modi-
obtained using structured interviews. · fied to include a delayed cued recall trial.
Form 1 of the test was administered fol- This might facilitate performance on the
lowing the standard procedure, with the ex- Recognition trial.
ception of addition of a cued recall (by
category) trial following delayed free i recall.
[WHO-UCLA AVLT.1] Ponton, Satz, Herrera,
Performance indices included Delayed Cued
Ortiz, Urrutia, Young, D'Eiia, Furst, and
Recall and Learning in addition to thf stan-
Namerow, 1996 (Table A19.43)
dard indices. :
The authors fmmd a moderately largJ effect The WHO-UCLA Auditory Verbal Learning
of age and moderate effects of educatifn and Test, Spanish version, was administered to
gender on test performance. Therefote, the Spanish-speaking volunteers as part of a Jarger
data for all performance indices are; parti- battery in a project designed to provide
tioned into two age levels and three Fduca- standardization of the Neuropsychological
tional levels and reported for males and Screening Battery for Hispanics (NeSBHIS).
females separately. In addition to raw .erfor- Volunteers were recruited through fliers and
mance indices, the authors provide tabifs that advertisements in community centers of the
convert raw data into percentiles, with )Corre- greater Los Angeles area over a period of
sponding education and gender adjus~ents 2 years. Exclusion criteria were a history
for the two age groups. , of neurological or psychiatric disorder, drug or
In Table A19.42, we reproduced the data alcohol abuse, and head trauma. Data for a
for selected performance indices: Total Recall sample of 300 participants with a median ed-
(sum of trials 1 through 3), Delayed 1\ecall, ucational level of 10 years were analyzed.
Percent Retention [Delayed Recalllhi~er of Participants ranged in age 16-75 years, with a
the scores on trials 2 and 3 x 100], an4 Rec- mean of 38.4 (13.5) years. Education ranged
ognition Discrimination Index (total . false- 1-20years, with a mean of10.7 (5.1) years. The
positive- total true-negative score). For the male to female ratio was 40%/60%. The aver-
normative data on other indices, see the age duration of residence in the United States
original article. was 16.4 (14.4) years. Seventy percent of the
sample were monolingual Spanish-speaking,
Study strengths and 30% were bilingual. The proportion of the
1. Large sample size. sample respective to their country of origin
2. The sample composition is described in closely approximates the 1992 U.S. Census
terms of age, education, gender, and distribution. Correlations between Marin and
geographic area. . Marin (1991) acculturation scale scores and
3. Test administration procedures are neuropsychological variables are provided.
specified. Five acquisition, interference, recaU after
4. Means and SDs for the test scores are interference, and 20-minute delayed recall
reported, as are tables for conver~on of trials were administered. The authors re-
raw scores into T scores. · ported data for recall on trial V, recall after
5. The data are partitioned by age, ~uca interference, and delayed recall.
tion, and gender.
Study strengths
Considerations regarding use of the study 1. Large overall sample with acceptable
1. Although information obtained in struc- size for most cells.
tured interviews is described, exclusion 2. The sample composition is well described
criteria are not specified. in terms of age, education, gender, ac-
2. Recruitment procedures are not re- culturation information, geographic area,
ported, although a reference is prqvided and recruitment procedures.
to another study. i 3. Adequate exclusion criteria.
'
4. Test administration and scoring proce- included into the analyses: eight studies,
dures are specified. which generated 24 data points based on a
5. Means and SDs for the test scores are total of 1,910 participants for trial I; eight
reported. studies, which generated 23 data points based
6. Data are partitioned by gender x age x on a total of 1,901 participants for trial V; se-
education. ven studies, which generated 20 data points
based on a total of 983 participants for RAI;
Considerations regarding use of the study four studies, which generated 14 data points
1. No information on IQ is reported. based on a total of 453 participants for Rec-
2. It is unclear which of the two educa- ognition; and six studies, which generated 20
tional groups included participants with data points based on a total of 1,699 partici-
10 years of education. pants for Total Recall.
Quadratic regressions of the test scores on
age yielded R2 of 0.842 for trial I, 0.877 for trial
V, 0.923 for RAI, 0.948 for Total Recall, as well
as R2 of 0.892 based on linear regression for
OF THE REY AVLT DATA
Recognition, indicating that 84%-95% of var-
(See Appendix 19m)
iance in the test scores for the five measures is
Data collected from the studies reviewed in accounted for by the models. Based on these
this chapter were combined in regression models, we estimated scores for the five mea-
analyses, to describe the relationship between sures for age intervals between 20 and 79
age and Rey AVLT performance and to predict years. If predicted scores are needed for age
test scores for different age groups. Effects of ranges outside the reported boundaries, with
other demographic variables were explored in proper caution (see Chapter 3), they can be
follow-up analyses. The general procedures for calculated using the regression equations in-
data selection and analysis are described cluded in the tables, which underlie calcula-
in Chapter 3. Detailed results of the meta- tions of the predicted scores. The predicted
analyses and predicted test scores across adult scores are relevant for the following adminis-
age groups are provided in Appendix 19m. tration sequence: five acquisition trials, inter-
Separate analyses were performed on the ference trial, recall after interference, and
data for trial I, Trial V, Recall after Interfer- recognition (immediate or after a short delay).
ence (RAJ), Recognition, and Total Recall for It should be noted in the context of across-
five acquisition trials. In addition, patterns of condition comparisons that mean age for the
learning (trial V- trial I) and forgetting (trial Recognition condition is considerably lower
V- RAI) were examined. Original data cor- than mean ages for other conditions because
responding to three points along the age data for several large studies based on the
continuum are presented in order to demon- older samples were not available for the
strate the slope of decline in learning capacity. Recognition condition.
A separate run of the regression analysis was Regressions of SDs on age for all five con-
performed on the forgetting scores. The pre- ditions suggest that age does not account for a
dicted values for the forgetting curve based on signiAcant amount of variability in SDs (R 2
linear regression (R2 = 0. 743, Fo.s) = 52.38, ranges 0.010-0.471). Though some increase in
p< 0.0004) are presented for comparison variability with advancing age is expected, this
purposes in the summary table (see Table trend was not statistically significant in the
Al9m.6). Supporting statistics for the forget- collected data. Therefore, we suggest that the
ting scores are not reported. Data for Delayed mean SD for the aggregate sample be used
Recall were not analyzed due to inconsistency across all age groups.
in terms of administration procedures and Examination of the effects of demographic
high variability in the delay interval. variables on Rey AVLT scores indicated that
After data editing for consistency and for education did not contribute significantly to
outlying scores, the following data were scores in the data available for analyses. The
effect of intelligence level on Rey AVLT per- Limitations of the analyses

formance was not explored due to a scarcity of 1. Postestimation tests for normality for
data available for review. RAI and Recognition were marginally
The effect of gender on test perfonnance significant. The Kdensity plot for RAI
was examined using a series of t-tests. A sta- demonstrated an asymmetrical curve,
tistically significant difference in fa\'or of with a nearly bimodal distribution of
males was found for trial I, trial V, and Rec- residuals, reflected in a small bump at
ognition. This is in contrast to some ev;,dence the lower extreme of the curve, possibly
of superiority of females available in the lit- pointing to a diagnostic significance of
erature. The significant findings of gender low scores on the RAI. The Kdensity
differences in our analyses are not reli.ble as plot for Recognition revealed a nega-
the numbers of data points included ·.in the tively skewed distribution of the resid-
analyses are very small for both male and uals. These deviations from normality do
female samples (seven and five for trial I; not affect the estimates of regression
seven and four for trial V; five and Rve for coefficients and accuracy of prediction
Recognition). Analyses for RAI, which in- but do influence the results of signifi-
cluded nine and eight data points, and for cance tests.
Total Recall, which included eleven and nine 2. Data for only a narrow range of higher
data points, did not yield significant sender levels of education are available for the
differences. This suggests that the jender analyses (12.~16.0 years or even more
differences found in the earlier analY!tes are narrow, depending on the variable).
likely due to individual differences between Mean educational levels for all variables
samples. range 13.92-14.10. We were unable to
fully explore the effect of education on
Strengths of the analyses the test scores because lower educa-
1. Total sample size of 1,910 for trial I; tional levels are not represented. Though
1,901 for trial V; 983 for RAI; 453 for reports on the relationship between ed-
Recognition; and 1,699 for Total Recall. ucation and test scores are equivocal, a
2. R2 of 0.842 for trial I, 0.877 for trial V, number of studies suggest that higher
0.923 for RAI, 0.892 for Recogpition, levels of education are associated with
and 0.948 for Total Recall, indicating a better test performance. Therefore, the
good model fit. predicted values might overestimate ex-
3. Postestimation tests for parametem: spec- pected scores for individuals with lower
ifications did not indicate problent; with educational levels.
normality or homoscedasticity, with the 3. Although the effect of intellectual level
exception of the marginally significant on Rey AVLT performance has been
tests for normality for RAI and l\ecog- reported in several studies, we could not
nition. include measures of intellectual level
4. Although the data available for diferent in our analyses due to scarcity of this
conditions vary considerably in tetms of information in the data available for
the number of studies included aitd re- analyses.
sulting sample sizes, a compari$>n of 4. Equivocal findings for gender differ-
weighted means for different concJitions ences in test performance are likely to be
suggests that inter-trial differences are due to small numbers of data points for
very close to those reported in the lit- each gender included in the analyses.
erature.
5. The predicted values match closely the
metanorms provided by Schmidt (1996)
CONCLUSIONS
for some variables and are similar in the
rate of age-related changes for most of A review of the literature o'n the list-learning
the variables. ; tests suggests high clinical utility of these tests
due to their sensitivity to disruption in different demographic characteristics should also be

memory mechanisms reSected in various indi- of concern for clinicians and investigators
ces derived from the test performance. Hence alike.
the importance of reporting a full range of test For future research on list-learning tests,
scores in clinical practice and in research, in- classification of data into age groups, educa-
cluding delayed recall and recognition (if ad- tionaVIQ levels, and gender would be desir-
ministered) to assure optimal use of these tests able. Development of new and validation of
in identifying faulty memory mechanisms. existing alternate forms for the Rey AVLT
Moreover, thorough description of administra- would aid clinicians and researchers in mini-
tion procedures, especially deviations from mizing carry-over effects in retesting situa-
standard procedures, is of utmost importance tions. The utility of different list-learning tests
for accurate interpretation of test scores. with different diagnostic groups should be
Due to the considerable effect of age on further explored. In addition, further studies
list-learning test scores, individual sets of data on different memory mechanisms using con-
should be referenced to an appropriate age cepts adapted from cognitive science would
group. Similarly, education, intelligence level, be of great value in understanding the un-
and gender are possible contributing fac- derlying cognitive processes associated with
tors to test score variance. Therefore, these performance on these tests.
20
Benton Visual Retention Test
BRIEF HISTORY OF THE TEST card. Six general categories of errors can be
scored: omissions, distortions, perseverations,
The Benton Visual Retention Test (BVRT) rotations, misplacements, and size errors. Some
measures immediate visual memory and visual authors have suggested that application of a
perceptual-constructional skill and has several background interference procedure to standard
administration formats and versions. Each administration of the BVRT may have utility
standard version contains 10 line drawings on (Crockett et al., 1983).
individual pages measuring 8.5" x 5.5" in an Further information regarding administra-
easel-style booklet. The patient uses a response tion and scoring can be found in the manual for
booklet that contains 10 blank pages on which the BVRT, 5th edition (Benton-Sivan, 1992).
the patient reproduces the designs. The first A 15-item recognition version was published
two designs contain single large stimuli, while in French by Benton in 1965 (in Fabrigoule
the remaining eight pages depict two larger et al., 1998), and in 1977 Benton and col-
figures and one peripheral figure. There are leagues produced an unpublished manuscript
three alternate versions of the 10 cards: forms in English detailing the recognition adminis-
C, D, and E. In the first administration format tration. In this administration, the usual BVRT
(A), each card is presented for 10 seconds and three-figure constellation, with two larger fig-
the patient is to draw the design immediately ures and one peripheral figure, is presented for
after removal of the stimulus. In administration 10 seconds. Immediately following removal of
B, the clinician presents each stimulus for only 5 the stimulus, the subject is presented with
seconds prior to the patient's attempt at repro- another card that has four similar arrays. Three
duction. In administration C, the patient copies of these choices are variations on the original
each design while designs remain in view. In stimuli, while the fourth design matches the
administration D, each stimulus is presented for target exactly. The patient is to choose the al-
10 seconds but a 15-second delay is interspersed ternative which matches the original stimulus.
between presentation and reproduction. Benton also reported an administration in
The scores derived from the test include which the target is presented simultaneously
total correct out of 10 and total errors. The with the four choices, to measure visual dis-
upper range of errors is approximately 24, crimination rather than memory; this version
although the possible range extends higher subsequently was renamed the Visual Form
because four errors can be committed on each Discrimination Test.
394
BENTON VISUAL RETENTION TEST 395
Some studies have indicated that patients of dementia is highly predictive of significant
with right hemisphere or diffuse lesions per- decline in visual memory performance 3 years
form worse than patients with left hemi- later (Small et al., 1995), and the BVRTwas one
spheredamage (Crockettetal., 1983), but other of three tests predictive of Alzheimer's disease
data have failed to confirm a relationship be- diagnosis 1-3 years later (Dartigues et al.,
tween lesion laterality and BVRT perfor- 1997). Individuals with Alzheimer's disease
mance (Arena & Gainotti, 1978; Benton, 1962). show larger declines in BVRT performance in
Soininen and colleagues (1994) reported a the 6 years prior to diagnosis (Zonderman et al.,
significant relationship between BVRT scores 1995), and a greater number of BVRT errors
and volume of the right hippocampus and is associated with an increased risk of Alzhei-
magnitude of asymmetry between the right and mer's disease up to 15 years later (Kawas
left hippocampi in patients with age-associated et al., 2003). The memory trial, but not the
memory impairment and that the lowest- copy trial, of the BVRT discriminates normals
scoring patients had reduced volume of both from patients with very mild Alzheimer's dis-
right and left amygdala. Further, patients with ease (Robinson-Whelen, 1992), and worse
anterior lesions have been reported to dem- performance on the BVRT copy but rela-
onstrate more perseverations than patients tively better performance on BVRT delayed
with posterior lesions (Vilkki, 1989). recall predicted a more rapid cognitive de-
A sample of patients with thalamic stroke cline (Rasmusson et al., 1996). BVRT perfor-
was described as exhibiting moderate to severe mance declined less rapidly than language
impairment on the BVRT (Radanovic et al., skills over a 2-year period in Alzheimer's
2003). Poorer BVRT scores have also been patients (Rebok et al., 1990).
associated with larger white-matter hyper- The BVRT was reported to differ in indi-
intensityvolume (Swan et al., 1998). The larger viduals with a family history of Alzheimer's
the lesions, the higher the percentage of errors, disease vs. those without (Smalley et al., 1992)
especially of distortion and laterality (Kasahara and to identify persons with one apolipoprotein
et al., 1993, 1995). In addition, lesions in basal E4allelle from those with none (Soininen et al.,
ganglia were associated with distortion errors, 1995). Palmer and colleagues (1994) reported
while lesions in the thalamus were related to that there are two distributions of BVRT scores
perseverations (Kasahara et al., 1995}. The in offspring of patients with Alzheimer's dis-
relationship between these subcortical lesions ease, with 6% of participants performing in a
and BVRT errors may at least partially explain low-scoring cluster hypothesized to be at-risk
the decline in BVRT performance observed carriers of a putative Alzheimer's gene.
with age; 70% of elderly participants are found In addition to the sensitivity of quantitative
to have these abnormalities on magnetic res- scores to Alzheimer's disease, qualitative anal-
onance imagery (MRI) (Kasahara et al., 1995). ysis of errors has shown promise in the de-
BVRT performance is also lowered in elderly tection of Alzheimer's disease. Specifically,
men with total brain volume below the median increases in omission and perseveration errors
(Carmelli et al., 2000). No relationship be- appear to selectively occur in Alzheimer's dis-
tween BVRT scores and resting regional ce- ease (La Rue et al., 1986; Vollant et al., 1986).
rebral glucose metabolism rates have been However, error types, with the exception of
observed (Haxby et al., 1986). distortion errors, for both the memory and
In mixed dementia patients (Eslinger et al., copy trials may be less useful in tracking the
1985) and in Alzheimer's disease patients progression of dementia due to the promi-
(Jacqmin-Gadda et al., 2000), the BVRT was nence of omission errors as dementia severity
one of the most sensitive tests included in a increases. BVRT scores are comparable in in-
dementia screening battery; it is also effective dividuals with Alzheimer's disease and Par-
for discriminating mild Alzheimer's disease kinson's disease with dementia but lower than
from Age-Associated Memory Impairment those observed in Parkinson's disease without
(Youngjohn, 1992). BVRT performance in in- dementia (Kuzis et al., 1999). BVRT scores
dividuals with memory loss without diagnosis were also lowered in individuals who were
carriers for Huntington's disease but npt clin- performance on the BVRT in schizophrenia
ically diagnosed with the condition (Witjes-Ane patients appears to be due to poor visual
et al., 2003). scanning of the figures (e.g., failure to look at
Patients with right or left temporal lobe peripheral figures) (Obayashi et al., 2003;
epilepsy score significantly below controls on Tsunoda et al., 1992). Identification of happy
the BVRT but not differently from each other faces was significantly associated with BVRT
(Helmstaedter et al., 1995; Mayeux , et al., performance in schizophrenic patients, al-
1980) or from patients with generalized epi- though identification of sad or neutral faces
lepsy (Mayeux et al., 1980). However, the per- was not (Silver et al., 2002).
formance of patients with right temporal lobe Bipolar patients also show deficient perfor-
epilepsy, but not that of left foci patifnts or mance on the BVRT (Loo et al., 1981), although
controls, appears to be mediated by verbal lithium treatment in this population has not
learning capacity; the selective impairdtent in been related to BVRT scores (Engelsmann et al.,
right temporal lobe seizure patients btcomes 1988). BVRT performance of depressed pa-
apparent when the complexity of the items tients is lower relative to controls (Crookes &:
exceeds the compensatory verbal mempry ca- McDonald, 1972; Dealberto et al., 1996; La
pacity (Helmstaedter et al., 1995). In a¥tion, Rue et al., 1986; Mormont, 1984) and further
patients with temporal lobe epilepsy, show suppressed in the presence of psychotic
distinct eye movements when completing features but not impacted by the effects of
the BVRT that vary with laterality of feizure medication treatment (Shipley et al., 1981).
foci and suggest functional overactivation of BVRT scores do not decline post-ECTor in
the epileptogenic hemisphere (Sonobe.f et al., response to raised blood pressure post-ECT
1991). Reduction of polypharmacy in epileptic (O'Donnell&:Webb,1986).BVRTperformance
patients has been associated with iiJt>rove- has been reported to be lowered in PTSD
ment in BVRT scores (Ludgate et al., 1985). (Kirling-Boden &: Sundbom, 2003) but not in
The BVRT was the only cognitive ~est of women traumatized by childhood sexual abuse
15 administered which was significantly low- (Stein et al., 1999). Children and adolescents
ered in HIV+ men with no or minor. symp- with learning disabilities have exhibited a high
toms (Collier et al., 1992). BVRT perfmtnance level of errors, especially distortions (Snow,
is also suppressed in multiple sclerosi~ (Rug- 1998), and children with attention-deficit hy-
gieri et al., 2003), with nearly 50% of this peractivity disorder (ADHD) receiving stimu-
population showing declines on this 'nstru- lant medication have also scored significantly
ment (Arias et al., 1991), although resulp have lower on the BVRT (Risser &: Bowers, 1993).
been contradictory regarding the relatipnship Abusers of cocaine and heroin and polydrug
between extent of decline on the BVRT and abusers exhibit lowered BVRT performance
duration of illness (Halligan et al., 198li; Rug- (Amir &: Bahri, 1994; Rosselli et al., 2001b).
gieri et al., 2003). BVRT scores are al*> low- Cannabis-dependent adolescents score lower
ered in Turner's syndrome (Downey ,et al., on the BVRT than controls but show some
1991), in females from fragile X families improvement in test scores after 6 weeks of
(Miezejeski et al., 1986), and in female (Jlrriers abstinence (Schwartz et al., 1989). Alcoholics
of the fragile X syndrome who inherited the with cirrhosis of the liver display particularly
fragile X chromosome from their ~thers depressed BVRT scores (Arria et al., 1991).
(Hinton et al., 1992). In 1983, at the behest of the World Health
The BVRT is suppressed in schizophrenic Organization and the U.S. National Institute
patients (Goldsmith&: Brengelmann, 1911) and for Occupational Safety and Health, experts
in patients with tardive dyskinesia (B~els &: proposed the Neurobehavioral Core Test Bat-
Themelis, 1983), with clinical impro~ment, tery (NCTB) to identify nervous system effects
as measured by the Positive and Negative of chemical exposures in humans worldwide.
Syndrome Scale (PANSS), significantly,corre- The BVRT, administered in a recognition for-
lated with fewer errors on the BVRT (*ollnik mat, was included in this battery and in this
et al., 2002). At least some of the lcf'ered context has been widely used to assess the
.
effects of workplace chemical exposures in 1988). Long-term liver transplant survi-

various cultures. BVRT scores were lower in vors show significantly depressed BVRT scores
mercury-exposed Norwegian workers (Elling- (Lewis & Howdle, 2003). Conflicting data re-
sen et al., 2001; Mathiesen et al., 1999) and was garding the effects of estrogen on BVRT per-
the most sensitive measure to the presence of formance have been reported, with some
mercury exposure in Chinese workers, regis- investigators citing a reduction in BVRT er-
tering an effect size of 6.0 (Zhou et al., 2002). rors with estrogen replacement therapy (Re-
BVRT performance was also lower in Chinese snick et al., 1997) and others finding no
workers exposed to lead (Hai-Wang et al., 1995) relationship between estrogen level and test
and in Indian petrol pump workers (Kumar scores (Portin et al., 1999). Individuals who
et al., 1988). Similarly, Indian workers exposed have never smoked perform better on the
to organophosphates (Misra et al., 1994) and BVRT relative to current smokers or recent
Egyptians who worked with organophospho- quitters, and light drinkers (::::;1 drink per day)
rous pesticides (Farahat et al., 2003) have perform better than nondrinkers (Carmelli
demonstrated lower BVRT scores. Japanese et al., 1999).
workers and female workers from Singapore Of more obscure interest, BVRT scores are
exposed to toluene, Singapore workers exposed higher in children who eat breakfast at school
to styrene, Korean painters and printers, and rather than at home or who go without, sug-
spray painters in Singapore have shown lower gesting that ingestion of food 30 minutes prior
BVRTscores (Chiaet al., 1994; Foo et al., 1990; to test administration enhances performance
Kishi et al., 1993; Lee & Lee, 1993; Ng et al., (Vaisman et al., 1996). The BVRT did not
1992), with loss of color vision in exposed differ between individuals with psychic expe-
populations particularly associated with lower riences and those without (Fenwick et al.,
BVRT scores (Dick et al., 2004). BVRT scores 1985), and brain MRI has no significant impact
were not lower in rescue workers 3 years after on BVRT scores (Sweetland et al., 1987).
the Tokyo subway sarin attack (Nishiwaki et al., BVRT scores were not lower in persons who
2001), in paint formulators in Singapore (Foo involuntarily retire vs. voluntary retirees (Swan
et al., 1994), or in Venezuelan workers exposed et al., 1991).
to organic solvents (Escalona et al., 1995).
Some data suggest that patients with ele-
vated diastolic and systolic blood pressure
readings show declines on the BVRT (Re- Factor analysis of BVRT and Benton Visual
inprecht et al., 2003), although some studies Form Discrimination (BVFD) scores has sug-
have failed to detect this relationship (Swan gested that BVRT free recall and recognition
et al., 1998). Patients post-cardiopulmonary scores load on a factor with BVFD while BVRT
bypass may show slight improvement on the copy scores load on a separate factor (Moses,
test (Zeitlhofer et al., 1993). Diabetes has not 1986). The fact that number correct and error
been found to suppress BVRT scores after scores are found on the same factor has led to
adjustment for age and education (Robertson- the suggestion that these two scores may be
Tchabo et al., 1986). BVRT performance ap- redundant and that no appreciable information
pears to be highly predictive of 3-year survival is gained by specifying error type over and
status in chronic obstructive pulmonary dis- above total errors (Moses, 1986). In a sample
ease (COPD) patients (Fix et al., 1985), but of mentally retarded individuals, one-third of
scores did not improve in COPD patients the variance in memory scores was accounted
administered long-term oxygen treatment for by copying ability, although free recall and
(Borak et al., 1996) and BVRT scores are not recognition scores were not significantly re-
lowered in snoring- or sleep-related breathing- lated (Silverstein, 1962). Factor analyses using
stoppage episodes (Dealberto et al., 1996). a larger battery of neuropsychological test
No difference in BVRT performance was scores has indicated that BVRT loads on visual
reported in lung cancer patients treated with processing (Snow, 1998) and visual perceptual-
chemotherapy vs. radiotherapy (Kaasa et al., motor (Larrabee et al., 1985) factors, rather
than a memory factor, along with visual re- RELATIONSHIP BETWEEN BVRT
production immediate, block design, and ob- PERFORMANCE AND DEMOGRAPHIC
ject assembly. BVRT performance has been FACTORS
significantly correlated with the Bender Ge-
stalt but only modestly associated with verbal Increasing age has been consistently reported
memory measures (Fabrigoule et al., 1998; to adversely impact BVRT scores (Arenberg.
Moses, 1986; Snow, 1998), digit span (Moses, 1978, 1982; Benton, 1974; Benton et al.,
1986), psychomotor speed (Digit Symbol; 1981; Chen et al., 1990; Coman et al., 1999,
Fabrigoule et al., 1998), and verbal abstraction 2002; Dartigues et al., 1992; Dealberto et al.,
(Fabrigoule et al., 1998). BVRT scores were a 1996; Duara et al., 1984; Giambra et al., 1995;
significant predictor of performance on Be- Jacobs et al., 1997; Lee & Lee, 1993; Mor-
chara's gambling task (Torralva et al., 2000) mont, 1984; Palmer et al., 1994; Prakash &
and more related to source memory than was Bhogole, 1992; Resnick et al., 1995;
the WCST (Dywan et al., 1993). Lowered Robinson-Whelan, 1992; Shichita et al., 1986;
BVRT scores are associated with commission Shipley et al., 1981; Snow, 1998; Youngjohn
errors on continuous performance tasks et al., 1993; Zappala et al., 1995), with age
(Dougherty et al., 2003). accounting for approximately 10% of test
Correlations between forms have generally score variance (Youngjohn et al., 1993). The
been respectable (0.79--0.84), although Amir most pronounced decline occurs in either the
(2001) reported a more modest relationship 65--74 year decade (Giambra et al., 1995) or
between versions D and E at a 2-week re- at age 75 (Coman et al., 2002). Increase in
test inteJVal (correct=0.455, errors=0.491). errors was moderate in men in their 50s and
Forms may not be of equivalent difficulty; 60s but large for men over age 70 (Arenberg,
Benton-Sivan (1992) indicates that under 1978), with no further decline detected in
administration A, form C may be slightly easier participants between 80 and 92 (Klonoff &
than forms D and E. Test-retest reliability Kennedy, 1965). Age-related losses in BVRT
may be somewhat low (number correct= 0.57, performance are less pronounced in very
number of errors= 0.53) (Youngjohn et al., healthy individuals (Haxby et al., 1986). Quali-
1993). tative analyses have shown that while all error
Randall et al. (1988) reported interrater re- types increase with age, greater age effects
liabilities of 0.85 for number correct and 0.93 have been observed for distortions, omissions,
for number of errors, while higher values were and rotations (Resnick et al., 1995). Some
cited by Prakash and Bhogle (1992, 0.95 for data suggest that BVRT procedures involv-
number correct) and Swan et al. (1990, 0.963 ing recognition trials may be less affected by
for total number correct and 0.974 for total age (Anger et al., 1993). In longitudinal
errors). Kappa values for each error type ran- studies, educational level and type of activities
ged from 0.976 for omissions to 0. 737 for size; of daily life somewhat attenuate age-related
agreements were lowest for misplacement on performance decline (Shichita et al., 1986).
design 9 and size errors on design 10 (0.440 BVRT performance is significantly affected
and 0.480, respectively; Swan et al., 1990). by education (Amir, 2001; Anger et al., 1993;
An initial attempt at developing a shortened Coman et al., 1999, 2002; Dartigues et al.,
version of the memory trials (i.e., <10 items) 1992; Dealberto et al., 1996; Jacobs et al.,
indicated that correlations between the full 1997; Le Carret et al., 2003; Lee & Lee, 1993;
and abbreviated versions ranged from 0.829 Palmer et al., 1994; Ritchie & Hallerman,
for five items to 0.987 for nine items. How- 1989; Shichita et al., 1986; Youngjohn et al.,
ever, the small savings in administration time 1993; Zappala et al., 1995; with the exception
was not judged adequate for the sacrifice in of Robinson-Whelen, 1992), and the rela-
interpretation accuracy (Benton, 1972). The tionship between BVRT scores and education
1992 manual does reference an eight-item is more pronounced in lower-education com-
version for the copy trial. pared to higher-education groups (Kang, 2000;
Shichita et al., 1986). Further, in samples figures and low education; the sample aver-
with <3 years of education, illiterate individ- aged 3 years of formal education (32% had no
uals perform worse than literate subjects formal education, 42% had 1-3 years, and
(Manly et al., 1999). The effect of education 26% had 4-19 years of education) and 74%
on BVRT performance appears to be medi- were illiterate or only marginally literate. In
ated by enhanced executive abilities rather fact, performance reported in a sample of
than visual discrimination skills (Le Carret Venezuelan controls averaging 8 years of ed-
et al., 2003); for example, participants with ucation was significantly higher (6.2 vs. 4)
more education use a more exhaustive explo- than that of the original Venezuelan data.
ration strategy on the BVRT recognition trial Of note, the effects of age and education on
(Le Carret et al., 2003). BVRT performance are moderated by level of
Separate from the effects of education, cognitive impairment; patients with moderate/
poorer BVRT scores have been observed in severe cognitive deficits do not register an
blue-collar workers compared to professionals/ effect of these demographic variables on test
managers (Dartigues et al., 1992); however, performance, although patients with more
other investigators have not found a relation- mild cognitive deterioration continue to show
ship between BVRT performance and occupa- a demographic impact on test performance
tion (Palmer et al., 1994). (Coman et al., 1999).
BVRT scores have been reported to be Some studies have reported a relationship
similar across a broad range of cultures, al- between IQ and BVRT scores (Amir, 2001;
though performance may be atypical for very Benton, 1945, 1962; Netherton et al., 1989),
poorly educated samples (Anger et al., 1991). although this appears to be confined to low IQ
For example, older native Spanish speakers levels; individuals of average or higher IQ
residing in the United States with an average perform comparably (Randall et a!., 1988).
of 8 years of education scored lower than Scores of mentally retarded individuals are
English speakers (Jacobs et al., 1997). Simi- impaired, with this group reproducing from
larly, French speakers showed a trend toward memory only two of 10 designs accurately
lower performance relative to English speak- (Silverstein, 1962) and committing an average
ers in Canada, although this finding was ofll-12 errors (Silverstein, 1963). Performance
confounded by the lower educational level in on memory trials improves in a stepwise fash-
the French speakers (Steenhuis & Ostbye, ion from mentally retarded to borderline, from
1995). Older African Americans scored lower borderline to low average, and from low av-
than non-Hispanic whites, but after covarying erage to average, while on the copy trial,
for reading level, group differences disap- performance improves from the mentally re-
peared (Manly et al., 2002). U.S.-bom vs. tarded to low average level and then plateaus
foreign-born older non-Hispanic white sub- (Randall et al., 1988).
jects who were very fluent in English did not Studies have generally indicated no effect
differ in BVRT performance (Touradji et al., of gender on BVRT scores (Amir, 2001; Anger
2001). Farahat et al. (2003) reproduces data et al., 1993; Chen et al., 1990; Coman et al.,
showing that performance of control partici- 1999, 2002; Dartigues et al., 1992; Youngjohn
pants in Austria, France, Italy, Poland, Hun- et al., 1993), although Shichita et al. (1986)
gary, the Netherlands, China, the United did observe lower scores in older Japanese
States, and Ecuador is comparable, although women relative to men, likely related to lower
scores of participants from Nicaragua and levels of education in women in this cohort.
Egypt were lower, which the authors suggest Some data also suggest that women commit
may be due to differences in experience with more errors (Giambra et al., 1995) and, in
testing and/or educational background. Anger particular, that older women may exhibit more
and colleagues (1993) suggest that factors lead- rotation and omission errors than older men
ing to poor performance in Nicaraguan partic- but that men show steeper increases in
ipants were lack of experience with geometric omission errors with age (Resnick et al., 1995).
400 · VERBAL AND VISUAL LEARNING AND MEMORY
METHOD FOR EVALUATING TttE Reporting of IQ Levels

NORMATIVE REPORTS 1 Given the probable relationship between BVRT
Our review of the literature located a few BVRT performance and IQ, information regarding
normative reports for adults and dozens "f other intellectual level should be provided.
studies which have reported control ~ubject
data, and we have confined our discu~ion to Procedural Variables
those investigations which involved a Eample
size of at least 50 and represented the s dard Specification of Test Version
test administration (i.e., studies in whi h only Given the four standard administration options
half of the items were administered or a elayed plus the recognition format as well as the three
recall trial was given, etc., were excludefl). stimuli form, studies should report which
To adequately evaluate the BVRT ~orma version/administration format was followed.
tive reports, seven key criterion variabl~s were
deemed critical. The first five of these rtlate to Data Reporting
subject variables, and the remaining tWp refer Means and standard deviations for total cor-
to procedural issues. . rect and/or total errors are required.
terion variables were as follows. !

Subject Variables OF THE NORMS
Sample Size Twenty-three data sets had total sample sizes
Fifty cases are considered a desirable Sample 2:100 (Alder et al., 1990; Amir, 2001; Aren-
size. Although this criterion is somewhat ar- berg, 1978; Benton, 1962; Benton-Sivan,
bitrary, a large number of studies sugg~st that 1992; Carmelli et al., 1999; Coman et al.,
data based on small sample sizes are; highly 1999; Dealberto et al., 1996; Giambra et al.,
influenced by individual differences ~d do 1995; Jacobs et al., 1997; Kawas et al., 2003;
not provide a reliable estimate of the! popu- Klonoff & Kennedy, 1965; Manly et al., 2002;
lation mean. I Palmer et al., 1994; Prakash & Bhogle, 1992;
Randall et al., 1988; Reinprecht et al., 2003;
Sample Composition Description Resnick et al., 1995; Robertson-Tchabo &
Arenberg, 1989; Robinson-Whelen, 1992;
As discussed previously, information regard-
Steenhuis & Ostbye, 1995; Touradji et al., 2001;
ing medical and psychiatric exclusion criteria
Youngjohn et al., 1993).
is important; it is unclear if geographic re-
Twenty-two of the studies summarized in
cruitment region, gender, socioecononpc sta-
this chapter present BVRT data according to
tus or occupation, ethnicity, and rec~tment
circumscribed age ranges and/or age sub-
procedures are relevant. Until determiped, it
groups (Alder et al., 1990; Arenberg, 1978;
is best that this information be providep.
Benton-Sivan, 1992; Carmelli et al., 1999; Co-
man et al., 1999; Dealberto et al., 1996; Eslin-
Age Group Intervals
ger et al., 1985; Giambra et al., 1995; Kawas
Given the association between age and iBVRT et al., 2003; Klonoff & Kennedy, 1965; Jacobs
performance, information regarding theiage of et al., 1997; Larrabee et al., 1986; Manly et al.,
the normative sample is critical and nonnative 1999, 2002; Palmer et al., 1994; Prakash &
data should be presented by age inte~s. Bhogle, 1992; Reinprecht et al., 2003; Resnick
et al., 1995; Robertson-Tchabo & Arenberg,
Reporting of Educational Levels 1989; Steenhuis & Ostbye, 1995; Touradji et al.,
Given the relationship between educational 2001; Youngjohn et al., 1993).
level and BVRT scores, it is preferable that Educational level was also indicated in all but
data be stratified by educational level. four studies (Alder et al., 1990; Arenberg, 1978;
Benton-Sivan, 1992; Reinprecht et al., 2003), 1986; Manly et al., 1999, 2002; Mathiesen et al.,
and Youngjohn et al. (1993) and Coman et al. 1999; Randall et al., 1988; Resnick et al., 1995;
(1999) stratify by age and educational level, Robertson-Tchabo & Arenberg, 1989; Young-
while Dealberto et al. (1996) and Robertson- john et al., 1993).
Tchabo and Arenberg (1989) present data by Geographic recruitment areas were speci-
age, gender, and educational level separately. fied in all but one publication (Youngjohn et al.,
Manly and colleagues (1999) stratify data by 1993). Six publications present data from the
literacy in a population with <3 years of formal Baltimore Longitudinal Study on Aging (Alder
education. Information on IQ level is reported et al., 1990; Arenberg, 1978; Giambra et al.,
in one study (Larrabee et al., 1986), with Vo- 1995; Kawas et al., 2003; Resnick et al., 1995;
cabulary raw score presented in two studies Robertson-Tchabo & Arenberg, 1989), three
(Alderetal., 1990; Mathiesenetal., 1999);Amir studies report data from the Washington
(2001) and Randall et al. (1988; also re- Heights Inwood Columbia Aging Project based
produced in Benton-Sivan, 1992) present in northern Manhattan, New York (Manly
BVRT data in IQ groupings. et al., 1999, 2002; Touradji et al., 2001), and
Information on gender composition was in additional data sets were collected in the
all but four reports (Benton, 1962; Benton- United States in Iowa (Benton, 1962; Benton-
Sivan, 1992; Robinson-Whelen, 1992; Tour- Sivan, 1992; Coman et al., 1999; Eslinger et al.,
adji et al., 2001); 10 data sets included only 1985), Florida (Larrabee et al., 1986), Mis-
male (Alder et al., 1990; Arenberg, 1978; sissippi (Randall et al., 1988), Missouri
Carmelli et al., 1999; Farahat et al., 2003; (Robinson-Whelen, 1992), New York (Jacobs
Klonoff & Kennedy, 1965; Lee & Lee, 1993; et al., 1997), and Massachusetts, Indiana, and
Mathiesen et al., 1999; Palmer et al., 1994; California (Carmelli et al., 1999). Data sets
Reinprecht et al., 2003) or nearly all-male were also gathered in Canada (Klonoff &
(Escalona et al., 1995) populations, and four Kennedy, 1965; Steenhuis & Ostbye, 1995),
data sets were composed primarily of females France (Dealberto et al., 1996), Norway (Ma-
(Jacobs et al., 1997; Larrabee et al., 1986; thiesen et al., 1999), Sweden (Reinprecht et al.,
Manly et al., 1999, 2002). Robertson-Tchabo 2003), Italy (Ruggieri et al., 2003), the Neth-
and Arenberg (1989), Dealberto et al. (1996), erlands (Witjes-Ane et al., 2003), the United
and Giambra et al. (1995) reported data Arab Emirates (Amir, 2001), Korea (Lee &
separately for males and females. Lee, 1993), Venezuela (Escalona et al., 1995),
Data regarding ethnicity were presented in India (Prakash & Bhogle, 1992), and Egypt
22 data sets (Amir, 2001; Arenberg, 1978; Car- (Farahat et al., 2003).
melli et al., 1999; Coman et al., 1999; Dealberto Administration A, form C, was the most com-
et al., 1996; Escalona et al., 1995; Farahat et al., mon version reported, appearing in 16 publica-
2003; Giambra et al., 1995; Jacobs et al., 1997; tions (Alder et al., 1990; Arenberg, 1978;
Larrabee et al., 1986; Lee & Lee, 1993; Manly Benton-Sivan, 1992; Carmelli et al., 1999; Co-
et al., 1999, 2002; Mathiesen et al., 1999; Pal- man et al., 1999; Eslinger et al., 1985; Giambra
mer et al., 1994; Prakash & Bhogle, 1992; Re- et al., 1995; Kawas et al., 2003; Klonoff &
inprecht et al., 2003; Resnick et al., 1995; Kennedy, 1965; Larrabee et al., 1986; Mathie-
Robinson-Whelen, 1992; Ruggieri et al., 2003; sen et al., 1999; Palmer et al., 1994; Prakash &
Touradji et al., 2001; Witjes-Ane et al., 2003). Bhogle, 1992; Resnick et al., 1995; Robertson-
Occupation or socioeconomic status was de- Tchabo & Arenberg, 1989; Robinson-Whelen,
scribed in five reports (Arenberg, 1978; Giam- 1992), and Randall et al. (1988) included not
bra et al., 1995; Klonoff & Kennedy, 1965; Lee only this administration but also administration
& Lee, 1993; Palmer et al., 1994). Exclusion D, form D, and administration C, form C,
criteria were judged to be adequate in 16 pub- while Robinson-Whelen (1992) also included
lications (Alder et al., 1990; Arenberg, 1978; administration C, form D. Four studies col-
Escalona et al., 1995; Eslinger et al., 1985; lected data for administration A but did not
Farahat et al., 2003; Giambra et al., 1995; Jacobs specify the form (although it is assumed to be
et al., 1997; Kawas et al., 2003; Larrabee et al., C; Carmelli et al., 1999; Coman et al., 1999;
Reinprecht et al., 2003; Youngjohn et al., clinical studies, presented in ascending chron-
1993). Amir (2001) employed administration ological order. The text of study descriptions
A, form D, followed by formE 2 weeks later, contains references to the corresponding ta-
while Prakash and Bhogle (1992) administered bles identified by number in Appendix 20.
forms C, D, and E under administration A Table A20.1, the locator table, summarizes in-
conditions. Jacobs et al. (1997) used adminis- formation provided in the studies described in
tration C, form C. Benton (1962) present data this chapter. 1
for forms C and E under administrations B and
C, and Benton-Sivan ( 1992) provide data for
four administrations (A, B, C, and D). Seven
studies report data for the 15-item recognition SUMMARIES OF THE STUDIES
trial only (Escalona et al., 1995; Jacobs et al.,
Test Manual
1997; Lee & Lee, 1993; Manly et al., 1999,
2002; Steenhuis & Ostbye, 1995; Touradji et [BVRT.1] Benton-Sivan, 1992
al., 2001), although the Jacob et al. (1997) data
The manual for the BVRT, 5th edition, con-
appear to be for form D rather than C; the
tains normative data for administration
Dealberto et al. (1996) data are assumed to be
(number correct and number of errors) on
for the recognition trial since the mean correct
600 participants from the third edition of the
exceeds 10. Three studies (Farahat et al., 2003;
manual (Benton, 1963), as well as the Aren-
Ruggieri et al., 2003; Witjes-Ane et al., 2003)
berg (1978) data on 769 men and the Randall
did not report any information regarding test
et al. (1988) data on 120 participants. Exclu-
administration version or format.
sion criteria for the Benton (1963) sample
Total mean number correct was reported in
included no history of psychosis, no cerebral
22 data sets (Amir, 2001; Coman et al., 1999;
injury or disease except for mental retarda-
Dealberto et al., 1996; Escalona et al., 1995;
tion, and no serious physical depletion as a
Eslinger et al., 1985; Farahat et al., 2003; Jacobs
consequence of somatic disease. A majority
et al., 1997; Klonoff & Kennedy, 1965; Lee &
of participants were inpatients and outpatients
Lee, 1993; Manly et al., 1999, 2002; Mathiesen
of hospitals in Iowa City and Des Moines,
et al., 1999; Palmer et al., 1994; Prakash &
Iowa.
Bhogle, 1992; Randall et al., 1988; Reinprecht
For administration B, data are reported for
e~ ~·· 2003; Robinson-Whelen, 1992; Rug-
103 medical patients aged 16-60 with no his-
gien et al., 2003; Steenhuis & Ostbye, 1995;
tory or evidence of brain disease. Performance
Touradji et al., 2001; Witjes-Ane et al., 2003;
was approximately 1 point lower for number
Youngjohn et al., 1993) and number of errors in
correct than for administration A. The rec-
14 data sets (Alder et al., 1990; Amir, 2001;
ommendation is to use the manual norms for
Arenberg, 1978; Benton, 1962; Eslinger et al.,
administration A, subtracting 1 point.
1985; Giambra et al., 1995; Kawas et al., 2003;
Administration C norms were obtained on
Klonoff & Kennedy, 1965; Mathiesen et al.,
200 medical patients with no history or evi-
1999; Randall et al., 1988; Resnick et al.,
dence of cerebral disease (Benton, 1962). Al-
1995; Robertson-Tchabo & Arenberg, 1989;
most half the group obtained perfect scores,
Robinson-Whelen, 1992; Youngjohn et al.,
and 88% made two or fewer errors. The manual
1993). Means were reported in all but one study
also presents norms for an eight-item abbre-
(Comanetal., 1999), andSDswereindicatedin
viated version, based on performance of
all but four studies (Arenberg, 1978; Benton,
100 controls (Benton, 1962).
1962; Carmelli et al., 1999; Coman et al., 1999).
. ~or administration D, performance in par-
Means and SDs for individual error scores are
ticipants younger than 60 is reported to be
provided in two publications (Resnick et al.,
comparable to that for administration A.
1995; Robinson-Whelen, 1992).
Values for administration A for expected
Below, information about the test manual
number correct are presented for three age
will be reported first, followed by summaries
of normative publications and control data from 'Norms for children are available in Baron (2004).
groups (15-49, 50-59, and 60--69) and six IQ 3. Test version/administration format speci-
levels (2:110, 95-109, 80-94, 70-79, 60-69, fied.
and ~59). Values for administration A for ex-
pected error scores are provided for four age Considerations regarding use of the study
groups (15-44, 45-59, 60-64, and 65-69) and 1. Questionable adequacy of exclusion
eight IQ levels (2:110, 105-109,95-104,90-94, criteria (while patients with seizures,
80-89, 70-79, 60-69, and ~59). Expected head injury, and psychiatric or mental
values for administration C number errors are deficiency-related hospitalizations were
reported for the 10-design and eight-design excluded, participants were inpatients on
versions. medical or neurologic wards).
Data are presented as expected number 2. Low educational level.
of correct responses and errors and apply to 3. No information regarding gender or IQ.
all three forms (C, D, E). Data are not re- 4. No SDs reported.
produced in this book. 5. Data for the differing forms were ap-
parently collapsed.
Manual strengths (Benton data) 6. Data not stratified by age.
1. Large overall sample size.
[BVRT.3] Klonoff and Kennedy, 1965
2. Data stratified by age and IQ levels.
(Table A20.3)
3. Information regarding geographic area.
4. Data reported for administrations A, B, BVRT data were obtained on 172 Canadian
C, and D and the eight-item abbreviated veterans aged 80-92 "managing well on their
version. own in the community;" 30% of War Veteran
Allowance recipients 2:80 and residing in the
Considerations regarding use of the manual Vancouver area were selected; 88 could not
1. Inadequate exclusion criteria for the participate due to major loss of visual or au-
Benton sample (mentally retarded in- ditory acuity or motor dysfunction. Testing
cluded, most were hospital inpatients or was completed during 1963-1964. Informa-
outpatients). tion on educational level was available
2. No information regarding gender, edu- for 155 participants; mean education was
cation, or recruitment strategies. 7.04 years. Seventy-three percent were born
in the British Isles, 19% were hom in Canada,
and 8% were hom elsewhere in Europe. The
[BVRT.2] Benton, 1962 (Table A20.2)
majority were unskilled workers (47%), while
BVRT data were obtained on 100 patients on 12% were semiskilled, 14% were skilled, 10%
medical and neurological wards in Iowa City were clerical, 10% were in a service industry,
hospitals who showed no evidence or history 5% were semiprofessional, and 2% were pro-
of cerebral disease or injury. Additional ex- fessionals. Forty-five percent carried cardio-
clusion criteria were seizures, head trauma vascular diagnoses, 22% had pulmonary
with loss of consciousness, or hospitalization disease, and 15% had psychiatric diagnoses;
for psychiatric disorder or mental deficiency. 25% rated their health as "very good," 51%
Mean age was 41 years (range 16--60), and judged their health to be "good," and 24%
mean educational level was 10 years. designated their health as "fair." The sample
Participants received form C or E, adminis- was divided into six age groupings: 80
tration C, and form C or E, administration B. (n = 35), 81 (n = 34), 82 (n = 23), 83 (n = 27),
Means for number of errors for the copy and 84-85 (n = 26), and 86+ (n =27).
memory trials are reported. Administration A, form C, was used and
scored according to the Benton method. Means
Study strengths and SDs for number correct and number errors
1. Large sample size. are reported; 34% committed omission errors,
2. Information regarding age, education, 33% produced distortions, 12% had rota-
and geographic area. tions, 11% exhibited perseverations, 8% of the
errors were misplacements, and 2% were size geographic area, and recruitment strat-
errors. egies.
4. Test administration format and version
Study strengths specified.
1. Large overall sample size. 5. Means reported for number of errors.
2. Data stratified by age.
3. Information regarding gender, educa- Considerations regarding use of the study
tional level, recruitment strategy, occu- 1. Exclusion criteria not specified, although
pational status, and geographic area. reported in Giambra et al. (1995).
4. Test administration, version, and scoring 2. All-male sample.
specified. 3. SDs and IQ level not reported.
5. Means and SDs reported.
[BVRT.S] Eslinger, Damasio, Benton, and
Considerations regarding use of the study Van Allen, 1985 (Table A20.5)
1. Questionably adequate exclusion criteria BVRT scores were collected on 53 normal
(those with unspecified psychiatric di- volunteers (25 men, 28 women) aged 60-88,
agnoses included). recruited through local senior-citizen and com-
2. Language in which testing conducted munity organizations in the Iowa City area.
not specified but assumed to be English. Mean age was 73.1, and mean educational level
3. All-male sample. was 12.0. All were independent, community-
4. No information regarding IQ. dwelling individuals who were screened for
5. Low educational level. neurologic disorder (including head injwy and
alcoholism), psychiatric illness requiring hospi-
[BVRT.4] Arenberg, 1978 (Table A20.4) talization, and any disabling medical or physical
BVRT data were obtained on 857 men aged 18- condition. Participants considered themselves to
102, as part of the Baltimore Longitudinal Study be in generally good physical and mental health.
of Aging, who were tested between 1960 and Form C, administration A. was used. Total
1973. All were volunteers who had agreed to correct (10 possible) and total errors (26 pos-
come to Baltimore city hospitals for physiologi- sible) were calculated. Means and SDs are
cal, biochemical, and behavioral testing. The reported.
sample was primarily Caucasian, well-educated,
and of high socioeconomic status, residing in Study strengths
the Baltimore-Washington, DC area. 1. Minimally adequate sample size.
Participants were divided into seven age 2. Information on age, gender, education,
groups (<30, 30s, 40s, 50s, 60s, 70s, ?80) and geographic area, and recruitment strat-
three subgroups based on date of test ad- egies.
ministration (1960-1964.9, 1965.0-1968.5, and 3. Adequate exclusion criteria.
1968.6-1973.5). 4. Test version/administration format speci-
Participants were given form C, adminis- fied.
tration A. Errors were scored according to the 5. Means and SDs for number correct and
1963 manual by two independent psycholo- errors reported.
gists, and infrequent disagreements were re-
solved by discussion or a third psychologist. Consideration regarding use of the study
Means for total errors are reported. 1. Data not stratified by age, although age
grouping is fairly narrow.
Study strengths 2. No data on IQ level.
1. Large overall sample size, although some
individual cells are small. [BVRT.6] Larrabee, Levin, and High, 1986
2. Data stratified by age. (Table A20.6)
3. Some information regarding gender, age, BVRT scores were gathered on 88 reportedly
education, socioeconomic status, ethnicity, healthy participants aged 60-90 recruited
from retirement apartments and organizations the mentally retarded. Participants were ad-
in Galveston County, Florida. Participants had ministered the Satz-Mogel version of the
at least 20/40 vision on screening, and hearing WAIS-R for placement into IQ ranges. A total
was adequate on audiometric assessment. In- of 120 participants were included (69 females,
dividuals with a history of psychiatric disorder, 51 males), with 20 in each of six IQ ranges:
stroke, head injury, or other neurologic dis- mentally retarded (60-69), borderline (70-79),
ease were excluded. No subject had a clinical low average (80-89), average (90-109), high
level of depression on the Zung self-report average (110-119), and superior (120+). Par-
inventory. Ten participants were identified as ticipants were not compensated. Exclusion
having senescent forgetfulness by low memory criteria were history or evidence of cerebro-
scores (on at least four scores out of 13), 1 or vascular illness, traumatic head injury, epi-
more SD below age-residualized means, and lepsy, alcoholism, or psychiatric illness. Mean
appearing in a memory disorder cluster on age across groups ranged 23.87-26.20.
cluster analysis. Administrations A (form C), D (form D),
Mean age of the remaining participants and C (form C) were conducted, although
(n = 78) was 72.9 (6.9) years, and mean years the order of presentation of the cards was
of education was 12.2 (3.3); 61 were female changed (i.e., 2, 6, and 10 and then 1, 3, 4, 5,
and 17 were male; 73 were white and five 7, 8, and 9). Means and SDs for number
were African American. Mean VIQ was 112.0 correct and errors are reported for the six IQ
(12.8), and mean PIQ was 114.4 (11.5). groups.
Form C, administration A, of the BVRT was
given. Means and SDs for number of errors Study strengths
reported. 1. Large overall sample size, although in-
dividual cell sizes are small.
Study strengths 2. Data stratified by IQ level.
1. Large sample size. 3. Adequate exclusion criteria.
2. Information regarding age, education, 4. Information regarding geographic area,
ethnicity, IQ, gender, geographic area, gender, and age.
vision and hearing acuity, and recruit- 5. Administration procedures and versions
ment strategy. specified.
3. Adequate exclusion criteria. 6. Means and SDs reported for number
4. Test administration format/version speci- correct and errors for three administra-
fied. tion versions.
5. Means and SDs reported for total er-
rors. Considerations regarding use of the study
1. No information regarding educational
Considerations regarding use of the study level, although this is obviated by the
1. Data not stratified by age. data on IQ.
2. Exclusion criteria likely "too stringent" 2. No data on gender for individual cells.
(i.e., 10 participants were excluded based 3. Data not stratified by age, although the
on memory test performance). As a re- age range appears adequately narrow.
sult, sample may not be characteristic of 4. Altered administration format (order of
healthy individuals in this age range. stimuli changed).
[BVRT.7] Randall, Dickson, and Plasay, 1988 [8VRT.8] Robertson-Tchabo and Arenberg,
(Table A20.7) 1989 (Table A20.8)
BVRT data were obtained on volunteers from BVRT data on 1,643 participants from the
psychology undergraduate classes and the Baltimore Longitudinal Study of Aging are
Honors College at the University of Southern presented. Data are grouped separately by
Mississippi, various church and community gender, college degree or no college degree,
organizations, and a residential care facility for and seven age decades (20s, 30s, 40s, 50s, 60s,
70s, 80s). Exclusion criteria are described 6. Adequate exclusion criteria.

below in Giambra et al. (1995).
Data were collected using form C, admin- Considerations regarding use of the study
istration A. Means and SDs (as well as fre- 1. No data regarding educational level, al-
quencies not reproduced in this book) are though this is somewhat obviated by data
reported for errors. on the Vocabulary subtest.
2. All-male sample.
Study strengths
[BVRT.10] Prakash and Bhogle, 1992
1. Large overall sample size, although some
(Table A20.1 0)
individual cells fall below 50.
2. Data stratified by age, gender, and edu- BVRT scores were collected on 660 partici-
cational level. pants in India. Exclusion criteria were "evident
3. Information regarding recruitment physical or psychological disorders," and only
strategies and geographic area. participants with a minimum level of higher
4. Test administration and scoring proce- secondary education were included. Partici-
dures specified. pants were aged 15-65; gender distribution was
5. Means and SDs reported for total approximately equal (331 male, 329 female).
errors. Participants were divided into 10 age
6. Adequate exclusion criteria. groupings: 15-19 (n = 90), 20-24 (n = 86), 25-
29 (n = 84), 30-34 (n =56), .35-39 (n =53),
Consideration regarding use of the study 40-44 (n = 62), 45-49 (n = 73), 50-54 (n =55),
1. No information regarding IQ level. 55-59 (n = 62), and 60-64 (n = 39) years.
Forms C, D, and E were used with ad-
ministration A. Means and SDs for number
[BVRT.9] Alder, Adam, and Arenberg, 1990
correct for each age grouping are provided.
(Table A20.9)
BVRT data on 277 men are reported from the Study strengths
Baltimore Longitudinal Study of Aging. Data 1. Large sample size, and all but one sub-
are presented for five age groupings: 25-34 group had >50.
(n = 27), 35-44 (n = 74), 45-54 (n = 101), 2. Data stratified by age.
55-64 (n = 42), and 65+ (n = 33). Total raw 3. Information regarding gender, educa-
scores for the Vocabulary subtest of the WAJS tional level, and geographic area.
are provided for each subgroup. Exclusion 4. Test administration format and forms
criteria are described below in Giambra et al. specified (although it is unclear if all ver-
(1995). sions were given to all participants and, if
Data were collected using form C, admin- so, whether administration was in a coun-
istration A. Each protocol was scored inde- terbalanced order).
pendently by two psychologists. Means and 5. Means and SDs for number correct
SDs for number of errors are reported. reported.

1. Large overall sample size, although some 1. No information regarding recruitment
individual cells fall below 50. strategy or IQ level.
2. Data stratified by age. 2. Three test forms were administered, but
3. Information regarding gender, recruitment only one set of means and SDs is reported;
strategies, geographic area, and Vocabu- it is unclear whether these represent
lary raw score. means for the three forms or just one.
4. Test administration and scoring proce- 3. Adequacy of exclusion criteria question-
dures specified. able (sample apparently consists of par-
5. Means and SDs reported for total errors. ticipants who appear normal).
BENTON VISUAl RETENTION TEST 407
4. Data provided on East Indians, which four-choice array. Means and SDs for number
may limit generalizability for clinical incorrect are provided.
terpretation in the United States.
Study strengths
[BVRT.11] Robinson-Whelen, 1992 1. Large sample size.
(Table A20.11) 2. Information regarding age, education,
BVRT data are reported on 122 Caucasian gender, occupational status, and alcohol
older normal individuals in good physical and cigarette use.
health recruited in Missouri as part of a study 3. Test version and administration format
on the BVRT in normal aging and dementia. reported.
Mean age was 72.23 (9.0) years, and mean 4. Means and SDs provided.
education was 13.61 (3.4) years.
Form C, administration A, followed by Considerations regarding use of the study
form D, Administration C (copy), were used. 1. Inadequate exclusion criteria.
Means and SDs for number correct, total er- 2. Data not stratified by age.
rors, and eight error types are reported. 3. Information on test translation not pro-
vided.
Study strengths 4. All-male sample.
1. Large sample size. 5. The stimuli version not reported.
2. Information regarding age, education, 6. Data collected on Korean workers, which
geographic area, and ethnicity. may limit generalizability for clinical in-
3. Form and administration version reported. terpretation in the United States.
4. Means and SDs for number correct as
[BVRT.13] Youngjohn, Larrabee, and Crook,
well as number of errors and various
1993 (Table A20.13)
error types provided.
BVRT scores were collected on 1,128 (464 male,
664 female) normal volunteers aged 17-84,
Considerations regarding use of the study recruited to participate in testing through news
1. Data not stratified by age. media. Participants were not compensated.
2. No information regarding gender, re- Mean age was 58.45 (11.43) years, and mean
cruitment strategies, or IQ. education was 16.01 (2.2), with a range of 12-25
3. No information regarding exclusion cri- years. Participants with a history of physical,
teria; participants are only reported as psychiatric, or neurological conditions that
being in good physical health. could affect memory (e.g., depression, head
trauma, or stroke) were excluded.
[BVRT.12] Lee and Lee, 1993 (Table A20.12) Administration A was used. Data are pre-
BVRT data were obtained on 81 male controls sented in five age groupings (18--39, 40--49,50-
as part of a study on the effects of organic 59, 60-69, and 70+) and three educational levels
solvents on cognition in Korean workers re- (12-14, 15-17, and 18+). Means and SDs for
siding in the Seoul area. Participants averaged number of designs correctly reproduced and
34.7 (8.16) years of age, with 12.9 (2.5) years of total number of errors are reported.
education. The group was primarily composed
of manual workers, guards, clerks, and tech- Study strengths
nicians. They consumed an average of 10.4 1. Large overall sample size, although some
(10.2) liters of alcohol per year and smoked an cells are relatively small.
average of 11.6 (33.1) cigarettes per day. They 2. Data stratified by age and educational
had not been exposed to organic solvents. level.
Participants were administered the NCTB, 3. Information regarding gender and re-
which includes a 10-second administra- cruitment strategy.
tion followed by selection responses from a 4. Adequate exclusion criteria.
5. Test administration format reported, 2. All male sample.

although not stimuli version. 3. Well-educated sample.
6. Means and SDs provided. 4. No exclusion criteria and no information
regarding IQ.
1. Well-educated sample, and no data for [8VRT.15] Escalona, Yanes, Feo, and Maizlish,
participants with < 12 years of education. 1995 (Table A20.15)
2. No information on IQ level. BVRT scores, collected as part of NCTB ad-
ministration, were obtained on 67 (56 male,
Other comments 11 female) controls as part of a study of the
1. The authors provide regression effects of organic solvents on cognition in Ven-
equations for predicted BVRT scores: ezuela. Mean age was 30 years, and mean ed-
a. Predicted BVRT number correct ucational level was 8 years. Participants were
(±1.57) = 7.87- 0.045(age) + 0.098 aged 16-45; could read at the junior high
(years of education) school level; had no previous history of mental
b. Predicted BVRT number of errors illness, drug abuse, head trauma, epilepsy, or
(±2.88) = 1.73 + 0.088 (age)- 0.126 neurotoxic exposure; had not consumed alco-
(years of education) hol within 24 hours of testing; and had ade-
(8VRT.14] Palmer, Wolkenstein, LaRue, Swan, quate sleep the night before testing.
and Smalley, 1994 (Table A20.14) An NCTB recognition version of the BVRT
was administered, involving exposure to the
BVRT data were obtained on 1,149 primarily
items for 10 seconds and selection of a correct
Caucasian (98%), community-dwelling older
match for a four-choice array. Means and SDs
males who resided in California and participated
for number correct are reported.
in the Western Collaborative Group Study from
1986 to 1989. Mean age was 71.4 (4.69) years.
Study strengths
The sample is described as well educated (ap-
1. Information on geographic area, age, ed-
proximately 62% attended college) and engaged
ucation, and gender.
in high-level occupations (approximately 40%
had managerial positions, nearly 50% had tech-
3. Means and SDs for number correct re-
nical jobs, and < 10% were laborers or clerical
ported.
staff). Approximately 30% were hypertensive,
4. Test format specified (NCTB recogni-
10% were diabetic, and 25% had coronary heart
tion trial with four-choice array).
disease. The rating for the group as a whole on a
depression screening measure was very low, and Considerations regarding use of the study
average MMSE score was 28.1. 1. Low educational level.
Participants were given form C, adminis- 2. Data not stratified by age.
tration A. Means and SDs for number correct 3. The stimuli version and IQ level not
were reported. reported.
4. Method of translation of test instructions
Study strengths
1. Very large sample size. not reported.
2. Information regarding age, ethnicity, ed- 5. Data obtained on Spanish speakers in
ucation, occupation, gender, recruitment Venezuela, which may limit generaliz-
strategy, and geographic area. ability for clinical interpretation in the
United States.
3. Test version and administration format
reported.
[BVRT.16] Giambra, Arenberg, Zonderman,
4. Means and SDs provided.
Kawas, and Costa, 1995 (Table A20.16)
Considerations regarding use of the study BVRT data are reported from the Baltimore
1. Data not stratified by age but appear to Longitudinal Study of Aging, collected on
reflect a fairly narrow age range. 1,163 men from 1960 and on 558 women from
1978, aged 28-87, until 1992. Exclusion cri- (n = 268), 30-39 (n = 353), 40--49 (n = 270),
teria were past or present psychosis, major 50-59 (n=306), 60-69 (n=334), 70-79
depression, organic brain syndrome, demen- (n = 340), and 80+ (n =-129) years.
tia, Parkinson's disease, stroke, or epilepsy. Administration A, form C, was employed.
Participants were recruited through invitation Errors were scored according to the 1974
or in response to learning of the project manual. Errors were classified into seven major
through the media. Participants were mostly categories: omissions, distortions, persevera-
white and highly educated: < 12 (1. 7% of men, tions, rotations, misplacement, size, and addi-
2.3% of women), 12 (7.1% of men, 13.7% of tion errors. Errors were scored by two
women), 13-16 (45.5% of men, 44.9% of experienced, independent examiners, and dis-
women), > 16 (53. 7% of men, 39.1% of women) agreements were resolved through consensus.
years of education. A large proportion were Errors across the 10 cards were summed for
currently in or retired from administrative or a total error score. Means and SDs for the seven
professional positions. error types are reported for men and women
Form C, administration A, was employed. separately and together in each age grouping.
Means and SDs for number of errors are
reported for men and women separately for Study strengths
10 age groupings. Each protocol was scored by l. Very large sample size and large indi-
two psychologists according to the 1974 man- vidual cell sizes.
ual, and disagreements were addressed through 2. Information regarding education, recruit-
consensus or a third, independent rater. ment procedures, ethnicity, and geo-
graphic area.
1. Very large overall sample size, although 4. Data stratified by age and gender.
the size of individual age and gender 5. Administration format and version
cells is not reported. specified.
2. Stratification of data by narrow age 6. Means and SDs for seven error types
groupings and gender. reported.
4. Information on age, gender, geographic
area, recruitment strategy, education, Considerations regarding use of the study
ethnicity, and occupation. 1. Well-educated sample.
5. Administration and scoring specified. 2. No information regarding IQ level.
6. Means and SDs for number of errors
reported. [8VRT.18] Steenhuis and Ostbye, 1995
(Table A20.18)
Considerations regarding use of the study BVRT recognition trial data were collected in
1. High educational level. Canada on 591 participants over age 65 in an
2. No information on IQ level. epidemiological study based on a representa-
tive sample of elderly Canadians. All partici-
[8VRT.17] Resnick, Trotman, Kawas, and pants received a final consensus diagnosis of
Zonderman, 1995 (Table A20.17) no cognitive impairment based on a dementia
BVRT data were analyzed from 2,000 (1,365 screening instrument, although a subset (ap-
men, 635 women) mostly Caucasian participants proximately one-fourth) also underwent sub-
of the Baltimore Longitudinal Study of Aging, sequent comprehensive testing. Mean age was
aged 20-102 years; 82% of the men and 69% of 78.5 (6.7) years, and mean education was
the women had completed college, and 61% of 9.8 (4.0); 61% were female, and 21% spoke
the men and 48% of the women had some French; 18% resided in an institution.
graduate school education. For exclusion crite- The multiple-choice version of the BVRT
ria, see Giambra et al. (1995). Participants were was administered. Means and SDs for number
partitioned into seven age groups: 20-29 correct are reported.
Study strengths men and 4.4% of women; and in the age group
1. Very large sample size. 65-69, 3.6% of men and 6.8% of women had
2. Information regarding age, education, scores in this range.
gender, geographic area, recruitment BVRT means and SDs for number correct
strategy, and language. are reported separately for two age groups
3. Exclusion of patients with dementia and (60-64, 65-70), gender, and three educational
nondementia cognitive loss. levels (<6, 6-12, and >12 years).
4. BVRT version specified.
5. Means and SDs for number correct Study strengths
reported. 1. Very large sample size.
Considerations regarding use of the study gender, geographic area, smoking his-
1. Data not stratified by age but appear to tory, alcohol and medication use, and
cover a relatively narrow age range. depressive symptoms.
2. Minimal exclusion criteria. 3. Means and SDs for number correct re-
3. Low educational level. ported.
4. Both English and French speakers 4. Data stratified by two age groups, gen-
included. der, and three educational levels.
5. No information regarding IQ level.
[BVRT.19] Dealberto, Pajot, Courbon, and 1. No exclusion criteria and participants
Alperovitch, 1996 (Table A20.19} included with substantial depressive
BVRT scores were collected on 1,389 (574 symptoms (14%) and MMSE scores <24
male, 815 female) French participants aged (3o/!Hl%).
60-70 as part of a study of the effects of sleep- 2. Test version and format not specified
related breathing disorders on cognition. Mean (mean score is number correct but ex-
age was 65 (3); 46% of men and 55% of women ceeds 10; therefore, this test appears to
had <6, 33% of men and 36% of women had 6- be an altered format).
12, and 21% of men and 10% of women had 3. Recruitment strategies not reported.
>12 years of schooling. The majority of women 4. Data on French speakers. Method of
had never smoked (82.8%), whereas only translation oftest instructions not specified.
23.2% of men were never-smokers; 62.2% of 5. No information regarding IQ level.
the men were former smokers, as were 12.8%
of women, and 14.6% of men and 4.4% of [BVRT.20] Jacobs, Sano, Albert, Schofield,
women were current smokers. Most partici- Dooneief, and Stern, 1997 (Table A20.20}
pants ingested 1-3 medications per day (44.6% BVRT scores were obtained on ll8 older En-
of men, 44.9% of women), while 28.6% of men glish speakers and l18 older Spanish speakers
and 18.7% of women took no medications; as part of a community-based epidemiological
26.8% of men and 36.4% of women took four study in northern Manhattan, New York City.
or more medications daily. The majority of English speakers averaged 75.07 (5.90) years of
participants drank 1-40 mL of alcohol per day age and 8.85 (3.78) years of education, while
(58.9% of men, 59.3% women), 12.9% of men Spanish speakers averaged 74.91 (5.71) years
and 38.0% of women drank no alcohol, and of age and 8.41 (3.98) years of education; 75%
28.2% of men and 2.6% of women imbibed of the English-speaking sample and 72% of the
>40 mL per day. No significant medical con- Spanish-speaking sample were female. Partici-
ditions were present in 42.6% of participants, pants were designated as English speakers or
one condition was present in 34.5%, and two or Spanish speakers based on which language they
more in 22.9% of the sample; 14.2% of the elected to use on the examination. Among
sample obtained scores suggesting a high level Hispanic participants, country of origin was
of depression on a self-report inventory. In the primarily the Dominican Republic, Cuba, or
age group 60-64, MMSE was <24 in 3.1% of Puerto Rico; most had resided in the United
States for > 15 years; 43% spoke English "not at 5. Data obtained on Spanish speakers. It is
all," 33% spoke English "not well,'' 11% spoke unknown how translation of the test in-
English "well," and 13% spoke it "very well;" structions affected performance.
97% indicated that Spanish was the primary 6. No information on IQ level.
language spoken in the home. Among the En-
glish speakers, 73% had been born and raised in [BVRT.21] Carmelli, Swan, Reed, Schellenberg,
the United States. Immigrants were primarily and Christian, 1999 (Table A20.21)
from European countries, and all had immi- BVRT data were obtained in 1985-1986 on 589
grated prior to 1980; 98% spoke English "well" white male World War II veterans 59-69 years
or "very well," and 89% spoke primarily English old as part of a study of the impact of smoking,
in the home. Participants with signs of dementia drinking, and Apo Eon cognitive function; 514
or cognitive impairment, based on neurological twin pairs were selected for study from a larger
mental status testing, were excluded, as were registry of 16,000 based on geographic location
participants with major depression or history of (within 200 miles of Framingham, MA; In-
Parkinson's disease, stroke, head injury with loss dianapolis, IN; and Davis, San Francisco, and
of consciousness, or alcohol abuse or who were Los Angeles, CA); 248 monozygotic twins, 242
less than age 65. fraternal twins, and 99 singletons were studied;
To measure visuoperceptual skills, form C 341 participants averaged ~ 12 years educa-
was administered, in which each target stim- tion, and 254 participants had >12 years of
ulus was presented along with a four-choice schooling. Patients with alcohol use, hyper-
array. For measurement of visual memory, a tension, diabetes, coronary heart disease,
multiple-choice version (form D) was admin- stroke, transient ischemic attack, myocardial
istered. Participants were shown each design infarction, congestive heart failure, and angina
for 10 seconds and, after removal, asked to were included. Standard testing administration
choose the design from a four-choice array. and scoring procedures were followed.
Spanish test instructions were translated by a Means for number correct are reported for
committee of native Spanish speakers from former smokers (quit 2::10 years, n = 222), for-
Cuba, Puerto Rico, Spain, and the Dominican mer smokers (quit <10 years, n = 72), current
Republic and then back-translated. smokers (n = 102), and never-smokers (n = 199)
Means and SDs for number correct are as well as for nondrinkers (n = 158), light
reported. drinkers (~1 drink per day, n =204), moderate
drinkers (<3 and >1 drinks per day, n = 150),
Study strengths and heavy drinkers (2::3 drinks per day, n = 83).
1. Large overall sample size. Means for BVRT correct reported.
gender, recruitment strategy, and geo- Study strengths
graphic recruitment area. 1. Large samples size, and most individual
3. Data stratified by language. cells have > 100.
4. Adequate exclusion criteria. 2. Narrow age range.
5. Test administration specified, and meth- 3. Information on education, gender, geo-
od of translating instructions reported. graphic area, ethnicity, and recruitment
6. Means and SDs for total correct for strategies.
forms C and D are reported. 4. Means reported for number correct.
5. Data stratified by cigarette and alcohol use.
1. Nonstandard administration (multiple- Considerations regarding use of the study
choice response). 1. No exclusion criteria.
2. Mostly female sample. 2. All-male sample.
3. Low educational level. 3. Test version not specified, and no SDs
4. Data not stratified by age, but the age range reported.
appears to be adequately narrow. 4. No information regarding IQ level.
[BVRT.22] Coman, Moses, Kraemer, Friedman, abuse or neurological or functional signs of

Benton, and Yesavage, 1999 (Table A20.22) delirium or dementia based on physician exam
Archival BVRT data on a total of 156 (31 male, (independent of neuropsychological scores).
125 female) normal participants obtained at Literacy was determined by self-report to the
the University of Iowa were analyzed. Partic- query "Did you ever learn to read and write?''
ipants were aged 61-97, with a mean of 77.7 Both groups were 74% female, and 72% of
(7.89) years. The sample was primarily Cau- the literate group were Spanish speakers com-
casian, with a mean of 12.67 (3.46) years of pared to 86% of the illiterate group. The lit-
education (range 4-20). erate group was 81% Hispanic, 9% African
All participants were given administration American, and 10% non-Hispanic white, while
A according to standard procedures. Based the illiterate group was 91% Hispanic and 9%
on the results of regression analyses, expected African American.
BVRT number correct scores are provided for Means and SDs for multiple-choice recog-
nine ages (55, 60, 65, 70, 75, 80, 85, 90, 95) by nition and matching trials are reported.
11 educational levels (8, 9, 10, 11, 12, 13, 14,
15, 16, 18, 20). Study strengths
1. Large overall sample size, and individual
Study strengths groups approach 50.
1. Relatively large sample size, although it 2. Information regarding age, education,
encompassed a 36-year age range. gender, ethnicity, geographic area, and
2. Data presented by age and education. recruitment strategy.
3. Information regarding gender, ethnicity, 3. Adequate exclusion criteria.
and geographic area. 4. Data stratified by literacy.
4. Test version and administration proce- 5. Test format specified, although not stim-
dures reported. uli version.
5. Expected number of correct scores for 6. Means and SDs reported.
age x education groupings are reported.
Considerations regarding use of the study 1. Data not stratified by age, but apparently
1. Sample was mostly female. a fairly narrow age range was used.
2. No information regarding recruitment 2. Subjects mostly female.
strategies or IQ level. 3. No information regarding IQ level.
3. No exclusion criteria.
4. Form not specified. [BVRT.24] Mathiesen, Ellingsen, and Kjuus,
5. Actual means and SDs not reported. 1999 (Table A20.24)
BVRT scores were gathered on 52 male con-
[BVRT.23] Manly, Jacobs, Sano, Bell, Merchant, trols aged <65 as part of a study on the ef-
Small, and Stern, 1999 (Table A20.23) fects of mercury vapor on cognitive function in
BVRT data were collected on 43 literate and Norway. Mean age was 45.5 (10.8) years, and
43 illiterate older participants in the Wash- mean education was 9.5 (1.8) years. Exclusion
ington Heights Inwood Columbia Aging Pro- criteria were alcohol abuse; major head injury
ject, a community-based epidemiological study (loss of consciousness >6 hours); metabolic
conducted in northern Manhattan, New York disorders; neurological, psychiatric, or other
City. Subjects were drawn from a random sam- diseases causing severe disability; and exposure
ple of older (>65 years) Medicare recipients to known occupational neurotoxicants; 15.4%
residing in selected census tracts of Wash- had experienced mild concussions. Mean
ington Heights and Inwood. The sample was WAIS-R Vocabulary scaled score was 8.7 (1.2).
restricted to subjects with 0-3 years of educa- Administration A of form C was given.
tion. Participants were excluded if they had a Means and SDs for number correct and num-
history of stroke, Parkinson's disease, or alcohol ber of errors are reported.

1. Adequate sample size. 1. No apparent exclusion criteria.
2. Information regarding age, education, 2. Data not stratified by age.
gender, Vocabulary scaled score, and 3. Recruitment strategy not specified.
geographic area reported. 4. Data on Arab-speaking sample, collected
3. Adequate exclusion criteria. in United Arab Emirates, which may
4. Test administration format and stimuli limit generalizability for clinical inter-
specified. pretation in the United States.
5. Means and SDs reported for number
correct and number of errors reported. [BVRT.26] Touradji, Manly, Jacobs, and Stern,
2001 (Table A20.26)
Considerations regarding use of the study BVRT test scores were obtained on 193 ran-
1. Data not stratified by age. domly selected English-speaking older com-
2. Low educational level. munity residents drawn from the Washington
3. All male sample. Heights Inwood Columbia Aging Project, an
4. It is assumed that the test was adminis- epidemiological study involving elderly Medi-
tered in Norwegian, but the method of care recipients residing in 13 census tracts of
translation is not reported. Washington Heights and Inwood. All self-
5. Data collected in Norway, which may identified as non-Hispanic white; 106 were
limit generalizability for clinical inter- U.S.-bom, and 87 were born outside of the
pretation in the United States. United States (39% in Western Europe, 14%
in Austria/Hungary, 10% in Poland, 6% in
[BVRT.25] Amir, 2001 (Table A20.25) England/Luxembourg, 6% in the former
BVRT data are reported for 260 participants USSR, 3% in Bulgaria/Romania, 2% in south-
(124 males, 136 females) recruited from various em Europe, 2% in Turkey, 1% in Iraq!Jordan,
educational institutions and workplaces in the 1% in other Eastern Europe, 1% in Scandi-
United Arab Emirates. The mean age of the navia, 1% in Canada, 1% in the Caribbean, and
males was 21.7 (5.89) years, with a range of 15- 1% in North Africa). Only subjects who rated
44 years, and the mean age of the females was themselves as speaking English "very well"
21.6 (4.66) years, with a range of 15--39 years. were included in the study. All subjects were
Mean educational level of the males was 11.6 rated as nondemented by a physician, inde-
(2. 71) years and of the females, 12.0 (2.80) years. pendent of neuropsychological test scores.
Data were grouped into four IQ (superior, above All subjects were >65 years of age. The
average, average, below average) by gender by U.S.-bom subjects averaged 75.7 (7.2) years
two education ($9 years, ;?:university) groups. of age and 12.9 (3.5) years of education.
Administration A, form D, was administered, Foreign-hom subjects averaged 77.9 (7.3)
followed 2 weeks later by form E. Means and years of age and 12.0 (3.7) years of education.
SDs are reported for number correct and errors. Means and SDs for multiple-choice recogni-
tion and matching trials are reported.
Study strengths
1. Large sample size Study strengths
2. Information regarding age, education, 1. Large sample sizes.
gender, IQ, and geographic area. 2. Information regarding age, education,
3. Data stratified by gender, IQ level, and English fluency, ethnicity, geographic
education. area and place of birth, and recruitment
4. Test administration format and stimuli strategies.
versions specified. 3. Data stratified by place of birth (U.S.-
5. Data presented on two forms at 2-week bom vs. foreign-hom).
retest interval. 4. Test format specified, although not stim-
6. Means and SDs for number correct and uli form.
total errors reported. 5. Means and SDs reported.
Considerations regarding use of the study neurological or functional signs of dementia

l. Data not stratified by age, but apparently based on a physician's clinical rating (inde-
a fairly narrow age range was used. pendent of neuropsychological test scores).
2. Minimal exclusion criteria (i.e., non- Means and SDs for number correct on the
demented). multiple-choice recognition and matching tri-
3. No information regarding gender or IQ. als of the BVRT are reported.
[BVRT.27] Coman, Moses, Kraemer, Friedman, Study strengths

Benton, and Yesavage, 2002 (Table A20.27) 1. Adequate sample size.
BVRT data are the same as those reported in 2. Information provided regarding age,
the 1999 study. Means and SDs for number education, gender, fluency in English,
correct are now reported for four age groups: reading scores, recruitment strategies,
55--64 (n=6), 65-74 (n=54), 75-84 (n=67), and geographic area.
and 85+ (n = 29). 3. Good exclusion criteria.
4. Data stratified by ethnicity.
Study strengths 5. Test administration format reported (but
l. Same as above, although now means and not stimuli form).
SDs for the actual data are reported. 6. Means and SDs for multiple-choice
recognition and matching trials reported.
1. Same as above, with the exception that
l. Data not stratified by age, but there ap-
actual means and SDs provided.
pears to be a fairly narrow age range.
[BVRT.28] Manly, Jacobs, Touradji, Small, 2. No information regarding IQ level.
and Stern, 2002 (Table A20.28) [BVRT.29] Farahat, Abdelrasoul, Amr, Shebl,
BVRT test scores were obtained on 192 older Farahat, and Anger, 2003 (Table A20.29)
African Americans and 192 older non-Hispanic BVRT data were collected in 2000 on 50 male
white subjects who participated in the Wash- controls in Egypt as part of a study on the
ington Heights-Inwood Columbia Aging Pro- cognitive effects of exposure to organophos-
ject, an epidemiological study which drew phorous pesticides. The sample was recruited
participants from northern Manhattan, New from clerks and administrators; the response
York City. Subjects were drawn from a random rate among invited participants was 79%. Mean
sample of Medicare recipients in selected age was 42.48 (5.54) years; eight had university
census tracts of Washington Heights and degrees, and 42 had secondary education. Ex-
Inwood. All subjects were >65 years of age; clusion criteria were <12 years of education or
mean age of the African-American sample was medical illnesses such as diabetes, liver or kid-
73.9 (5.8) years, and mean educational level ney disease, peripheral neuropathy, vitamin
was 12.8 (2.8) years; mean age of the white deficiency, anemia, drug addiction, long-term
sample was 74.6 (5.9) years, and mean educa- treatment with psychotropic drugs, history of
tional level was 13.0 (3.0) years. Each sample head injury including loss of consciousness, or
was 68.2% female. Mean Wide Range recent exposure to neurotoxic agents. Nineteen
Achievement Test, 3rd edition (WRAT-3) participants were smokers, and the sample av-
Reading score was 44.2 (7.2) for the African- eraged five cups of coffee or tea per day.
American sample and 49.3 (4.1) for the white Means and SDs for number correct are
sample. Testing was conducted in English, and reported.
only those participants who indicated that they
spoke English "very well" were included. Study strengths
Exclusion criteria consisted of Parkinson's 1. Adequate sample size.
disease, stroke, head injury with loss of con- 2. Information provided regarding age, ed-
sciousness, alcohol abuse, serious mental ill- ucation, gender, recruitment strategies,
ness such as depression or schizophrenia, or and geographic area.
3. Good exclusion criteria. 2. The exact BVRT stimuli used are

4. Mean for number correct reported. not specified (although it is assumed to
be form C, given that this is Baltimore
Considerations regarding use of the study Longitudinal Study of Aging data).
1. Data not stratified by age. 3. Mostly male sample.
2. Test version/administration format not 4. Well-educated sample.
specified.
[BVRT.31] Reinprecht, Elmstahl, Janzon, and
Andre-Petersson, 2003 (Table A20.31)
not specified.
4. SDs not reported. BVRT data were obtained on 141 81-year-old
5. Data collected on Arabic speakers in men in a prospective study of the effects of hy-
Egypt, which may limit generalizability pertension on cognition. All men born in the
for clinical interpretation in the United even months of 1914 and residing in the munic-
States. ipality of Malmo, Sweden, were contacted; 500
6. No information on IQ level. of the 560 identified men agreed to participate in
an examination in 1982-1983, and 281 surviving
[BVRT.30] Kawas, Corrada, Brookmeyer, men were reinvited to participate in 1995-1996.
Morrison, Resnick, Zonderman, and Of these, 185 agreed to the reevaluation, and
Arenberg, 2003 (Table A20.30) BVRT data were available on 141.
The men were classified into three groups:
These BVRT data were collected as part of the no hypertension at ages 68 and 81 (n = 22),
Baltimore Longitudinal Study of Aging. The hypertension at 81 but not 68 (n = 11), hy-
sample consisted of 1,425 participants (1,004 pertension at 68 and 81 (n = 108).
men, 421 women); 72.4% had a college edu- Administration A was followed; stimuli
cation or higher. Data are presented for six age version was not specified. Means and SDs for
groupings: <50 (n = 298), 50-59 (n = 546), 60- number correct are reported.
69 (n = 815), 70-79 (n = 760), 80-89 (n = 380),
and 90+ (n = 40). Mean raw Vocabulary scores
Study strengths
from the WAIS are provided for each age 1. Large overall sample size, although some
grouping. Exclusion criteria are described in individual cells were small.
Giambra et al. (1995); 144 participants subse- 2. Information regarding age, gender, geo-
quently developed Alzheimer's disease. graphic area, and ethnicity reported.
Administration Awas employed (presumably 3. Information on recruitment strategy;
form C). Means and SDs for total number of sample apparently very representative of
errors are reported. this cohort.
4. Data on blood pressure readings.
Study strengths 5. Test administration format reported, al-
1. Very large sample size. though not test stimuli version.
2. Data stratified by age. 6. Means and SDs for number correct
3. Adequate exclusion criteria. provided.
4. Information regarding gender, educa- 7. Narrow age range.
tion, geographic area, recruitment strat-
egy, and WAIS Vocabulary. Considerations regarding use of the study
5. Means and SDs for total number of er- 1. No exclusion criteria.
rors reported. 2. No data on educational level or IQ.
3. All-male sample.
Considerations regarding use of the study 4. Method of translation of test instructions
1. Approximately 10% of participants were not specified.
subsequently diagnosed with Alzheimer's 5. Data collected in Sweden, which may
disease (although this would be true of limit generalizability for clinical inter-
any older sample). pretation in the United States.
[BVRT.32] Ruggieri, Palermo, Vitello, Gennuso, Study strengths

Settipani, and Piccoli, 2003 (Table A20.32) 1. Large overall sample size.
BVRT data were obtained on 50 Italian con- 2. Information regarding gender, age, ed-
trols (24 males, 26 females) as part of a study ucation, geographic area, and recruit-
on cognitive function in relapsing-remitting ment strategy.
multiple sclerosis. Mean age was 30.08 (8.37) 3. Means and SDs for number correct
years, with a range of 17-45 years, and mean reported.
education was 11.76 (3.04) years, with a range
of 8 to 18 years. BVRT means and SDs for Considerations regarding use of the study
number correct are reported. 1. Data not stratified by age.
2. No exclusion criteria aside from being
Study strengths non-gene carriers.
1. Minimally adequate sample size. 3. Test format and version not specified.
2. Information available on age, education, 4. Method of translation of test instructions
and gender. not specified.
3. Means and SDs reported. 5. Data collected on Dutch speakers in the
Netherlands, which may limit generaliz-
Considerations regarding use of the study ability for clinical interpretation in the
1. Data not stratified by age, and age range United States.
unacceptably large. 6. No information on IQ level.
2. No exclusion criteria listed or IQ level.
3. Test format and version not specified.
4. Recruitment strategies not specified.
CONCLUSIONS
not specified. The BVRT, more than any other neu-
6. Data collected in Italy, which may limit ropsychological measure, has gained wide in-
generalizability for clinical interpretation ternational usage for a wide range of clinical
in the United States. conditions, with the test used not only in
English-speaking populations but also in
[BVRT.33] Witjes-Ane, Vegter-van der VIis, France, Italy, Sweden, the Netherlands,
van Vugt, Lanser, Hermans, Zwinderman, van Egypt, United Arab Emirates, Korea, Vene-
Ommen, and Roos, 2003 (Table A20.33) zuela, China, Sinapore, Japan, and India.
BVRT data were obtained between 1993 and Adequate normative data are available for
1998 on 88 non-gene carriers as part of a study older male and female English-speaking
on cognition in Huntington's disease at Lei- samples, but data are generally sparse for
den University Medical Center in the Neth- younger age groups and for individuals
erlands. The group consisted of 40 men and with <12 years of education. Considerable
48 women and averaged 42 years of age (range data have been accumulated for administra-
18--64); six had less than a high school edu- tion A, form C, but considerabl~ less data are
cation, 56 had completed high school, and 26 available for the other versions.
had post-high school education.
Means and SDs for number correct are
reported.
'Meta-analyses were not perfonned for the BVRT as the

data available for review are heterogeneous in tenns of
measures reported (e.g., number correct vs. total errors),
test form/administration used, and country where data
were collected. In addition, data from the same studies
were reported in several articles, which considerably re-
duces the number of data points available for analyses.
VI
MOTOR FUNCTIONS
21
Finger Tapping Test
BRIEF HISTORY OF THE TEST a 30-degree angle from the counter. It has
been reported that 400 g of pressure is re-
The Finger Tapping Test (FTT) is one of the quired to depress the lever (Knights and
original tests introduced by Halstead and is Moule, 1967). This tapper is available from
commonly used as a simple measure of motor Psychological Assessment Resources and from
speed and motor control. Originally, it was the Reitan Neuropsychology Laboratory (see
called the Finger Oscillation Test, and the Appendix 1 for ordering information).
number of taps was recorded for the dominant A second finger tapping apparatus, the
hand only (Russell et al., 1970). Reitan mod- Digital Tapping Test, is also available. This
ified the administration of this test to include device consists of an electronic, self-contained
the performance of both hands. Tapping timer with digital readout, which automatically
speed with the dominant hand was one of the begins timing with the first depression of the
10 measures used in computing the Impair- tapping key and allows for no further recording
ment Index (Halstead, 1947; Reitan, 1955b). of taps after exactly 10 seconds have elapsed.
To take into account impaired performance This digital finger tapping device requires a
with the hand contralateral to the brain static weight of 80 g of pressure to depress the
damage, Rennick modified the procedures for lever and 0.13 inch of travel in the key to
computing the Average Impairment Rating to change the counter. The majority of the stud-
include tapping speed with the most impaired ies summarized in this chapter used the stan-
hand, rather than with the dominant hand dard manual, key-driven tapping device that is
only (Russell et al., 1970). part of the Halstead-Reitan Battery (HRB).
Interpretation of the norms available in the The most common administration and
literature is complicated by heterogeneity of scoring technique used in the reviewed stud-
the tapping devices and administration tech- ies is based on the instructions for the HRB
niques. The most frequently used device is (Rennick method), in which the Finger Tap-
a tapping lever mounted with a key-driven ping score for each hand is the mean of five
counter. The counter rotates when the tap- consecutive 10-second trials within a range of
ping key is depressed 0.50 inch. The counter five taps. A maximum of 10 trials with each
and key are mounted on a board, with the hand is allowed, and if the above criterion
tapping lever located approximately 1. 75 is not met, the score is the mean of the
inches above the surface of the board and at best five trials (see Lezak, 1995, pp. 680--682;
419
420 MOTOR FUNCTIONS
Lezak et al., 2004; Spreen & Strauss, 1998, for further converted into T scores, based on data
further information). First, all five trials with for a sample of neurologically normal partici-
the preferred hand are completed, followed pants stratified by age, education, and gender.
by the nonpreferred hand trials. The majority Golden et al. (1981b) reported that the FTT
of authors also referred to the standard de- is a measure of fine motor controL which is
scription of the procedures specified by Re- based on motor speed as well as kinesthetic
itan and Wolfson (1985). and visual-motor abilities. It has been sug-
Some studies used fewer than five trials per gested that the FTT is one of the most sen-
hand, did not enforce the procedure of obtain- sitive tests in the HRB for determining brain
ing five consecutive trials within a range of five impairment (Russell et al., 1970). In a factor-
taps, or alternated hands after each trial. There analytic study, Lansdell and Donnelly (1977)
was variability in the terms of data recording as reported that tapping performance may be
well. Some studies reported performance for impaired with most, but not necessarily all,
the dominant hand only, worse hand only, total types of cerebral damage. Other authors have
for both hands, average of both hands, etc. also noted that brain impairment generally,
A modification of the FTT administration but not always, results in a compromise in fin-
procedure was introduced by Russell and ger tapping speed (Dodrill, 1978a; Haaland
Starkey (1993) as a result of the inclusion et al., 1977; Lezak, 1995; Lezak et al., 2004;
of the Finger Tapping Test in their Halstead- Prigatano & Borgaro, 2003).
Russell Neuropsychological Evaluation System Golden and colleagues (Golden, 1978;
(HRNES). According to their instructions, the Golden et al. 1981b) suggested that at the
subject is to tap, using just the index finger, as cortical level impaired performance may re-
fast as possible. The subject is to keep the Hect dysfunction of the pJ."emotor and motor
"heel" of his or her hand on the board and strip regions of the frontal lobes or abnormal-
avoid using the whole hand, wrist, or arm. ities of sensory feedback secondary to parietal
After a brief practice, the subject is instructed lobe dysfunction. These authors reported that,
to perform six 10-second trials with each hand, in general, the greater the deficit on the finger
in sets of three trials, alternating hands be- tapping score, the nearer the lesion is to the
tween sets, starting with the dominant hand. A area of the precentral gyrus. In addition, they
score at least four taps faster or slower than indicated that subcortical disruption of sensory
the next highest or lowest score is considered or motor tracts, as well as peripheral damage to
to be an "outlier." This score is eliminated the extremities, may result in compromised
from the calculation of the total score and is performance on this measure.
replaced with an alternate trial to make a total In spite of the well-documented sensitivity
of six valid trials. This substitution is allowed of the FTT to brain dysfunction, psychometric
for two trials only. Fifteen-second rest periods issues related to the optimal balance of sen-
are allowed between trials. Total scores rep- sitivity and specificity of the test have been
resent the average speed for valid trials with widely disputed in the literature. Wheeler and
each hand and with both hands. The score Reitan (1963) have reported 79% hit rates
used in computing an overall index for the in brain-damaged populations when using
entire battery is based on the average perfor- a level-of-performance criterion. However,
mance with the dominant and nondominant while hit rates across different studies appear
hands instead of the worse-hand performance to be quite good, the rates of false-positive
used in the earlier versions. misclassifications based on the cutoffs pro-
Some studies provide data to allow con- posed by Halstead are unacceptably high
version of raw scores into other units, which (especially for older age groups) (Bornstein,
facilitates comparison between different tests. 1986a; Bomstein et al., 1987b; Heaton et al.,
For example, a normative system for the ex- 1986; Trahan et al., 1987). In McKeever and
panded HRB developed by Heaton et al. Abramson's (1991) study on college students,
(1991, 2004) allows conversion of raw scores only 10% of left-handed vs. 14% of right-
into scaled score equivalents, which can be handed females and 39% of left-handed vs.
FINGER TAPPING TEST 421
57% of right-handed males scored within the manual measures, whereby "a high percent-
nonimpaired range, using the original Hal- age (approximately 25 percent) of the normal
stead cutoff criteria. The authors emphasized sample obtained scores more than one stan-
the need for revising the norms due to the dard deviation from the control mean on a
high rate of false-positive "diagnoses," espe- single measure" (p. 719). Thus, Bomstein
cially for females and left-banders. emphasized the importance of consistency in
In addition to its utility in determining the performance pattern across tasks, rather
the presence of brain dysfunction, the FIT than use of a "rigid application of 'cookbook'
provides an index of lateralized dysfunction formulas or 'rules of thumb' " in FIT inter-
due to the contralateral effect of cerebral le- pretation (p. 723).
sions since independent measures of domi- More recently, FIT has been used in studies
nant and nondominant hand performance are exploring hemispheric specialization and neu-
customarily obtained. ral asymmetry, as reflected in handedness and
Bornstein (1986d) has systematically stud- intermanual differences in motor speed (Corey
ied the magnitude and variability of these et al., 2001; Nalcaci et al., 2001).
intermanual differences. His results suggest Data on test-retest reliability of the FTf
that in a normal sample, for approximately vary widely, with reliability coefficients rang-
30% of males and 20% of females, the non- ing from 0.04 to almost perfect values. A
preferred hand was found to be superior to majority of the studies, however, report high
the preferred hand. Trahan et al. (1987) found reliability ratings for different interprobe in-
nonpreferred hand performance to be faster tervals (Bomstein et al., 1987a; Ruff & Parker,
than preferred-hand performance in 14.7% of 1993). Reliability in Charter et al.'s (1987)
subjects. Similar findings are reported in study, which was expressed as the average
clinical studies. According to Massman and item-test correlations, was 0.99 for both hands
Doody's (1996), 26% of their patients with for normal and mixed samples with over
probable Alzheimer's disease displayed an 300 participants, with standard errors of
exaggerated right-hand advantage (associated measurement of 0.83 and 0.79 for the pre-
with higher educational level), whereas 37% ferred and nonpreferred hands, respectively.
demonstrated a reversal of expected asym- Data on repeated administration of the FIT
metry. The authors emphasized that asym- are presented by McCaffrey et al. (2000).
metry in motor speed correlated significantly For further information on the psychomet-
with cognitive asymmetries. ric properties of the FIT, see Franzen (2000),
In a follow-up on his previous study, Born- Lezak et al. (2004), and Spreen and Strauss
stein (1986b) noted that this variability in (1998).
preferred-hand performance frequently re-
sults in interpretive difficulties when the com-
monly used guideline of a 10% preferred-hand
superiority is employed. Fromm-Auch and
FTT PERFORMANCE AND
Yendall (1983) reported only a 5% difference in
DEMOGRAPHIC FACTORS
favor of the preferred hand in males.
Bornstein (1986b) suggests that, in the Empirical investigations report the effect of
evaluation of lateralized hemisphere lesions, demographic and situational variables on fin-
FIT findings have to be supported by non- ger tapping speed, such as age (Bomstein,
motor tasks and by additional instruments 1986a; Elias et al., 1993; Fromm-Auch &
measuring motor performance. In this study, Yendall, 1983; Heaton et al., 1986; McCurry
he evaluated the pattern of motor performance et al., 2001; Ruff & Parker, 1993; Trahan et al.,
on three motor tests (FIT, Grooved Pegboard 1987), education (Bernard, 1989; Bomstein,
Test, and Hand Dynamometer), which were 1985; Finlayson et al., 1977; Fromm-Auch &
administered to normal and unilateral brain Yendall, 1983; Heaton et al., 1986; Vega &
lesion samples. Interestingly, a large degree Parsons, 1967), order of test administration
of variability was observed across these inter- (Harris et al., 1981; Neuger et al., 1981),
422 MOTOR FUNCTIONS
anxiety (King et al., 1978), and personality Age Group Intervals

characteristics (Heaton, 1985). Intelligence This criterion refers to grouping of the data
level was not found to be related to finger into limited age intervals. This requirement is
tapping speed (Tremont, 1998). relevant for this test since an effect of age on
Many studies have questioned the feasibil- FTT performance has been demonstrated in
ity of restricting test interpretation for both the literature.
males and females to an identical level-
of-performance cutoff. Dodrill (1979) has Reporting of Educational Levels
reported considerable gender differences in
samples of neurological patients and neuro-
cation and FTT scores, information regarding
logically intact individuals. Other authors
educational level should be reported for each
have also reported notable gender differences
subgroup.
across demographically diverse samples, with
males consistently outperforming females
Grouping by Gender
by three to five taps (Bomstein, 1985; Ech-
temacht, 1981; Filskov & Catanese, 1986; Given the strong association between gender
Fromm-Auch & Yeudall, 1983; Harris et al., and FTT performance, normative data should
1981; Heaton et al., 1991, 2004; Hoffman, be reported for males and females separately.
1969; King et al., 1978; McKeever & Abram-
son, 1991; Morrison et al., 1979; Ruff & Par- Description of Hand Preference
ker, 1993; Trautt et al., 1983). Assessment
To address the issue of lateralization in test
performance, assessment procedures for hand
preference should be fully described. Without
METHOD FOR EVALUATING THE this assessment, assumptions regarding func-
NORMATIVE REPORTS tional lateralization cannot be made.
To adequately evaluate the FTT data, eight key
criterion variables were deemed to be critical. Procedural Variables
The first six relate to subject variables, and the
remaining two refer to procedural variables. Description of Administration Procedures
Administration procedures for the FTT differ
Subject Variables widely among studies. Detailed description of
the procedures allows selection of the most
Sample Size appropriate norms or corrections to account
Fifty cases are considered a desirable sample for deviations in administration procedures.
Data Reporting
data based on small sample sizes are highly Group means and standard deviations for the
influenced by individual differences and do number of taps averaged over all trials for the
not provide a reliable estimate of the popu- dominant and nondominant hands should
lation mean. be presented at minimum.

OF THE NORMS
order of test administration, geographic re-
cruitment region, socioeconomic status, occu- As discussed above, there is a great deal of
pation, ethnicity, handedness, and recruitment variability in sample composition, administra-
procedures are relevant. Until determined, it tion, scoring, and interpretation of the FTT.
is best that this information be provided. Some of these factors are outlined below.
In addition to nonnative studies based on In this chapter, nonnative publications and

"normal" samples, a number of clinical com- control data from clinical studies are reviewed
parison studies have explored differences in ascending chronological order. The text of
in FIT performance between clinical groups study descriptions contains references to the
and "normal control" groups (which are some- corresponding tables identified by number in
times matched on demographic characteris- Appendix 21. Table A21.1, the locator table,
tics). Unfortunately, normal control groups are summarizes information provided in the stud-
frequently comprised of medical or psychiat- ies described in this chapter. 1
ric patients. These samples cannot be consid-
ered truly "normal" due to possible effects of
their illnesses and medications on FIT
performance. SUMMARIES OF THE STUDIES
The majority of studies present data as the Original Studies
number of taps averaged over five trials for
each hand. Some studies, however, present Halstead's (1947) original control group in-
average scores for both hands; total score cluded only 28 participants (eight of whom
across both hands; total score across both were females), although 30 sets of scores were
hands over all trials, cumulatively; raw data presented. Apparently, the reason for this in-
converted to T scores; or scores for the dom- consistency was that two participants took the
inant hand only. Such deviations from the tests twice (pre- and postlobotomy), and the
standard method of data reporting are iden- results of both test administrations were in-
tified in our review of the nonnative data in the cluded in the data pool.
context of each pertinent table. In addition to Concerning the nonnative performance of
providing nonnative data for each hand, sev- Halstead's control group, Lezak (1995, p. 680)
eral studies report the proportion of partici- reported that the mean number of finger taps
pants falling in the impaired range or rates of (per 10-second trial) was 50 for the right hand
intennanual differences. and 45 for the left. Nonnative cut scores were
Several authors stratified their samples by obtained by comparing the perfonnance of
age, education, and/or gender. Procedures for controls with that of neurologically impaired
assessment of handedness are thoroughly de- individuals. On the basis of this comparison,
scribed in some studies. Furthermore, some Halstead (1947) recommended that a cutoff
authors divide their samples into groups based score of 50 (tapping scores of 50 and below are
on handedness pattern. in the brain-damaged range) be used for the
The majority of investigators recruited dominant hand in differentiating between nor-
mostly young and middle-aged participants. mal and impaired participants. The correspond-
Only a few studies present data for elderly ing cutoff for the nondominant hand was 44.
individuals. Several publications provide test- Methodological concerns are apparent
retest data over varying interprobe intervals. when using these data as a normative refer-
Some studies provide data for left-handed ence. For example, the age group interval
samples. Several studies report data on spe- being assessed was too broad (range 14-50
cific ethnic groups (e.g., Japanese Americans) years), with an unequal sampling across age
or collected abroad, including Canada, Aus- ranges. The sample consisted of mostly young
tralia, Italy, Holland, and Colombia. people, with an average age of 28.3 years. In
Among all the studies available in the liter- addition, gender was not adequately rep-
ature, we selected for review those based on resented. Moreover, the sample consisted of
well-defined samples or that offer some in- inmates and individuals being treated for
formation not routinely reported. psychiatric disturbances, thereby confounding
It should be noted that nonnative data for nonnative interpretations. For example, Hal-
airline pilots on FIT and several other tests stead (1947) noted the following:
are available in Kay (2002), (data are not re- 'Norms for children are available in Baron (2004) and
produced in this book). Spreen and Strauss (1998).
424 MOTOR FUNCTIONS
Several gave abnonnal scores on the personality These data laid the foundation for extensive
tests, thus supporting the psychiatric W,.gnoses. research concerning the psychometric prop-
Their symptoms at the time of testing ran~d from erties and clinical utility of the FIT. In spite of
mild to severe headaches; from loss of appetite, their historical value, their use as a normative
easy fatiguability, acute or chronic gastroiotestinal standard for clinical comparison is not re-
disturbance to insomnia and minor disturbances in commended due to the idiosyncratic demo-
memory functions. (p. 37) graphic composition of Halstead's and Reitan's
samples.
Reitan (1955b) reported the results of studyr
designed to assess the validity of HaJstead's
Cutoffs for the FIT based on four perfor-
mance levels, from "perfectly normal" to
(1947) Impairment Index. The sample in- "severely impaired," were published by Re-
cluded 50 non-brain-damaged contrhls (35 itan in his update on the HRB (Reitan, 1985).
males, 15 females) who were matched ~ pairs The FIT has enjoyed wide popularity among
with 50 brain-damaged participants bn the researchers and clinicians. Since its introduction
basis of race, gender, and as closely as possible by Halstead and Reitan, over 100 studies have
chronological age and years of formal ;educa- addressed performance on the FIT in normal
tion. The mean age of controls wrut 32.36 and clinical samples (usually along with other
(SD = 10. 78), and mean education wa; 11.58 HRB tests or as part of a battery comprising
(SD = 2.85). Participants received ne'frologi- various neuropsychological tests). The most
cal examinations before testing and .bowed relevant of those studies are reviewed below.
"no signs or symptoms of cerebral 4amage
or dysfunction" (p. 29). The majority pf par-
[FT.1] Vega and Parsons, 1967 (Table A21.2)
ticipants comprising the control grmlp did,
however, have various diagnoses, sue~ as de- The HRB performance of brain-damaged and
pression (n = 17), paraplegia (n = 13); acute control groups recruited in Oklahoma was
anxiety state (n = 6), and obsessivEH!ompulsive compared. The control group included
neurosis (n = 2). The author noted th• these 43 patients hospitalized for causes other than
patients were included to "minimize ~e pos- central nervous system (CNS) dysfunction and
sibility that differences in the test resultS for the seven nonhospitalized participants.
brain-damaged and control groups could be
attributed to hospitalization, chronic illness, Study strengths
and possible affective disturbances" (p. 29). 1. Group composition was described in terms
Administration procedures follow~d the of age, education, gender, IQ, clinical set-
Halstead (1947) format. Testing and scoring ting, and geographic area.
were completed before the groups were com- 2. Relatively large sample size.
posed or the participants were matched. 3. Means and SDs for the test scores are
According to the results of this study (for reported.
males only), the mean number of finger taps
for the preferred hand was 50.74 (SD ~ 7.29) Considerations regarding use of the study
for the control group, while for the brain- 1. Data are presented only for the domi-
damaged group it was 45.58 (SD = 7.32). The nant hand.
difference between the means was statistically 2. Data are not partitioned by age group.
significant. Data for the nonpreferrecl hand 3. Procedures used to determine hand
were not provided. Reitan stated that although preference are not described.
further validity studies were needed, jhe re- 4. Control sample is primarily composed of
sults of this investigation suggested tbttt "the hospitalized patients.
Halstead battery is sufficiently sensitive to 5. Data are collapsed across genders.
the effects of organic brain damage to frovide
an objective and quantitative basis for tletailed [FT.2] Goldstein and Shelly, 1972 (Table A21.3)
study of relationships between brain ~nction Inpatients at the Topeka Veterans Adminis-
and behavior" (p. 35). · tration Hospital were used. Participants came
from different services in the hospital and The authors point to a steady decrease in
were tested on a referral basis. Most partici- mean speed for both hands as an expected
pants were male adults. Participants were consequence of the aging process, which
classified into brain-damaged and control might be related to diminished function of
groups. Controls included general medical and interhemispheric neural transfer. The number
psychiatric patients. Participants for whom of reversals from the expected difference be-
definitive diagnostic differentiation could not tween the two hands became substantial in
be made were dropped. the group of 70-year-olds.
The administration procedure conforms
with the original Halstead instructions. Study strengths
1. Overall sample size is quite large, al-
Study strengths though some individual cells are small.
1. Sample size is large. 2. Sample is divided into six age groups.
2. Administration procedure is described. 3. Hand preference and method for deter-
3. Means and SDs for the test scores are mining handedness are indicated.
reported. 4. Test administration procedure is de-
4. Geographic area is indicated. scribed.
5. Gender is reported.
Considerations regarding use of the study 6. Minimally adequate exclusion criteria.
1. Demographic characteristics for the 7. Means and SDs for the test scores are
sample such as age, education, and gen- reported.
der are not reported.
2. Procedures used to determine hand Considerations regarding use of the study
preference are not described. 1. Sample is comprised primarily of males;
3. Control sample consists of medical and male and female data are collapsed.
psychiatric inpatients. 2. No data on educational level or geo-
4. Sample is comprised primarily of males; graphic area.
male and female data are collapsed.
[FT.4] Finlayson, Johnson, and Reitan, 1977
[FT.3] Goldstein and Braun, 1974 (Table A21.4) (Table A21.5)
The study explores changes in speed of per- The study compared male brain-damaged and
formance on bilateral motor tasks as a func- control samples. Controls included healthy
tion of increased age. Participants were individuals as well as hospitalized medical and
considered representative of a "normal" pop- psychiatric patients. The sample was divided
ulation, without reported history of neuro- into three groups based on education: uni-
logical difficulties. A sample consisting of versity groups included persons with at least
201 men and eight women was divided into six 3 years of college; high school groups included
age groups. those who had completed grade 12 but had
Preference of the right hand was endorsed not attended college; grade school groups in-
by all but five participants and confirmed by a cluded persons with <10 years of education.
lateral dominance examination. Participants Raw scores were converted into T scores,
who indicated mixed dominance were not with a mean of 50 and SD of 10.
included.
The procedure generally conforms with the Study strengths
original Halstead instructions. The first set of 1. Sample composition is well described in
trials established the mean for the preferred terms of age, education, gender, and IQ.
hand. Then, the procedure was repeated for 2. Data are presented by education groupings.
the nonpreferred hand. Means and SDs for
each hand as well as percent reversal (percent Considerations regarding use of the study
of participants who tap faster with the non- 1. Data are not partitioned into age groups.
preferred hand) are presented. 2. All participants are male.
426 MOTOR FUNCTIONS
3. Raw scores are not reported. group included participants from the commu-
4. Procedures used to determine hand nity with no evidence of neurological disorder.
preference are not described. Of the controls, nine were students, six were
5. Control group includes hospitalized housewives, 20 were unemployed, and 15 were
medical and psychiatric patients. employed. Controls were recruited through
6. Small individual cell sizes. employment facilities, churches, a community
college, a public high school, a volunteer ser-
[FT.S] Wiens and Matarazzo, 1977 vice agency, and a semisheltered workshop.
(Table A21.6)
The authors collected FIT data on 48 male Study strengths
applicants to a patrolman program ill Port- 1. Sample composition is well described in
land, Oregon, as part of an investigation terms of age, education, gender, occu-
of the WAIS and MMPI correlates .of the pation, geographic area, ethnicity, and
Halstead-Reitan Battery. All partieipants recruitment procedures.
passed a medical exam and were judged to be 2. Data are stratified by gender.
neurologically normal. Participants were di- 3. Means and SDs for the test scores are
vided into two equal groups, which were reported.
comparable in age, education, and W AJS full-
scale IQ. Group 1 ranged in agf 21- Considerations regarding use of the study
27 years and group 2, 21-28 years. Me~s and 1. Procedures used to determine hand
SDs are provided for each hand. : preference are not described.
Correlations for the two groups b'tween 2. Apparently adequate exclusion criteria,
W AIS FSIQ and FIT scores were -0.05 and although some controls were recruited
0.03 for the preferred hand and -0.11 from semisheltered workshops.
and -0.44 for the nonpreferred hand. The 3. Undifferentiated age range.
authors concluded that, for the top half of the 4. Relatively small sample sizes in the gen-
population in terms of education and IQ, der subgroupings.
individual differences in scores on the W AIS
do not influence performance on the HRB [FT.7] Dodrill, 1978b (Table A21.8)
measures. Performance on motor tests was compared for
a control and three brain-damaged groups.
Study strengths The 25 control participants were Caucasian,
1. Demographic characteristics of the right-handed adults over 15 years of age, re-
sample are pJ;"esented in terms of gender, cruited from community resources in Wash-
age, education, IQ, recruitment proce- ington, who had no history of injury or disease
dures, and geographic area. that involved the CNS.
2. Adequate medical exclusion criteria.
3. Means and SDs for the test scoies are Study strengths
reported. 1. Sample composition is well described in
4. Data are provided in a restricted age range. terms of age, education, gender, hand-
edness, and geographic area.
Considerations regarding use of the study 2. Minimally adequate exclusion criteria for
1. Procedures used to determine hand controls.
preference are not described. 3. Means and SDs are reported.
2. High IQ level.
3. Relatively small sample size. Considerations regarding use of the study
4. All-male sample. 1. Procedures used to determine hand
preference are not described.
[FT.6] Dodrill, 1978a (Table A21.7) 2. Small sample size.
The study compares epileptic and ~ntrol 3. Undifferentiated age range.
groups in the state of Washington. The eontrol 4. Data are collapsed across genders.
[FT.8] Morrison, Gregory, and Paul, 1979 structural brain lesions and normal controls
(Table A21.9) were compared. The control group consisted of
11le study explored interexaminer reliability volunteers with no medical or psychiatric
for test-retest conditions of the F'Tf. Partici- problems and no history of head trauma, brain
pants were 60 volunteers from introductory disease, or substance abuse. 11le study was
psychology courses with a modal age of 19. All conducted in Colorado.
participants were white and from rural back-
grounds in Idaho; half the sample were male Study strengths
and half, female. 1. Information regarding education, IQ,
Two conditions were used: age, and geographic area is provided.
1. Participants were tested by the same 3. Adequate exclusion criteria.
examiner twice with a 1-week interval 4. Means and SDs for the test scores are
(test-retest condition). reported.
2. Participants were tested by different
examiners twice with a 1-week interval Considerations regarding use of the study
(interexaminer condition). 1. Procedures used to determine hand pref-
erence are not described.
Means and SDs for the test-retest and in- 2. Undifferentiated age grouping.
terexaminer conditions are reported. 3. No information regarding gender; data
11le authors found significant gender dif- are collapsed across genders.
ferences, with males performing about three
[FT.1 0] Bak and Greene, 1980 (Table A21.11)
taps faster than females.
11le study investigated the effects of age
on different neuropsychological measures in
Study strengths
healthy, active older adults in Texas aged
1. Data are presented for males and females
50-86 years. Two age groups were compared:
separately.
50-62 and 67-86 years. All participants were
2. Information on ethnicity, occupation
right-handed. Participants were fluent in En-
(college students), age, and geographic
glish and denied history of CNS disorders,
area is presented.
uncorrected sensory deficits or illnesses, or
"incapacities" which might affect test results;
reported.
participants in poor health were excluded.
4. Restricted age grouping.
Four W AIS subtests were administered (In-
formation, Arithmetic, Block Design, Digit
Considerations regarding use of the study Symbol); mean scores on these measures sug-
1. Results are reported only for the domi- gested that IQ levels were within the high
nant hand. Data are averaged over the average range or higher.
test and retest conditions. Means and SDs were reported for each
2. No exclusion criteria. hand.
3. Relatively small individual cell sizes.
4. Procedures for assessment of hand dom- Study strengths
inance are not described. 1. 11le study provides data on a very elderly
age cohort not found in other published
[FT.9] Anthony, Heaton, and Lehman, 1980 normative data.
(Table A21.10) 2. Adequate exclusion criteria.
11le purpose of the study was to cross-validate 3. Sample composition is well described in
two computerized programs designed to de- terms of age, gender, education, hand-
termine the presence, location, and process of edness, and geographic area.
brain lesions using scores from the Halstead- 4. Means and SDs for the test scores are
Reitan Battery and the WAIS. Patients with reported.
428 MOTOR FUNCTIONS
Considerations regarding use of the study Normal volunteers were matched on age,
1. Procedures for assessment of handed- gender, and education to the patient group
ness are not described. and satisfactorily completed a brief screening
2. Sample sizes are small. exam by a neurologist.
3. High IQ and educational level for the
older age grouping. Study strengths
4. The older age grouping spans nearly two 1. Sample sizes are sufficiently large.
decades and may be too broad for opti- 2. Minimally adequate exclusion criteria.
mal clinical interpretation. 3. Means and SDs for the test scores are
5. Data are collapsed across genders. reported.
[FT.11] Eckardt and Matarazzo, 1981 Considerations regarding use of the study
(Table A21.12) 1. No data are reported for mean age, ed-
Performance on neuropsychological tests for ucation, or gender distribution for con-
drug-free alcoholic inpatients and nonalco- trols, although it is assumed that controls
holic medical inpatients was compared. All approximate the age, gender, and edu-
participants were male inpatients at V.A. hos- cation of patients given that the groups
pitals in California aged 21-60. No psychoac- were matched.
tive medication had been ingested by the 2. Method for determining handedness is
patients during the 48 hours prior to testing. not reported.
The nonalcoholic group consisted of medical 3. Undifferentiated age range.
inpatients from the same hospital who were [FT.13] Rounsaville, Jones, Novelly,
referred from a variety of services and were and Kleber, 1982 (Table A21.14)
neurologically stable during the study. They
The study compared performance of opiate
were assumed to have no recent drinking prob-
addicts, epi1epsy patients, and a control group.
lem. Sixty percent of participants had some
A sample of 29 Comprehensive Employment
college education. They were tested twice with
Training Act (CETA) participants was used
an interval of 12-22 days between probes.
for a normal comparison group. Participants
with a history of drug or alcohol abuse or of a
Study strengths
neurological disorder were excluded. Partici-
pants were screened for alcohol and i1licit
terms of age, geographic area, and gender,
psychoactive substances, and urine specimens
with cursory information on education.
were taken at the time of testing.
2. Test-retest data are available.
reported. 1. Controls are described in terms ofgender,
age, education, and percent right-handed.
Considerations regarding use of the study 2. Adequate exclusion criteria.
1. Procedures used to determine hand
preference are not described. Considerations regarding use of the study
2. Data are provided only for males and 1. SDs are not provided.
only for the dominant hand. 2. The procedures for assessment of hand
3. Small sample size. dominance are not described.
4. Controls were medical inpatients. 3. Age range is not provided.
5. Undifferentiated age range. 4. Sample size is small.
5. Data are collapsed across genders.
[FT.12] Pirozzolo, Hansch, Mortimer, Webster,
and Kuskowski, 1982 (Table A21.13) [FT.14] Yeudall, Fromm-Auch, and Davies,
The study compares performance on neu- 1982 (TableA21.15)
ropsychological measures by Parkinson's dis- The study compares performance on the HRB
ease patients and normal controls. for delinquent and nondelinquent adolescents
in Canada. The delinquent group included 2. Procedures for assessment of hand dom-
adolescents admitted to the primary residen- inance are not described.
tial treatment resource for persistent delin- 3. Sample size is relatively small.
quents with severe behavioral disturbances. 4. Data are collapsed across genders.
The nondelinquent group included adoles- 5. High IQ level.
cents from regular classrooms.
Handedness was measured by the Annett [FT.16] Prigatano, Parsons, Levin, Wright,
(1970) Handedness Questionnaire: 88% of the and Hawryluk, 1983 (Table A21.17)
delinquent sample and 83% of the control The study, conducted in Oklahoma and Can-
sample were right-handed. ada, explores neuropsychological functioning
in mildly hypoxemic patients with chronic
Study strengths obstructive pulmonary disease (COPD). The
1. Samples are described in terms of age, 25 control participants, free of physical or
gender, IQ, handedness, and geographic emotional illnesses, were matched to patients.
area.
2. Procedure for assessment of handedness Study strengths
is identified. 1. Control sample is described in terms of
3. Means and SDs for the test scores are age, education, gender, IQ, handedness,
reported. and geographic area.
4. Sample sizes are relatively large. 2. Minimally adequate exclusion criteria.
Considerations regarding use of the study reported.
2. No apparent exclusion criteria. Considerations regarding use of the study
1. Procedures for assessment of hand dom-
inance are not described.
Other comments
2. Small sample size.
1. Data were collected in Canada.
[FT.15] O'Donnell, Kurtz, and Ramanaiah,
1983 (Table A21.16)
Other comments
The study compares neuropsychological test 1. Data were partially collected in Canada.
performance of normal, learning-disabled, and
brain-damaged young adults. The normal [FT.17] Fromm-Auch and Yeudall, 1983
control group consists mostJy of college student (Table A21.18)
volunteers without a history of learning prob- The authors obtained data on 193 Canadian
lems, blows to the head, or seizures. participants (111 male, 82 female) recruited
through posted advertisements and personal
Study strengths contacts. Participants' mean education was
1. Composition of the sample is well de- 14.8 (3.0) years and mean FSIQwas 119.1 (8.8).
scribed in terms of age (college students), Participants are described as "nonpsychiatric"
education, gender, handedness, and and "non-neurological." Handedness was de-
occupation. termined by the writing hand; 83.4% of the
2. Minimally adequate exclusion criteria. sample were right-handed. Strength of hand-
3. Narrow age range. edness was determined by the Annett (1970)
4. Means and SDs for the test scores are Handedness Questionnaire. Means and SDs
reported. for each hand are reported for the entire sam-
ple and for five age groups stratified by gender.
Considerations regarding use of the study The authors observed a pronounced effect
1. Data are presented for the dominant of gender on all motor tests, with females
hand only. appearing weaker and slower than males.
430 MOTOR FUNCTIONS
The relationship between age and perfor- stratified by age group (20--39, 40-59, ~9),
mance appears to be curvilinear for both gen- level of education (<high school, ~high school),
ders, with peak performance in the 33-40 year and gender. Normative data are presented for
old age range. the preferred and nonpreferred hands for each
demographic group separately, as well as for
Study strengths different combinations of demographic strata.
1. Sample composition is described in terms Individual group sample sizes range 13--86.
of gender, IQ, education, geo!raphic Hand preference was determined as the hand
area, and recruitment proceduresj used for signing the consent form. Participants
2. Some psychiatric and neurolo~al ex- were recruited from the general population of
clusion criteria were used. , a large city in western Canada.
3. Large overall sample size.
4. Method for determining handedpess is Study strengths
specified. ~ 1. Very large overall sample size.
5. Normative data are presented for the 2. Data are stratified by age, gender, and
entire sample and separately for: differ- educational level.
ent age and gender groups. ; 3. This data set is unique in that it reports
6. Means, SDs, and ranges for ~e test data for participants with less than a high
scores are reported. school education.
4. Information on recruitment procedures
Considerations regarding use of the sttfdy and geographic area is provided.
1. Sample sizes for some age grot4>s are 5. Method for determining handedness is
very small. ; specified.
2. High intellectual and educational level of 6. Means and SDs for the test scores are
the sample. : reported.
7. Test administration procedures are spec-
Other comments , ified.
1. The article provides a summary of pre-
viously published normative data.; Considerations regarding use of the study
2. Calculation of the educational level in- 1. Individual sample sizes of some cells are
cluded technical or vocational training. small. It is unclear whether the youngest
3. Data were collected in Alberta, danada. age included was 18 or 20.
2. No reported exclusion criteria.
[FT.18] Bornstein, 1985
(Tables A21.19, A21.20) Other comments
The author collected data on 365 Capadian 1. Data were collected in Canada.
individuals (178 males, 187 females) re~ruited
through posted notices on college campuses [FT.19] Villardita, Cultrera, Cupone,
and unemployment offices, newspaper ads, and and Mejia, 1985 (Table A21.21 )
senior-citizen groups. Participants wei!e paid All participants were healthy volunteers resid-
for their participation. Participants we~ aged ing in Catania, Italy, with 8--13 years of school-
18--69, with a mean of 43.3 (17.1) years, and ing and scored >23 on the Mini-Mental State
had completed 5-20 years of education; with a Exam (MMSE). The score is the total number
mean of 12.3 (2.7) years; 91.5% of the .ample of taps recorded for each hand on two trials. The
were right-handed. No other demographic data are presented in four age groupings.
data or exclusion criteria are reported. i
Means and SDs are reported for eacH hand. Study strengths
Scores are based on the mean of fiv. trials 1. Administration procedure is described.
within five taps of each other to a maxirlJum of 2. Means and SDs for the test scores are
10 trials. When not accomplished, the score is reported.
the mean of the best five trials. The sa~ple is 3. Data are presented by age groupings.
Considerations regarding use of the study Study strengths

1. Demographic characteristics such as gen- 1. Sample is described in terms of age,
der distribution and mean educational education, and geographic area.
level are not presented. 2. Adequate sample size.
2. Data for 25--45 years of age are not 3. Means and SDs for the test scores are
presented, and data are not stratified by reported.
gender.
3. Sample sizes are small. Considerations regarding use of the study
4. Administration procedures and data re- 1. Procedures used to determine handed-
porting deviate from the standard in- ness are not described.
structions. 2. Data are reported in T scores rather than
5. Method for determining handedness is number of taps.
not reported. 3. Control group consists of medical and
6. Exclusion criteria are not adequate. psychiatric patients.
4. No information on gender; presumably,
Other comments the majority of participants are males.
1. Data were collected in Italy. 5. Undifferentiated age range.
[FT.20] Heaton, Nelson, Thompson, Burks,
and Franklin, 1985 (Table A21.22) [FT.22] Heaton, Grant, and Matthews, 1986
(Table A21.24)
The authors compared performance of mul-
tiple sclerosis patients and normal controls The authors obtained data on 553 normal
recruited in Colorado. The control group in- controls in Colorado, California, and Wiscon-
cluded 100 participants with no history of sin as part of an investigation into the effects
neurological illness, significant head trauma, of age, education, and gender on Halstead-
or substance abuse. Reitan Battery performance. The sample
consisted of 356 males and 197 females. Ex-
Study strengths clusion criteria were history of neurological
1. Control sample size is large. illness, significant head trauma, and substance
2. Information regarding age, education, abuse. Participants were aged 15-81 years,
gender, and geographic area is reported. with a mean age of 39.3 (17.5) years, and ed-
3. Exclusion criteria are adequate. ucation ranged 0-20 years, with a mean of 13.3
4. Means and SDs for the test scores are (3.4) years; 7.2% were left-handed. The sam-
reported. ple was divided into three age groups and
three education groups.
Considerations regarding use of the study Testing was conducted by trained techni-
1. Data are provided as total for both hands cians, and all participants were judged to have
only. expended their best effort on the task.
2. Undifferentiated age range. The chapter provides a review of different
3. No information regarding handedness. studies exploring the relationship of neuro-
4. High educational level of controls. psychological test performance with age, ed-
ucation, and gender. The authors concluded
[FT.21] Kane, Parsons, and Goldstein, 1985 that different sets of norms should be used
(Table A21.23) for participants of different ages, educational
The study compares performance of brain- levels, and genders when determining whe-
damaged and control participants on neuro- ther an individual's performance is normal or
psychological tests. The control group consists abnormal.
of 46 medical and nonschizophrenic psychiatric
V.A. patients with a mean age of 38.9 (11.3) Study strengths
years, recruited in Oklahoma and Pittsburgh. 1. Large overall sample size and sizes of
Data for two hands are reported in T scores. individual cells.
432 MOTOR FUNCTIONS
2. Information regarding age, education, criteria and test administration procedures are
gender, handedness, and geographic specified (Tables A21.25, A21.26).
area is provided.
[FT.24] Polubinski and Melamed, 1986
4. Data are grouped by age and educational
(Table A21.27)
level.
Participants were students taking introductory
Considerations regarding use of the study psychology classes. All participants were right-
1. SDs are not provided, which limits utility handed. The Crovitz-Zener test (1962) was
of the norms. used to assess the degree of hand dominance
2. Procedures for assessment of hand based on consistency in hand preference for
dominance are not described. five unimanual tasks. Participants with scores
3. Age groupings are quite large in terms of of 25 on this test formed the firm right-
ranges. handed groups, whi1e those with scores of $24
formed the mixed right-handed groups.
[FT.23] Bornstein, 1986a A switchback design was used, though it
(Tables A21.25, A21.26) remains unclear how many tria1s were used
This study expands the analysis of the data per hand. Number of taps in 15-second tria1s
provided in Bomstein (1985). The author ex- was averaged for each hand.
amined cutoff levels for impairment and the
proportion of participants falling in the im- Study strengths
paired range. For the preferred and nonpre- 1. Assessment of handedness is well de-
ferred hands, the clinically employed cutoff scribed.
criteria for males were 50 and 44 taps, re- 2. Information on age, education, handed-
spectively, and those for females were 46 and ness, gender, and occupation (college
40, respectively. Performance below these students) is provided.
criteria placed participants into the impaired 3. Relatively large sample for restricted
range. The high proportions of impaired age, education, and handedness groups.
scores obtained are viewed by the authors 4. Means and SDs for the test scores are
as suggesting caution in using standard cutoff reported.
scores. Base rate issues are discussed
from the perspective of the validity of test Considerations regarding use of the study
interpretation. 1. No exclusion criteria are reported.
The scores are based on the mean of five 2. Nonstandard test administration (IS-
trials within five taps of each other to a max- second tria1s), and exact test procedures
imum of 10 trials. When not accomplished, are not specified.
the score is the mean of the best five trials.
The sample was stratified by age group [FT.25] Trahan, Patterson, Quintana,
(18--39, 40--59, ~9), level of education and Biron, 1987 (Table A21.28)
(<high school, ~high school), and gender. A Participants were 713 neurologically intact
proportion of participants obtaining scores in adu1ts (382 males, 331 females) aged 18-91.
the impaired range for each of the above The score was the mean for three tria1s for
strata was provided for the preferred and each hand; trials began with the dominant
nonpreferred hands. hand and alternated between hands.
Participants were recruited from the gen- The authors found no difference between
eral population of a large city in western Ca- three-trial and five-trial scores in a subgroup
nada. Those with a history of neurological or of 102 adu1ts.
psychiatric illness were excluded. The authors concluded that the data re-
For the strengths and considerations on use vealed significant age-related differences. In
of the study, see [FT.l8] Bomstein, 1985. addition, males performed faster than females
In addition, in the current study, exclusion at all age levels.
Use of the traditional cutoff suggested by Study strengths

Halstead and Reitan resulted in false-positive 1. Large sample size.
rates ranging 28.0%--89.5% in the various 2. Data are stratified by age and gender.
groups. The percentage of participants dis- 3. Data availability for a 1~20 year age
playing an intermanual difference >10% group.
ranged 22.9%-55.2% in the various groups. 4. Adequate medical and psychiatric ex-
Reversals (nonpreferred hand faster than clusion criteria.
preferred hand) were observed in 14.7% of 5. Information regarding age, education,
the sample. The data challenge the traditional IQ, gender, occupation, recruitment
hypothesis regarding "normal" adult tapping procedures, and geographic area is pro-
performance. vided.
6. Method for determining handedness is
1. Administration procedure is described. 7. Means and SDs for the test scores are
2. Data are stratified by age and gender. reported.
4. Means and SDs for the test scores are Consideration regarding use of the study
reported. 1. High educational level of the sample.
5. Large sample sizes for younger age
groupings. Other comments
1. IQ was measured by the WAIS and
Considerations regarding use of the study WAIS-R. WAIS IQ scores were linearly
1. Sample composition is minimally de- equated to WAIS-R IQ scores.
scribed. 2. Data were collected in Canada.
2. Procedures for assessment of hand dom- 3. Correlations of FI'T scores with age and
inance are not described. education were 0.20 and 0.06 for the
preferred hand and 0.22 and 0.08 for
[FT.26] Yeudall, Reddon, Gill, and Stefanyk, the nonpreferred hand, respectively. The
1987 (Table A21.29) effect of gender on performance was also
The authors obtained data on 225 Canadian explored. The authors concluded that
participants recruited from posted advertise- age effects were not significant for either
ments in workplaces and personal solicita- hand, but there were gender effects for
tions. Participants included meat packers, both the preferred and nonpreferred
postal workers, transit employees, hospital lab hands. Therefore, they suggest using the
technicians, secretaries, ward aides, student gender norms collapsed across age.
interns, student nurses, and summer students.
In addition, high school teachers identified for [FT.27] Alekoumbides, Charter, Adkins,
participation average students in grades and Seacat, 1987 (Table A21.30)
10-12. Participants (127 males, 98 females) did The authors report data on 123 medical and
not report any history of forensic involvement, psychiatric inpatients and outpatients from
head injury, neurological insult, prenatal or V.A. hospitals in southern California without
birth complications, psychiatric problems, or cerebral lesions or history of alcoholism or
substance abuse. Handedness was determined cerebral contusion. The sample included
by the writing hand. Data were gathered by 82 participants not suffering from psychiatric
experienced technicians who "motivated illness and 32 neurotic and 9 psychotic par-
the participants to achieve maximum perfor- ticipants. In addition to psychiatry services,
mance" partially through the promise of de- participants were drawn from medicine, neu-
tailed explanations of their test performance. rology, spinal cord injury, and surgery units.
The results are presented for the whole Mean IQ was within the average range;
sample and stratified by four age groups x means and SDs for individual age-corrected
gender. subtest scores are also reported. This group,
434 MOTOR FUNCTIONS
characterized as "normal participants," was two groups was examined to determine the
recruited from the patient population of a optimal cutoff score resulting in the best overall
large general hospital and consisted mostly of classification rates, with emphasis on accurate
inpatients. Ages ranged 19-82 years, and ed- classification of normal participants.
ucation ranged 1-20 years. Seven percent of
participants were black, and all but one were Study strengths
male. Most were urban residents. Data were 1. Large sample size.
collected in southern California as part of 2. Data are provided for males and females
a project on development of standardized separately.
scores corrected for age and education for the 3. Cutoff scores are provided, as well as
Halstead-Reitan Battery. means and SDs.
4. Information on age, gender, and educa-
Study strengths tional level is provided.
1. Sample composition is well described.
Information regarding age, IQ, educa- Considerations regarding use of the study
tion, ethnicity, gender, occupational 1. Procedures for determination of hand
attainment, and geographic area is preference are not identified.
provided. 2. No exclusion criteria are identified.
2. Sample size is large. 3. Undifferentiated age range.
3. Regression equation for computation of
age- and education-corrected scores is
Other comments
provided.
4. Means and SDs for the test scores are 1. Data were collected in Canada.
reported.
[FT.29] Bornstein, Baker, and Douglass, 1987a
(Table A21.33)
1. Procedures used to determine hand The study assessed the test-retest reliability of
preference are not described. FIT over a period of 3 weeks. Participants
2. Sample was heterogeneous in terms of were 14 women and nine men without a
medical diagnoses. Psychiatric patients positive history of neurological or psychiatric
were included in this sample, which was illness. Ages ranged 17-52 with a mean of
supposedly representative of "normal" 32.3 (10.3) years; mean Verbal IQ was 105.8
participants. (10.8), and mean Performance IQ was 105.0
3. Wide age range is not partitioned by age (10.5).
groups. Participants were administered the Hal-
4. All participants but one are male. stead-Reitan Battery in standard order on
both initial testing and again 3 weeks later.
[FT.28] Bornstein, Paniak, and O'Brien, Means and SDs for raw score change over
1987b (Tables A21.31, A21.32) 3 weeks are -1.1 (5.3) and 0.9 (2.7) for the
The authors compared performance of a right and left hands, respectively.
subset of 134 healthy Canadian participants Data for the whole sample for both testing
(49 males, 85 females) from a control sample probes are provided.
described in Bomstein (1985, see above)
and 94 brain-damaged patients (47 males, 47 Study strengths
females). Both groups were matched on 1. Sample composition is described in
age and education. Performance of brain- terms of age, VIQ, PIQ, and gender.
damaged patients was considerably lower 2. Information on short-term (3-week) re-
compared to controls. Classification rates ob- test data were provided.
tained with conventional cutoff scores pre- 3. Minimally adequate exclusion criteria.
sented in Russell et al. (1970) were examined. 4. Means and SDs for the test scores are
In addition, the distribution of scores for the reported.
Considerations regarding use of the study [FT.31] Thompson, Heaton, Matthews,

1. Sample size is small. and Grant, 1987 (Table A21.35)
2. Age range is wide; the effect of age on The article presents a percentage of 426 nor-
test-retest change is not explored. mal participants (279 males, 147 females)
3. Educational levels are not specified. scoring in the lateralized lesion range using
4. Procedures for assessment of hand Golden's (1978) guidelines. Dominant hemi-
dominance are not described. It is un- sphere dysfunction was defined as superiority
clear whether the authors used domi- of nonpreferred hand performance over pre-
nant/nondominant comparisons or right/ ferred hand performance. Nondominant hemi-
left comparisons. sphere dysfunction was identified as preferred
5. Data are collapsed across genders. hand performance at least 20% better than
nonpreferred hand performance.
[FT.30] Russell, 1987 (Table A21.34) Lateral preference type was assessed based
The study explored test parameters for the on participants' performance on the Reitan-
Rennick Index of the Halstead-Reitan Battery. Klove Lateral Dominance Exam and the Miles
Brain-damaged and control groups were com- ABC Test of Ocular Dominance (Reitan &
pared. The control group consisted of patients Wolfson, 1985). The following groups were
seen in V.A. medical centers in Miami and identified:
Cincinnati who were suspected of having a
neurological condition but who had negative 1. All right-participants who wrote with
neurological findings. No other exclusion cri- their right hand and manifested right
teria are described. lateral preference on all hand, eye, and
Tests were administered and scored ac- foot measures.
cording to the standard directions given 2. Mixed right-participants who wrote
by Russell (1984), with some modification for with their right hand but manifested left
the FIT. preference on one or more hand, eye, or
foot measures.
Study strengths 3. Left-left-handed participants.
1. Information regarding IQ, education,
ethnicity, gender, age, and geographic Intermanual percent difference scores were
area is provided. calculated as preferred hand minus nonpre-
2. Large sample. ferred hand divided by preferred hand.
3. Means and SDs for the test scores are Participants' mean age is 40.59 (18.27) years
reported. and mean education is 13.15 (3.49) years.
4. Test administration procedures are re- They had been screened for history of head
ported. trauma, neurological illness, substance abuse,
serious psychiatric illness, and peripheral in-
juries that might affect test performance.
The authors concluded that age, education,
1. Sample is not partitioned into age
and gender are not significantly related to in-
groups.
termanual difference scores.
2. Data are averaged for two hands.
3. Control sample consists of medical
and psychiatric inpatients, who were sus- Study strengths
pected of having a neurological condi- 1. Large sample size.
tion but had negative neurological 2. Lateral preference was thoroughly as-
findings. sessed, and three groups are identified.
4. Method for determining handedness is 3. Intermanual differences and a percent-
not reported. age of participants scoring in the later-
5. Data are collected on mostly males and alized lesion range are reported.
collapsed across genders. 4. Adequate exclusion criteria.
436 MOTOR FUNCTIONS
5. Information on age, education, and were paid for their participation. Participants
gender is reported. represent a subset of those used in Bornstein
(1985). The sample was partitioned into three
Considerations regarding use of the study education groups. Nearly two-thirds of the
1. Means and SDs for each group are not sample are female (n = 85). Average age is
reported. 62.7 (4.3) years, and mean ages of the three
2. Age range is wide; data are not parti- education groups are comparable. Exclusion
tioned into age groups, which precludes criteria were history of neurological or psy-
consideration of the effect of age on in- chiatric disorder.
termanual differences.
Study strengths
[FT.32] van den Burg, van Zomeren, 1. Large overal1 sample size, and adequate
Minderhoud, Prange, and Meijer, 1987 individual cell sizes.
(Table A21.36) 2. Data are partitioned into three educa-
The study compares a group of patients with tion groups; the study is unique in
multiple sclerosis and demographical1y mat- terms of representation of participants
ched contro]s in northern Holland. The con- with <12 years of education.
trol group consists of 40 healthy participants 3. Information regarding gender, age, and
without a history of neurological disease, who geographic area is provided.
had never been administered psychological 4. Means and SDs for the test scores are
tests prior to their participation in the study. reported.
The total number of taps in three 10-second 5. Minimal1y adequate exclusion criteria.
trials for both hands constitutes the total score. 6. Reasonably restricted age grouping.

1. Sample is described in terms of gender, 1. Procedure for determination of hand
age, education, and geographic area, al- preference is not identified.
though education is reported as 5 years 2. The >12 years of education category is
of schooling, according to the Dutch too large.
educational system. 3. Data are collapsed across genders.
2. Administration procedure is identified.
3. Means and SDs for the test scores are Other comments
reported. 1. Data were collected in Canada.
4. Sample size is re1atively large.
5. Minimally adequate exclusion criteria. [FT.34] Ardila and Rosselli, 1989
(Table A21.38)
Considerations regarding use of the study The sample included 346 normal older
1. Sample is not partitioned into age in- Colombian adults. Participants had a score
tervals. of ;:::23 on the MMSE, had no neurological
2. Participants' handedness is not identified. or psychiatric background as determined
3. Data are reported as totals for both by a neurological and psychiatric screening,
hands over three trials. and performed adequately in everyday life
4. Data are collapsed across genders. activities.
Data are presented by age x education
Other comments groups.
1. Data were collected in northern Holland.
Study strengths
[FT.33] Bornstein and Suga, 1988 1. Large overal1 sample size.
(Table A21.37) 2. Sample is partitioned by age x education
The authors reported data on 134 healthy groups.
older Canadian volunteers aged 55-70, who 3. Adequate exclusion criteria.
Considerations regarding use of the study educational level. A total of 32% of test score
1. Demographic characteristics are curso- variance was accounted for by demographic
rily described. variables. A similar effect was described for
2. Sample size for each cell is small. nondominant hand performance, where 20%
3. SDs are not reported. of score variance was accounted for by gender,
4. Procedures used to determine hand while 9% was attributable to age and 6% to
preference are not described. educational level. A total of 34% of test score
5. Data are collapsed across genders. variance was accounted for by demographic
variables.
Other comments For the sample as a whole, mean number of
1. Data were collected in Bogota, Colombia. taps for the dominant hand was 49.9 (7.9) and
for the nondominant hand, 45.2 (7.3).
[FT.35] Heaton, Grant, and Matthews, 1991 The interested reader is referred to the
The authors provided normative data from 486 Fastenau and Adams (1996) critique of the
(378 in base sample, 108 in validation sample) Heaton et al. (1991) norms and Heaton et al.'s
urban and rural participants recruited in sev- (1996a) response to this critique.
eral states (California, Washington, Colorado, In 2004, the authors published the revised
Texas, Oklahoma, Wisconsin, Illinois, Michi- norms, which are based on a sample of over
gan, New York, Virginia, and Massachusetts) 1,000 normal adults. In addition to age, edu-
and Canada. Data were collected over a cation, and gender stratification, the data are
15-year period through multicenter collabo- partitioned by race/ethnicity (African Ameri-
rative efforts. can and Caucasian).
Sixty-five percent of the sample were male.
Mean age for the total sample was 42.06 (16.8) Study strengths
years, and mean educational level was 13.6 1. Large sample size.
(3.5) years. Mean FSIQ, VIQ, and PIQ were 2. Comprehensive exclusion criteria.
113.8 (12.3), 113.9 (13.8), and 111.9 (11.6), 3. Detailed description of demographic
respectively. Exclusion criteria were history characteristics in terms of age, educa-
of learning disability, neurological disease, tion, IQ, geographic area, and gender.
illnesses affecting brain function, significant 4. Administration procedures were out-
head trauma, significant psychiatric distur- lined.
bance (e.g., schizophrenia), and alcohol or 5. Normative data are presented in com-
other substance abuse. prehensive tables in T-score equivalents
The FTf was administered according to the for males and females separately in
procedures described by Reitan and Wolfson 10 age groupings by six education
(1985). Participants were generally paid for groupings.
their participation and judged to have pro-
vided their best efforts on the tasks. Average Consideration regarding use of the study
number of taps for five trials per hand are 1. No information regarding how hand
reported. preference was determined.
The normative data, which are not re-
produced here, are presented in comprehen- [FT.36] Ruff and Parker, 1993
sive tables in T-score equivalents for scaled (Tables A21.39, A21.40)
scores for males and females separately in 10 The FTf was administered as part of a com-
age groupings (20-34, 35-39, 40-44, 45-49, prehensive test battery to 358 normal volun-
50-54, 55-59, 60--64, 65-69, 7~74, 75--80 teers recruited in California, Michigan, and
years) by six education groupings (6-8, 9-11, the eastern seaboard aged 16-70 years
12, 13-15, 16-17, ~18 years). with 7-22 years of education. Participants
For dominant hand performance, 19% of were screened for psychiatric hospitaliza-
score variance was accounted for by gender, tions, chronic polydrug abuse, or neurological
while 9% was attributable to age and 6% to disorders.
438 MOTOR FUNCTIONS
The score was the mean number of taps and addressing its psychometric properties.
over five 10-second trials with alternating The normative sample consisted of veterans
hands, starting with the dominant hand If the treated at the Cincinnati V.A. Hospital be-
criterion was not met, up to two additional tween 1968 and 1971 and the Miami V.A.
trials were given per hand and the ~ghest/ Medical Center between 1971 and 1989. All
lowest scores eliminated from comput~on of participants received neurological examina-
the mean score. tions. Those who were administered the Hal-
Data are stratified by gender x age. Data for stead tests and the WAIS or WAIS-R were
a left hand-dominant sample are also relorted. included in the study. Nine percent of
The authors reported test-retest re .ability the sample were representatives of minority
for a 6-month interval based on data forfive or groups.
more participants from each of the 12 ~emo The total sample was divided into a com-
graphic cells (30% of sample). Reliability coef- parison group and a brain-damaged group.
ficients for women, men, and the total Sample The comparison group included "normal" in-
were 0.63, 0.70, and 0.70 for the dotJ,inant dividuals, all males. No subject in this group
hand and 0.68, 0.75, and 0.76 for th, non- had a diagnosis of CNS pathology. Presenting
dominant hand, respectively. Effect .,f age symptoms for the majority of these partici-
and gender on motor speed was sp~fically pants were neurosis with memory or somatic
addressed. The authors explored the ~tio of complaints or personality disorders with epi-
dominant/nondominant hand perfollllance sodes of explosive behavior.
rate. Patients diagnosed with schizophrenia or
severe depression requiring hospitalization, as
Study strengths well as those with evidence of systemic vas-
1. Sample composition is identified in cular disease, were not included in the sample.
terms of age, education, gendet, and Test scores can be corrected for age and
geographic area. · IQ and converted into scaled scores to fa-
2. Assessment of handedness is well cilitate comparison with other tests. Statistics
described. are reported for four groups of patients:
3. Test administration procedure is thor- comparison, left hemisphere damage, right
oughly described. hemisphere damage, and diffuse brain dam-
4. Data are stratified according to gender age. Data only for the comparison group,
and age. stratified by gender, are reproduced in this
5. Data for a left hand-dominant sample chapter.
are reported. The authors published an appendix to the
6. Means and SDs for the test scores are manual (HRNES-R; Russell & Starkey, 2001),
reported. which contains tables of scaled scores based
7. Adequate exclusion criteria. on the original HRNES norms, demographic
8. Sample sizes for each demographic cell corrections, and regression-based predicted
are quite large. scores.
9. Good exclusion criteria.
Study strengths
Consideration regarding use of the study 1. Sample composition is identified in
1. Educational levels for each demographic terms of age, education, gender, eth-
cell are not reported. nicity, and geographic area.
2. Control sample size for males is large.
[FT.37] Russell and Starkey, 1993 3. Means and SDs for the test scores are
(Table A21.41) reported.
This study describes the standardization
sample used by the authors in their manual Considerations regarding use of the study
introducing the Halstead Russell Neu- 1. Procedures used to determine hand
ropsychological Evaluation System (HRNES) preference are not described.
2. Classification of participants in the combined. Demographic information for

comparison group as normal is ques- all groups combined is also provided. The
tionable since they were suspected of mean WAIS FSIQ (Wechsler, 1955) on the
having neurological conditions and re- initial testing for the three groups combined
ferred for neurological evaluation, which was 108.8 (12.3).
yielded negative results. The FIT for the dominant and nondomi-
3. Undifferentiated age range. nant hands was administered according to the
procedures specified by Reitan and Wolfson
[FT.38] Dikmen et al., 1999 (Table A21.42) (1993). Scores represent average number of
The FIT was used in a study on the psycho- taps for five trials within 5 points of each other.
metric properties of a broad range of neu- The authors provide raw scores for perfor-
ropsychological measures, based on a sample mance at two time probes, as well as various
of 384 normal or neurologically stable adults measures of test-retest reliability and magni-
who were tested twice as part of several lon- tude of practice effect. Test-retest reliabilities
gitudinal studies. A group of friend controls for the FIT were r=0.77 for the dominant
consisted of 138 individuals who had no his- hand and r= 0.78 for the nondominant hand.
tory of recent trauma and were friends of
head-injured patients. Their mean age was Study strengths
28.5 (12.2) years, and mean education was 1. Large sample sizes for the three groups.
12.2 (1.9) years; 60% of the sample were 2. Sample composition is well described in
males, and the test-retest interval was 11.1 terms of age, education, gender, IQ,
(.6) months. A group of trauma controls con- geographic area, and setting.
sisted of 121 individuals who had a recent 3. Test administration procedures are
traumatic injury that did not involve the head. specified.
They were tested at baseline 1 month after the 4. Means and SDs for the test scores are
trauma and then 11 months later. Their mean reported.
age was 31.2 (13.6) years, and mean education 5. Information on test-retest reliability is
was 12.0 (2.6) years; 70% of the sample were provided.
males, and the test-retest interval was 10.7
(0.6) months. Both of these groups were tes- Considerations regarding use of the study
ted at the University of Washington under the 1. Exclusion criteria are not clearly de-
direction of one of the authors. Twenty per- scribed. As the authors pointed out, 20%
cent of friend controls and 46% of trauma of friend controls and 46% of trauma
controls had preexisting conditions that might controls had preexisting conditions that
affect test performance, the most significant might affect test performance, the most
being alcohol abuse or a significant traumatic significant being alcohol abuse and a
brain injury. The rest of the participants in significant traumatic brain injury.
these samples denied any history of conditions 2. Data are not partitioned by age group.
that might affect brain function. The third 3. No information regarding how hand
group, mixed normal controls, consisted of preference was determined.
125 participants who had no history of trauma
or disease involving the brain. They were en- [FT.39] McCurry, Gibbons, Uomoto,
rolled in longitudinal research projects at Thompson, Graves, Edland, Bowen,
multiple sites under the supervision of the McCormick, and Larson, 2001 (Table A21.43)
neuropsychology laboratories at the University The FIT was administered as part of the
of Colorado and the University of California battery used in a prospective study examining
at San Diego. Their mean age was 43.6 the effects of age and other demographic
(19.6) years, and mean education was 12.0 factors on cognitive test performance in a
(3.3) years; 68% of the sample were males, sample of older Japanese American adults.
and the test-retest interval was 5.4 (2.5) The sample consisted of 201 nondemented,
months. Data are reported for all groups community- and institution-dwelling elderly
440 MOTOR FUNCTIONS
2::70 years of age, who were of at least 50% 3. Adequate exclusion criteria.
Japanese heritage and enrolled in the Kame 4. Test administration procedures are
Project, a study of aging and dementia in specified.
Seattle-King County, Washington. A two- 5. Means and SDs for the test scores are
stage stratified design was used for follow-up reported.
sampling.
All participants underwent additional in- Considerations regarding use of the study
terviews with proxy informants and a physical 1. Procedures for assessment of hand
and neurological examination. The test battery dominance are not described.
was administered by a trained psychometri- 2. Weighted statistics are reported.
cian and interpreted by a geriatric psycholo- 3. Data are collapsed across genders.
gist. Participants who were judged to be
nondemented, based on the consensus of a Other comments
multispecialty team of clinicians, were in- 1. Data were collected on Japanese-American
cluded in the study. Participants were aged participants.
70-101 years, with a mean of 79.6 (7.2) years,
and had 5-20 years of education, with a mean [FT.40] Sackellares and Sackellares, 2001
of 11.0 (3.0) years; 56.2% of the sample were (Table A21.44)
female; 47.9% were Japanese-speaking or
The authors compared manual motor speed in
spoke mixed English/Japanese; 72.9% were
40 patients with psychogenic pseudoseizures
born in the United States; and 94% were
and matched controls. The control group in-
right-handed. Assessment was conducted in
cluded 40 healthy adults (28 right-handed,
either English or Japanese, based on the
12 left-handed) 18-50 years of age, with a
participants' primary language and speaking
mean of 33.2 years. Participants had no history
preference. Test administration instructions
of neurological or psychiatric disorder.
and materials were originally translated by
The dominant hand was defined as the
bilingual interviewers, then back-translated
preferred writing hand. The test was admin-
into English by two professional translators for
istered by a trained professional. The Asym-
content comparisons.
metry Index was calculated as the ratio of
An average score for five 10-second FIT
dominant minus nondominant hand to domi-
trials was computed for each hand. To adjust
nant hand performance multiplied by 100.
for the stratified sampling design, each indi-
Performance rates for both hands and
vidual's test score had to be weighted by the
asymmetry indices are provided for right-
inverse of the sampling fraction for that stra-
handed and left-handed samples.
tum (for further procedure, see the original
article). Weighted means, SDs, medians, and
25th and 75th percentile scores are presented, Study strengths
stratified into two age groups: 70-79 and 2::80. 1. Adequate sample size.
Only means, SDs, and demographic informa- 2. The sample composition is described in
tion are reproduced in this chapter. terms of age and clinical setting.
The authors found that age significantly 3. Minimally adequate exclusion criteria.
affected rate of finger tapping, whereas edu- 4. Method for determining handedness is
cation did not influence FIT performance. specified.
5. Means and SDs for the test scores
Study strengths are reported for right-handed and left-
1. Large sample. handed samples separately.
terms of age, education, gender, pre- Considerations regarding use of the study
ferred language, handedness, setting, 1. Wide age range; data are not partitioned
and geographic area. by age group.
2. Sample is cursorily described; no infor- RESULTS OF THE META-ANALYSES

mation on education or gender. OF THE FINGER TAPPING TEST DATA
3. Data are collapsed across genders. (See Appendix 21 m)
[FT.411 Prigatano and Borgaro, 2003 Data collected from the studies reviewed in
(Table A21.45) this chapter were combined in regression an-
The authors investigated "normal" and "ab- alyses, to describe the relationship between
normal" finger tapping patterns in traumatic age and test performance and to predict ex-
brain injury patients and normal contro1s. The pected test scores for different age groups.
control group included 15 participants from Effects of other demographic variables were
the general population, who were friends of explored in follow-up analyses. The general
either the patients' families or the experiment- procedures for data selection and analysis are
ers. They had no reported history of psychi- described in Chapter 3. Detailed results of the
atric or neurological disease. All participants meta-analysis and predicted test scores across
were interviewed and administered the Bar- adult age groups are provided in Appendix
row Neurological Institute Screen for Higher 21m.
Cerebral Functions. The test was not admin- Only studies that stratify results by gender
istered to any subject with a peripheral or were used in the prediction analyses. After
orthopedic injury to the hand or arm. initial data editing for consistency and for
The FTT was administered by trained outlying scores, the following data were in-
professionals. Three test trials with the dom- cluded in the analyses: eight studies, which
inant hand were followed by three trials with generated 20 data points, based on a total of
the nondominant hand. Additional consecu- 963 participants for males, dominant hand;
tive trials with each hand followed, until five seven studies, which generated 19 data points,
trials were obtained in which the mean num- based on a total of 933 participants for males,
bers of taps were within 5 points of each nondominant hand; four studies, which gen-
other. erated 10 data points, based on a total of 560
In addition to comparing finger tapping participants for females, dominant hand; and
rates for traumatic brain injury and control three studies, which generated nine data
participants, the authors provided qualitative points, based on a total of 530 participants for
interpretation of FTT performance patterns. females, nondominant hand.
Rates of tapping for the control group are It should be pointed out that the integrity of
presented in Table A21.45. the results is undermined by the lack of con-
sistency in data reporting. A majority of studies
report data for the "dominant hand" and
Study strengths "nondominant hand," while some report for the
1. The sample composition is described in "right hand" and "left hand." Some of the latter
terms of age, education, gender, setting, studies include left hand-dominant partici-
and recruitment procedures. pants. Though the percent of left-banders is
2. Adequate exclusion criteria. typically small (1%-7.5%), their inclusion
3. Test administration procedures are confounds the outcome. Also, determination of
specified. the dominant hand was based on a wide range
4. Means and SDs for the test scores are of criteria, ranging from comprehensive ques-
reported. tionnaires to self-report of the writing hand.
Quadratic regressions of FTT scores on age
Considerations regarding use of the study were used for both male and female data for
1. Small sample size. the dominant hand and linear regressions, for
2. Procedures for assessment of hand the nondominant hand. R2 ranged 0.622-
dominance are not described. 0.937. Based on the derived models, we esti-
3. Data are collapsed across genders. mated FTT scores for age intervals between
442 MOTOR FUNCTIONS
20 and 74 years. If predicted scores are needed with the effect of gender on test per-
for age ranges outside the reported bound- formance described in the literature,
aries, with proper caution (see Chapter 3) they with males expected to outperform fe-
can be calculated using the regression equa- males by three to five taps. It should be
tions included in the tables, which underlie noted in respect to this comparison for
calculations of the predicted scores. the dominant hand, that average age
Regressions of SDs on age yielded R2 for males is about 5 years greater than
ranging 0.663--0.800, indicating increase in for females.
variability with advancing age, consistent with
the literature. Predicted SDs, based on these Limitations of the analyses
models, are reported. 1. R2 of 0.686 and 0.622 for the dominant
Only a few studies reported data on edu- and nondominant hands for males and of
cational level. Therefore, the effect of educa- 0.779 for females, nondominant hand,
tion on test performance was not examined. are acceptable. However, these values
indicate that only 62%-78% of variance
Strengths of the analyses in FIT scores is accounted for by the
1. Total sample sizes range 530-963 models.
participants. 2. Number of data points for females is
2. R2 of 0.937 for females, dominant hand, small.
indicates a good model fit. However, the 3. Postestimation tests for parameter speci-
number of data points in this analysis is fications indicated marginally acceptable
only 10, which might result in somewhat homoscedasticity for males. Variability in
inflated R2 values. scores across age groups is greater than
3. Postestimation tests for parameter spec- expected by chance, with a considerable
ifications did not indicate problems with increase in variability in the older age
normality. There were no problems with groups, as reflected in the size of the
homoscedasticity for female data. confidence intervals. Therefore, the pre-
4. Differences in mean predicted scores for dicted scores for the older age ranges are
the dominant vs. nondominant hands are less accurate than for the younger ranges.
4.85 (51.61 vs. 46. 76) for males and 3. 73
(47.47 vs. 43.74) for females, which are
consistent with the guideline of a 10%
CONCLUSIONS
preferred-hand superiority (particularly
for males, somewhat smaller for females). The large number of studies focusing on the
It should be noted in the context of this psychometric properties of the FIT reflect its
comparison that data only for the domi- wide clinical use. In fact, a survey of neuro-
nant hand for a group of 19-year-olds psychologists identified the FIT as one of the
were reported in one study. This did not two neuropsychological tests (along with the
affect considerably the mean age for Category Test) most frequently used in the as-
the aggregate sample of males. However, sessment of adults (Sellers & Nadler, 1992).
the mean weighted age for the aggre- A review of the FIT research suggests con-
gate sample of females used in the data siderable consistency in the data across dif-
analysis for the non-dominant hand, in- ferent studies. A decreasing rate of tapping as a
creased by approximately 3.5 years be- function of advancing age and lower levels of
cause of a small number of data points for education are well demonstrated. Gender dif-
this group available for analysis. ferences, with males outperforming females,
5. Differences between males and females are also unequivocal. In addition, the literature
in mean predicted scores are 4.14 (51.61 review suggests that test performance is highly
vs. 47.47) for the dominant hand and affected by disruption of sensory or motor
3.02 (46.76 vs. 43.74) for the non- tracts and by peripheral damage to the upper
dominant hand, which are consistent extremities. Because FIT performance is
affected by many factors, the interpretation of caution should be taken in the interpretation
tapping speed as indicative of cortical dys- of dominant-nondominant hand comparisons.
function should be made with great caution. Despite the large number of empirical
Following recommendations by Bomstein studies exploring the psychometric properties
(1986c), the accuracy of F1T interpretive of the F1T accumulated to date, some aspects
conclusions must be confirmed by findings of F1T performance are not sufficiently ad-
from other motor and nonmotor tasks. dressed. For example, normative data for
The importance of clinical judgment, rather older age groups are scarce. Test-retest con-
than "rule of thumb," is especially apparent in cordance should be further explored, to assess
view of the controversy surrounding cutoffs the magnitude of the practice effect and to
for brain impairment. Unacceptably high address the issue of test reliability over dif-
false-positive rates with use of the original ferent interprobe intervals. Similarly, very few
Halstead cutoffs warrant further research di- studies provide norms for left-banders. In-
rected at the formulation of revised cutoff vestigation of left-handed groups is a chal-
scores, assuring an optimal balance of sensi- lenging task due to the great variability in
tivity and specificity, which would differ for cerebral dominance among left-handed indi-
the various demographic groups. Similarly, an viduals, which obscures lateralization as-
issue of intermanual differences remains sumptions in the interpretation of test results.
highly disputed. A 10% dominant hand supe- Since F1T interpretation is based on laterali-
riority criterion is clearly consistent with the zation assumptions, it is of the utmost im-
average performance across the studies pre- portance to report the criteria for assessment
sented above. However, a wide range of in- of handedness, cutoff scores for subject se-
dividual differences documented in numerous lection on the basis of handedness pattern,
studies (including high rates of reversal in and the number of left-handed individuals in
intermanual difference) suggests that great the sample, if they are included.
22
Grip Streng~ Test (Hand Dynamometer)
BRIEF HISTORY OF THE TEST dynamometer, allow three trials with each
hand, alternating right and left hands, with a
The Smedley Hand Dynamometer o. Grip 10-second rest between trials. Only the high-
Strength Test is a part of the Lateral Domi- est record for each hand is used in subsequent
nance Examination added by Reitan to the computations. Additional information on the
Halstead battery. It is a measure of pure motor administration of this test is provided in Lezak
ability (Russell & Starkey, 1993). According to (1995), Lezak et al. (2004), and Spreen and
Spreen and Strauss (1998), this test mwsures Strauss (1998).
strength or intensity of voluntary grip move- Grip strength is most commonly reported in
ments of each hand. The dynamometer is kilograms, averaged across all trials for each
available from Lafayette Instruments. Psy- hand. Some studies, however, provide data al-
chological Assessment Resources, and the lowing conversion of raw scores into other units
Reitan Neuropsychological Laboratory (see which facilitate comparison between different
Appendix 1 for ordering instructions). tests. For example, a normative system for the
There are several variations in administra- expanded Halstead-Reitan Battery (HRB) de-
tion and scoring of the test that should be taken veloped by Heaton et al. (1991, 2004) converts
into consideration when interpretin* the raw scores into scaled score equivalents, which
norms. The majority of authors refer to the can be further converted into T scores adjusted
standard description of the procedures speci- for age, education, and gender.
fied by Reitan and Wolfson (1985). The most Performance on the Hand Dynamometer
common procedure is as follows. Parti<ipants Test reflects the integrity of the motor strip
take the test while standing. After the length of (Swiercinsky, 1978). Sensitivity of this test to
the stirrup is adjusted to the size of the subject's brain dysfunction has been demonstrated in
hand, one practice trial is allowed. Parti~pants many clinical comparison studies (Bomstein,
grip the dynamometer with the arm fully ex- 1986c; Dodrill, 1978b; Strauss & Wada, 1988).
tended and pointing toward the floor. The score The Hand Dynamometer allows comparison of
is the average of two consecutive trials within grip strength between both hands and therefore
5 kg for each hand, alternating hands after each is sensitive to a lateralized lesion in the hemi-
trial, starting with the dominant hand. Ten- sphere contralateral to the hand demonstrating
second rests are allowed between the trif}s. deviant performance. Generally, the preferred
Instructions provided by the Lafayette In- hand is expected to be 10% stronger than the
strument Company, which manufactur.s the nonpreferred hand (Reitan & Wolfson, 1985),
444
GRIP STRENGTH TEST 445
with intermanual differences in excess of 20% variables. The effect of age on test perfor-
being suggestive of brain impainnent (Golden, mance is reported by Anstey and Smith (1999),
1978). Use of this criterion in nonnative stud- Bomstein (1986a), Christensen et al. (2001),
ies, however, yielded unacceptably high rates of Fromm-Auch and Yeudall (1983), Heaton
false-positive misclassification, which was es- et al. (1996b), Koffier and Zehler (1985), and
pecially true for left-handed individuals (Bom- Yeudall et al. (1987), with equivocal findings
stein, 1986c; Koffier & Zebler, 1985; Thompson regarding the timing of the onset of decline in
et al., 1987). The large number of misclassifi- grip strength (after age 40 vs. 60). The effect
cations is due to a high rate of variability in of education on test performance is ques-
intermanual differences reported in the above tionable: Bomstein (1985) found a consider-
studies, which obscures the interpretive accu- able effect, whereas Ernst (1988) and Heaton
racy of the results. et al. (1991) reported negative findings. Spreen
Bomstein (1986c) suggests that in the eval- and Strauss (1991) relate Hand Dynamome-
uation of left hemisphere lesions, interpretive ter performance to participants' height and
consistency of performance on motor tasks weight, among other variables.
should be supported by nonmotor tasks and Performance on this test is affected by
by additional instruments measuring motor gender difference, more than any other motor
performance. In this study, the author evalu- test (Heaton et al., 1991, 2004). Superiority of
ated the pattern of performance on three males in test performance is documented by
motor tests (Finger Tapping Test, Grooved Fromm-Auch and Yeudall (1983), Koffier and
Pegboard Test, and Hand Dynamometer), Zebler (1985), Morehouse et al. (2000), Pey-
which were administered to normal and uni- nircioglu et al. (2000), and Yeudall et al.
lateral brain lesion samples. Interestingly, a (1987). Dodrill (1979) related the gender
large degree of variability was observed across difference on tests that have a strong motor
these intennanual measures, whereby "a high component to hand size. The effect of gender
percentage (approximately 25 percent) of the on intennanual differences is questionable;
normal sample obtained scores more than one Borod et al. (1984), Ernst (1988), Lewan-
standard deviation from the control mean on a dowski et al. (1982), Morehouse et al. (2000),
single measure" (p. 719). Thus, the author has and Thompson et al. (1987) reported nega-
emphasized the importance of consistency in tive findings, whereas Bomstein (1986d)
performance pattern across tasks, rather than found that males had larger intennanual
use of a "rigid application of 'cookbook' for- differences.
mulas or 'rules of thumb' " (p. 723) in the test Moffoot et al. (1994) relate grip strength to
interpretation. affective state, with lower strength in patients
High test-retest reliability of the Hand suffering from major depression with melan-
Dynamometer is well documented, with co- cholia. Furthermore, Burton et al. (2002) un-
efficients ranging 0. 79-0.94 across different derscore high intraindividual variability in grip
studies (see Lezak et al., 2004). Data on re- strength as a function of physical and emo-
peated administration are also presented by tional condition, especially negative affect,
McCaffrey et al. (2000). which is more pronounced in individuals with
For further information on the psychometric traumatic brain injuries.
properties of the Hand Dynamometer, see Le-
zaket al. (2004) and Spreen and Strauss (1998).
NORMATIVE REPORTS
RELATIONSHIP BETWEEN HAND
To adequately evaluate the Hand Dynamom-
DYNAMOMETER PERFORMANCE AND
eter data, eight key criterion variables were
DEMOGRAPHIC FACTORS
deemed critical. The first six of these relate to
Performance on the Hand Dynamometer subject variables, and the remaining two refer
varies as a function of several demographic to procedural issues.
446 MOTOR FUNCTIONS
Subject Variables description allows selection of the most ap-

propriate norms or corrections to account for
Sample Size deviations in administration procedures.
size. Although this criterion is somewhat arbi- Data Reporting
trary, a large number of studies suggest that data Group means and standard deviations for grip
based on small sample sizes are highly influ- strength measured in kilograms averaged over
enced by individual differences and do not pro- all trials, for the dominant and nondominant
vide a reliable estimate of the population mean. hands, should be presented at minimum.
exclusion criteria is important. It is unclear if SUMMARY OF THE STATUS
OF THE NORMS
status, occupation, ethnicity, handedness, or The information presented in studies report-
recruitment procedures are relevant. Until de- ing data for the Hand Dynamometer differs
termined, it is best that this information be considerably. Some of these differences will
provided. be summarized below.
In addition to normative studies based
Age Group Intervals
on "normal" samples, there are a number of
This criterion refers to grouping of the data clinical comparison studies that explore dif-
into limited age intervals. This requirement is ferences in test performance between clinical
especially relevant for this test since an effect groups and "normal control" groups (which are
of age on grip strength has been unequivocally sometimes matched on demographic charac-
demonstrated in the literature. teristics). "Normal control" groups are fre-
Reporting of Educational Levels
quently comprised of medical or psychiatric
patients. These samples cannot be considered
Given the possible association between edu- truly "normal" due to possible effects of their
cational level and grip strength, information illnesses and medication intake on test per-
regarding educational level should be re- formance. Administration and scoring proce-
ported for each subgroup. dures vary among studies. The number of
Reporting by Gender trials with each hand varies between two and
four, with the majority of studies using two
A strong relationship between gender and grip
alternating-hands trials. The majority of the
strength has been unequivocally demonstrated
authors report data in kilograms, averaged
in the literature. Therefore, it is imperative
across all trials for each hand; however, some
that normative data are reported for males
studies present data in T scores, report
and females separately.
strength at the best attempt, or provide scores
Description of Hand Preference Assessment for the dominant hand only. Starting hand
varies, although the majority of studies start
To address the issue of lateralization in test
with the dominant hand. Such deviations from
performance, assessment procedures for hand
the standard method of data reporting are
preference should be fully described. Without
identified in our review of the normative data
this assessment, assumptions regarding func-
in each pertinent table. In addition to provid-
tionallateralization cannot be made.
ing normative data for each hand, several
studies report the proportion of participants
Procedural Variables falling in the impaired range or rates of inter-
manual differences.
Description of Administration Procedures Several authors stratify their samples by
Administration procedures for the Hand Dy- age, education, and/or gender. Procedures for
namometer differ among studies. A detailed assessment of handedness are thoroughly
described in some studies. Furthermore, 2. Sample size is relatively small.

some authors divide their samples into groups 3. All-male sample.
based on handedness pattern.
The majority of studies recruited mostly [0.2] Wiens and Matarazzo, 1977
young and middle-aged participants. Only a (Table A22.3)
few studies present data for elderly individu- The authors collected data on 48 male appli-
als. Several studies provide test-retest data cants to a patrolman program in Portland,
over varying interprobe intervals ranging from Oregon, as part of an investigation of the
14 weeks to 6 months. WAIS and MMPI correlates of the Halstead-
Among all the studies available in the liter- Reitan Battery. All participants passed a
ature, we selected for review those based on medical exam and were judged to be neuro-
well-defined samples or that offer some infor- logically normal. Participants were divided
mation not routinely reported. into two equal groups, which were compara-
In this chapter, normative publications and ble in age, education, and WAIS FSIQ. Group
control data from clinical studies are reviewed 1 ranged in age 21-27 years and group 2,
in ascending chronological order. The text of 21-28 years.
study descriptions contains references to the Correlations for the two groups between
corresponding tables identified by number in W AIS FSIQ and dynamometer scores were
Appendix 22. Table A22.1, the locator table, -0.38 and 0.03 for the preferred hand
summarizes information provided in the and -0.13 and -0.36 for the nonpreferred
studies described in this chapter. 1 hand, respectively. The authors inferred that
for the top half of the population in education
and IQ, individual differences in scores
on the WAIS do not influence performance on
SUMMARIES OF THE STUDIES the test.
[0.1 1 Matarazzo, Wiens, Matarazzo,
and Goldstein, 1974 (Table A22.2) Study strengths
1. Demographic characteristics of the sam-
Participants were 29 normal young men who
ple are presented in terms of gender,
met strict selection criteria for the Portland
age, education, IQ, recruitment proce-
Police Department. Participants were aged
dures, and geographic area.
21-28 years, educational level ranged 12-
2. Adequate medical exclusion criteria.
16 years, and mean full-scale IQ was 118.
Participants were retested 14-24 weeks later,
reported.
with a median of 20 weeks.
4. Data are provided in a restricted age
range.
Study strengths
1. Sample composition is described in Considerations regarding use of the study
terms of age, gender, education, IQ, and 1. Procedures used to determine hand
geographic area. preference are not described.
2. Administration procedure is outlined. 2. High IQ level.
3. Data on test-retest are presented. 3. Relatively small sample size.
4. Adequate exclusion criteria. 4. All-male sample.
provided. [0.3] Dodrill, 1978b (Table A22.4)
Performance on motor tests was compared for
Considerations regarding use of the study a control and three brain-damaged groups.
1. Procedures for assessment of hand The 25 control participants were Caucasian,
dominance are not described. right-handed adults over 15 years of age, re-
'Nonns for children are available in Baron (2004) and cruited from community resources in Wash-
Spreen and Strauss (1998). ington, who had no history of injury or disease
448 MOTOR FUNCTIONS
that involved the central nervous system status, gender, ethnicity, and geographic
(CNS). area.
The Smedley Hand Dynamometer was 2. Data are presented for males and fe-
used. Two trials were given in altematiQ.g fash- males separately.
ion for each hand, beginning with ~ right 3. Sample sizes are adequate.
hand. The average of the two trials wis used 4. Means and SDs for the test scores are
as the final score for each hand.
The authors concluded that the :$.
eter correctly identified lateralization
!
om-
brain
reported.

lesions in more instances than other mot r tests. 1. Administration procedures are not
clearly described.
Study strengths . 2. Samples were not divided into age groups.
1. Sample composition is descri'~ed in 3. Procedures for assessment of hand dom-
terms of age, education, genderJhand- inance are not described.
edness, and geographic area. : 4. No apparent exclusion criteria.
2. Administration procedures are ~ll de-
scribed. I [D.51 Rounsaville, Jones, Novelly, and
3. Minimally adequate exclusion criteria. Kleber, 1982 (Table A22.6)
4. Means and SDs for the test scotes are The study compared performance of opiate
reported. , addicts, epilepsy patients, and a control group.
A sample of 29 Comprehensive Employment
'
Considerations regarding use of the st~dy Training Act (CETA) participants was used
1. Procedures used to determine hand for a normal comparison group. Participants
preference are not described. with a history of drug or alcohol abuse or of a
2. Small sample size. neurological disorder were excluded. Partici-
3. Data are collapsed across genders. pants were screened for alcohol and illicit
4. Undifferentiated age range. psychoactive substances, and urine specimens
were taken at the time of testing.
[D.41 Dodrill, 1979 (Table A22.5) The Stoelting Dynamometer was used;
The study explored gender differences on no other test administration information is
various neuropsychological measure$. The provided.
control group included 47 matched pairs of
nonneurological males and females recruited Study strengths
in Washington. Within each pair, participants 1. Control participants are described in
were matched for age (±5 years) and ~duca terms of gender, age, education, and per-
tion (±2 years). All participants were ~auca cent right-handed.
sian and older than 16 years. In addition, 2. Adequate exclusion criteria.
groups were matched for Hollingshead~s two-
factor index of social position. ; Considerations regarding use of the study
It is assumed, but not stated by the authors, 1. SDs were not provided.
that the original Halstead procedure was fol- 2. Testing procedure is scarcely described.
lowed, and the score for the dominant hand 3. Procedures for assessment of hand
was reported. ~ dominance are not described.
The authors reported considerable gender 4. Age range is not provided.
differences on the tests that have very :strong 5. Sample size is small.
motor components, which they rela~d to 6. Data are collapsed across genders.
hand size.
[D.61 Yeudall, Fromm-Auch, and Davies,
Study strengths 1982 (Table A22.7)
1. Sample composition is descrilJFd in The study compares performance on the HRB
terms of age, education, socioeccromic for delinquent and nondelinquent adolescents
in Canada. The delinquent group included 2. Data are presented for the dominant
adolescents admitted to the primary residen- hand only.
tial treatment resource for persistent delin- 3. Small sample size.
quents with severe behavioral disturbances. 4. Data were collapsed across genders.
The nondelinquent group included adoles- 5. Undifferentiated age range.
cents from regular classrooms.
Handedness was measured by the Annett [0.8] Fromm-Auch and Yeudall, 1983
(1970) Handedness Questionnaire; 88% of the (Table A22.9)
delinquent sample and 83% of the control The authors obtained data on 193 Canadian
sample were right-handed. participants (111 male, 82 female) recruited
through posted advertisements and personal
Study strengths contacts. Mean education was 14.8 (3.0) years,
1. Samples are described in terms of age, and mean FSIQ was 119.1 (8.8). Participants
gender, IQ, handedness, and geographic are described as "nonpsycbiatric" and "non-
area. neurological." Handedness was determined
2. Procedure for assessment of handedness by the writing hand; 83.4% of the sample were
is identified. right-handed. Strength of handedness was
3. Means and SDs for the test scores are determined by the Annett (1970) Handedness
reported. Questionnaire. Means and SDs for each hand
4. Sample sizes are relatively large. are reported for the entire sample and for five
age groups stratified by gender.
Considerations regarding use of the study The authors concluded that a pronounced
1. Data were collapsed across genders. effect of gender was seen on all motor tests,
2. No apparent exclusion criteria. with females appearing weaker and slower
than males. The relationship between age and
Other comments performance appears to be curvilinear for both
1. Data were collected in Canada. genders, with peak performance in the 33-40
year range.
[0.7] Prigatano, Panons, Levin, Wright,
and Hawryluk, 1983 (Table A22.8) Study strengths
The study, conducted in Oklahoma and Can- 1. Sample composition is described in terms
ada, explores neuropsychological functioning of age, gender, IQ, education, geographic
in mildly hypoxemic patients with chronic area, and recruitment procedures.
obstructive pulmonary disease (COPD). The 2. Some psychiatric and neurological ex-
25 control participants, free of physical or clusion criteria were used.
emotional illnesses, were matched to patients 3. The large overall sample size.
on age, education, handedness, gender ratio, 4. Handedness was established.
and social class rating. Data only for the 5. Data are partitioned into five age groups
control group are reproduced in this chapter. and gender groups.
6. Means, SDs, and ranges for the test
Study strengths scores are reported.
1. Control sample was described in terms
of age, education, gender, IQ, handed- Considerations regarding use of the study
ness, and geographic area. 1. Sample sizes for some age groups are
2. Minimally adequate exclusion criteria. very small.
3. Means and SDs for the test scores are 2. High intellectual an:d educational level of
reported. the sample.
Considerations regarding use of the study Other comments

1. Procedures for assessment of band dom- 1. Provides a summary of previously pub-
inance are not described. lished normative data.
450 MOTOR FUNCTIONS
2. Calculation of the educational level in- Other comments

cluded technical or vocational training. 1. Data were collected in Canada.
3. Data were collected in Alberta, Canada.
[0.10] Koffler and Zehler, 1985
[0.9] Bomstein, 1985 (Table A22.12)
(Tables A22.10, A22.11) In this study, 206 normal (by self-report)
The author collected data on 365 Canadian adults (100 males, 106 females) aged 20-77
individuals (178 males, 187 females) recruited were administered the Hand Dynamometer to
through posted notices on college c~puses obtain normative data. The Stoelting Dyna-
and unemployment offices, newspapt:!r ads, mometer was used, and Reitan's procedure of
and senior-citizen groups, who we~ paid using the highest reading for each hand was
for their participation. Participants were aged followed. After adjustment for hand size, two
18--69, with a mean of 43.3 (17.1) ye.-s, and alternating trials were given, beginning with
had completed 5-20 years of educatioi1J with a the dominant hand. Determination of hand
mean of 12.3 (2.7) years; 91.5% of the pample dominance was based on self-report; 87% of
were right-handed. No other demopphic participants were right-handed, 9. 7% were
data or exclusion criteria are reported.' left-handed, and 3.3% reported mixed domi-
Means and SDs are reported for eacll hand. nance. The data were stratified by age and
The sample is stratified by age group t20--39, gender.
40--59, 60--69), level of education ~<high The authors concluded that greater strength
school, ~high school), and gender. Notmative of grip is demonstrated by males at all ages.
data are presented for the preferred and They cautioned that the use of commonly
nonpreferred hands, for each demographic accepted criteria for detection of lateralized
group separately as well as for different com- motor dysfunction leads to a large number of
binations of demographic strata. Individual false-positive errors.
group sample sizes ranged 13--86. ' Hand
preference was determined as the hand used Study strengths
for signing the consent form. Partipipants 1. Administration procedure is well de-
were recruited from the general popul~on of scribed.
a large city in western Canada. : 2. Data are presented by age and gender.
3. Method for determining handedness is
1. Very large overall sample size. 4. Overall sample size is adequate, though
2. Data are stratified by age, gender, and sizes of individual cells are small.
educational level. 5. Means and SDs for the test scores are
3. Sample is unique in that it includes reported.
participants with less than a high school
education. Considerations regarding use of the study
4. Information on recruitment procedures 1. Demographic characteristics of the
and geographic area is provided. sample are only cursorily described.
5. Method for determining handedness is 2. Medical exclusion criteria are unclear.
specified.
6. Means and SDs for the test scores are [0.11] Heaton, Nelson, Thompson, Burks,
reported. and Franklin, 1985 (Table A22.13)
Considerations regarding use of the stUdy tiple sclerosis patients and normal control
1. Individual sample sizes of some c•lls are participants recruited in Colorado. The con-
small. It is unclear whether the yo)lngest trol group included 100 participants with no
age included in the study was 18 br 20. history of neurological illness, significant head
2. No reported exclusion criteria. trauma, or substance abuse.
Study strengths effects of age, education, and gender on

1. Control sample size is large. Halstead-Reitan Battery performance. The
2. Information regarding age, education, sample consisted of356 males and 197 females.
gender, and geographic area is reported. Exclusion criteria were history of neurological
3. Exclusion criteria are adequate. illness, significant head trauma, and substance
4. Means and SDs for the test scores are abuse. Participants were aged 15-81 years,
reported. with a mean age of39.3 (17.5) years, and had
education of 0-20 years, with a mean of 13.3
Considerations regarding use of the study (3.4) years; 7.2% were left-handed. The sam-
1. Data are provided as total for both hands ple was divided into three age groups and
only. three education groups.
2. Undifferentiated age range. Testing was conducted by trained techni-
3. No information regarding handedness. cians, and all participants were judged to have
4. High educational level of control expended their best effort on the task.
participants. The chapter reviews different studies that
explore the relationship of neuropsychological
[0.12] Kane, Parsons, and Goldstein, 1985 test performance with age, education, and gen-
(Table A22.14) der. The authors conclude that different sets
The study compares performance of brain- of norms should be used for participants of
damaged and control participants on neuro- different ages, educational levels, and genders
psychological tests. The control group consists when determining whether performance is
of 46 medical and nonschizophrenic psychi- normal or abnormal.
atric Veterans Administration patients, with a
mean age of 38.9 (11.3) years, recruited in Study strengths
Oklahoma and Pittsburgh. 1. Large overall sample size and sizes of
Data for two hands are reported in T scores. individual cells.
Study strengths gender, handedness, and geographic area
1. Sample is described in terms of age, is provided.
education, and geographic area. 3. Adequate exclusion criteria.
2. Adequate sample size. 4. Data are grouped by age and educational
3. Means and SDs for the test scores are level.
reported.
Considerations regarding use of the study 1. SDs are not provided, which limits the
1. Procedures used to determine handed- utility of the norms.
ness are not described. 2. Procedures for assessment of hand dom-
2. Data are reported in T scores rather than inance are not described.
in kilograms. 3. Age groupings are quite large in terms of
3. Control group consisted of medical and ranges.
psychiatric patients.
4. No information regarding gender; pre- [0.14] Yeudall, Reddon, Gill, and
sumably, the majority of participants are Stefanyk, 1987 (Table A22.16)
male. The authors obtained data on 225 Canadian
5. Undifferentiated age range. participants recruited from posted advertise-
ments in workplaces and personal solicita-
[0.13] Heaton, Grant, and Matthews, 1986 tions. The sample included meat packers,
(Table A22.15) postal workers, transit employees, hospital lab
The authors obtained data on 553 normal technicians, secretaries, ward aides, student
controls in Colorado, California, and Wis- interns, student nurses, and summer students.
consin as part of an investigation into the In addition, high school teachers identified for
452 MOTOR FUNCTIONS
participation average students in grades [0.15] Thompson, Heaton, Matthews,

10-12. Participants (127 males, 98 females) and Grant, 1987 (Table A22.17)
did not report any histo:ry of forensic in- The article presents a percentage of 426
volvement, head injury, neurological insult, normal participants (279 males, 147 females)
prenatal or birth complications, ps~hiatric scoring in the lateralized lesion range using
problems, or substance abuse. Han.edness Golden's (1978) guidelines. Dominant hemi-
was determined by the writing hand. : sphere dysfunction was defined as superior-
Data were gathered by experiencefl tech- ity of nonpreferred hand performance over
nicians who "motivated the participflnts to preferred hand performance. Nondominant
achieve maximum performance" t)artially hemisphere dysfunction was identified when
through the promise of detailed explapations preferred hand performance was at least 20%
of their test performance. Standard ~st ad- better than nonpreferred hand performance.
ministration procedures were used. Lateral preference type was assessed based
The results are presented for the whole on performance on the Reitan-Klove Lateral
sample and stratified by four age groups x Dominance Exam and the Miles ABC Test of
gender. Ocular Dominance (Reitan & Wolfson, 1985).
The following groups were identified:
Study strengths
1. Large sample size. 1. All right-participants who wrote with
2. Data are stratified by age and gender. their right hand and manifested right
3. Data available for a 15-20 age group. lateral preference on all hand, eye, and
4. Adequate medical and psychiatric ex- foot measures.
clusion criteria. 2. Mixed right-participants who wrote
5. Administration procedure is well de- with their right hand but manifested left
scribed. preference on one or more other hand,
6. Information regarding age, education, IQ, eye, or foot measures.
gender, occupation, recruitment proce- 3. Left-left-handed participants.
dures, and geographic area is provided.
7. Method for determining handedness is Intermanual percent difference scores were
specified. calculated as preferred hand minus non-
8. Means and SDs for the test scores are preferred hand divided by preferred hand.
reported. Mean age was 40.59 (18.27) years, and mean
education was 13.15 (3.49) years. Participants
Consideration regarding use of the stu~y had been screened for histo:ry of head trauma,
1. High educational level of the sample. neurological illness, substance abuse, serious
psychiatric illness, and peripheral injuries that
Other comments might affect test performance.
1. IQ was measured by the WAlS and The authors concluded that neither age nor
WAIS-R. WAIS IQ scores were linearly education nor gender was significantly related
equated to WAIS-R IQ scores. to intermanual difference scores.
2. Data were collected in Canada.
3. Correlations of dynamometer scores with Study strengths
age and education were 0.25 a~ 0.16 1. Large sample size.
for the preferred hand and 0.27 atid 0.17 2. Lateral preference was thoroughly as-
for the nonpreferred hand, respettively. sessed, and three groups were identified.
The effect of gender on perfmjmance 3. Intermanual differences and percentage
was also explored. The author$ con- of participants scoring in the lateralized
cluded that age and gender effects were lesion range are reported.
significant for both hands. Therefore, 4. Adequate exclusion criteria.
they suggest using gender norms strati- 5. Information on age, education, and gen-
fied by age. der is reported.
Considerations regarding use of the study 3. Ratio of dominantlnondominant hands

1. Means and SDs for each group are not was computed for the entire sample, to
reported. indicate the strength of lateralization.
2. Data are presented for a wide age range 4. Relatively large sample size for con-
not separated into age groups, which stricted age range.
precludes consideration of the effect of 5. Good medical and psychiatric exclusion
age on intermanual differences. criteria.
[0.16] Ernst, 1988 (Table A22.18) reported.
The author collected data on 85 Brisbane
(Australian) uncompensated volunteers aged
1. Procedures used to determine hand
65-75, recruited from the Queensland State
preference were not described.
electoral roll. All but one were Caucasian and
2. Approximately half of the participants
right-handed. Thirty-nine were males and
had at least one chronic illness, and over
46 were females. The sample was derived from
half were taking prescribed medications.
518 names randomly selected based on date of
birth and residence. Potential participants
Other comments
were sent information regarding the project
1. Data were collected in Australia.
and a health questionnaire and asked to par-
ticipate. Individuals with histories of sub- [D. 17] Heaton, Grant, and Matthews, 1991
stance abuse, head trauma, stroke, psychiatric
The authors provide normative data from 486
hospitalization, or epiJepsy were excluded. A
(378 base sample, 108 validation sample) ur-
large minority of participants (42%) had a
ban and rural participants recruited in several
history of at least one treated and/or well-
states (California, Washington, Colorado,
controlled chronic illness (10 heart disease,
Texas, Oklahoma, Wisconsin, Illinois, Michi-
17 hypertension, 5 asthma, 2 emphysema, 10
gan, New York, Virginia, and Massachusetts)
hypo- or hyperthyroidism, 2 diabetes). A ma-
and Canada. Data were collected over a 15-
jority of participants were currently using
year period through multicenter collaborative
prescribed medications (55%) for the above
efforts.
chronic diseases or as a hypertensive preven-
Sixty-five percent of the sample were male.
tative. Mean educational level of 10.4 was
Mean age for the total sample was 42.06
comparable to the modal educational level for
(16.8), and mean educational level was 13.6
that age range according to the Australian
(3.5). Mean FSIQ, Verbal IQ, and Perfor-
Bureau of Statistics. A wide range of occupa-
mance IQ were 113.8 (12.3), 113.9 (13.8), and
tions was represented, including unskilled
111.9 (11.6), respectively. Exclusion criteria
laborers, homemakers, business persons,
were history of learning disability, neurologi-
teachers, etc.
cal disease, illnesses affecting brain function,
The authors concluded that a significant
significant head trauma, significant psychiatric
gender difference was demonstrated for both
disturbance (e.g., schizophrenia), and alcohol
hands. Superiority of the preferred hand was
or other substance abuse.
approximately 10%. No gender differences
The Hand Dynamometer Test was admin-
were demonstrated on intermanual ratio for
istered according to the procedures provided
grip strength.
by Reitan and Wolfson (1985). Participants
were paid and judged to have provided their
Study strengths best efforts on the tasks. Average number
1. Ethnic characteristics, gender, education, of kiJograms for two trials for each hand sep-
handedness, age, recruitment proce- arately is reported.
dures, and geographic area are reported. The normative data, which are not repro-
2. Data are presented for males and fe- duced here, are presented in comprehensive
males separately. tables in T-score equivalents for scaled scores
454 MOTOR FUNCTIONS
for males and females separately in 10 age [0.18] Russell and Starkey, 1993
groupings (20-34, 35-39, 40-44, 45-49, 50- (Table A22.19)
54,55-59,60-64,65-69,70-74, 75-BO)bys~ This study describes the standardization
education groupings (6-8, 9--11, 12, 13-15, sample used by the authors in their manual
16-17, 2::18 years). introducing the Halstead-Russell Neuropsy-
For dominant hand performance, 58% of chological Evaluation System (HRNES) and
the score variance was accounted for by gen- addressing its psychometric properties. The
der, while age and educational level accounted normative sample consisted of veterans
for a negligible amount of unique variance treated at the Cincinnati V.A. Hospital be-
in performance (2% and 1%, respectively). A tween 1968 and 1971 and the Miami V.A.
total of 63% of test score variance was ac- Medical Center between 1971 and 1989. All
counted for by demographic variables. For participants received neurological examina-
nondominant hand performance, 55% of score tion. Those participants who were adminis-
variance was accounted for by gender, while tered the Halstead tests and the W AIS or
age and educational level also accounted for a WAIS-R were included in the study. Nine
negligible amount of unique variance in per- percent of the sample were representatives of
formance (4% and 1%, respectively). A total of minority groups.
62% of test score variance was accounted for The total sample was divided into a com-
by demographic variables. parison group and a brain-damaged group.
For the sample as a whole, mean scores in The comparison group included "normal"
kilograms were, for the dominant hand, 43.4 individuals. No subject in this group had a
(13.1) and, for the nondominant hand, 39.7 diagnosis of CNS pathology. Presenting symp-
(12.7). toms for the majority of these participants
The interested reader is referred to the were neurosis with memory or somatic com-
Fastenau and Adams (1996a) critique of the plaints or personality disorders with episodes
Heaton et al. (1991) norms, and Heaton et al.'s of explosive behavior.
(1996) response to this critique. Patients diagnosed with schizophrenia or
In 2004, the authors published the revised severe depression requiring hospitalization as
norms, which are based on a sample of over well as those with evidence of systemic vas-
1,000 normal adults. In addition to age, edu- cular disease were not included in the sample.
cation, and gender stratification, the data are Test scores can be corrected for age and IQ
partitioned by race/ethnicity (African Ameri- and converted into scaled scores to facilitate
can and Caucasian). comparison with other tests. Statistics are
reported for four groups of patients: compar-
Study strengths ison, left hemisphere damage, right hemi-
1. Large sample size. sphere damage, and diffuse brain damage.
2. Comprehensive exclusion criteria. Data on1y for the comparison group, stratified
3. Detailed description of the demographic by gender, are reproduced in this chapter.
characteristics of the sample in terms of The authors published an appen~ to the
age, education, IQ, geographic area, and manual (HRNES-R, Russell & Starkey, 2001),
gender. which contains tables of scaled scores based on
4. Administration procedures are outlined. the original HRNES norms, demographic cor-
5. Normative data are presented in com- rections, and regression-based predicted scores.
prehensive tables in T-score equivalents
for males and females separately in 10 age
groupings by s~ education groupings. Study strengths
Consideration regarding use of the study terms of age, education, gender, ethnic-
1. No information regarding how hand ity, and geographic area.
preference was determined. 2. Control sample size for males is large.
reported. 1. Procedures used to determine hand
preference are not described.
Considerations regarding use of the study 2. Data should be used with caution since
1. Procedures used to determine hand pref- they were collected on patients sus-
erence are not described. pected of having neurological conditions
2. Classification of control participants as and referred for neurodiagnostic proce-
"normal" is questionable since they had dures, which yielded negative results.
been suspected of having neurological 3. Undifferentiated age range.
conditions and were referred for neuro- 4. It is unclear how many males and fe-
logical evaluation, which yielded nega- males are in each IQ group.
tive results.
[0.20] Dikmen et al., 1999 (Table A22.21)
The Hand Dynamometer Test was used in a
study on the psychometric properties of a
[0.19] Tremont, Hoffman, Scott, and
broad range of neuropsychological measures,
Adams, 1998 (Table A22.20)
based on a sample of 384 normal or neuro-
The Halstead-Reitan Battery was used in a logically stable adults who were tested twice as
study examining the effect of intelligence part of several longitudinal studies. A group of
level on neuropsychological test performance. "friend controls" consisted of 138 individuals
The sample included 157 patients (71 males, who had no history of recent trauma and were
86 females) aged 16-74 years; 143 patients friends of head-injured patients. Their mean
were Caucasian, 8 African American, 4 other, age was 28.5 (12.2) years and mean education
and 2 undetermined. Patients were referred was 12.2 (1.9) years; 60% of the sample were
to the University of Oklahoma Neuropsycho- males, and the test-retest interval was 11.1
logical Laboratory for neuropsychological (0.6) months. A group of "trauma controls"
evaluation but determined to be neurologi- consisted of 121 individuals who had a recent
cally normal based on neurodiagnostic pro- traumatic injury that did not involve the head.
cedures. The sample was devided into below They were tested at baseline 1 month after
average (FSIQ ~89), average (FSIQ 90-109), trauma and then 11 months later. Their mean
and above average (FSIQ ~110) ranges, based age was 31.2 (13.6) years and mean education
on performance on the WAIS-R. Data for the was 12.0 (2.6) years; 70% of the sample were
dominant hand performance on the Hand males, and the test-retest interval was 10.7
Dynamometer for males and females parti- (0.6) months. Both groups were tested at the
tioned by intelligence level are presented in University of Washington under the direction
Table A22.20. of one of the authors. Twenty percent of
The authors concluded that the Grip friend controls and 46% of trauma controls
Strength Test failed to differentiate between had preexisting conditions that might affect
different IQ levels. test performance, the most significant being
alcohol abuse or a significant traumatic brain
Study strengths injury. The rest of the participants denied any
1. Relatively large sample. history of conditions that might affect brain
2. Sample composition is well described in function. The third group, "mixed normal con-
terms of age, education, gender, IQ, trols," consisted of 125 participants who had
geographic area, and clinical setting. no history of trauma or disease involving the
3. Adequate exclusion criteria. brain. They were enrolled in longitudinal re-
4. Means and SDs for the test scores are search projects at multiple sites under the
reported. supervision of the neuropsychology laborato-
5. Data are presented for males and ries at the University of Colorado and the
females separately. University of California at San Diego. Their
456 MOTOR FUNCTIONS
mean age was 43.6 (19.6) years and. mean hand performance asymmetries. The sample
education was 12.0 (3.3) years; 68% of the included 30 right-handed and 30 left-handed
sample were males, and the test-retest inter- healthy volunteers aged 21-57 years, with a
val was 5.4 (2.5) months. The data ·are re- mean age of 37 (9) years, recruited primarily
ported for all groups combined. Demographic from hospital staff. The two groups contained
information for all groups combined is also equal numbers of men and women. Exclusion
provided. The mean WAIS FSIQ (Wechsler, criteria were history of brain injury or any
1955) on the initial testing for the three medical condition expected to affect perfor-
groups combined was 108.8 (12.3). mance on the study tasks. All participants had
The test was administered accordinc to the completed high school and most had com-
procedures specified by Reitan and Wolfson pleted at least 4 years of college.
(1993). The scores represent average eumber Handedness was assessed using the Annett
of kilograms across two trials for the ddminant (1970) Handedness Inventory, modified by
and nondominant hands. Briggs and Nebes, and the Edinburgh Hand-
The authors provide raw scores for perfor- edness Inventory. Grip strength was measured
mance at two time probes, as well as ovarious with a dynamometer, using three trials with
measures of test-retest reliability and ·magni- each hand.
tude of practice effect. The test-ret~st reli-
ability was r = 0.90 for the dominant h$11d and Study strengths
r=0.91 for the nondominant hand. · 1. Adequate sample size.
2. Sample composition is described in
Study strengths terms of age, education, and recruitment
1. Large sample sizes for the three kroups. procedures.
2. Sample composition is well desc~bed in 3. Adequate exclusion criteria.
terms of age, education, gender, IQ, 4. Means and SDs for the test scores are
geographic area, and setting. reported.
3. Test administration procedures ate spec- 5. Method for determining handedness is
ified. specified.
4. Means and SDs for the test sc~es are 6. Test administration procedures are spec-
reported. ified.
5. Information on test-retest reliability is
provided.
1. Gender distribution within each group is
Considerations regarding use of the stfldy
not provided.
1. Exclusion criteria are not clearly de-
scribed. As the authors pointed out, 20%
of friend controls and 46% of trauma
controls had preexisting conditioos that [0.22] Christensen, Mackinnon, Korten,
might affect test performance, the most and Jorm, 2001 (Table A22.23)
significant being alcohol abuse and a The authors evaluated the "common cause
significant traumatic brain injury. hypothesis" in a longitudinal study of cogni-
2. Data are not partitioned by age group. tive aging on a large probability sample of
3. Data for males and females are not healthy participants aged 70 years and older
presented separately. drawn from an electoral roll in Australia. Ex-
4. No information regarding how hand clusion criteria were not described; a refer-
preference was determined. ence was made to earlier studies.
The dominant hand was used first, after a
[0.21] Triggs, Calvanio, Levine, Heaton, practice trial. Participants swapped hands after
and Heilman, 2000 (Table A22.22) two trials and then repeated the procedure.
Grip strength was assessed in a study exaprlning The measure of grip strength is presented as
the relationship between hand preference and the mean of the scores for two hands averaged
over four trials, for four age groups, for males 454 participants for females, dominant hand;
and females separately. and four studies, which generated 10 data
The authors underscored the effect of age points, based on a total of 407 participants for
on grip strength after controlling for all other females, nondominant hand.
variables. It should be pointed out that the integrity of
the results is undermined by the lack of con-
Study strengths sistency in data reporting. A majority of studies
1. Large sample. report data for the "dominant hand" and
2. Sample composition is described in terms "nondominant hand," while some report for
of age, education, gender, geographic the "right hand" and "left hand." Some of the
area, and recruitment procedures. latter studies include left hand-dominant
3. Test administration procedures are thor- participants in their samples. Though the
oughly described. percent of left-banders is typically small
4. Data are stratified into four age groups. (1%-7.5%), their inclusion confounds the
5. Means and SDs for the test scores are outcome. Also, determination of the dominant
reported. hand was based on a wide range of criteria,
ranging from comprehensive questionnaires
Consideration regarding use of the study to self-report of the writing hand.
1. Procedures used to determine hand Quadratic regressions of the scores on age
preference are not described. were used for female and linear regressions
for male data. R2 ranged from 0.630 to 0.833.
Other comments Based on the derived models, we estimated
1. Data were collected in Australia. scores for age intervals between 25 and 69
years. If predicted scores are needed for age
ranges outside the reported age boundaries,
with proper caution (see Chapter 3) they can
be calculated using the regression equations
THE HAND DYNAMOMETER TEST DATA
included in the tables, which underlie calcu-
(See Appendix 22m)
lations of the predicted scores.
Data collected from the studies reviewed in this Regressions of SDs for Hand Dynamome-
chapter were combined in regression analyses in ter scores on age suggest that age does not
order to describe the relationship between age account for a significant amount of variability
and test performance and to predict expected in SDs (R2 ranges between 0.076 and 0.229).
test scores for different age groups. Effects of Though some increase in variability with ad-
other demographic variables were explored in vancing age is expected, this trend was not
follow-up analyses. The general procedures for present in the collected data. Therefore, we
data selection and analysis are described in suggest that the mean SDs for the aggregate
Chapter 3. Detailed results of the meta-analysis sample be used across all age groups.
and predicted test scores across adult age Education did not contribute to the test
groups are provided in Appendix 22m. scores in the data available for analyses.
Only those studies that stratifY the results
by gender were used in the prediction anal- Strengths of the analyses
yses. After initial data editing for consistency 1. Total sample sizes of 407-713 participants.
and for outlying scores, the following data 2. R2 of 0.833 for females, nondominant
were included in the analyses: nine studies, hand, indicates a good model fit. How-
which generated 15 data points, based on a ever, the number of data points in this
total of 713 participants for males, dominant analysis is only 10, which might result in
hand; seven studies, which generated 13 data somewhat inflated R2 values.
points, based on a total of 641 participants for 3. Postestimation tests for parameter spec-
males, nondominant hand; five studies, which ifications did not indicate problems with
generated 11 data points, based on a total of normality or homoscedasticity.
458 MOTOR FUNCTIONS
4. Differences in mean predicted scores for reported and supported by the meta-analyses
the dominant vs. nondominant hands are described in this chapter. Although high sen-
2.39 for males (48.10 vs. 45.71) and 3.25 sitivity of this test to brain impairment is well
(30.81 vs. 27.56) for females, which, in documented, interpretation of test results
case of females, is consistent with the from the perspective of lateralization of brain
guideline of a 10% preferred-hand su- damage should be made with caution. A 10%
periority. The intermanual difference for dominant hand superiority criterion is clearly
males is approximately 5%. consistent with the average performance
5. Differences between males and fe- across the studies presented above. However,
males in mean predicted scores are 17.29 a wide range of individual differences docu-
(48.10 vs. 30.81) for the dominant hand mented in numerous studies warrants great
and 18.15 (45.71 vs. 27.56) for the non- caution in the interpretation of dominant/non-
dominant hand in favor of males, which dominant hand comparisons. In addition to
is consistent with the strong effect of high variability of intermanual differences,
gender on test performance described in peripheral dysfunction might influence test
the literature. performance (e.g., arthritis, hand or arm or-
thopedic problems, etc.). Unacceptably high
Limitations of the analyses false-positive misclassification rates using
I. R2 of 0.747 and 0.630 for the dominant standard criteria for lateralized impairment
and nondominant hands for males and of warrant further research directed at revision
0.673 for females, dominant hand, are of these criteria.
acceptable. However, these values indi- Some aspects of grip strength variability
cate that only 63%-75% of the variance in have not received sufficient attention in the
test scores is accounted for by the models. literature. For example, normative data for
2. Number of data points for females is older age groups are scarce. Test-retest con-
small. cordance should be further explored to assess
the magnitude of any practice effect and to
address the issue of test reliability over dif-
ferent interprobe intervals. Since interpreta-
CONCLUSIONS
tion of the results is based on assumptions of
A review of the above studies suggests high cerebral lateralization, it is of the utmost im-
consistency in the data across different re- portance to report the criteria for assessment
ports. Pronounced gender differences, with of handedness, cutoff scores for subject se-
males outperforming females, represent an lection on the basis of handedness pattern,
unequivocal finding. Decline in grip strength and number of left-handed individuals in the
associated with advancing age is also frequently sample, if they are included.
23
Grooved Pegboard Test
BRIEF HISTORY OF THE TEST b. 1he subject is allowed to place a cer-

tain number of pegs (the number
1he Grooved Pegboard Test (GPT) acquired its varies for different studies) prior to
popularity over 30 years ago as part of two neu- the actual trial, as practice.
ropsychological batteries. It consists of a metal 2. Beginning of timing:
board with a mabix of slotted holes angled in a. An examiner starts timing when he or
different directions. 1he task is to insert 25 metal she cues the subject to start working
pegs with ridges along the sides into each hole in on the test.
sequence. A further description of the test, its b. Timing starts as the subject drops the
applications, and references to the original sour- first peg into a slot.
ces are provided in Lezak (1995) and Lezak et al. 3. Number of trials:
(2004). Administration and scoring instructions a. According to the test manual, one trial
are also provided by Lafayette Instrument Com- is administered per hand, starting with
pany, which manufactures the pegboard. the dominant hand.
Scores represent time in seconds required b. Two or more trials are administered
to complete the matrix with each hand, with per hand, alternating dominant and
higher scores reflecting lower levels of per- nondominant hands to counteract the
formance. Russell and Starkey (1993) propose practice effect which confounds per-
a limit of 180 seconds, after which the trial is formance with the nondominant
discontinued. According to their modification, hand; the score is the mean of all trials
the number of pegs not placed within the time for each hand.
limit is prorated into the time score. c. Two trials are administered per
Although instructions for the test adminis- hand in a switchback design, where
tration are relatively simple, review of the the first trial is performed with the
literature suggests that there is considerable dominant hand, the next two trials
variability in the following aspects of admin- with the nondominant hand, and
istration and scoring: the last trial with the dominant hand;
the score is the mean of all trials for
1. Administration of practice trials: each hand.
a. 1he trial starts after instructions are 4. Assessment of laterality: This test is
given to the subject; no opportunity known as a sensitive measure of lateralized
for practice is offered. brain damage. As such, accurate identification
459
460 MOTOR FUNCTIONS
of handedness is of utmost importance. The a "rigid application of 'cookbook' formulas or

majority of studies do not provide a description 'rules of thumb' "(p. 723) in test interpretation.
of laterality assessment. It is based in most Schmidt et al. (2000) studied the transfer of
cases on participants' self-report of the hand training between hands by counterbalancing
preferred for writing. the order of the starting hand in right-banders
vs. left-banders. The authors found an effect
Unfortunately, precise test administration of opposite-hand training only in left-handed
procedures are not clearly described by the men, which they attributed to a larger corpus
majority of authors, which hampers the com- callosum in left-handed men.
parability of the norms generated from dif- Reliability of the GPr has been addressed in
ferent studies. several studies. For instance, Ruff and Parker
The GP'f measures psychomotor s~d. fine (1993) report test-retest reliability coefficients
motor control, and rapid visual-motor coordi- of0.69-0.76 for the dominant hand and 0.68-
nation. Motor abilities measured by this test are 0. 78 for the nondominant hand over a 6-month
more complex than those measured: by the period.
Finger Tapping Test and the Hand Dynamom- For further information on the psychomet-
eter. Essentially, the GPr is a cognitive-motor ric properties of the GPr, see Lezak et al.
task. In contrast, the finger tapping and dyna- (2004) and Spreen and Strauss (1998).
mometer tasks require less task-specific cogni-
tive effort and concentration and can be
performed passively. Performance on the GPr RELATIONSHIP BETWEEN GPT
is also highly dependent on psychomotor speed PERFORMANCE AND DEMOGRAPHIC
(Miller et al., 1990; Lezak, 1995; Lezak et al., FACTORS
2004). Axelrod and Milner (1997) fo,nd the
The effect of age on GPrperformance was quite
GPr to be a good predictor of low psychomotor
pronounced across different studies, with slow-
speed in veterans of Operation Desert Storm
ing associated with advancing age (Bornstein,
and Operation Desert Shield who dJsplayed
1985: Concha et al., 1995: Heaton et al., 1991,
cognitive problems. Harnadek and ,Rourke
2004; Ruff & Parker, 1993; Ryan et al., 1987;
(1994) report its sensitivity to nonverbal )earning
Seines et al., 1991). The effects of education and
disability. However, most commonly the GP'f is
gender have been reported but are much weaker
used for assessment of lateralized cerebral dys-
(Bomstein, 1985; Concha et al., 1995; Heaton
function. Use of the original cutoff scores for
et al., 1991, 2004; Ryan et al., 1987; Seines et al.,
impairment (Heaton et al., 1986) and inter-
1991). Polubinski and Melamed (1986) and
manual differences of about 10% in determi-
Schmidt et al. (2000) found that females per-
nation of brain dysfunction yields high rates of
formed faster than males, and Thompson et al.
false-positive misclassification (Bomstein et al.,
(1987) report greater intermanual differences
1987b). Revised cutoffs have been proposed
for females compared to males. However,
(Bomstein et al., 1987b; Ryan et al., 1987).
Strenge et al. (2002) did not find an effect of age
Bomstein (1986c) evaluated the pattern of
or gender on GPr performance in their sample
motor performance on three motor tests (Fin-
of students 19--30 years old, which is probably
ger Tapping Test, GPr, and Hand Dynamom-
due to the restricted age range of the sample.
eter), which were administered to normal and
Ryan and colleagues (1987) proposed a regres-
unilateral brain lesion samples. A large degree
sion equation to control for the effects ofage and
of variability was observed across these inter-
education on GP'f performance.
manual measures, whereby "a high percentage
(approximately 25 percent) of the normal sam-
ple obtained scores more than one standard
deviation from the control mean on a single
NORMATIVE REPORTS
measure" (p. 719). Thus, the author bas em-
phasized the importance of consistency in per- To adequately evaluate the GPr normative
formance pattern across tasks, rather than use of reports, seven criterion variables were deemed
GROOVED PEGBOARD TEST 461
critical. The first five of these relate to subject this assessment, assumptions regarding func-
variables, and the remaining two refer to tionallateralization cannot be made.
procedural issues.
Subject Variables
Description of Administration Procedures
Sample Size
Administration procedures for the GPT differ
Fifty cases are considered a desirable sample among studies. Detailed description of the
size. Although this criterion is somewhat ar- procedures allows selection of the most ap-
bitnuy, a large number of studies suggest that propriate norms or corrections to account for
data based on small sample sizes are highly deviations in administration procedures.
not provide a reliable estimate of the popu- Data Reporting
lation mean. Group means and standard deviations for the
number of seconds required to complete the
matrix for the dominant and nondominant hand
Information regarding medical and psychiatric should be presented at minimum.
status, occupation, ethnicity, handedness, or
recruitment procedures are relevant. Until SUMMARY OF THE STATUS
determined, it is best that this information be OF THE NORMS
provided.
A number of studies have reported normative
Age Group Intervals data for the GPT. Studies vary in subject se-
lection, description of procedural and subject
variables, and grouping of data into categories.
In addition to normative studies based on "nor-
relevant for this test since a strong effect of
mal" samples, there are a number of clinical
age on GPT performance has been demon-
comparison studies that explore differences
strated in the literature.
in GPT performance between clinical groups
Reporting of Educational levels
and "normal control" groups (which are some-
times matched on demographic characteris-
Given the association between education and tics). Unfortunately, normal control groups are
GPT performance, information regarding ed- frequently comprised of medical or psychiat-
ucational level should be reported for each ric patients. These samples cannot be consid-
subgroup, and preferably normative data ered truly "normal" due to possible effects of
should be presented by educational levels. their illnesses and medications on test perfor-
mance.
Reporting of Gender Composition
The majority of studies present the data
Given the possible association between gender in number of seconds required to complete
and GPT performance, information regarding the matrix with each hand. Several studies
gender composition should be reported for report the proportion of participants falling in
each subgroup, and preferably normative data the impaired range or rates of intermanual
should be presented for males and females difference.
separately. Several authors stratify their samples by
age, education, and/or gender. Procedures for
Description of Hand Preference Assessment the assessment of handedness are described in
To address the issue of lateralization in test some studies. Furthermore, some authors di-
performance, assessment procedures for hand vide their samples into groups based on a
preference should be fully described. Without subject's handedness pattern.
462 MOTOR FUNCTIONS
The majority of studies include. mostly 4. Age ranges were not provided. It is dif-
young and middle-aged participants.; Only a ficult to extrapolate age limits for use of
few studies present data for elderly ~dividu the presented norms.
als. Several studies provide test-retest data 5. Sample size is small.
over varying interprobe intervals. Some stud-
ies provide data for left-handed samples. [GPT.2] Bornstein, 1985 (Tables A23.3, A23.4)
Among all the studies available in! the lit- The author collected data on 365 Canadian
erature, we selected for review those based on individuals (178 males, 187 females) recruited
I
well-defined samples or that offer some in- through posted notices on college campuses
formation not routinely reported. and unemployment offices, newspaper ads,
In this chapter, normative publicatjpns and and senior-citizen groups. Participants were
control data from clinical studies are reviewed paid; ranged in age from 18 to 69, with a mean
in ascending chronological order. Th~ text of age of 43.3 (17.1) years; and had completed
study descriptions contains references to the 5-20 years of education, with a mean of 12.3
corresponding tables identified by nutnber in (2.7); also, 91.5% of the sample were right-
Appendix 23. Table A23.1, the locat<k table, handed. No other demographic data or ex-
summarizes information provided in dae stud- clusion criteria are reported.
ies described in this chapter. 1 Means and SDs are reported for each hand.
The sample is stratified by age group (20-39,
40--59, 60--69), level of education (<high
school, ~high school), and gender. Normative
SUMMARIES OF THE STUDIH; data are presented for the preferred and
nonpreferred hands for each demographic
[GPT.1] Rounsavile, Jones, Novelly, an4
group separately, as well as for different
Kleber, 1982 (Table A23.2)
combinations of demographic strata. Individ-
The study compared GPT performance of ual sample sizes range 13-86. Hand prefer-
opiate addicts, epilepsy patients, and controls. ence was determined as the hand used for
A sample of 29 Comprehensive Emp~yment signing the consent form. Participants were
Training Act (CETA) workers was used for a recruited from the general population of a
normal comparison group. Participants with a large city in western Canada.
history of drug or alcohol abuse or a neuro-
logical disorder were excluded. Participants Study strengths
were screened for alcohol and illicit psycho- 1. Very large overall sample size.
active substances, and urine specime's were 2. Stratification of the data by age, gender,
taken at the time of testing. and educational level.
3. This study is unique in that it reports
Study strengths data for participants with less than a high
1. Sample is described in terms of gender, school education.
age, education, and percent of right- 4. Information on recruitment procedures
handed participants. and geographic area is provided.
2. Adequate exclusion criteria. 5. Method for determining handedness is
specified.
1. SDs were not provided. reported.
2. Testing procedure was scarcely de-
scribed. Considerations regarding use of the study
3. Procedures for assessment of hand 1. Individual sample sizes of some cells
dominance were not described. are small. It is unclear whether the youn-
gest age included in the study was 18
'Norms for children are available in Baron (2004) and or 20.
Spreen and Strauss (1998). 2. No reported exclusion criteria.
Other comments Testing was conducted by trained techni-

1. It has been established in several studies cians, and all participants were judged to have
that performance on GPf varies as a expended their best effort on the task.
function of age; however, it has only a The chapter reviews different studies that
weak relationship with education and explore the relationship of neuropsychological
gender. Therefore, norms broken down test performance with age, education, and gen-
by age group are most appropriate for use. der. The authors concluded that different sets
2. Data were collected in Canada. of norms shou1d be used for participants of
different ages, educational levels, and genders
[GPT.3] Heaton, Nelson, Thompson, Burks, when determining whether performance is
and franklin, 1985 (Table A23.5) normal or abnormal.
tiple sclerosis patients and normal control Study strengths
participants recruited in Colorado. The con- 1. Large overall sample size and sizes of
trol group included 100 participants with no individual cells.
history of neurological illness, significant head 2. Information regarding age, education,
trauma, or substance abuse. gender, handedness, and geographic area
is provided.
Study strengths
4. Data are grouped by age and educational
1. Control sample is large.
level.
gender, and geographic area is reported.
3. Exclusion criteria are adequate.
1. SDs are not provided, which limits utility
of the norms.
reported.
2. Procedures for assessment of hand
dominance are not described.
Considerations regarding use of the study 3. Age groupings are quite large in terms of
1. Data are provided as total for both hands ranges.
only.
[GPT.5] Bornstein, 1986a
3. No information regarding handedness.
(Tables A23.7, A23.8)
4. High educational level of control
participants. This report expands the analysis of the data
provided in Bomstein (1985). The author ex-
[GPT.4] Heaton, Grant, and Matthews, 1986 amined cutoff levels for impairment and the
(Table A23.6) proportion of participants falling in the im-
paired range. For both preferred and non-
The authors obtained data on 553 normal
preferred hands, the clinically employed cutoff
controls in Colorado, California, and Wis-
criterion was 66 seconds. Performance time
consin as part of an investigation into the ef-
fects of age, education, and gender on >66 seconds placed participants into the im-
paired range. The high proportion of impaired
Halstead-Reitan Battery performance. The
scores is viewed by the authors as suggesting
sample consisted of 356 males and 197 fe-
caution in using standard cutoff scores. Base
males. Exclusion criteria were history of neu-
rate issues are discussed from the perspective
rological illness, significant head trauma, and
of the validity of test interpretation.
substance abuse. Participants were aged 15-
81 years, with a mean of 39.3 (17.5), and had The administration and scoring were as
follows:
education ranging 0-20 years, with a mean of
13.3 (3.4); 7.2% were left-handed. The sample The score for the Grooved Pegboard Test was
was divided into three age groups and three the time required to fill the board according to
education groups. standard instructions, described in a privately
464 MOTOR FUNCTIONS
published manual developed by Matth~. The 4. Relatively large sample for restricted
preferred hand trial was administered Jrst, and age, education, and handedness groups.
timing of the trial was not interrupted in fhe event 5. Means and SDs for the test scores are
of a dropped peg. (p. 414) : reported.
The sample was stratified by age grqup (18-
39, 40-59, 60-69), level of educatioq (<high Consideration regarding use of the study
school, ~high school), and gender. A~propor 1. No exclusion criteria.
tion of participants obtaining scorel' in the
impaired range for each of the abo .e strata [GPT.7] Ryan, Morrow, Bromet, and
was provided for the preferred a.Jt.d non- Parkinson, 1987 (Table A23.10)
preferred hands.
Participants were recruited from fie gen- This report describes the development of
eral population of a large city in Jwestem the Pittsburgh Occupational Exposures Test
Canada. Those with a history of neutological (POET) battery. It explored the factor struc-
or psychiatric illness were excluded. 1 ture of the battery and interrelations of test
For the strengths and considerationt> on use scores with age and education.
of the study, see [GPT.2] Bomstei,, 1985, The article provides norms for 182 blue-
collar workers who do not have a history of
above. In addition, in the curre1 study,
exclusion criteria and test admi stration exposure to industrial toxins, to be used to
procedures are specified. assess the effect of industrial toxins on neu-
I ropsychological functioning in clinic.
All participants were white, native English-
[GPT.6] Polubinski and Melamed, 1986' speaking males who had been employed at a
(Table A23.9) heavy industrial plant in eastern Pennsylvania
for at least 1 year. Participants had no previ-
Participants were 120 students taim:t intro- ous exposure to industrial toxins and no his-
ductory psychology classes. All we~ right- tory of neurological or psychiatric disorder or
handed. The Crovitz-Zener Test (Ciovitz & renal or hepatic disease; in addition, they had
Zener, 1962) was used to assess dtfgree of restrained from alcohol consumption in the
hand dominance based on consistencylin hand 12 hours prior to testing.
preference for five unimanual tasks. j Partici- Since age and education were highly re-
pants with a score of 25 on this test foqned the lated to test performance, the authors devel-
firm right-handed groups, while thQse with oped a linear regression procedure that
scores of 524 formed the mixed righ~handed controls for the confounding effect of these
groups. l
variables. The prediction of a test score for
A switchback design was used, in which the each individual is based on the following
first and fourth trials were performed ~th the equations:
right hand and the second and thitd trials
were performed with the left hand.
The authors found that women petformed Dominant hand predicted score =
faster than men and mixed right-banders 71.233 + 0.301 (age)- 0.904 (education)
performed faster than firm right-han<frs. Nondominant hand predicted score=
85.929 + 0.151 (age) -1.347 (education)
Study strengths
1. Assessment of handedness is well The authors also reported the percent pre-
identified. dicted ratio score (ratio of the actual score to
2. Test administration procedure. is de- the predicted score x 100) that falls at or be-
scribed. : low the fifth centile for this population. The
3. Sample composition is descrlbed in cutoff values for impairment are as follows:
terms of age, education, gend~. occu- 130 for the dominant hand and 128 for the
pation, and geographic area. nondominant hand.
Study strengths whether the authors used dominant/

1. Sample composition is described in nondominant comparisons or right/left
terms of gender, occupation, education, comparisons.
ethnicity, and geographic area. 4. Information on educational level is not
2. Sample size is large, and most individual reported.
cell sizes approach 50. 5. Data are collapsed across genders.
3. Testing procedure is well described.
4. Sample is divided into four age groups. [GPT.9] Thompson, Heaton, Matthews,
5. Means and SDs for the test scores are and Grant, 1987 (Table A23.12)
reported. The article presents a percentage of 426 nor-
6. Adequate exclusion criteria. mal participants (279 males, 147 females)
scoring in the lateralized lesion range using
Considerations regarding use of the study Golden's (1978) guidelines. Dominant hemi-
1. Procedures for assessment of hand sphere dysfunction was defined as superior-
dominance are not described. ity of nonpreferred hand performance over
2. All-male sample. preferred hand performance. Nondominant
hemisphere dysfunction was identified when
[GPT.8] Bornstein, Baker, and Douglass, preferred hand performance was at least 20%
1987a (Table A23.11) better than nonpreferred hand performance.
The study assessed test-retest reliability of the Lateral preference type was assessed based
CPT over a period of 3 weeks. Participants on performance on the Reitan-Klove Lateral
were 14 women and nine men recruited from Dominance Exam and the Miles ABC Test of
a university community without a positive Ocular Dominance (Reitan & Wolfson, 1985).
history of neurological or psychiatric illness. The following groups were identified:
Their ages ranged 17-52, with a mean of 32.3
(10.3) years; mean VIQ was 105.8 (10.8), and 1. All right-participants who wrote with
mean PIQ was 105.0 (10.5). their right hand and manifested right
Participants were administered the Halstead- lateral preference on all hand, eye, and
Reitan Battery in standard order both on initial foot measures.
testing and again 3 weeks later. Means and SDs 2. Mixed right-participants who wrote
for raw score change over 3 weeks are -2.8 with their right hand but manifested left
(6.1) and 0.3 (6.4) for the right and left hands, preference on one or more other hand,
respectively. eye, or foot measures.
Data for the whole sample for both testing 3. Left-left-handed participants.
probes are provided.
Intermanual percent difference scores
Study strengths were calculated as preferred hand minus non-
1. Sample composition is described in preferred hand performance divided by pre-
terms of age, VIQ, PIQ, and gender. ferred hand.
2. Information on short-term (3-week) re- Participants' mean age was 40.59 (18.27)
test data is provided. years, and mean education was 13.15 (3.49)
3. Minimally adequate exclusion criteria. years. They had been screened for history of
4. Means and SDs for the test scores are head trauma, neurological illness, substance
reported. abuse, serious psychiatric illness, and periph-
eral injuries that might affect test performance.
Considerations regarding use of the study The authors concluded that females tend to
1. Sample size is small. show greater disparity between preferred and
2. Age range is wide; the effect of age on nonpreferred hand performance (M = 9.8%,
test-retest change is not explored. SD = 11.8) than males (M =6.7%, SD = 12.1).
3. Procedures for assessment of hand dom- Age and education were not significantly re-
inance are not described. It is unclear lated to intermanual difference scores.
466 MOTOR FUNCTIONS
Study strengths 2. The ~ 12 years of education category is

1. Large sample size. too large.
2. Lateral preference was thoroughly as- 3. Data are collapsed across genders.
sessed, and three groups were identified.
3. Intermanual differences and a percent- Other comments
age of participants scoring in the later- 1. Data were collected in Canada.
alized lesion range are reported.
4. Adequate exclusion criteria. [GPT.11] Miller, Seines, McArthur, Satz,
5. Information on age, education, and gen- Becker, Cohen, Sheridan, Machado, Van Gorp,
der is reported. and Visscher, 1990 (Tables A23.14, A23.15)
Considerations regarding use of the study The article provides data for homosexual!
1. Means and SDs for each group are not bisexual males recruited in the Multi-Center
reported. AIDS Cohort Study (MACS), an epidemio-
2. Data are presented for a wide age range logical project designed to assess the natural
not separated into age groups, which history of HIV-1 infection. The study uses
precludes consideration of the effect of large sample sizes to explore the effect of HIV
age on intermanual differences. serostatus and symptom status on cognitive
and motor functioning. Handedness was es-
[GPT.10] Bornstein and Suga, 1988 tablished based on self-report. The test ad-
(Table A23.13) ministration procedure followed that outlined
The authors report data on 134 healthy older by Lezak (1983).
Canadian volunteers aged 55-70, who were The paper reports the percentage of right-
paid for their participation. Participants rep- handed, ambidextrous, and left-handed
resent a subset of those used in Bomstein individuals; race composition (white, black,
(1985). The sample is partitioned into three Hispanic, other); CES Depression Scale
education groups. Nearly two-thirds of scores (with SDs); and CD4 cell!mm3 count
the sample are female (n = 85). The average (with SDs).
age for the sample is 62.7 (4.3) years, and the
mean ages of the three education groups Study strengths
are comparable. Exclusion criteria were history 1. Sample sizes are large.
of neurological or psychiatric disorder. 2. The demographic characteristics of each
sample are meticulously described in
Study strengths terms of gender, sexual orientation,
1. Large overall sample size and adequate handedness, ethnicity, age, education,
individual cell sizes. and geographic area.
2. Data are partitioned into three educa- 3. Method for determining handedness is
tion groups; the study is unique in described.
terms of representation of participants 4. Means and SDs for the test scores are
with <12 years of education. reported.
3. Information regarding gender, age, and 5. Test administration procedures are
geographic area is provided. reported.
reported. Considerations regarding use of the study
5. Minimally adequate exclusion criteria. 1. The study recruited participants aged
6. Reasonably restricted age grouping. 21-72. Data are presented for all ages
combined.
Considerations regarding use of the study 2. No exclusion criteria reported.
1. Procedure for determination of hand 3. All-male sample.
preference is not identified. 4. High educational level.
[GPT.12] Heaton, Grant, and Matthews, 1991 For the sample as a whole, mean time in
seconds was 67.3 (16.1) for the dominant hand,
The authors provide normative data from 486 and 72.3 (17.5) for the nondominant hand.
(378 base sample, 108 validation sample) ur- The interested reader is referred to the
ban and rural participants recruited in sev- Fastenau and Adams (1996) critique of the
eral states (California, Washington, Colorado, Heaton et al. (1991) norms, and Heaton et al.'s
Texas, Oklahoma, Wisconsin, Illinois, Michi- (1996a) response to this critique.
gan, New York, Virginia, and Massachusetts) In 2004, the authors published revised
and Canada. Data were collected over a 15- norms, which are based on a sample of over
year period through multicenter collaborative 1,000 normal adults. In addition to age, edu-
efforts. cation, and gender stratification, the data are
Sixty-five percent of the sample were males. partitioned by race/ethnicity (African Ameri-
Mean age for the total sample was 42.06 can and Caucasian).
(16.8), and mean educational level was 13.6
(3.5). Mean full-scale IQ, VIQ, and PIQ were Study strengths
113.8 (12.3), 113.9 (13.8), and 111.9 (11.6), 1. Large sample size.
respectively. Exclusion criteria were history of 2. Comprehensive exclusion criteria.
learning disability, neurological disease, ill- 3. Detailed description of demographic
nesses affecting brain function, significant characteristics in terms of age, educa-
head trauma, significant psychiatric distur- tion, IQ, geographic area, and gender.
bance (e.g., schizophrenia), and alcohol or 4. Administration procedures are outlined.
other substance abuse. 5. Normative data are presented in com-
The GPT was administered according to prehensive tables in T-score equivalents
procedures provided by the test manufacturer. for males and females separately in
Participants were paid and judged to have 10 age by six education groupings.
provided their best efforts on the tasks. Time
in seconds to complete the 25-peg placement Consideration regarding use of the study
with each hand separately is reported. 1. No information regarding how hand
The normative data, which are not re- preference was determined.
produced here, are presented in comprehen-
sive tables in T-score equivalents for scaled [GPT.13] Seines, Jacobson, Machado, Becker,
scores for males and females separately in Wesch, Miller, Visscher, & McArthur, 1991
10 age groupings (20-34, 35-39, 40-44, 45- (Tables A23.16-A23.18)
49, 50-54, 55-59, 60-64, 65-69, 70-74, 75- The article presents data for seronegative
80) by six education groupings (6--8, 9-11, 12, homosexual and bisexual males from the
13-15, 16-17, ~18 years). MACS, who were earlier described by Miller
For dominant hand performance, 40% of et al. (1990), for the purpose of establishing
the score variance was accounted for by age, normative data for neuropsychological test
while 17% was attributable to educational performance based on a large sample. Partic-
level; gender accounted for a negligible ipants with a history of head injury with loss of
amount of unique variance in performance consciousness >1 hour and who reported
(4%). A total of 47% of test score variance was drinking ~21 drinks per week in the previous
accounted for by demographic variables. For 6 months were excluded. The paper reports
nondominant hand performance, 39% of the percentage of right-handed, ambidextrous,
score variance was accounted for by age, while and left-handed individuals as well as the race
13% was attributable to educational level; composition (Caucasian, African American)
again, gender accounted for a negligible for age and education strata.
amount of unique variance (3%). A total of Handedness was established based on self-
42% of test score variance was accounted for report. Standard procedures described by the
by demographic variables. Lafayette Instrument Company were used.
468 MOTOR FUNCTIONS
The authors point out a significant effect The authors report test-retest reliability for
of age on GPT performance. Education, how- a 6-month interval, based on data for five or
ever, was not significantly related to perfor- more participants from each of 12 demo-
mance. graphic cells (30% of sample). Reliability co-
efficients for women, men, and the total
Study strengths sample were 0.76, 0.69, and 0.72 for the
1. Normative data are stratified by flge and dominant hand and 0.78, 0.68, and 0.74 for
education. ' the nondominant hand, respectively. The ef-
2. The demographic composition · of the fect of age and gender on motor speed was
sample is described in terms Jof age, specifically addressed. The authors explore
gender, sexual orientation, hantdness, the ratio of dominant/nondominant hand per-
ethnicity, and geographic area; demo- formance.
graphic composition is described reach
age and education cell separately. Study strengths
3. Means, SDs, as well as scores t>r per- 1. Sample composition is described in
centiles 5 and 10 are presented. ; terms of age, gender, education, hand-
4. Method for determining handeclness is edness, and geographic area.
reported. : 2. Assessment of handedness is well de-
5. Very large overall sample and individual scribed.
cells. 3. Test administration procedure is de-
6. Administration procedures are specified. scribed.
7. Minimally adequate exclusion cr(teria. 4. Data are presented in gender x
I
education x age groupings.
Considerations regarding use of the s'tfldy 5. Data for a left hand-dominant sample
1. All-male sample. r are reported.
2. High educational level. 6. Means and SDs for the test scores are
reported.
Other comments 7. Adequate exclusion criteria.
The authors provide an update (~rsonal 8. Large overall sample size, although some
communication) on the data reported Jn their cells are relatively small.
1991 paper, reflecting ongoing data coUection
[GPT.15] Russell and Starkey, 1993
and analysis for their longitudinal ep~emio
(Table A23.21)
logical study of HIV infection. They present
the data in a combined age and ec41cation This study describes the standardization
grouping. All data are from healthy HIV- sample used by the authors in their manual
negative gay and bisexual males. · introducing the Halstead-Russell Neuropsy-
chological Evaluation System (HRNES) and
addressing its psychometric properties. The
[GPT.14] Ruff and Parker, 1993 normative sample consisted of veterans trea-
(Tables A23.19, A23.20) ted at the Cincinnati Veterans Administration
The GPT was administered as part of a com- Hospital between 1968 and 1971 and the
prehensive test battery to 360 normal 1volun- Miami V.A. Medical Center between 1971
teers recruited in California, Michig~, and and 1989. All participants received a neuro-
the eastern seaboard, aged 1~70 ye~. with logical examination. Those participants who
education of 7-22 years. Participant$ were were administered the Halstead tests and the
screened for psychiatric hospit~tions, WAIS or WAIS-R were included in the study.
chronic polydrug abuse, and neur.logical Nine percent of the sample were representa-
disorders. ' tives of minority groups.
Data are stratified by education x g$-tder x The total sample was divided into a com-
age. Data for a left hand-dominant saq>le are parison group and a brain-damaged group.
also reported. ' The comparison group included "normal"
individuals. No subject in this group had a this group had no history of trauma or disease
diagnosis of central nervous system pathology. involving the brain. They were enrolled in
Presenting symptoms for the majority of these longitudinal research projects at multiple sites
participants were neurosis with memory or under the supervision of the neuropsychology
somatic complaints or personality disorders laboratories at the University of Colorado and
with episodes of explosive behavior. the UDiversity of California at San Diego.
Patients diagnosed with schizophrenia or Their mean age was 43.6 (19.6) years and
severe depression requiring hospitalization as mean education was 12.0 (3.3) years; 68% of
well as those with evidence of systemic vas- the sample were males, and the test-retest
cular disease were not included in the sample. interval was 5.4 (2.5) months. The other two
Test scores can be corrected for age and IQ groups do not have data for the GPT and
and converted into scaled scores to facilitate therefore will not be described in this chap-
comparison with other tests. Statistics are ter. The mean WAIS FSIQ (Wechsler, 1955)
reported for four groups of patients: compar- on the initial testing for the three groups
ison, left hemisphere damage, right hemi- combined was 108.8 (12.3).
sphere damage, and diffuse brain damage. The GPT was administered and scored ac-
Data only for the comparison group are recording to the instructions provided by the
produced in this chapter. Lafayette Instrument Company. The scores
The authors published an appendix to the represent the numbers of seconds required to
manual (HRNES-R; Russell & Starkey, 2001), place all 25 pegs on the board, separately for
which contains tables of scale scores based the dominant and nondominant hands.
on the original HRNES norms, demographic The authors provide raw scores for perfor-
corrections, and regression-based predicted mance at two time probes, as well as various
scores. measures of test-retest reliability and the
magnitude of the practice effect. The test-
Study strengths retest reliability for the GPT was r = 0.86 for
1. Sample composition is described in both dominant and nondominant hands.
terms of age, education, gender, ethnic-
ity, and geographic area. Study strengths
2. Control sample size is large. 1. Large sample size.
3. Means and SDs for the test scores are 2. Sample composition is well described in
reported. terms of age, education, gender, IQ,
geographic area, and setting.
Considerations regarding use of the study 3. Test administration procedures are spec-
1. Procedures used to determine hand ified.
preference are not described. 4. Means and SDs for the test scores are
2. Classification of participants in the reported.
comparison group as normal is ques- 5. Information on test-retest reliability is
tionable since they were suspected of provided.
having neurological conditions and re-
ferred for neurological evaluation, which Considerations regarding use of the study
yielded negative results. 1. Exclusion criteria are not clearly de-
3. Undifferentiated age range. scribed.
2. Data are not partitioned by age group.
[GPT.16] Dikmen, Heaton, Grant, and 3. No information regarding how hand
Temkin, 1999 (Table A23.22) preference was determined.
The GPT was used in a study on the psycho-
metric properties of a broad range of neu- [GPT.17] Strenge, Niederberger, and
ropsychological measures based on a sample Seelhorst, 2002 (Table A23.23)
of 125 normal or neurologically stable adults, The GPT was administered as part of a battery
121 of whom were tested twice. Participants in in a study examining the relation between
470 MOTOR FUNCTIONS
tests of manual dexterity and attentional It should be pointed out that the integrity
functions. The sample consisted of 49 right- of the results is undermined by the lack of
handed medical students (26 women, 23 men) consistency in data reporting. A majority of
19-30 years of age, with a median age of studies report data for the "dominant hand"
23 years. Handedness was assessed with the and "nondominant hand," while some report
Annett (1970) inventory. The GPT was ad- for the "right hand" and "left hand." Some of
ministered according to standard instructions. the latter studies include left hand-dominant
The authors controlled for finger size in their participants. Though the percent of left-
analyses. banders is typically small (5%-7.5%), their
No significant effect of age or ge.der on inclusion confounds the outcome. Also, deter-
test performance was found. ~ mination of the dominant hand was based on a
wide range of criteria, ranging from compre-
Study strengths hensive questionnaires to self-report of the
1. Sample size approaches 50. writing hand.
2. Sample composition is desc~ed in linear regressions of GPT scores on age
terms of age, gender, and setting. yeilded R2 of 0.936 and 0.907 for the domi-
3. Test administration proceduJlls are nant and nondominant hands, respectively.
specified. j Based on the derived models, we estimated
4. Means and SDs for the test scqres are GPT scores for age intervals between 20 and
reported. 64 years. If predicted scores are needed for
5. Method for determining handecJness is age ranges outside the reported age bound-
specified. · aries, with proper caution (see Chapter 3)
they can be calculated using the regression
Considerations regarding use of the .J,dy equations included in the tables, which un-
1. Exclusion criteria are not identified. derlie calculations of the predicted scores.
2. Recruitment procedures are not rprted. Regressions of SDs on age yielded R2 of
3. Data were obtained on German [Partici- 0.489 and 0.468 for the two hands, indicating
pants, which may limit their usefulhess for increase in variability with advancing age,
clinical interpretation in the Unite4 States. consistent with the literature. Predicted SDs,
based on these models, are reported.
The effect of education on the test scores
was analyzed on the data set described above
RESULTS OF THE META-ANALYS~ OF
and on a separate data set that included data
THE GROOVED PEGBOARD TEST DATA
broken down by education groups rather than
(See Appendix 23m)
by age groups. Whereas the former analysis
Data collected from the studies reviawed in indicated no effect of age on the test scores,
this chapter were combined in regression anal- the latter analysis revealed a significant effect
yses, to describe the relationship between age of education on test performance. The t value
and test performance and to predict e:1pected for education was-2.71 (p=0.030) for the
test scores for different age groups. Effects of dominant hand and -3.96 (p = 0.005) for the
other demographic variables were explored in nondominant hand. Because of the inconsis-
follow-up analyses. The general proqedures tency in findings from the two data sets and
for data selection and analysis are d~cribed the marginally significant effect for the dom-
in Chapter 3. Detailed results of th~ meta- inant hand, education-correction tables were
analysis and predicted test scores across adult not reported.
age groups for both genders combi~d are
provided in Appendix 23m. Strengths of the analyses
After initial data editing for consisteflcy and 1. Total sample size is 2,382 for both hands.
for outlying scores, six studies, which .gener- 2. R2 of 0.936 for the dominant hand and of
ated 15 data points based on a total ci 2,382 0.907 for the nondominant hand indicate
participants, were included for each hand. a good model fit.
3. Postestimation tests for parameter spec- CONCLUSIONS

ifications did not indicate problems with
Despite the wide variability in administration
homoscedasticity.
procedures for the GPT, there is high con-
4. The difference in the mean predicted
sistency in the data across different studies.
time for the dominant vs. nondominant
Decline in performance is clearly associated
hands is 6.23 seconds (66.63 vs. 72.86),
with advancing age. The effects of education
which is consistent with the guideline of
and, specifically, gender are more equivocal.
a 10% preferred-hand superiority.
Peripheral (orthopedic and muscular) prob-
lems impacting performance on the test must
Limitations of the analyses also be considered. Because GPT perfor-
1. Postestimation tests for parameter spec- mance is affected by several factors, its inter-
ifications indicated positive skew in the pretation as an indicator of cortical
test scores. This non-normality of the dysfunction should be made with great cau-
score distribution does not affect the es- tion. Following the recommendations of
timates of regression coefficients and Bomstein (1986c), diagnostic accuracy rests
accuracy of prediction but does influ- on the consistency of findings across different
ence the results of significance tests. tasks and different functional domains.
2. Educational level of the aggregate sam- Given the unacceptably high false-positive
ple is approximately 13.5 years. Effect of rates with the original cutoffs for impairment,
education on GPT performance has further research should be directed at the
been reported in several studies and was formulation of revised cutoff scores to im-
also evident in our analysis on the addi- prove specificity.
tional data set. Time to completion is Despite the large body of empirical studies
expected to increase as a function of exploring the psychometric properties of the
decrease in educational level. Therefore, GPT, some aspects of performance are not
predicted values are likely to underesti- sufficiently addressed. For example, norma-
mate expected time to completion for tive data for older age groups are scarce.
individuals with lower educational levels. Since interpretation of the results is based on
3. Possible gender differences in time to assumptions of cerebral lateralization, it is of
completion and intermanual differences utmost importance to report the criteria for
in favor of females have been reported in the assessment of handedness, cutoff scores
the literature. However, we were unable for subject selection on the basis of hand-
to examine these differences because the edness pattern, and number of left-handed
data available for review were not strat- individuals in the sample (if they are in-
ified by gender. cluded).
VII
CONCEPT FORMATION
AND REASONING
24
Category Test
BRIEF HISTORY OF THE TEST of different aspects of reasoning (Johnstone

et al., 1997; Kelly et al., 1992; Perrine, 1993;
The Category Test was developed by Halstead Shute & Huertas, 1990). However, Leonber-
(1947) to assess the ability to "abstract" cate- ger et al. (1991) emphasize visual concentra-
gorization parameters such as size, shape, tion and visual memory as important
number, position, brightness, and color. The determinants of Category Test performance,
original test apparatus consisted of a boxed while Boyle (1988) views this test as a measure
screen placed in front of the subject, on which of intelligence.
were presented visual stimuli in groups of Golden and colleagues (1981a) suggest that
four; the task was to identify which of the four the Category Test is sensitive to prefrontal
stimuli differed from the other designs by lobe disturbance as well as diffuse dysfunc-
pressing one of four corresponding keys lo- tion. Adams and colleagues (1993, 1995) ex-
cated on a pad below the screen. Feedback plored the association between Category Test
was provided in the form of a "chime" for performance and frontal tissue glucose me-
correct responses and a "raspberry buzzer" tabolism rates. The authors attribute the test's
for incorrect answers. Halstead's (1947) orig- sensitivity to reasoning, concept formation,
inal test included 336 items organized into and abstraction to involvement of three frontal
nine subtests, while the version employed by subdivisions in information processing: the
Reitan and Wolfson (1985) was reduced to cingulate gyrus and the dorsolateral and or-
208 stimuli in seven subgroups. The Halstead bitomedial aspects of the frontal lobes (Adams
version is no longer in use, and the data pre- et al., 1995). Other authors, however, do not
sented in this chapter refer to the Reitan relate Category Test performance to any
edition. specific brain area (Bomstein, 1986c; Choca
The Category Test involves several different et al., 1997).
abilities, including attention and concentra- Lowered Category Test performance has
tion, learning and memory, and visuospatial been observed in moderate brain injury (Rimel
skills, as well as concept formation, abstraction et al., 1982), multiple sclerosis (but not chronic
of similarities and differences among stimuli, fatigue syndrome; Krupp et al., 1994), and
and modification of problem-solving hypoth- psychiatrically hospitalized male prisoners
eses in responses to feedback. The majority (Young & Justice, 1998). In addition, schizo-
of recent studies on the construct validity phrenic patients show decreased performance
of the Category Test describe it as a measure (Goldstein & Zubin. 1990; Gottschalk & Selin,
475
476 CONCEPT FORMATION AND REASONING
1991), with nondelusional patients show- variability in Category Test error scores was
ing higher scores than delusional patients accounted for by spatial memory, while 15%
(Steindl & Boyle, 1995). Category Test per- was also explained by response time (Rattan
formance is also depressed in alcoholics (Lo- et al., 1986). Interestingly, administering the
berg, 1980; O'Leary et al., 1977), and scores Category Test prior to the Wisconsin Card
significantly correlate with MMPI clinical Sorting Test results in higher scores on the
scales in this population, suggesting a rela- latter; however, administration of the Wis-
tionship between emotional distress and ex- consin Card Sorting Test first is associated
ecutive function in alcoholism (Johnson- with poorer scores on the Category Test
Greene et al., 2002). Of interest, the presence (Franzen et al., 1993).
of antisocial personality disorder in alcoholics Specific test administration instructions are
is associated with higher Category Test per- provided in Reitan's (1979) Manual for Ad-
formance in men compared to women (Hes- ministration of Neuropsychological Test Bat-
selbrock et al., 1985). Category Test scores teries for Adults and Children and in Reitan
have also been significantly correlated with and Wolfson's (1985) The Halstead-Reitan
ventricular size in depressed patients (Kellner Neuropsychological Test Battery. Snow (1987)
et al., 1986), and they are lower in older cites concerns regarding lack of standardiza-
compared to younger bipolar patients (Savard tion of test administration. He points out
et al., 1980). Finally, Category Test perfor- that the test manual dictates that the examiner
mance is suppressed in diabetes (Skenazy & assist the subject, but the nature and extent
Bigler, 1984) and hypertension (Pentz et al., of the help to be offered is not specified.
1979). Several recent studies have indicated He quotes the following passage from the
that the Category Test may show particular manual: "it may become necessary to urge
promise for the detection of malingered cog- [patients] to study the pictures carefully, to
nitive symptoms (DiCarlo et al., 2000; Sweet & ask for . . . description . . . to urge them
King, 2002; Tenhula & Sweet, 1996). to try to notice and remember how the pic-
Of some concern, a low relationship be- tures change . . . and to try to think of the
tween Category Test scores and activities of possible reason when a correct answer occurs"
daily living (as measured by the Scale of (pp. 26--27).
Competence in Independent living Skills) in Halstead (1947) comments that although
a sample of geriatric patients was reported by several methods of scoring were considered,
Searight et al. (1989). The authors report he settled on a single score: total number of
correlation coefficients between Category errors. He recommended a cutoff of >80 er-
Test scores and measures of 16 activities of rors as the criterion score in computing his
daily living ranging between -0.03 and 0.37. impairment index, while Reitan and Wolfson
However, little and colleagues (1996) indi- (1985) used a criterion score of 50 for their
cate that neuropsychological tests, including shortened test. Given the significant associa-
the Category Test, were more predictive of tion between Category Test scores and de-
outcome in their sample of brain-injured pa- mographic factors and IQ, a single cutoff
tients than were intelligence tests. would not appear to be appropriate, particu-
The Category Test has been only modestly larly in older participants. For example, Ernst
correlated with the Wisconsin Card Sorting (1987) documented in his sample (65-75 years
Test (Macinnes et al., 1983a), another mea- old) a misclassification rate of 84% on a
sure of problem-solving/abstraction; and in booklet version of the Category Test. Dodrill
factor analyses, the Category Test has been (1987) documented a 22.5% misclassification
shown to load with general intellectual and rate in a young control sample. Logue and
memory measures (Boyle, 1988) and, among Allen (1971) recommend that "ultimate in-
WAIS subtests, to primarily correlate with terpretation of the significance of critical data
Block Design, Digit Symbol, and Similarities on the Category Test rests often not only on
(Stone et al., 1988). In a study in which the standard cutoff of 50 but also on a sub-
response time was measured, 15% of the jective evaluation by the psychologist as to
CATEGORY TEST 477
whether the score is to be reasonably expected required over 1 hour. Some patients may
by a normal subject of this general level of complete the task in 2 hours or more, and the
intelligence" (p. 1095). Similarly, Bomstein associated fatigue may result in random re-
and colleagues (1987a) emphasize that cutoff sponding. Rest periods alleviate fatigue, but
scores may be useful but only if considered in these breaks may compromise performance
the context of other neuropsychological in- on subtest VII because it involves recall of
formation obtained in a test battery and if age, earlier strategies.
education, and other appropriate adjustments To address the problem of lengthy admin-
are made. istration, at least 10 shortened versions of the
An item analysis reported by Laatsch and Category Test have been developed. Some
Choca (1991) revealed an uneven progression formats have involved administration of three
of item difficulty in successive subtests. In or four of the seven subtests (Calsyn et al.,
addition, all items on subtests I and II were 1980; Moehle et al., 1988), while other for-
found to be too easy to yield useful information mats have used selected items from five or six
(see Choca et al., 1997). Charter (1994) ex- subtests (Boyle, 1975, 1986; Gregory et al.,
amined the frequency of random responding 1979; LaBreche, 1983; Russell & Levy, 1987)
on the Category Test using the formula ap- or have split the test in half using even vs. odd
proximating binomial distribution for a large items (Kilpatrick, 1970). Kilpatrick docu-
sample, based on observed scores and proba- mented high correlations between number of
bility of guessing. Frequencies of random errors on odd items, even items, and all items
responses for the 90%, 95%, and 99% confi- (r=0.90-0.99). Calsyn et al. (1980; cf. Dunn
dence intervals presented in this article are et al., 1985; Golden et al., 1981a; Taylor et al.,
summarized in Table 24.1. Charter's review 1984) found that scores based on administra-
of different studies reported in the literature tion of the first four subtests had high corre-
suggested that no normative participants' lations with the total score (r=0.83--0.89) and
scores fall in the random range. accounted for 77%-79% of total score vari-
ance. They suggested that full test scores can
be approximated by multiplying the shortened
Alternate Formats
version score by 1.4 and adding 14. Inclusion
Several authors have observed that a major of information regarding age, education, and
negative aspect of the Category Test for clin- gender has not increased the predictive ac-
ical use is the extensive amount of time which curacy of this short form (Pierce et al., 1989).
may be required for administration (Golden Taylor and colleagues (1984) corroborated
et al., 1981a). Finlayson and colleagues (1986) a high correlation between the Calsyn short
note that a quarter of their sample of severely version and the complete test format (r = 0.91),
brain-injured patients completed the test in but they noted that use of the short form re-
29 minutes or less; however, a third required sulted in a substantially higher misclassification
30-39 minutes, another third completed the of normals as brain-damaged and that partici-
test in 40-59 minutes, and nearly 10% pants with right-sided focal lesions tended to
be misidentified as normal. Dunn et al. (1985)
suggested that the following equation may lead
to more accurate estimates of full Category
Table 24.1. Frequencies of Random Responses
for 90%, 95%, and 99% Confidence Intervals, Test scores in a geriatric population: (short
According to Charter (1994) form score x 1.6) + 22.
Moehle and colleagues (1988), using mul-
90% 95% 99% tiple regression analyses, reported that a short
Low High Low High Low High form composed of subtests N, VI, and III
accounted for the highest percentage of long-
Full test 146 167 145 169 141 173 form score variance (77%) and is the "psy-
Subtests II, VII 13 19 12 19 12 21
chometrically soundest short form." Their
Subtests III-VI 26 35 26 36 24 38
analyses indicated that the Calsyn short form
accounted for only 62% of full-fo~ score subtests V and VI were reorganized so that
variance. Moehle et al. (1988) recomm~nded a subtest V included only pure quantitative items
cutoff of 26 (which represents the Score 1 and subtest VI consisted of complex counting
standard deviation [SD] below the m~an for items. A full Category Test score was calcu-
their brain-damaged sample) for use with lated by multiplying the short-version score by
their short form. They suggested some ~aution 2.2. A correlation of 0.97 was obtained be-
in the use of their version because !fl their tween the abbreviated and full test scores in a
participants on whom the analyses wer+ based neurological population. Mean Cronbach's oc
were administered the entire Category Test. was 0.71, suggesting moderate item homoge-
The authors questioned whether there are un- neity across subtests (Boyle et al., 1994).
1
known order effects or if some subte!¢5 (i.e., LaBreche (1983) deleted memory items
I and II) have an influence on sub~quent (subtest VII), discarded the formal scoring
subtest performance. : of subtests I and II (because they simply ori-
Other investigators have retained ~t least ent the patient to the task), and took out re-
five or six of the seven subtests from te Re- dundant items, resulting in 81 items, which
itan version of the Category Test bub short- he called the "Victoria Revision." Correlation
ened some of the subtests (Boyle, 197ft 1986; with the full form was 0.96, and the revised
Gregory et al., 1979; Russell & Levy, j1987). test produced a classification rate of 84%,
Taylor et al. (1990) argue that "becaqse the which was comparable for the 83% obtained
Category Test is designed to be a test! of ab- for the full version. Sherrill (1985) concluded
stract reasoning and requires the su~ect to that "having the smallest number of scored
make conceptual shifts among several princi- items of any of the attractive short forms, the
ples, the same number of principles apd the Victoria Revision probably offers the best
same number of conceptual shifts sh~d be combination of relatively short administration
required on a shortened version of thq Cate- time and relatively good predictive accuracy
gory Test" (p. 486). ' for routine clinical use" (p. 350). Kozel and
Boyle (1975, 1986) created two ~arallel Meyers (1998) reported that short-form and
84-item forms of the Category Test, inyolving full-form scores obtained on matched groups
half the items from subtests I-IV and \!II and of head-injured and dementia patients did not
20 items from subtests V and VI. The 84-item
I
significantly differ.
version discriminated between a norn:+U and Finally, Wetzel and Boll (1987) published
neurological population, and using cu.Pffs of the Short Category Test, which contains 100
38 and 39 errors, only 6%-13% of the, brain- Category Test items (the first 20 items from
damaged participants were miscla$sified, subtests II-VI). Subtest I was eliminated be-
while 20%-22% of the non-brain-daJnaged cause the first 12 items in subtest II were
participants were misidentified. , judged to be adequate to introduce patients to
Gregory et al. (1979) developed a 12J)-item the task. Subtest VII was deleted because it
test version employing all subtest I it~'s, the repeated items from previous subtests and
first 16 items from subtest II, and first was thought to tap memory more than ab-
32 items from subtests III-VI. While . egory straction ability. This shortened form was ad-
and colleagues suggest that a cutoff s<i>re of ministered to a clinical sample, and the same
35 best corresponded to the full test ;cutoff pattern of correlations as observed between
of 51 in their brain-damaged participapts vs. the full Category Test and various neu-
college students, Sherrill (1985) found ,that a ropsychological measures was found for the
cutoff of 29 was a better predictor of th~ long- Short Category Test, with the exception that
form cutoff in his heterogeneous, neuropsy- the Short Category Test was more strongly
chiatric population. , associated with dominant finger tapping
Russell and Levy (1987) devised a 9~-item speed. Normative data were derived from 120
Category Test format composed of fivej items control participants, who averaged 15 years of
from subtest I, 10 items from subtest •· and education. Correlations between the Short
20 items from subtests III-VI. Item~ from Category Test and full test ranged 0.80--0.93,
i
CATEGORY TEST 479
depending on which test was administered subtests (i.e., I and II) have an influence on
first. The classification accuracy rate of 83% subsequent subtest performance or if unique
matches that found for the full version of the order effects emerge if subtests are given out
test. Gontkovsky and Souheaver (2002) com- of order (Moehle et al., 1988). Moehle and
pared performance on the Short Category colleagues (1988) rightly express caution in
Test and Booklet Category Test in a sample of the use of their shortened version for these
neuropsychology clinic referrals and found reasons. Snow (1987) reminds that any short-
that T scores were comparable between the ening of a test necessitates compilation of new
two, although they observed that the short- normative data:
form cut-off of 46 for older participants needed
to be lowered to 41 to match the sensitivity for example, the process of shortening the Category
values of the full test version. Test may make it less fatiguing, and hence less
Sherrill (1985) found that the Gregory et al. demanding. Short forms may therefore be less able
(1979) 120-item version and the Calsyn et al. to discriminate patients with subtle brain dysfunc-
(1980) 108-item version were highly corre- tion. Further, when short forms are developed, it is
lated with each other (0.968) and that, while often the case that little work is done in validating
both were highly correlated with the standard the findings obtained with the newer versions of the
format (0.943-0.981), the 120-item version test. Instead, the old cutting scores continued to be
had the highest correlation with the full test used, with short-form scores merely being prorated
and the smallest standard error of estimate to their equivalent lengthier versions. Clearly, when
(+7.5), suggesting that it was overall the best a test is shortened, new norms will be required.
predictor of the full test score. Sherrill (1985) (p. 258)
suggested that the 120-item version is the
most attractive short-form alternative to the In addition to the length of time required
208-item test, particularly for high-functioning for administration, the lack of portability and
participants, since it includes subtest V; a wide cumbersome nature of the Category Test ap-
variation in scores occurs on this subtest in paratus have been drawbacks (Slay, 1984;
this population. Wood & Strider, 1980). Slay (1984) provides
Taylor et al. (1990) compared the Gregory instructions for constructing a portable Cate-
et al. (1979), Calsyn et al. (1980), and Russell gory Test for the clinician "with a modicum of
and Levy (1987) short forms and reported that workshop skills." Several investigators have
the Russell and Levy version and the Gregory developed paper-and-pen, card, or booklet
et al. version were better predictors of total forms of the Category Test (Adams & Tren-
test scores than the Calsyn et al. version and ton, 1981; DeFilippis et al., 1979; Kimura,
that while the Russell and Levy version had 1981; Wood & Strider, 1980). Adams and
only 95 items, it performed comparably to the Trenton (1981) devised a laminated booklet
120-item Gregory et al. format. Taylor and test form, which provides visual feedback as to
colleagues (1990) recommend the following the correctness of a response, although the
equation for the Russell and Levy format in practicality of this method is somewhat in doubt:
calculating the predicted total score: (Number "the answer sheet was treated by touching each
of short-form errors x 2.73)- 4.49. These au- correct answer with a swab containing di-
thors suggest that the Russell and Levy formula menthylglyoxine. The subject was given a felt
of short-form errors x 2.2 tends to overesti- tip pen that was treated chemically with nickel
mate total test errors. chloride in aqueous ammonia. If the subject
While efforts to shorten the Category Test responded to an item correctly, the circle
are commendable, a major concern involves on the answer sheet would immediately tum
the fact that the majority of the shortened red. An incorrect response resulted in
versions have been based on analyses derived a green circle" (p. 299). The high modified
from administration of the entire Category split-half Spearman-Brown coefficient docu-
Test. Few studies have actually administered a mented between the slide-format Category
shortened version, and it is unlmown if some Test and the paper-and-pen version did not
significantly differ from the coefficient ob- and the interested reader is referred to this
tained for the slide version, and the two test publication.
forms were judged to be equivalent. A highly consistent relationship has been
Wood and Strider (1980) report a similar documented between age and Category Test
shortened version of the Category Test using a scores in both normal and patient (brain-
latent image transfer sheet. When a developer damaged, psychiatric, medical) samples (Ale-
pen is applied to the correct rectangle on the koumbides et al., 1987; Anthony et al., 1980;
answer sheet, the rectangle darkens, providing Bigler et al., 1981a; Boyle, 1986; Boyle et al.,
feedback as to the correctness of the response. 1994; Choca et al., 1997; Corrigan et al., 1987;
No significant difference in performance across Elias et al., 1990, 1993; Ernst et al., 1987;
psychiatric groups was found between this Fitzhugh et al., 1964; Goldstein & Zubin,
version and the original Reitan format. 1990; Heaton et al., 1986, 1991; Mack &
Kimura (1981) also developed a card version Carlson, 1978; Moses et al., 1999; Prigatano &
of the Category Test, in which the patient Parsons, 1976; Query, 1979; Reed & Reitan,
provides verbal responses to which the exam- 1963b; Seidenberg et al., 1984; Vega & Par-
iner responds "right" or "wrong." No signifi- sons, 1967). However, Willis et al. (1988) did
cant differences in test performance were not find a relationship between age and per-
noted between groups of neurological patients. formance on the original version of the Cate-
Finally, DeFilippis et al. (1979) created a gory Test in a sample of 154 healthy elderly
booklet form in which the Category Test slides individuals, but the age range of their partic-
were reproduced on 8.5" x 11" sheets, which ipants was quite narrow (65--79 years).
were then placed into notebooks. A piece of A negative correlation frequently has been
cardboard with the numbers 1-4 was placed noted between education and Category Test
below the notebook, and participants were scores, particularly in normal compared to
instructed to point to the appropriate number patient groups. Several investigators have re-
for each sheet. The examiner provided feed- ported associations between education and
back as to the correctness of a response by Category Test performance in normal indi-
saying "correct" or "incorrect." The booklet viduals (Anthony et al., 1980; Boyle, 1986;
format was highly correlated with the original Choca et al., 1997; Ernst, 1987; Finlayson
slide version, and no effect of test version was et al., 1977; Heaton et al., 1986, 1991; Priga-
documented in patients and normals admin- tano & Parsons, 1976; Vega & Parsons, 1967;
istered both formats. Macinnes et al. (1983b) Warner et al., 1987; Yeudall et al., 1987).
report data validating the Calsyn et al. (1980) However, the relationship between education
short form in conjunction with the DeFilippis and Category Test performance has been
booklet format. The DeFillippis et al. (1979) equivocal in patient groups, with some authors
version in particular appears to be in much documenting a significant correlation (Ale-
wider usage than the original slide projector koumbides et al., 1987; Boyle, 1986; Lin &
format. Rennick, 1974; Seidenberg et al., 1984) and
A comprehensive summary of the history others failing to detect an association (Corri-
and current perspectives on the Category Test gan et al., 1987; Finlayson et al., 1977; Priga-
is offered by Choca et al. (1997). tano & Parsons, 1976; Vega & Parsons, 1967).
Seidenberg et al. (1984), using a multivariate
approach, reported that education was a more
influential variable in performance than age;
RELATIONSHIP BETWEEN CATEGORY
however, the age range of their participants
TEST PERFORMANCE AND
was very attenuated: 15-52 years.
DEMOGRAPHIC FACTORS
A consistent negative relationship has been
Corrigan et al. (1987) summarize much of the observed between Category Test scores and IQ
available literature on the relationship of IQ, in both normal and patient groups, although
education, and age to Category Test scores; reports have differed as to whether VIQ or
CATEGORY TEST 481
PIQ is more tied to performance. For exam- published since 1965 (Dodrill, 1987; Ernst,
ple, several studies have indicated that scores 1987; Fromm-Auch & Yeudall, 1983; Harley
are more associated with PIQ (Corrigan et al., et al., 1980; Heaton et al., 1991; Pauker, 1980;
1987; Cullum et al., 1984; Goldstein & Shelly, Yeudall et al., 1987), as well as the original
1972; Lansdell and Donnelly, 1977), while Halstead (1947) and Reitan (1955b, 1959) nor-
other publications have documented a stron- mative data and three interpretive guides
ger association with VIQ or WAIS verbal (Logue & Allen, 1971; Reitan & Wolfson, 1985;
factor scores (Shore et al., 1971; Yeudall et al., Russell et al., 1970). Hundreds of other studies
1987). Corrigan et al. (1987) point out that the have also reported control subject data, and we
correlations with PIQ have been relatively will discuss several of those investigations which
consistent across studies, suggesting a reliable involved some unique feature, such as large
relationship; however, the correlation with sample size (> 100), retest data, elderly popu-
VIQ has varied widely and inexplicably across lation, or non-English-speaking sample (Ale-
studies. koumbides et al., 1987; Anthony et al., 1980;
In general, no significant gender differences Barrett et al., 2001; Bomstein et al., 1987a; Elias
have been noted in Category Test perfor- et al., 1990, 1993; El-Sheikh et al., 1987; Hea-
mance (Dodrill, 1979; Elias et al., 1990, 1993; ton et al., 1986; Klove & Lochen, cited in Klove,
Fromm-Auch & Yeudall, 1983; Heaton et al., 1974; Mack & Carlson, 1978; Russell, 1987;
1986, 1991; Kupke, 1983; Pauker, 1980; Wiens & Matarazzo, 1977).
Seidenberg et al., 1984; Yeudall et al., 1987), Russell and Starkey (1993) developed the
although Ernst (1987) reported that in his Halstead-Russell Neuropsychological Evalua-
elderly sample, men performed slightly better tion System (HRNES), which includes the
than women on the booklet version of the test. Category Test among 22 tests. In this system
Handedness (Gregory & Paul, 1980; Sei- and its revised version (HRNES-R), individual
denberg et al., 1984) and socioeconomic status performance is compared to that of 576 brain-
(Seidenberg et al., 1984) do not appear to be damaged participants and 200 participants who
related to Category Test scores. However, were initially suspected ofhaving brain damage
health status is moderately related, as reported but had negative neurological findings. Data
by Willis et al. (1988). were partitioned into seven age groups and
Arnold et al. (1994) documented a signifi- three educationai!IQ levels. This study will not
cant effect of acculturation on performance be reviewed in this chapter because the "nor-
for the original version of the Category Test in mal" group consisted of V.A. patients who
a sample of Mexican Americans, with more presented with symptoms requiring neu-
acculturated individuals demonstrating higher ropsychological evaluation. For further dis-
performance. Similarly, Manly et al. (1998) cussion of the HRNES, see Lezak et al. (2004).
found that performance of 30% of healthy Of note, few relevant manuscripts have
African Americans was in the impaired range emerged since the 1980s, perhaps due either
using the norms of Heaton (1992) and Heaton to the publication of Heaton et al.'s (1991)
et al. (1991), although significant differences comprehensive normative tables or to the es-
between African Americans and Caucasian calating use in research and clinical practice of
(HIV+) individuals disappeared when level of flexible neuropsychological test protocols
acculturation (including use of black English) which include newer tasks rather than tradi-
was considered. tional fixed neuropsychological batteries.
To adequately evaluate the Category Test
normative reports, six key criterion variables
were deemed critical. The first five of these
relate to subject variables, and the one re-
NORMATIVE REPORTS
maining dimension refers to a procedural issue.
Our review of the literature located seven Minimal requirements for meeting criterion
Category Test normative reports for adults variables were as follows.
Subject Variables 1987; Halstead, 1947; Klove &: Lochen, cited in

Klove, 1974; Mack &: Carlson, 1978; Reitan,
Sample Size 1955b, 1959; Wiens &: Matarazzo, 1977). Only
As discussed in previous chapters, a minimum three publications consistently had at least
of at least 50 participants per groupin!i inter- 50 participants in individual subject groupings
val is optimal. I (Elias et al., 1993; Ernst, 1987; Heaton et al.,
1986), although some reports had some sub-
Sample Composition Description groups which met this criterion (Dodrill, 1987;
As discussed previously, information re~arding Fromm-Auch &: Yeudall, 1983; Harley et al.,
medical and psychiatric exclusion crite~ is im- 1980; Pauker, 1980; Yeudall et al., 1987).
portant; it is unclear if geographic recruitment Eleven of the studies summarized in this
region, gender, socioeconomic status ot occu- chapter present Category Test data according
pation, ethnicity, handedness, and rec~tment to circumscribed age ranges (Elias et al.,
procedures are relevant. Until determin~, it is 1990, 1993; Ernst, 1987; Fromm-Auch &:
best that this information be provided. . Yeudall, 1983; Harley et al., 1980; Heaton
j
et al., 1986, 1991; Mack &: Carlson, 1978;
Age Group Intervals : Pauker, 1980; Wiens &: Matarazzo, 1977; Yeu-
Given the association between age an~Cegory dall et al., 1987). Information on IQ levels is
Test performance, information reg · g the reported in all but six studies (Barrett et al.,
age of the normative sample is critical nor- 2001; Elias et al., 1993; El-Sheikh et al., 1987;
mative data should be presented by age in rvals. Ernst, 1987; Heaton et al., 1986; Russell, 1985),
and one report presented Category Test data in
Reporting of IQ Levels ! age-by-IQ groupings (Pauker, 1980). Similarly,
Given the relationships between C4egory educational level was also indicated in all but
Test performance and IQ, data shmpd be two studies (Bomstein et al., 1987a; Pauker,
i
presented by IQ intervals, or at least infor- 1980), and Heaton et al. (1986, 1991) organized
mation regarding intellectual level sho..ld be data by educational levels. Information on the
provided. In addition, given some e~ence gender composition of the samples was avail-
that PIQ may be more related to CaJ:egory able in all but three reports (Anthony et al.,
Test performance than VIQ, informatbn on 1980; Harley et al., 1980; Klove &: Lochen, cited
PIQ and VIQ separate from FSIQ is deslrable. in Klove, 1974); four data sets included only
male (Barrett et al., 2001; Wiens &: Matarazzo,
Reporting of Educational Levels 1977) or nearly all-male (Alekoumbides et al.,
1
Given the possible, although minor, a$Socia- 1987; Russell, 1987) populations, and one data
tion between educational level and C~egory set was composed primarily of females (Mack&:
Test scores, it is preferable that info~ation Carlson, 1978). Ernst (1987) and Heaton et al.
regarding highest educational level completed (1991) presented data separately for males and
be reported. females.
Information on other subject variables was
Procedural Variables provided less frequently; data on handedness
were indicated in three studies (Dodrill, 1987;
Data Reporting Fromm-Auch &: Yeudall, 1983; Yeudall et al.,
Means, standard deviations, and preferably ran- 1987); occupation or socioeconomic status was
ges for total Category Test errors are reqtPred. described in seven reports (Alekoumbides et al.,
1987; Barrett et al., 2001; Dodrill, 1987; Elias
et al., 1990; Halstead, 1947; Wiens&: Matarazzo,
1977; Yeudall et al., 1987); and information re-
garding ethnicity was presented in four data sets
OF THE NORMS
i (Alekoumbides et al., 1987; Barrett et al.,
All but eight data sets had total sample sizes 2001; Dodrill, 1987; Russell, 1987). Exclusion
>100 (Bomstein et al., 1987a; El-Sheikh:et al., criteria were judged to be adequate in only
CATEGORY TEST 483
10 publications (Anthony et al., 1980; Bomstein Battery (HRB), the Halstead (1947) and Re-
et al., 1987a; Dodrill, 1987; Elias et al., 1990, itan (1955b, 1959; Reitan & Wolfson, 1985)
1993; Fromm-Auch & Yeudall, 1983; Heaton data and interpretation recommendations will
et al., 1991; Pauker, 1980; Wiens & Matarazzo be reported first, followed by a summary of
1977; Yeudall et al., 1987). Geographic re~ the other interpretation formats. Then, the
cruitment areas were specified in all but six normative publications and control groups
publications (Barrett et al., 2001; Bomstein from clinical comparison studies will be re-
et al., 1987a; Dodrill, 1987; Elias et al., 1990, viewed in ascending chronological order.
1993; Mack & Carlson, 1978). Twelve data sets
were obtained in the United States (Ale-
Original Studies
koumbides et al., 1987; Anthony et al., 1980;
Barrett et al., 2001; Halstead, 1947; Harley [CT.l] Halstead, 1947 (Table A24.2)
et al., 1980; Heaton et al., 1986, 1991; Klove &
The author obtained Category Test data on
Lochen, cited in Klove, 1974; Reitan, 1955b,
28 control participants in Chicago, more than
1959; Russell, 1987; Wiens & Matarazzo, 1977),
half of whom had psychiatric diagnoses. The
three in Canada (Fromm-Auch & Yeudall
1983; Pauker, 1980; Yeudall et al., 1987), one i~
14 participants without psychiatric diagnoses
were nine male and five female civilians aged
Norway (Klove & Lochen, cited in Klove,
15-50 (mean= 25.9) without history of brain
1974), one in Egypt (El-Sheikh et al., 1987),
injury. The eight participants with diagnoses of
and one in Australia (Ernst, 1987).
mild psychoneurosis were male soldiers aged
. Total mean number of errors was reported
.2_2--38 (mean= 29.6); some had combat expe-
m all data sets, and SDs were indicated in all
nence but none had a history of head injury.
but four studies (Barrett et al., 2001; Halstead,
The last six participants were aged 27-39 and
1947; Heaton et al., 1986; Klove & Lochen
included a depressed military prisoner facing
cited in Klove, 1974). Ranges for number of
execution, a severely depressed female with
errors were presented in four publications
suicidal and homicidal impulses tested prior to
(Bomstein et al., 1987a; Fromm-Auch &
lobotomy, and a suicidal/homicidal female and
Yeudall, 1983; Halstead, 1947; Harley et al.,
a suicidal male tested pre- and post-lobotomy.
1980), and means and SDs for individual
Educational level ranged 7-18 years, and the
subtest scores are provided in two publications
following occupations were represented: artist,
(Ernst, 1987; Mack & Carlson, 1978). Some
entertainer, farmer, housewife, semiskilled
studies reported supplementary Category Test
and unskilled laborers, professional, secretary,
scores, such as IQ-equivalent scores (Dodrill,
teacher, technician, trade, and student. Ethnic
1987), test-retest data (Bomstein et al., 1987;
background included American, Balkan, En-
El-Sheikh et al., 1987), T-score equivalents for
glish, French, German, Irish, Polish, Scandi-
raw scores (Harley et al., 1980), and T-score
navian, and Scottish. IQ level ranged 70-140.
equivalents corrected for age, education, and
Mean errors are reported for the total
gender (Heaton et al., 1991).
group and each control subgroup, as well as
The text of study descriptions contains
individual scores for each subject. The Cate-
references to the corresponding tables identi-
gory Test criterion score used in calculating
fied by number in Appendix 24. Table A24.1,
the Impairment Index was >80 errors.
~e l<><:ator table, summarizes information pro-
VIded m the studies described in this chapter. 1
Study strength
1. Information provided regarding IQ, ed-
SUMMARIES OF THE STUDIES ucation, occupation, ethnicity, geographic
recruitment area, age, and gender.
Given that the Category Test has typically been
used within the context of the Halstead-Reitan Considerations regarding use of the study
'Nonns for children are available in Baron (2004) and 1. Small sample size, including use of two
Spreen and Strauss (1998). participants twice.
2. Inclusion of participants with psychiatric gory Test raw error scores based primarily on
diagnoses and postlobotomy. "rules of thumb" recommended by P. M. Re-
3. No reporting of SDs. nnick. Russell (1984) subsequently modified
4. Undifferentiated age range. the ratings as reHected in Table 24.2.
Logue and Allen (1971) published a pre-
[CT.2] Reitan, 1955b (see also Reitan, 19$9) dictor table plotting the expected number
(Table A24.3) of Category Test errors for nine W AIS FSIQ
The author obtained Category Test s~es on va1ues based on Reitan's 1959 Wechsler-
50 participants in Indiana who had app;rently Bellevue and Category Test scores on control
been referred for neuropsychologicaJ testing participants (see Table 24.3).
and "who had received neurologicaJ txami-
Use of the table . . . allows a direct comparison of
nations before testing and showed nq signs
scores actually obtained from a given client on the
or symptoms of cerebra] damage or d~func
WAIS and on the Category Test. Where the rela-
tion . . . . None . . . had positive ~am
tionship is not at the predicted level, the examiner
nestic findings" (p. 29); some were hospi~d
can have more confidence that the obtained cate-
with paraplegia or neurosis. The sam~le in-
gory score is not an artifact of limited or superior
cluded 35 men and 15 women, mean ase was
intelligence. (p. 1096)
32.36 (10.78), and mean educationaJle~l was
11.58 (2.85). Mean WAIS VIQ, PIQ, anclFSIQ The authors caution that the expected Cate-
were 110.82 (14.46), 112.18 (14.23), and l12.64 gory Test scores at the highest IQ va1ues are
(14.28), respectively. ' probably unrealistically low.
Study strengths
1. Information regarding IQ, educationaJ [CT.3] Reitan and Wolfson, 1985
level, gender, age, and geographic re- The authors provide generaJ guidelines for Cat-
cruitment area. · egory Test score interpretation in the form of
2. Adequate sample size. "severity ranges:" "perfectly normaJ (or better
3. Means and SDs are reported. than average)," "normal," "mildly impaired," and
"seriously impaired." They list the number of
Considerations regarding use of the stuf.y errors which correspond to each severity range:
2. Insufficient medica] and psychiatric ex- 0-25: perfectly normaJ (or better than
clu~ion criteria; sample included partic- average)
ipants hospitaJized with spinaJ cord 26-45: normaJ
injuries and psychiatric disorders.
46--65: mildly impaired
3. High average IQ.
2':66: seriously impaired
Interpretive Guides
No other information is provided, such as
In constructing their neuropsychologic41 key score means or SDs, or any data regarding the
approach, Russell, Neuringer, and Golilstein normative sample on which these guidelines
(1970) devised six rating equivaJents of Pate- were developed.
Table 24.2. Rating Equivalents of Catetory Test Raw Scores, • According to Russell et al. (1970)
Rating Equivalent
0 2 3 4 5 6
Errors 26-52 53-78 79-104 105-130 131-156 >5 Er-col II
• A score of 156 is considered random.

CATEGORY TEST 485
Table 24.3. Predictor Table, Accord- damage. Mean age, educational level, and IQ
ing to Logue and Allen (1971) for the American participants were 31.6, 11.1,
and 109.3, respectively; and mean age, edu-
WAIS Predicted
FSIQ" Category Test Score cational level, and IQ for the Norwegian
participants were 32.1, 12.2, 111.9, respe~
140 10 tively. Category Test scores are presented m
130 15
terms of mean errors for each group.
120 21
110 26
100 32 Study strengths
90 37 1. This publication is unique in providing
80 43 Category Test data on a Norwegian
70 48
population.
60 54
2. Information regarding educational level,
•wAIS FSIQ, Wechsler Adult Intelligence IQ, age, and geographic recruitment area
Scale full-scale intelligence quotient. is reported.

Considerations regarding use of the study 1. Small sample size.
The authors argued that these norms were
2. Undifferentiated age ranges.
meant as "general guidelines" and that "exact
3. No SDs reported.
percentile ranks corresponding with each
4. No exclusion criteria specified, and no
possible score are hardly necessai)' because
information regarding gender distribu-
the other methods of inference are used to
tion of the sample is provided.
supplement normative data in clinical inter-
pretation of results of individual participants" [CT.S] Wiens and Matarazzo, 1977
(p. 977). However, we maintain that m~re (Table A24.5)
precise scores as well as separate normative
The authors collected Category Test data on
data for different age, IQ, and educational
48 male applicants to a patrolman program in
levels are necessary to avoid false-positive er-
Portland, Oregon, as part of an investigation
rors in diagnosis.
of the WAIS and MMPI correlates of the
It is not clear how the cutoffs were devel-
HRB. All participants passed a medical exam
oped; they do not match the cutoffs cited by
and were judged to be neurologically normal.
Halstead (1947). The authors report that a
Participants were divided into two equal
cutoff of 50 was recommended by Halstead in
groups, which were comparable in age (23.6
computing the Impairment Index, but exami-
vs. 24.8), education (13.7 vs. 14.0), and WAIS
nation of Halstead's (1947) manuscript revealed
FSIQ (117.5 vs. 118.3). A random subsample
that his cutoff was >80, not >50. It appears that
of 29 applicants was readministered the Cat-
Reitan derived his cutoff by computing the ratio
egory Test 14-24 weeks following the original
of errors to total items for the Halstead ad-
administration. Means and SDs for total er-
ministration and applied the same ratio to his
rors are reported for both the original testing
208-item version (e.g., 80/336=50/208).
and retest. One of the 29 participants obtained
a score higher than Reitan's suggested cutoff
Normative Studies and Control Groups of 50/51 errors. No significant correlations
from Clinical Comparison Studies were observed between Category Test scores
and FSIQ, VIQ, or PIQ in either control
[CT.4] Klove and Lochen (Cited in Klove, group (Wiens & Matarazzo, 1977).
1974) (Table A24.4)
The authors obtained Category Test data on Study strengths
22 American controls from Wisconsin and 1. Information on test-retest performance.
22 Norwegian controls as part of a validation 2. Relatively large sample size for the small
study on the ability of the HRB to detect brain age range.
3. Adequate medical exclusion criteria. Total mean errors and SDs are provided as
4. Information provided regarding educa- well as mean errors and SDs for subtests III,
tional level, IQ, gender, and geographic IV, V, VI, and VII. The elderly participants
recruitment area. performed significantly worse than young con-
5. Means and SDs are reported. trols and comparably to a middle-aged brain-
damaged sample. The elderly sample showed
Considerations regarding use of the study particular difficulty on subtests II and IV rel-
1. High IQ level. ative to the younger sample.
2. High educational level.
3. SDs differ markedly between the two Study strengths
control groups, suggesting either unusual 1. Data are presented in age groupings.
variability in scores for the first group 2. Adequate exclusion criteria in older
or unusual lack of variability in the sec- participants.
ond group or an error in reporting the 3. Information regarding education, IQ,
data. gender, and (in the younger participants)
4. All-male sample. occupation.
4. Data for several individual subtests as
[CT.6] Mack and Carlson, 1978 well as total errors.
(Table A24.6) 5. Means and SDs are reported.
The authors obtained Category Test data on
41 old (range 60-80) and 40 young (range 20- Considerations regarding use of the study
37) participants as part of an investigation into 1. No exclusion criteria for the younger
the neuropsychological effects of aging. Older participants.
participants with histories of neurological im- 2. IQ data not available for all participants
pairment or "signs or symptoms of diseases with and two IQ measures used.
neurological significance (or which predispose 3. Minor alterations in test administration
participants to possible neurological disorder)" format (computer-assisted).
were excluded. No screening for neurological 4. Relatively small sample size and rela-
impairment was conducted in the younger tively large age range within each age
sample, which was drawn from a university grouping.
student body and hospital staff. The older 5. High educational and intellectual levels.
sample included three men and 38 women; 6. Samples are primarily female.
mean educational level was 14.05 (3.39) years,
and mean WAIS FSIQwas 119.90 (15.14). The [CT.7] Anthony, Heaton, and Lehman, 1980
young sample consisted of 31 female and nine (Table A24.7)
male participants, and mean years of education The authors amassed Category Test data
and mean IQ (based on Shipley scores in 17 on 100 normal volunteers from Colorado as
participants) were 15.43 (2.65) years and part of a cross-validation of two objective,
113.76 (4.89), respectively. computerized interpretive programs for the
The Category Test was computer- HRB. Participants had no history of medical
administered according to standard instruc- or psychiatric problems, head trauma, brain
tions, with the exception that after participants disease, or substance abuse. In addition, for
made a response by pressing a button, they 85% of the controls, normal EEGs and neu-
could change their response. Once they were rological exams were obtained; in the re-
satisfied with a response, they pushed the "0" maining 15% of participants, it appeared that
key, which was followed by feedback on the this information was not available. Mean age
correctness of the response, and the slide pro- was 38.88 (15.80) years, and mean education
jector advanced to the next trial. The authors was 13.33 (2.56) years. Mean W AIS FSIQ,
concluded that this modification had little ef- VIQ, and PIQ were 113.54 (10.83), 113.24
fect given' that few participants corrected an (11.59), and 112.26 (10.88), respectively. Cate-
initial choice. gory Test data are presented in terms of mean
CATEGORY TEST 487
number of errors and SD. Participants incor- raw scores are reported as well as "percentage
rectly identified as brain-damaged (according to of best raw score," which indicated where a
the Russell et al., 1970, system) were older, less raw score fell within the range of raw scores
educated, and less intelligent than participants for that age grouping. We reproduce only the
correctly classified as non-brain-damaged. mean, SD, and range for total errors due to
space considerations.
Study strengths
1. Large sample size. Study strengths
2. Adequate exclusion criteria. 1. Large sample size, and many individual
3. Information regarding education, IQ, cells approximate 50.
age, and geographic recruitment area. 2. Reporting of data on IQ, educational
4. Means and SDs are reported. level, and geographic recruitment area.
3. Data presented in age groupings.
Considerations regarding use of the study 4. Means and SDs are reported.
1. Large undifferentiated age grouping.
2. IQ range is high average. Considerations regarding use of the study
3. No information regarding gender. 1. Presence of substantial neurological
(chronic brain syndrome), substance
[CT.8l Harley, Leuthold, Matthews, abuse, and major psychiatric disorders in
and Bergs, 1980 (Table A24.8) the sample.
The authors collected Category Test data on 2. Low educational level, although IQ lev-
193 V.A.-hospitalized patients in Wisconsin els are average.
aged 55-79. Exclusion criteria were FSIQ 3. No information regarding gender, al-
<80, active psychosis, unequivocal neurologi- though given that data were obtained in
cal disease or brain damage, and serious visual a V.A. setting, the sample is likely all or
or auditory acuity problems. Patients with a nearly all male.
diagnosis of chronic brain syndrome were 4. Odd variability across scores, with those
included. Patient diagnoses were as follows: 75-79 years old outperforming the youn-
chronic brain syndrome unrelated to alcohol- ger age groups and those 60-64 years old
ism, psychosis, and alcoholic, neurotic, or outperforming those 55-59 years old.
personality disorder. Mean educational level
was 8.8 years. The sample was divided into five Other comments
age groupings: 55-59 (n =56), 60-64 (n = 45), The scores for the two oldest age groups are
65-69 (n =35), 7~74 (n =37), and 75-79 identical in the whole sample and alcohol-
(n = 20). Mean educational level and percent equated group because these two groups did
of sample included in each diagnostic classifi- not have overrepresentation of alcoholics, so
cation are reported for each diagnostic classi- they did not need to be adjusted.
fication and for each age grouping. The authors
also provide test data on a subgroup of 160
participants equated for percent diagnosed [CT.9] Pauker, 1980 (Table A24.9)
with alcoholism across the five age groupings. The author obtained Category Test scores
The "alcohol-equated sample" was developed on 363 Toronto citizens Ruent in English, re-
"to minimize the influence that cognitive or cruited through announcements and notices.
motor/sensory differences uniquely attribut- Participants were aged 19-71 and included
able to alcohol abuse might have upon group 152 men and 211 women. Exclusion criteria
test performance levels" (p. 2). This subsample consisted of significant physical disability,
remained heterogeneous regarding represen- sensory deficit, current medical illness, using of
tation of the other diagnostic categories. medication that might affect test performance,
Mean errors and SDs are provided by age history of actual or suspected brain disorder,
groupings for the total and alcohol-equated and alcoholism. MMPI profiles "could not
samples. In addition, T-score equivalents for suggest severe disturbance" or include more
than three clinical scales with T scores 2':70 or sample were right-handed, and mean age was
an F scale score >80. · 25.4 (8.2) years, with a range of 15-64 years.
The Category Test was administered ac- Mean education was 14.8 (3.0) years, with a
cording to Reitan's guidelines. Means aJld SDs range of8-26 years, and included technical and
for total errors are reported for the sample as a university training. Mean WAIS FSIQ, VIQ,
whole and for three age groupings (~9-34. and PIQ were 119.1 (8.8, range 98-142), 119.8
35-52, 53-71) by four WAIS IQ j levels (9.9, range 95-143), and 115.6 (9.8, range
(89-102, 103-112, 113-122, 123-143)! Indi- 89-146), respectively. No participants obtained
vidual cell sizes ranged 4-60. Age-by-·Q ca- an FSIQ lower than the average range. Partic-
tegories were determined "in a compjomise ipants were classified into five age groupings:
between what would be desirable and w~at the 15-17 (n = 32),18-23 (n = 75), 24-32 (n =57),
obtained sample characteristics and s* dic- 33-40 (n = 18), and 41-64 (n = 10).
tated" (p. 1). No differences in perforynance Mean errors, SDs, and ranges are reported
between men and women were docum.nted. for each age grouping. No gender differences
were documented, and male and female data
Study strengths 1 were collapsed. The authors suggest that a
1. Large sample size, although ind(vidual cutoff of 50 errors is appropriate only for
cell sizes are substantially less thap 50. participants <40 years of age.
2. Presentation of data in ageiby-IQ
groupings. Study strengths
3. Adequate medical and psychiatric 1. Large overall sample size, and some in-
exclusion criteria. . dividual cells approximate 50.
4. Information regarding gender, r+cruit- 2. Presentation of data by age groupings.
ment procedures, and geographJc re- 3. Information regarding mean IQ, educa-
cruitment area. I tional level, handedness, gender, re-
5. Means and SDs are reported. cruitment procedures, and geographical
recruitment area.
Considerations regarding use of the stuf,y 4. Some psychiatric and neurological ex-
1. No information regarding educatiqn. clusion criteria.
2. Participants were recruited in Cioada, 5. Means and SDs are reported.
raising questions regarding usef\tlness
for clinical interpretation in the Qnited Considerations regarding use of the study
States. 1. High intellectual and educational level.
3. The age-by-IQ cell representing partici- 2. An age grouping of 41-64 with 10 par-
pants aged 53-71 with IQ of 8~102 ticipants would not appear to be partic-
contained only four participants; ~uker ularly useful.
comments that this category "shouJd not 3. Participants were recruited in Canada,
be considered to be of any mord than raising questions regarding usefulness for
interest value" (p. 2). · clinical interpretation in the United States.
4. IQ levels below the average ran~ not 4. At least one subject in the 18-23 age
represented. group scored particularly poorly, causing
the mean to be artificially low and the
[CT.10] Fromm-Auch and Yeudall, 1983 SD to be excessively large for this group.
(Table A24.1 0)
The authors obtained Category Test data [CT.11] Heaton, Grant, and Matthews, 1986
on 193 Canadian participants (111 male, 82 (Table A24.11)
female) recruited through posted advcfrtise- The authors obtained Category Test data on
ments and personal contacts. Participants are 553 normal controls in Colorado, California,
described as "nonpsychiatric" and "non- and Wisconsin as part of an investigation into
neurological. " Eighty-three percent <t
the the effects of age, education, and gender on
!
CATEGORY TEST 489
HRB performance. Nearly two-thirds of the employed, 26 unemployed, 11 homemakers,

sample were male (356 males, 197 females). and one retiree. Participants were recruited
Exclusion criteria were history of neurological from various sources, including schools,
illness, significant head trauma, and substance churches, employment agencies, and com-
abuse. Participants were aged 15--81 years, munity service agencies; and they were either
with a mean of 39.3 (17.5) years, and had paid for their participation or offered an in-
mean education of 13.3 (3.4) years, with a terpretation of their abilities. Exclusion crite-
range of 0-20 years. The sample was divided ria were history of "neurologically relevant
into three age categories ( <40, 40-59, and disease (such as meningitis or encephalitis),"
~60) with sizes of 319, 134, and 100, respec- alcoholism, birth complications "of likely
tively, and classified into three education ca- neurological significance," oxygen deprivation,
tegories (<12, 12-15, ~16 years) with sizes of peripheral nervous system injury, psychotic or
132, 249, and 172, respectively. psychotic-like disorders, or head injury asso-
Testing was conducted by trained techni- ciated with unconsciousness, skull fracture,
cians, and all participants were judged to have persisting neurological signs, or diagnosis
expended their best effort on the task Mean of concussion or contusion. One-third of po-
errors are reported for the six subgroups, as tential participants failed to meet the above
well as percent classified as normal using Rus- medical and psychiatric criteria, resulting in
sell et al.'s (1970) criteria. Approximately 30% a final sample of 120. Mean age was 27.73
of the test score variance was accounted for by (11.04) years, and mean education was 12.28
educational level. Significant group differences (2.18) years. Participants tested in the 1970s
were found across the three age groups and the were administered the WAIS, and those as-
three educational levels, and a significant age- sessed in the 1980s were administered the
by-education interaction was documented. No WAIS-R; WAIS scores were converted to
significant differences in performance were WAIS-R equivalents by subtracting 7 points
found between males and females. from the VIQ, PIQ, and FSIQ. Mean FSIQ,
VIQ, and PIQ scores were 100.00 (14.35),
Study strengths 100.92 (14.73), and 98.25 (13.39), respec-
1. Large size of overall sample and indi- tively. IQ scores ranged 60-138 and reftected
vidual cells. a normal distribution.
2. Information regarding education, gender, Mean errors and SDs are reported as well
age, and geographic recruitment area. as IQ-equivalent scores for various levels of
3. Data are grouped by age as well as ed- intelligence. Using Reitan's cutoff of 50/51
ucational level. errors, 22.5% of a subgroup of the sample
were misclassified as brain-damaged.
1. No reporting of SDs. Study strengths
2. Means reported for individual WAIS 1. Large sample size.
subtest scaled scores but not for overall 2. Comprehensive exclusion criteria (al-
IQ scores. though the appropriateness of including
individuals with WAIS-R scores falling
[CT.12] Dodrill, 1987 (Table A24.12) in the mentally deficient range could be
The author collected Category Test data on questioned).
120 participants in Washington during the 3. Information regarding age, education,
years 197~1976 (n = 81) and 1986-1987 (n = IQ, occupation, gender ratio, handed-
39). Half of the sample was female, and 10% ness, ethnicity, recruitment procedures,
were minorities (six African American, three and geographic area.
Native American, two Asian American, one 4. IQ equivalent scores provided.
unknown). Eighteen were left-handed, and 5. Data for different IQ levels provided.
occupational status included 45 students, 37 6. Means and SDs are reported.
Considerations regarding use of the stt¥iy documented between test scores and VIQ and
1. Undifferentiated age range. : PIQ, again particularly in males; but only VIQ
2. On the IQ-equivalent scores, the two accounted for >10% of score variance. Because
highest IQ groups have poorer :scores no significant differences were found between
than the 115-120 IQ groups. men and women, only the combined sample
data are reproduced below.
[CT.13] Yeudall, Reddon, Gill, and Stefanyk,
1987 (Table A24.13) 1 Study strengths
The authors obtained Category Test ~ta on 1. Large sample size, and individual cells
225 Canadian participants recruitedi from approximate 50.
posted advertisements in workplace$ and 2. Grouping data by age.
personal solicitations. Participants included 3. Data availability for a 15-20 age group.
meat packers, postal workers, transit e~ploy 4. Adequate medical and psychiatric ex-
ees, hospital lab technicians, secretaries; ward clusion criteria.
aides, student interns, student nurse,, and 5. Information regarding handedness, edu-
summer students. In addition, high ;chool cation, IQ, gender, occupation, geographic
teachers identified for participation average recruitment area, and recruitment proce-
students in grades 10-12. Exclusion ~teria dures.
were evidence of "forensic involve.pent," 6. Means and SDs are reported.
head injury, neurological insult, pren*al or
birth complications, psychiatric proble~s. or Considerations regarding use of the study
substance abuse. Participants were claSsified 1. Sample was atypical in terms of its high
into four age groupings: 15-20, 21-25, $6--30, average intellectual level and high level
and 31-40. Information regarding p'rcent of education.
right-banders, mean years of educatio~, and 2. Data were obtained on Canadian par-
mean WAIS/WAIS-R FSIQ, VIQ, and PIQ ticipants, which may limit their useful-
are reported for each age grouping for :males ness for clinical interpretation in the
and females separately and combined. Fbr the United States due to possible subtle
sample as a whole, 88% were right-handed cultural differences.
and had completed an average of 14.87 (2.99) 3. Examination of the data reveals odd,
years of schooling. The mean FSIQ, VIQ, and unpredicted variability, with those 21-
PIQ were 113.98 (9.83), 114.77 (10.34), and 25 years old performing worse than
108.50 (10.34), respectively. those 26--30 years old.
Category Test data were gathered by .expe-
rienced testing technicians who "mottyated [CT.14] Ernst, 1987 (Table A24.14)
the participants to achieve maximum ~rfor The author obtained Category Test data on 110
mance" partially through the promise +f de- primarily Caucasian (99%) residents of Bris-
tailed explanations of their test perform4nce. bane, Australia, aged 65-75. Fifty-nine were
Means and SDs for total errors ar~ pre- female and 51 were male, and mean educa-
sented for each age grouping and eacij age- tional level was 10.3 years; men and women did
by-gender grouping. not differ in years of education. Participants
No significant relationships were found were recruited primarily through random se-
between Category Test scores and gender. lection from the Queensland State electoral
Age effects were also not significant, although roll (n = 97), with the remainder (n = 13) soli-
the authors note that variance effects' with cited through senior-citizen centers. Exclusion
age were significant and recommend use of criteria were history of significant head trauma
age norms. A significant negative assocJation or neurological disease. Nearly one-half of the
was found between Category Test !fores sample were diagnosed with at least one
and education, particularly in males; bu~ edu- chronic disease (hypertension = 33, heart dis-
cation accounted for <10% of test scor~ vari- ease = 9, thyroid dysfunction = 7, asthma= 5,
ance. Significant negative correlations :were emphysema= 2, diabetes= 1), for which they
CATEGORY TEST 491
were receiving treatment and which was de- rected for age and education for the HRB.
scribed as well controlled. Sixty-six participants Among the 41 psychiatric patients, nine were
were taking medications, primarily for the diagnosed as psychotic and 32 were neurotic.
diseases listed above. In addition to psychiatry services, patients
All participants were administered the Trail- were drawn from medicine (n =57), neurol-
making Test first, and half of the participants ogy (n = 22), spinal cord injury (n = 9), and
were also administered the Tactual Perfor- surgery (n = 6) units. Mean age was 46.85
mance Test (TPT) prior to the Booklet Cate- (7.17) years, ranging 19--82 years, and mean
gory Test. Mean errors and SDs for each of the education was 11.43 (3.20) years, ranging 1-20
seven subtests as well as total errors are re- years. Frequency distributions for age and
ported. Using a cutoff of 51 errors, 84% of the years of education are provided. Mean WAIS
sample were classified as impaired. Men ob- FSIQ, VIQ, and PIQ were within the average
tained fewer errors than women on subtests range: 105.89 (13.47), 107.03 (14.38), and
III and IV and on total errors. No differences 103.31 (13.02), respectively. Means and SDs
in scores emerged between participants with for individual age-corrected subtest scores are
and without chronic disease. Educational level also reported. All participants except one were
appeared to be related to scores on subtests IV, male; the majority were Caucasian (93% ), with
V, and VII and total errors. No differences in 7% African American. The mean score on a
test performance were documented between measure of occupational attainment was 11.29.
those participants who were or were not ad- No differences were found in test perfor-
ministered the TPT prior to the Category Test. mance between the two psychiatric groups and
the nonpsychiatric group, and the data were
Study strengths collapsed. Mean errors and SDs are presented.
1. Large sample size in a restricted age Both age and educational level had significant
range. associations with Category Test scores in the
2. Presentation of the data by gender. expected direction, and regression equation
3. Information regarding education, geo- information to allow correction of raw scores
graphic recruitment area, recruitment for age and education is included.
procedures, and ethnicity.
4. Information regarding test administra- Study strengths
tion order effects. 1. Large sample size.
5. Means and SDs are reported. 2. Information regarding age, IQ, education,
ethnicity, gender, occupational attain-
Considerations regarding use of the study ment, and geographic recruitment area.
1. Approximately half of the participants 3. Regression equation data for computation
had at least one chronic illness, and over of age- and education-corrected scores.
half were taking prescribed medications. 4. Means and SDs are reported.
3. Low mean educational level. Considerations regarding use of the study
4. Data were collected in Australia and may 1. Data were collected on medical and
be unsuitable for clinical use in the psychiatric patients.
United States. 2. Undifferentiated age range (mitigated by
the regression equation information).
[CT.15] Alekoumbides, Charter, Adkins, 3. Nearly all-male sample.
and Seacat, 1987 (Table A24.15)
The authors report Category Test data on [CT.16] Bornstein, Baker, and Douglass, 1987a
135 medical and psychiatric inpatients and (Table A24.16)
outpatients without cerebral lesions or histo- The authors collected Category Test-retest
ries of alcoholism or cerebral contusion, from data on 23 volunteers (14 women, nine men)
V.A. hospitals in southern California as part of aged 17--52 years, with a mean age of 32.3
their development of standardized scores cor- (10.3) years, as part of an examination of the
short-term retest reliability of the HRB. Ex- differences in performance were found be-
clusion criteria consisted of a positive history tween the English and Arabic administrations.
of neurological or psychiatric illness. Mean A significant practice effect was documented
VIQ was 105.8 (10.8), with a range of 88-128, over the 2-week interval.
and mean PIQ was 105.0 (10.5), with a range
of 85-121. Study strengths
Participants were administered the HRB in l. Data obtained on an Arabic sample.
standard order both on initial testing and again 2. Information on test-retest scores.
3 weeks later. Means, SDs, and ranges for total 3. Information regarding educational level,
errors for both testing sessions are provided, as age, and geographic recruitment area.
well as raw score change and SD, median raw 4. Means and SDs are reported.
score change, and mean percent of change.
Significant improvement in performance over Considerations regarding use of the study
the 3-week period was documented. Correla- 1. Small sample size.
tions of such demographic variables as age and 2. Minimal exclusion criteria.
education with mean percent of change and 3. No information regarding intellectual
mean change were small, with education ac- level.
counting for up to 7% of variance and age 4. Undifferentiated age range, although it
accounting for up to 4% of variance. can be assumed it is fairly restricted.
Study strengths [CT.18] Russell, 1987 (Table A24.18)

l. Information on short-term (3-week) re- The author obtained Category Test data on
test data. 155 controls during the years 1968-1982 in V.A.
2. Information on IQ level, gender, and age. hospitals in Cincinnati and Miami for devel-
3. Minimally adequate exclusion criteria. opment of a reference scale method for neu-
4. Means and SDs are reported. ropsychological test batteries. The 148 male
and seven female participants were suspected
Considerations regarding use of the study of having neurological disorders but had "neg-
l. Undifferentiated age range. ative neurological findings." No other exclusion
2. Relatively small sample size. criteria were described. Mean age was 46.19
3. No information regarding education. (12.86) years, and mean education was 12.29
(3.00) years. All but eight of the participants
[CT.17] EI-Sheikh, EI-Nagdy, Townes, and were Caucasian; the remainder were African
Kennedy, 1987 (Table A24.17) American. Mean WAIS FSIQ, VIQ, and PIQ
The authors report Category Test data on were 111.9, 112.3, and 109.90, respectively.
32 undergraduate and graduate Egyptians at the Mean errors and SDs are provided.
American University in Cairo as part of their
cross-cultural investigation of the Luna- Study strengths
Nebraska Neuropsychological Battery and the 1. Large sample size.
HRB. No subject had a history of known brain 2. Information regarding IQ, education,
damage. Participants were described as "Arabic ethnicity, gender, age, and geographic
and English-speaking." Category Test instruc- recruitment area.
tions were translated in Egyptian colloquial Ar- 3. Means and SDs are reported.
abic by the first author and checked by two
independent judges fluent in both Arabic and Considerations regarding use of the study
English. In the case of disagreement between 1. Undifferentiated age range.
these two judges, a third judge was consulted. 2. Insufficient exclusion criteria; all partic-
The Category Test was administered in ipants were suspected of having neuro-
English to 23 participants and in Arabic to logical disorders.
nine participants and readministered 2 weeks 3. High mean intellectual level.
later. Mean errors and SDs are reported. No 4. Mostly male sample.
CATEGORY TEST 493
[CT.19] Elias, Podraza, Pierce, and Robbins, were collected over a 15-year period through
1990 (Table A24.19) multicenter collaborative efforts; the authors
Participants were 183 community-dwelling trained the test administrators and supervised
individuals (76 men, 107 women) recruited data collection. Exclusion criteria were history
from church groups, businesses, professional of learning disability, neurological disease, ill-
organizations, and community service organi- ness affecting brain function, significant head
zations for older persons as part of a study on trauma, significant psychiatric disturbance (e.g.,
the impact of hypertension on cognition. Ex- schizophrenia), and alcohol or other substance
clusion criteria were no major chronic or abuse. Mean age for the total sample was 42.0
acute disease including hypertension, treat- (16.8), and mean educational level was 13.6
ment for a neurological disorder, brain trauma, (3.5). Sixty-five percent of the sample were
mental illness, or any cardiovascular or cere- males. Mean WAIS FSIQ, VIQ, and PIQ were
brovascular disease. Skilled clerical, super- 113.8 (12.3), 113.9 (13.8), and 111.9 (11.6),
visory, blue-collar, and professional-executive respectively.
occupations were represented. Participants were generally paid and judged
Participants were divided into three age to have provided their best efforts on the tasks.
groupings: 20-31 (41 men, 47 women), 37-49 The Category Test was administered accord-
(23 men, 38 women), and 55--67 (12 men, ing to Reitan and Wolfson's (1985) instruc-
22 women). Mean educational levels for the tions. A T-score system with demographic
three groups were 15.4, 15.7, and 14.9, re- correction was developed on 378 participants
spectively (range 12-20 for each age group), and cross-validated on 108 participants. Total
and mean WAIS VIQ and PIQ were 119 and number of errors was the performance pa-
116 for the youngest group, 122 and 122 for rameter employed. Age accounted for 38% of
the middle-aged group, and 124 and 121 for the variance in test scores, and education was
the oldest group, respectively. associated with 20% of score variance; gender
Means and SDs for number of errors are did not account for any score variance. These
reported. demographic variables in combination were
associated with 43% of score variance. Ex-
Study strengths tensive T-score tables corrected for age, edu-
1. Large overall sample size (with individ- cation, and gender are provided; and the
ual subgroup sizes of 88, 61, and 34). interested reader is referred directly to the
2. Adequate exclusion criteria. handbook for these data. The comprehensive
3. Information regarding gender, educa- tables present T-score equivalents for test
tional level, IQ, and recruitment strate- scaled scores for males and females separately
gies; data stratified by age. in 10 age groupings (20-34, 35-39, 40-44, 45-
4. Means and SDs reported. 49,50-54,55-59,60-64,65-69,70-74, 75-80)
by six education groupings (6--8, 9-11, 12, 13-
Considerations regarding use of the study 15, 16-17, 18+years). For the sample as a
1. No information regarding ethnicity or whole, mean errors was 39.6 (25.6).
geographic area (although it can be as- In 2004, the authors published revised norms
sumed it was Maine, given the academic based on a sample of over 1,000 normal adults.
affiliations of the authors). In addition to age, education, and gender
2. High educational and IQ levels. stratification, the data are partitioned by race/
ethnicity (African American and Caucasian).
[CT.20] Heaton, Grant, and Matthews, 1991
The authors provide normative data on the Study strengths
Category Test from 486 urban and rural 1. Large sample size.
participants recruited in several states (Cali- 2. T scores corrected for age, education,
fornia, Washington, Colorado, Texas, Oklahoma, and gender. The 2004 edition presents
Wisconsin, Illinois, Michigan, New York, data for two race/ethnicity groups.
Virginia, and Massachusetts) and Canada. Data 3. Adequate exclusion criteria.
4. Infonnation regarding IQ and geographic assumed it was Maine, given the aca-
recruitment area. demic affiliations of the authors).
2. High IQ level.
l. Above average mean intellectual level [CT.22] Barrett, Morris, Akhtar, and Michalek,
(which is probably less of an issue given 2001 (Table A24.21)
that this is WAIS rather than WAIS-R Test data were obtained on 1,052 Air Force
IQ data). veteran controls who served in Southeast Asia
from 1962 to 1971 in a study examining the
Other comments effects of Agent Orange on cognition. Partici-
1. The interested reader is referred to the pants averaged 43.9 (7.6) years of age, and
Fastenau and Adams (1996) critique of 5.3% were African American (with the rest
the Heaton et al. (1991) norms, and "nonb1ack"); 37% were officers, 16.6% were
Heaton et al.'s 1996 response to this enlisted Ryers, and 46.0% were enlisted ground
critique. crew; most of the officers were college-edu-
cated, and most enlisted personnel were high
[CT.21] Elias, Robbins, Walter, and Schultz, school-educated. No exclusion criteria are
1993 (Table A24.20) listed aside from epilepsy and low exposure to
Category Test data on 427 participants, in- dioxin. Participation was voluntary.
cluding those from the 1990 study and Mean number of errors is reported.
reRecting the same exclusion criteria, are
provided for men and women separately for Study strengths
six age groupings: 15-24 (37 men, 24 women), 1. Huge sample size.
25-34 (40 men, 56 women), 35-44 (36 men, 2. Infonnation regarding age, occupational
56 women), 45-54 (25 men, 46 women), status, and recruitment strategy, with
55-64 (25 men, 35 women), and 65 and over some limited data regarding ethnicity,
(24 men, 23 women). Participants with <12 educational level, and gender (assumed
and > 19 years of education were excluded all were male).
because participants outside this range were
disproportionately distributed across the age Considerations regarding use of the study
and gender groupings. Mean WAIS Vocabu- l. No exclusion criteria or infonnation re-
lary and Infonnation subtest scores ranged garding IQ.
13.9-14.7 and 13.2-13.7, respectively, across 2. No stratification of data by age or edu-
the age groups. cation.
Means and SDs for number of errors are 3. No SD reported for number of errors.
provided.
Study strengths
l. Large overall sample size, although most
OF THE CATEGORY TEST DATA
individual age x gender cells were <50.
(See Appendix 24m)
3. Information regarding educational level Data collected from the studies reviewed in
(although only ranges provided) and this chapter were combined in regression anal-
WAIS subtests (Information, Vocabu- yses, to describe the relationship between age
lary) provided and data stratified by age and test perfonnance and to predict estimated
and gender. test scores for different age groups. Effects of
4. Means and SDs reported. other demographic variables were explored in
follow-up analyses. The general procedures
Considerations regarding use of the study for data selection and analysis are described
l. No infonnation regarding ethnicity or in Chapter 3. Detailed results of the meta-
geographic area (although it can be analysis and predicted test scores across adult
CATEGORY TEST 495
age groups are provided in Table A24m.l in Limitation of the Analyses

Appendix 24m. 1. An effect of education on test scores
After initial data editing for consistency and in normal groups has been reported in
for outlying scores, 11 studies, which gener- several studies. Examination of this re-
ated 25 data points based on a total of 1,579 lationship did not yield meaningful re-
participants, were included in the analyses. sults in our data due to close association
Linear regression of the number of errors on of education with age.
age yielded an R2 of0.839, indicating that 84%
of variance in the scores is accounted for by the
model. Based on this model, we estimated CONCLUSIONS
number of errors for age intervals between 16
A large number of studies document the popu-
and 79 years. If predicted scores are needed
larity of the Category Test in clinical assessment.
for age ranges outside the reported age
The major drawbacks of the original version of
boundaries, with proper caution (see Chapter
the test are its length and lack of portability. To
3) they can be calculated using the regression
overcome these problems, more recent modifi-
equations included in the tables, which un-
cations focus on development of short forms and
derlie calculations of the predicted scores.
creation of booklet formats of the test. Despite
Linear regression of SDs on age yielded R2 of
the greater convenience offered by these mod-
0.469, indicating increase in variability with ad-
ified versions, their psychometric properties are
vancing age, consistent with the literature. Pre-
not yet sufficiently assessed. Most of the studies
dicted SDs, based on this modeL are reported.
address the reliability and validity of the short
Examination of the effects of demographic
forms based on analyses of extrapolated items
variables on the number of errors indicated
from the full version of the test, rather than on
that education did not contribute to test
data for the actual short version of the test. This
scores beyond its association with age in the
suggests caution in using these data in diagnostic
data available for analyses.
decision making and prompts further investi-
IQ had a significant effect on number of er-
gation of the psychometric properties of these
rors. However, the limited number of studies
versions. Correspondence between the original
that reported IQ does not allow close exami-
version and the booklet version of the test also
nation of this relationship.
deserves more attention. Further research also
The effect of gender was not examined as
needs to focus on standardization of instructions
information on gender distribution was not
for the test.
available in the data reviewed.
The clinical utility of the Category Test
would also be improved by adjusting cutoff
Strengths of the Analyses criteria relative to participants' age, education,
1. Total sample size of 1,579 participants. and intelligence level. Consideration of de-
2. R2 of 0.839, indicating a good model fit. mographic factors in assigning participants to
3. Postestimation tests for parameter spec- impaired vs. nonimpaired groups would im-
ifications did not indicate problems with prove the specificity of the Category Test. This
normality or homoscedasticity. would reduce the excessively high rates of
4. Significant effect of IQ is evident in the misclassification of "normal" participants in
data, consistent with the literature. the impaired range reported in the literature.
25
Wisconsin Card Sorting Test
BRIEF HISTORY OF THE TEST 64 cards to four key stimulus cards (Grant &
Berg, 1948; Heaton, 1981; Heaton et al., 1993).
The original version of the Wisconsin Card In their revised WCST manual, Heaton et al.
Sorting Test (WCST) was developed by Berg (1993) present a comprehensive review of the
and colleagues (Berg, 1948; Grant & Berg, variations in testing materials and administra-
1948). It was designed to assess abstract tiog. procedures used in previous studies.
reasoning and the ability to adapt cognitive Briefly, past versions have varied in the number
strategies to one's changing environment. For of response cards employed (e.g., two 48-card
this reason, the WCST is believed to measure a decks, two 60-card decks, or two 64-card decks),
complex range of executive functions. includ- type of designs used (e.g., standard figures or
ing planning. organizing, abstract reasoning, modified figures), and presentation style of the
concept formation, cognitive set maintenance stimuli (e.g., systematic or nonsystematic con-
and shifting ability, and inhibiting impulsive figurations in random or standardized order).
responses (Lezak, 1995; Lezak et al., 2004; Furthermore, test administration procedures,
Spreen & Strauss, 1998). such as discontinuation rules and scoring crite-
The WCST was primarily based on abstract ria for the specific outcome measures, have
reasoning and learning research conducted with varied widely among past studies.
primates (Zable & Harlow, 1946) and the Weigl Heaton (1981) published the first compre-
Color-Form Sorting Test, designed for assess- hensive administration, scoring, and norma-
ment of reasoning skills in humans (Weigl. tive manual on the two-deck, 64-card version
1941). The original version of the WCST in- of the WCST. In the first manual, he stan-
volves sorting either 60 response cards (Berg. dardized the Grant and Berg (1948) testing
1948) or 64 response cards (Grant & Berg, 1948) procedures, documented precise scoring pro-
to four key stimulus cards. If participants do not cedures, and presented normative data based
achieve the criterion number of sorts, the ex- on relevant demographic factors. Heaton et al.
aminer rearranges the deck of cards and testing (1993) later revised and updated the manual
is continued until the expected number of cat- to include wider normative age ranges (6.5-
egorical sorts is achieved or the second deck of 89 years), further clarify scoring criteria with
cards is sorted. This original format has been explicit examples, and revise the scoring forms
modified over time, and the most popular ver- to facilitate recording responses and calculat-
sion of the test now involves sorting two decks of ing outcome measures.
496
WISCONSIN CARD SORTING TEST 497
Since the publication of Heaton's manuals The WCST has a number of useful outcome
(Heaton 1981; Heaton et al. 1993) the 64-card, measures, which can be derived with use of
two-deck version of the WCST has gained (sometimes complex) sets of scoring rules (see
popularity and is the most frequently admin- the revised WCST manual for details; Heaton
istered format of this test for both clinical and et al., 1993). Briefly, some of the outcome
research purposes. In this version, the partic- measures include the following: (1) Trials to
ipant is presented with two decks of 64 stim- Complete First Category refers to the number
ulus cards (one deck at a time) and asked to of trials taken to complete sorting to the first
sort each card in the deck, one card at a time, principle; (2) Categories Completed (some-
to one of four key reference cards presented in times referred to as Categories Achieved) is an
a predetermined order. A single red triangle is indication of the total number of correct
printed on the first reference card, two green principles sorted, and credit for obtaining one
stars are on the second card, three yellow category sort is given when the participant
crosses are printed on the third card, and four completes 10 consecutive correct matches to
blue circles are printed on the fourth card. one principle; (3) Failure to Maintain Set
Each of the stimulus cards in the two decks is (sometimes labeled Set Failure) refers to the
unique in terms of design form (triangle, star, inability to complete the category sort and
cross, or circle), color (red, green, yellow, or occurs when participants have obtained at
blue), and number of items (one, two, three, or least five accurate categorical matches before
four designs per card). The participant is re- shifting their sorting strategy; (4) Percent
quired to match stimulus cards to response Conceptual Level Responses is a measure of
cards based on one of three sorting principles the participant's insight into the correct sort-
(color, form, or number). However, there are ing principle and is based on "consecutive
times when a response card matches a key card correct responses occurring in runs of three or
on two or more principles (e.g., color and more" (Heaton et al., 1993); (5) Learning to
form), and it is up to the examinee to decide Learn indicates the participant's "average
the principle he or she must follow. The first change in conceptual efficiency across the
sorting principle is color, followed by form, consecutive categories (stages) of the WCST"
and then by number. Testing is concluded (Heaton et al., 1993; calculating this measure
when this sequence of sorting is completed requires a number of sequential steps that are
two times by the participant or when all best described in the Heaton et al. manual);
128 cards have been sorted. (6) Total Errors refers to the number of
Participants are provided with very little sorting errors made throughout the task; and
specific information as to how to execute the (7) Perseverative Errors is the number of re-
task. Essentially, they are told to take one card petitive errors (i.e., sorting to a single, wrong
from the top of the deck and place it beneath principle repeatedly despite negative feed-
one of the four key cards. They are not given back) while (8) Nonperseverative Errors re-
any indication as to what principle to use flects errors but not those that are repetitively
when sorting and are provided only with the made to a single, wrong sorting princi-
feedback "right" or "wrong" after each card ple. Other outcome measures include Percent
sort. Thus, it is the task of the participant to Conceptual Level Responses, Percent Persev-
determine, from this ambiguous feedback, erative Responses, Percent Perseverative Er-
what principle he or she is to use for sorting. rors, Percent Nonperseverative Errors, and
Additionally, when the appropriate number of "Other" Responses (also referred to as Unique
sorts has been made to a given category, the Errors, matches that are not based on any of
examiner changes the sorting principle with- the three sorting principles). A more detailed
out alerting the participant. Other modified description of these outcome variables as well
versions of the WCST, in which the partici- as specific scoring criteria can be found in the
pant is notified when the sorting principle is WCST revised manual (Heaton et al., 1993).
about to change, will be discussed in later As mentioned above, Heaton et al. (1993)
parts of this chapter (Nelson, 1976). have very clearly presented the scoring criteria
and supplemented them with detailed exam- ries and commit more perseverative, but not
ples in the revised manual. Prior to the nonperseverative, errors than patients with
development of the revised manual, Axelrod nonfrontal damage. The results of the meta-
et al. (1992b) created supplementary scoring analysis further indicated that the lateralization
material and found that the interrater reli- of the lesion did not affect WCST performance
ability among novice raters did not change in frontallobEHlamaged patients. Nonetheless,
with the additional material but that scor- several clinical studies have found that the
ing time was reduced by 41%. Nonetheless, WCST is incapable of discriminating patients
Greve (1993) rightly points out that scoring with frontal lesions from those with lesions
errors are still likely to occur due to the com- in other brain regions (Anderson et al., 1991;
plex nature of calculating some of the WCST Axelrod et al., 1996; Drewe, 1974; Grafman
outcome measures. Fortunately, clinicians and et al., 1990; Hermann et al., 1988; Homer et al.,
researchers can now use computer scoring 1996; Mountain & Snow, 1993; Stuss et al.,
software, which very precisely and quickly 1983).
produces accurate scores as well as normative More recently, functional neuroimaging
data corrected by demographic factors such as studies have confirmed that WCST perfor-
age and education. However, scoring data with mance is related to activation of dorsolateral
this software requires that the stimuli cards be prefrontal regions (Berman et al., 1995; Ko-
sorted in a predetermined order. In a study nishi et al., 1998; Nagahama et al., 1998).
using the computerized scoring method, Additionally, associations between persevera-
Greve (1993) found that a novice could score tive behavior on the WCST and prefrontal
at the "expert" level. lesions, prefrontal volumes, and prefrontal cor-
tex activation have been solidly established
(Cabeza & Nyberg, 2002; Raz et al., 1998;
Anatomical Correlates and Effect of
Rogers et al., 2000; Stuss et al., 2000). In
Brain Pathology on the WCST
young healthy subjects, WCST performance
In one of the earliest clinical studies using the has been shown to activate the dorsolateral
WCST, McFie and Piercy (1952) found that prefrontal cortex (Berman et al., 1995), and in
the laterality of the lesion, rather than its site older adults it has been associated with pre-
within the hemisphere, was the cause of poor frontal cortex volume (Head et al., 2002).
sorting performance. However, later studies Esposito et al. (1999) used PET to study the
by Milner (1963; 1964) found that patients relationship between regional cerebral blood
with dorsolateral prefrontal cortex damage flow (rCBF) and the WCST and Raven Pro-
performed worse on the WCST than those gressive Matrices in individuals aged 18-80.
with damage to the orbitomedial or other Among other findings, they report that the left
brain regions. Using her version of the WCST, dorsolateral prefrontal cortex was activated by
Nelson (1976) found that patients with frontal both tasks in the young group, but in the older
lobe lesions obtained fewer categories and group activation decreased in this region for
made more perseverative errors than those the WCST but not the Raven Progressive
with lesions elsewhere in the brain. Following Matrices.
these early findings, numerous other studies Frontal lobe dysfunction has been specu-
have demonstrated similar results (Anderson lated to be the primary reason for the neu-
et al., 1991; Bomstein, 1986c; Drewe, 1974; ropsychological and psychiatric symptoms in
Heaton 1981; Robinson et al., 1980). A re- schizophrenia (Weinberger et al., 1994). Ac-
cent meta-analysis by Demakis (2003) of the cordingly, functional MRI and single-photon
WCST using published studies on patients emission computed tomography (SPECT)
with damage to frontal and "nonfrontal" brain studies have repeatedly shown less prefrontal
regions as well as left frontal vs. right frontal activity in schizophrenics compared to controls
brain-damaged patients reported similar on the WCST (Berman et al., 1993; Catafau
findings. Demakis found that patients with et al., 1994; Kawasaki et al., 1993; Parellada
frontal damage achieve far fewer catego- et al., 1994; Sagawa et al., 1990a,b; Seidman
et al., 1994; Weinberger and Berman, 1986; studies have found increased rCBF in the
Weinberger et al., 1986, 1989), with some data prefrontal cortex in patients with Huntington's
indicating an absence of activation for the right disease (Goldberg et al., 1990; Weinberger
frontal region in schizophrenics (Volz et al., et al., 1988) and Down syndrome (Schapiro
1997). Using transcranial Doppler sonography, et al., 1999), despite poor performance on the
a noninvasive method designed to measure the task. Based on these findings, Schapiro et al.
CBF velocity of the basal cerebral arteries, (1999) caution that prefrontal activation may
Schuepbach et al. (2002a,b) found that there be an indication of mental effort exerted ra-
was an increase in blood flow in controls, ther than level of task performance alone.
but not in schizophrenics, after shifting set on
the WCST. Haut et al. (1996) found that pa-
Brief Overview of Clinical Findings
tients with schizophrenia performed similarly
Using the WCST
on the WCST to patients with right frontal lobe
tumors but not those with left frontal lobe tu- In the past several years, the WCST has been
mors, nonfrontal tumors, and healthy controls. used in hundreds of clinical studies on pa-
Evoked-potential studies of healthy indi- tients with various psychiatric and neurologi-
viduals link WCST perseverative errors to the cal disorders to characterize frontal systems
frontal-extrastriate network (Barcelo, 1999). dysfunction. These studies have added valu-
In patients with Parkinson's disease, P300 la- able information regarding the clinical utility
tency response (a physiological measure of of this test and have assisted in determining the
cognitive processing) was correlated with num- expected pattern of performance for clinical
ber of WCST categories completed (lijima disorders. Due to the large volume of such
et al, 2000); and in both depressed and non- studies, not all will be reviewed in this chapter.
depressed elderly, longer P300 latency was Instead, a brief overview of the clinical findings
related to greater total errors and poorer is provided below.
ability to maintain set but not to perseverative As described earlier, patients with pre-
errors (Kindermann et al., 2000). dominantly prefrontal lobe lesions perform
While studies have demonstrated that re- poorly on the WCST (Anderson et al., 1991;
gions of the frontal lobes are activated during Bomstein, 1986c; Heaton, 1981; Milner, 1963;
performance of the WCST, the test's specifi- Milner, 1964; Nelson, 1976; Robinson et al.,
city to activation confined exclusively to the 1980).
prefrontal cortex has not been demonstrated Patients with various neurodegenerative
in all studies. This is not surprising given that diseases, such as Alzheimer's disease (Binetti
the WCST most likely requires multiple, et al., 1995; Bondi et al., 1993), Huntington's
parallel cognitive processes (Anderson et al., disease (Paulsen et al., 1995b), frontotemporal
1991; Dehaene & Changeux, 1991). Func- dementia (FTD; Boone et al., 1999; Razani
tional imaging studies have found the WCST et al., 2001), and Parkinson's disease (Alevria-
task to activate other brain regions, such as the dou et al., 1999; Beatty & Monson, 1990; Paolo
bilateral inferior parietal lobe and the inferior et al., 1995; Tomer et al., 2002), also exhibit
posterior temporal lobe (Berman et al., 1995), impaired WCST performance. Bondi et al.
the bilateral supramarginal gyrus and the an- (1993) found that Nelson's version of the
terior cingulate cortex (Konishi et al., 1998), WCST was quite sensitive and specific in dif-
and the bilateral inferior parietal lobe and the ferentiating patients with Alzheimer's disease
left superior occipital gyrus (Nagahama et al., (even those in the mild stages) from normal
1996) in addition to the frontal regions. Test controls. Using receiver operating character-
version and administration procedures might istic (ROC) curves, the authors found that the
explain some of the variability observed (Stuss number of categories completed best classified
et al., 2000). Alzheimer's patients and controls (sensitivity
Interestingly, poor performance on the of 94% and specificity of 87% using optimal
WCST does not always lead to lack of CBF cutoff scores) but that number of perseverative
activity in the prefrontal cortex. In fact, some errors best distinguished Alzheimer's patients
with mild symptoms from controls (sensitivity perseverative responses, and perseverative
of 74% and specificity of 87% using optimal errors. In this same study, Parkinson's patients
cutoff scores). Paulsen et al. (1995b), ~so us- with dementia committed more total errors
ing ROC curves, found the WCST to be quite than patients with Alzheimer's disease but the
effective at discriminating patients wtth Alz- two groups performed equally poorly on all
heimer's and Huntington's dementia from nor- other WCST measures. Alevriadou et al.
mal controls (90% classification accuracy) but (1999) found that, in addition to greater per-
that accuracy rates for discriminating among severations and set failures, patients with
the two dementia groups were far lower (63% Parkinson's disease who did not display cog-
classification accuracy). Razani et al. • (2001) nitive impairment required more trial ad-
examined WCST performance in F:fD pa- ministrations relative to controls, a finding
tients with asymmetrical left- or right-sided that is in contrast to Taylor et al. (1986)
anterior hypoperfusion and in those wlth Alz- and Cooper et al. (1991). Overall, the findings
heimer's disease. In general, while all three on patients with Parkinson's disease suggest
dementia groups performed poorly on the that they require more trials to develop a
WCST relative to normal controls, the FTD problem-solving strategy but that even after
patients with right-sided hypoperfusion dis- developing a strategy they tend either to
played the greatest difficulty. Patients with have difficulty switching from one sorting
right-FTD committed significantly more per- principle to another (perseverative) or to aban-
severative errors than patients with left-FTD don their strategy prematurely (failure to
and Alzheimer's disease, and the rig~t-FTD maintain set).
patients scored significantly below the left- A number of researchers have demon-
FTD patients on Percent Conceptual Level strated that patients with traumatic brain in-
Responses. jury (TBI) commit more perseverative errors
Documented deficits in patients wi~ Par- relative to controls (Ferland et al., 1998; Se-
kinson's disease include inability to niamtain galowitz et al., 1992; Stuss, 1987). Little et al.
set (Lees & Smith, 1983; Flowers & Ro- (1996) explored the implications of WCST
bertson, 1985; Taylor et al., 1986) and in- performance for daily life activities in patients
creased number of total and perseverative, but with TBI and found that the WCST correlated
not nonperseverative, errors (Paolo ·et al., moderately, and better than the Category
1995). Similar to patients with frontal lobe Test, with a daily living functional measure.
lesions, patients with Parkinson's disease seem Coelho (2002) analyzed the speech patterns
to benefit from cues to shift set (Fimll} et al., of patients with closed head injuries and
1994; Hsieh et al., 1995), even when those found that performance on the WCST may
cues do not explicitly direct them to the cor- predict sentence complexity, organization, and
rect solution (van Spaendonck et al., 1995); content. Sherer et al. (2003) found that Per-
but compared to patients with frontal lobe severative Errors, Perseverative Responses,
damage, the performance of patients with Categories Completed, and Conceptual Level
Parkinson's disease improves after the second Responses (which clustered on one princi-
shift in set (van Spaendonck et al., 1995), pal component) on both the full and 64-card
while patients with frontal lesions continue to versions of the WCST predicted the func-
have difficulty throughout the task (Heaton, tional level of patients with closed head
1981). Interestingly, Parkinson's patient5 with- injuries at the time of discharge from
out dementia also appear to perform worse rehabilitation.
than healthy controls (Bondi et al., 1993; There has been an abundance of research
Caltagirone et al., 1989; Tsai et al., 1994). In a with the WCST in the past two decades on
study using Nelson's modified version of the patients with schizophrenia and schizophrenia-
WCST, Paolo et al. (1996b) found that Par- related disorders. Among other WCST out-
kinson's patients who were matched Qn age, come measures, most studies report that
education, and overall intellectual abUity to patients with schizophrenia exhibit increased
normal controls committed more total errors, perseverative errors and/or complete few
categorical sorts (Beatty et al., 1994; Bustini spatial working memory may be responsible
et al., 1999; Ismail et al., 2000; Martinez et al., for number of perseverative errors and cate-
2002; Morice, 1990; Parellada et al., 1994, gories achieved but not nonperseverative
2000; Raine et al., 1992; Rossi et al., 2000; errors (Gooding & Tallent, 2002).
Schuepbach et al., 2002b; Weinberger et al., Individuals with schizotypal personality
1986), and these deficits do not appear to be disorder display some of the same deficits on
a result of inadequate effort or motivation the WCST as patients with schizophrenia, in-
(Ilonen et al., 2000). In fact, factor-analytic cluding completing fewer categorical sorts,
studies have found that WCST scores that more perseverative errors, and more set fail-
cluster on a perseveration-type factor are ures compared to controls (Lenzenweger &
the most diagnostically useful in characteriz- Korfine, 1994; Trestman et al., 1995). Most
ing executive dysfunction in schizophrenics studies suggest that nonschizophrenic siblings
(Koren et al., 1998). However, WCST per- of schizophrenic patients tend to perform
formance appears to vary by schizophrenia relatively normally on the WCST (Scarone
subtype and/or symptoms (Braff et al., 1991; et al., 1993; Ismail et al., 2000; Yurgelun-
i.e., those with greater frontal lobe abnor- Todd & Kinney, 1993), but at least one study
malities display greater WCST dysfunction). has found that schizophrenics and their heal-
Support for this notion comes from studies thy siblings committed greater percentages of
that have found a relationship between chron- perseverative errors relative to controls (Saoud
icity of the illness and WCST perseverative et al., 2000). Another study found that the
errors (Braff et al., 1991; Butler et al., 1992), offspring of schizophrenic patients achieved
between symptom severity and increased fewer categorical sorts and committed more
perseverative errors (Bomstein et al., 1990), perseverative errors and responses relative to
between disorganized symptoms and fewer healthy controls (Wolf et al., 2002).
categorical sorts as well as greater persevera- Studies examining WCST performance in
tive errors (Dahan et al., 2002), and between other psychiatric populations have produced
level of insight and WCST performance interesting, but at times mixed, results. Some
(Chen et al., 2001; Lysaker & Bell, 1994) and studies have reported impairment on the
studies that have shown alterations in WCST WCST in patients with obsessive-compulsive
performance as a result of the type of neuro- disorder, particularly decreases in categories
leptic medication used (i.e., respiredol was completed and/or increases in number of
better than olanzapine; Rybakowski & Bor- errors committed (Christensen et al., 1992;
kowaska, 2001). Additional studies have Harvey, 1986; Head et al., 1989; Malloy, 1987);
shown that more perseverative errors are but others have not found such impairments
committed by paranoid relative to non- (Boone et al., 1991; Gross-Isseroff et al., 1996).
paranoid patients (Abbruzzese et al., 1996), by Similarly, some investigators have reported
patients with negative symptoms compared poor performance on the WCST in pa-
to positive symptoms (Braff, 1989), and tients with depression, with patients typically
by those with early-onset (mean age 27.6) achieving fewer categorical sorts, more errors,
relative to late-onset (mean age 54.7 years) and set failures (Austin et al., 1992; Axelrod
schizophrenia (Jeste et al., 1995). In contrast, et al., 1994a,b; Boone et al., 1995; Channon,
other investigators have not found differences 1996; Martinet al., 1991, while others have not
in WCST scores between patients with posi- found marked deficits (Fossati et al., 1999,
tive and negative symptoms or a relationship 2001). Ishikawa et al. (2001) found that
between WCST performance and chronicity "successful" (i.e., nonconvicted) psychopaths
of illness or neuroleptic treatment (Parellada committed fewer total errors and fewer
et al., 2000). Gambini et al. (1992) found nonperseverative errors and achieved a
that factors such as education affect WCST greater number of category sorts relative to
scores in schizophrenic patients and suggest "unsuccessful" (i.e., convicted) psychopaths
that studies control for this demographic and normal controls. No differences, how-
factor. In schizophrenic patients, deficits in ever, were noted between the unsuccessful
psychopaths and controls on any of the WCST by cardiovascular health. Takashima et al.
measures. (2003) found that factors such as age, greater
Patients with multiple sclerosis (MS) have multiple lacunar infarcts, and lower HDL
been shown to attain fewer categories and cholesterol best explained scores on tests of
commit more perseverative responses and executive functioning, including a modified
perseverative errors on the WCST, secondruy computer version of the WCST. Hanninen
to frontal lobe dysfunction (Beatty et al., 1989, et al. (1997) found that individuals with age-
1995; Heaton et al., 1985; Rao et al., 1991b). associated memory impairment also perform
Poor performance on the WCST is correlated poorer than controls on the number of cate-
with frontal lobe lesion volumes in patients gories achieved, overall correct responses, and
with MS (Arnett et al., 1994). perseverative errors.
Studies of the effects of chronic and acute The identification of malingering has been
alcohol use on the WCST have provided in- studied with the WCST. A number of inves-
triguing results. Early studies found that tigators have developed regression equations
chronic alcoholics tend to be quite persever- based on various WCST outcome measures
ative on the WCST (Parson, 1975; Tarter and that discriminate between malingerers and
Parsons, 1971). Sullivan et al. (1993) com- nonmalingerers (Bernard et al., 1996; Suhr &
pared the WCST performance of patients with Boyer, 1999). The basic premise behind de-
chronic alcoholism to patients with schizo- riving such formulas is that it is more difficult
phrenia, frontal lobe lesions, and normal for individuals to feign a specific pattern of
controls and found that the alcoholic group performance. Bernard et al. (1996) used Ca-
scored highest on an "insufficient sorting" tegories Completed and Perseverative Errors
factor (i.e., failed to maintain set) relative to in deriving their malingering formulas. Cate-
the schizophrenic and frontal lobe groups, gories Completed was considered the "ob-
who performed worse on a perseverative fac- vious" measure since most individuals who
tor score. Other studies have demonstrated attempt to perform poorly would know not to
that WCST scores are predictors of risk for score well on this outcome measure. In con-
alcoholism (i.e., WCST scores correlate with trast, Perseverative Errors were considered
MacAndrew Alcoholism Scale; Deckel, 1999). the "subtle" measure since most individuals
Interestingly, acute alcohol intoxication can would not intuit the significance of commit-
also lead to increased perseverative errors on ting this type of error. Using this method,
the WCST (Lyvers & Maltzaman, 1991). Bernard et al. (1996) adequately classified
Poor WCST performance has also been brain-injured individuals (sensitivity 86%) and
linked with specific health factors. Fewer simulated malingerers (specificity 94%). Clas-
categorical sorts and increased perseverative sification rates were somewhat lower when
errors have been reported for those with discriminating simulated malingerers from a
chronic obstructive pulmonary disease (Crews mixed group of neurological patients (i.e.,
et al., 2001). Boone et al. (1992) found that an sensitivity 58%, specificity 90% ). Bernard et al.
increase in the size of white-matter lesions in (1996) found a false-positive rate of approxi-
older adults, particularly when the lesions mately 5%, which was replicated by Donders
exceeded a total volume of 10 cm2, led to an (1999b). Given the high correlation between
increase in the number of perseverative errors Categories Completed and Perseverative Er-
and a decrease in the number of categories rors, Suhr and Boyer (1999) chose failure to
completed. Boone (1999) also found that risk maintain set and Categories Completed in
for vascular illness was a significant predictor their set of logistic regression formulas used
of Categories Completed, Total Errors, Per- for the detection of malingering. These in-
severative Responses, Percent Conceptual vestigators found that these formulas classified
Level Response, and Trials to First Category. undergraduate simulating malingerers and un-
This finding is consistent with those of Dywan dergraduate nonmalingerers at relatively high
et al. (1992), who reported that 28% WCST rates, 70.7% and 87.1% respectively. Patients
performance in older adults was accounted for suspected of malingering were discriminated
from brain-injured patients at even higher (1976) suggested obtaining the following out-
rates (i.e., sensitivity= 82.4% and specific- come measures: Categories Completed, Total
ity=87.5%). Subsequent studies, however, Errors, Nonperseverative Errors, Persevera-
have found that classification with either Ber- tive Errors, and Percent Perseverative Er-
nard et al.'s (1996) or Donders' (1999b) for- rors ([Perseverative Errors/fotal Errors] x
mulas can produce false-positive rates as high 100). Additional scores, such as Conceptual
as 41% depending on the clinical sample used Level Responses and Trials to Complete First
and the severity of illness (Greve &: Bianchini, Category, can also be obtained using Heaton
2002). It is clear that further studies using ei- et al.'s (1993) scoring criteria (Nagahama et al.,
ther regression-based formulas or cutoff scores 2003).
on the WCST are needed to adequately evalu- Investigators who argue that this shortened,
ate whether this test is useful in detecting simplified version of the WCST is not sensi-
malingerers. tive to detecting mild cognitive impairment
have altered the MCST to include another
Modifications and Alternate deck of 24 cards (total of 72 unambiguous
Formats of the WCST cards) and have changed the instructions so
that participants are no longer alerted when
Modified Card Sort Test (Mcsn the sorting principle is changed (Hart et al.,
This version has also been referred to as the 1988; Jenkins &: Parsons, 1978).
modified WCST (mWCST; Bondi et al., 1993; Direct comparison of the MCST and the
Lineweaver et al., 1999; Nagahama et al., WCST is difficult given the substantial differ-
2003). Nelson (1976) altered the WCST sub- ences between the two tests; however, some
stantially by reducing the number of stimulus studies have demonstrated the MCST's sensi-
cards, eliminating ambiguous responses (i.e., tivity to detection of brain damage (Bondi
cards with more than one shared attribute et al., 1993; Nelson, 1976; Vanden Broek et al.,
with the stimulus cards), and altering exam- 1993). Lineweaver et al. (1999) reported
iner feedback to participants. Because this moderate test-retest reliability for this version
version eliminates the ambiguity in respond- (see Psychometric Properties of the WCST
ing, it is thought to be best for patients with below) and provided extensive normative data
severe impairment. De Zubicaray and Ashton as well as raw score to standard score conver-
(1996) offer an excellent review of the studies sion tables that correct for age and educational
that have used the MCST. level.
In this version, the same four reference
(key) cards as the 128-card version are used; 64-Card WCST Version
however, the stimulus cards that share more This shorter version of the 128-card WCST is
than one attribute with the reference card gaining popularity among clinicians and re-
(e.g., color and number) have been elimi- searchers given the shortened administration
nated, thereby leaving two decks of 24 re- time as well as the similarity in administration
sponse cards (Nelson, 1976). Thus, the procedures and scoring criteria to the 128-
response cards share only one attribute with card version. Throughout the literature, this
the stimulus cards (i.e., color with the first version has also been referred to as the "ab-
card, form with the second card, and number breviated WCST" or the 'WCST-64.'' Test
with the third card) and no attribute with the administration is virtually identical to the 128-
fourth card. Whichever category the partici- card version, except that only the first deck (64
pant matches is determined to be the first cards) is presented to participants (Kongs
principle, and each sorting principle changes et al., 2000). As with the full version, 10 cor-
after six, not ten, consecutive correct sorts. rect sorts to the predetermined principle
Once participants obtain six correct sorts to a constitute a complete categorical sort and the
category, they are instructed to find another task is discontinued when all of the 64 cards
rule to which to sort. The test is discontinued have been sorted. Axelrod et al. (1992a,b)
after completion of six categories. Nelson found a moderate correlation between this
version and the full 128-card version for all these are not norms that are used in individual
measures (most coefficients ranging 0.64- cases. For the demographically corrected
0.74), except for total correct response$, which T scores, they found that correlations between
was low (coefficient 0.22) yet statistically sig- the WCST-64 and the full version were
nificant. The authors argued for the validity of relatively high (r= 0.75-0.88) but that the
this shortened version for use with ;healthy WCST-64 captured only approximately 59%
individuals but noted that its clinical utility of the full WCST scores when a 5-point
would have to be further studied. margin of error was set. Merrick et al. (2003)
Studies have since found this shortened reported similar findings for a group of head-
version to yield clinically meaningful data for injured patients. They found that T scores for
patient groups such as those with Parkinson's perseverative responses were on average over
disease (Paolo et al., 1996a,b; Robinson et al., half a standard deviation (SD) lower for the
1991). However, Axelrod et al. (1996) argue WCST-64 relative to the full version for these
against simple conversion of scores to per- patients. In fact, for a quarter of the sample,
centages to obtain demographically c<tJ"ected T scores for perseverative responses obtained
standard scores based on the full W'CST. for the WCST-64 were over 10 points (1 SD)
When they examined the data by ~vel of lower relative to the full version. These find-
performance, they found that the best con- ings suggest that for brain-damaged patients
sistency between the 64- and 128-card ver- the 64-card version may not be equivalent to
sions could be obtained when five cat,gorical the full WCST.
sorts had been made within the fir~ deck.
Consistency between the two versi~s de- Computerized Administration Version
creased when fewer categorical sortS were A number of commercial and noncommercial
achieved within the first deck and ~ quite computer administration and scoring software
low for Perseverative Errors and Perseterative programs exist for the WCST (Harris, 1988;
I
Responses. Kongs et al. (2000) develbped a Heaton, 1993, 2003a,b; Keller & Davis, 1998).
normative manual for this 64-card ve~ion by Most computerized versions are designed
essentially reanalyzing responses from the first to administer the 128-card or the 64-card
deck according to the data used in the Heaton WCST. In most cases, the testing procedures
et al. (1993) normative manual (see review of are kept similar, if not identical, to the ad-
the manual in Summaries of the Studies, be- ministration manual (Heaton, 1981; Heaton
low). However, two follow-up studies (~elrod et al., 1993). Thus, the four reference cards
2002; Merrick et al., 2003) raise concerns re- appear at the top of the computer screen, and
garding using the 64-card version· inter- the task of the participant is to sort each card
changeably with the full 128-card veJliion in that appears at the bottom (typically center)
both healthy and patient groups. Using 332 of the screen to one of the reference cards by
clinical protocols, Axelrod (2002) co*verted pressing a computer key or using the mouse.
the 128-card responses into standard scores Written feedback is provided to the partici-
based on the Heaton et al. (1993) nonnative pant (i.e., the word right or wrong appears
data, converted the first deck (64-card) after each match). In some computerized
responses into standard scores based !on the versions, a combination of written and audi-
Kongs et al. (2000) normative data, an«! trans- tory cues (i.e., different tones for correct and
formed the responses from the first d'ck (64 incorrect responses) are provided.
cards) into percentages and then into squtdard Artiola i Fortuny and Heaton (1996) com-
scores based on normative data frqm the pared the standard (manual) WCST adminis-
full (128-card) WCST. Additionally, the same tration to a computerized version in a sample
scores were compared to census scores re- of healthy adults in Madrid, Spain, and found
ported in the full WCST (Heaton et al.1 1993) no differences between any of the outcome
and the WCST-64 (Kongs et al., 2000) mimuals. measures, except trials to complete first cate-
The authors noted that while the WCST and gory. Specifically, it appears that individuals
the WCST-64 census data were comPru-able, require an average of six more trials to obtain
the first category in the computerized com- similar to that of the Heaton et al. (1993)
pared to the manual administration. The au- sample, suggesting that their findings can be
thors speculate that individuals may need more generalized to that normative data set. Fer-
time to become familiar with the computer- land et al. (1998) also found significant im-
testing format. Conversely, Feldstein et al. provement in WCST performance (e.g., 38%
(1999) found that central tendency measures for Perseverative Errors and 40% for Perse-
and variability scores differed for one of the verative Responses) of patients with TBI when
computerized versions (Keller & Davis, 1998) testing probes were separated by 5 months.
relative to the manual administration of the Although WCST scores improved for the nor-
WCST. mal control group as well, these differences
did not reach statistical significance.
Lineweaver et al. (1999) reported modest
test-retest (probes separated by 1 year) cor-
There is surprisingly sparse information relation coefficients for Nelson's MCST. Co-
available regarding the reliability of the 128- efficients were 0.46 for Nonperseverative
card WCST. The Heaton et al. (1993) manual Errors, 0.56 for Categories Completed, and
reports adequate internal reliability coeffi- 0.64 for Perseverative Errors. However, no
cients (range 0.37-0.72) that are based on the significant practice effects, particularly for cat-
generalizability theory (i.e., how well the test egories completed and perseverative errors,
depicts a participant's true score) for healthy were observed.
children and adolescents. The manual, how- Relatively strong intra- and interrater scor-
ever, does not report test reliability data for ing reliabilities have been obtained for hand
adults who are over 18 years of age. scoring of the WCST. Axelrod et al. (1992a)
Paolo et al. (1996a) found moderate to low reported interrater scoring reliability coeffi-
test-retest reliability for testing probes sepa- cients of 0.93 for Perseverative Responses,
rated by 1 year. They found correlation coef- 0.92 for Perseverative Errors, and 0.88 for
ficients ranging from as low as 0.12 (for Nonperseverative Errors. Additionally, they
Learning to Learn) to 0.65 (for Categories found excellent consistency in the way raters
Completed). They also reported stability co- applied the scoring rules, with correlation
efficients for a 1-year retesting period to be coefficients ranging 0.91-0.96. Despite ade-
rather low, range from 0.55 (Nonperseverative quate inter- and intrascorer reliability, Ax-
Errors) to 0.66 (Total Errors). These authors elrod et al. (1994b) found that accuracy in
expressed concern over the reliability and poor scoring can improve even more for novice
stability of the WCST measures and indicated scorers if they use the written supplements
that the WCST may not accurately measure produced by Axelrod et al. (1992a). Greve
change in problem-solving skills in normal (1993) found that clinical neuropsychologists
adults. Ingram et al. (1999) reported test- who used the Heaton (1981) manual were less
retest (testing probes separated by 1-71 days) accurate than "experts" (neuropsychologists
reliability coefficients of 0.34 (total correct with at least 5 years of experience with the
responses) to 0.83 (perseverative responses) WCST) and novices. Finding that the com-
for 11 WCST outcome measures in a group of puter scoring method assisted novices in their
patients with untreated sleep apnea. scoring accuracy, Greve (1993) recommended
Paolo et al. (1996a) also found practice efits use.
fects (for testing probes separated by 1 year), Heaton et al. (1993) cite two studies that have
ranging from one-third to nearly one-half of a demonstrated adequate concurrent validity for
standard deviation, for all outcome measures the WCST. Shute and Huertas (1990) found
of the 128-card version of the WCST, with the that in a group of college students Perseverative
exception of Categories Completed, Trials to Errors loaded on a Piagetian measure of formal
Complete First Category, Learning to Learn, operational reasoning. Similarly, Perrine (1993)
and Nonperseverative Errors. They note that found that Total Errors, Categories Com-
the performance of their sample was quite pleted, and Perseverative Responses were
correlated with an attribution identification Investigators have found a weak relation-

test (a measure of concept formation). The ship between the WCST and another popular
validity of the WCST as a measure of execu- test of reasoning, the Category Test (Crockett
tive functioning and frontal lobe dysfunction et al., 1986; Donders & Kirsch, 1991; Pen-
has also been demonstrated in various patient dleton & Heaton, 1982; Perrine, 1993). Perr-
populations (see Brief Overview of Clinical ine (1993) attributes this weak relationship (of
Findings Using the WCST, above). approximately 30% shared variance) to the
Wildgruber et al. (2000) examined the fact that the two tests measure different un-
sensitivity and specificity of the WCST Per- derlying aspects of conceptual processes. He
severative Responses scores of patiet.ts with argues that the WCST measures "attribute
frontal lobe damage, nonfrontallobe damage, identification" (selection of critical features
a mixed group of brain-damaged individuals, for encoding and categorizing), while the
and controls. They found that WCS~ Perse- Category Test assesses "rule learning" (relat-
verative Responses was not very sen.itive at ing two or more concepts with a logical rule).
detecting frontal lobe patients (only i~ntified Interesting test order effects have been ob-
65.4%) or very specific at classifying pontrols served by Brandon and Chavez {1985) when
(60.9%). Other studies, however, ha~ found administering the Category Test and the
the WCST to be effective at discri~inating WCST. Essentially, these authors found that
healthy controls from brain-damaged .ndivid- while administering the WCST first did not
uals (discrimination accuracy was .pproxi- alter participants' scores on the Category Test,
mately 71% with all outcome measure~). Rossi perseverative responses and total errors ten-
et al. (2000) found that the overall clilssifica- ded to decrease for the WCST when the
tion rate of the WCST was 60.59% when Category Test preceded the WCST. This ef-
discriminating between patients with: schizo- fect was present across various administration
phrenia, bipolar disorder, and healthy con- delays, including back-to-hack test adminis-
trols. However, the classification ~tes of trations and when administration of the tests
patient groups (schizophrenia 48.5%,~bipolar was separated by 1 hour or by 24 hours.
40%) were far lower than that of ~on trois Recent investigations into the ecological
(85.9%). Additional studies with clinical pop- validity of the WCST have demonstrated that
ulations have found that discrimination be- it can predict the ability to carry out activities
tween various psychiatric and neurt>logical of daily living and the type of occupational
conditions (e.g., schizophrenia, moocJ disor- position one is likely to hold (Kibby et al.,
ders, and head injury) is difficult ~th the 1998; Little et al., 1996). Furthermore, accu-
WCST (Axelrod et al., 1994a); however, this is racy on a shopping task (Rempfer et al., 2003)
not unexpected given that all of these disor- can be predicted from the number of per-
ders involve frontal lobe dysfunction. severative responses. Trials to First Category
Feldstein et al. (1999) assessed whether and total correct responses can predict task
the normative data reported by Heaton et al. orientation at a vocational work placement in
(1993), in which the standard (man•al) ad- a sample of schizophrenics (Lysaker et al.,
ministration procedure was used, are equiva- 1995).
lent to various WCST computer administration
procedures (e.g., keyboard, mouse, or touch Factor Structure of WCST
screen). The authors found that while there Several studies have examined the latent
were no statistical difference between the structure of the WCST and have produced
mean scores for all WCST measur~ (with varying results. There appears to be a dis-
the exception of failure to maintain ~t), sig- crepancy in the number of factors obtained,
nificant differences existed in cential ten- and this may be due to the number of WCST
dency, dispersion, and distribution ~ shapes outcome measures used in a particular study
between the published normative datil (man- and whether these selected measures contain
ual administration) and the compOterized redundancy (e.g., perseverative errors and
administration. percent perseverative errors) or to the type of
sample used (type of clinical sample or control "Failure to Maintain Set" (since this outcome
groups). measure predicted the most variability). Greve
In general, fewer factors seem to be ob- et al. (1997) again found essentially the same
tained when nonclinical samples are used. Sev- two factors in a group of 135 college students
eral investigators have found that in healthy and a mixed clinical sample of 139. These fac-
adults a unitary factor best accounts for the tors are quite similar to Sullivan et al.'s (1993)
WCST outcome measures (Boone et al., 1998; Perseveration and Inefficient Sorting factors.
Bowden et al., 1998; Goldman et al., 1996; Greve et al. (1996) suggested that factor 1
Pineda &: Merchan, 2003). However, Salt- reflected problem solving, while factor 2 was
house et al. (1996) examined a group of interpreted as a measure of attentional pro-
healthy adults aged 18-94 years and reported cesses, which was later confirmed with a color
the existence of two factors, with the majority overlay study. In follow-up studies of etiologi-
of the WCST measures (e.g., Total Errors and cally mixed groups of patients with TBI and
Perseverative Errors/Responses, Categories chronic, severe TBI, however, Greve and col-
Completed) loading on the first factor. Paolo leagues found a three-factor solution which
et al. (1995) reported three underlying factor they labeled "Cognitive Flexibility" (contained
structures for healthy older controls, with Total Correct Responses, Percent Conceptual
Categories Completed, Total Errors, Perse- Level Responses, Categories Completed, Per-
verative Responses/Errors, Nonperseverative severative Errors, and Perseverative Responses),
Errors, and Conceptual Level Responses "Problem-Solving" (contained Nonperseverative
loading on the first factor, Failure to Maintain Errors), and "Response Maintenance" (con-
Set loading on the second factor, and Learn- tained Failure to Maintain Set; Greve et al.,
ing to Learn and Trials to First Category 1999, 2002). Weigner and Donders (1999), us-
loading on the third factor. These findings are ing a higher-functioning group of TBI patients
essentially the same as those reported than Greve and colleagues, found a virtually
by Greve et al. (1999) for a sample of head- identical three-factor solution to that of Greve
injured patients. et al. (1999).
Studies using various clinical samples have Nagahama et al. (2003) also found a three-
most typically found three factors that best factor structure for Nelson's MCST. Using
account for the WCST outcome measures patients with Alzheimer's disease, patients
(Greve et al., 1998; Sullivan et al., 1993; Weig- with mild cognitive impairment, and healthy
ner &: Donders, 1999). Sullivan et al. (1993) controls, Nagahama et al. (2003) found that
examined 11 WCST scores of 58 individuals Perseverative Errors, Categories Completed,
(schizophrenics, alcoholics, normal controls) Trials to Complete First Category, and Con-
and found a three-factor solution that ac- ceptual Level Responses loaded on the first
counted for 91% of the total variability. They factor, termed "Perseveration," and accounted
labeled factor 1 "Perseverations" (7 of the for 57% of the common variance. Failure to
11 scores), factor 2 "Inefficient Sorting" (2 Maintain Set loaded on the second factor,
scores), and factors 3 "Nonperseverative Er- termed "Insufficient Sorting," and accounted
rors" (2 scores). According to Sullivan et al., for 23% of the common variance. Nonper-
Factors 1 and 3 required executive and mem- severative Errors loaded on the final factor,
ory skills, while factor 2 appeared unrelated to termed "Nonperseverative Error," and ac-
either type of skill. Koren et al. (1998) repli- counted for 10% of the common variance.
cated the same three-factor structure in a These factor structures are quite similar to
group of patients with schizophrenia and con- those reported by Greves and colleagues and
trols and found the perseveration factor to be Sullivan et al. (1993).
best for distinguishing patients from controls. While it is clear from years of research that
Greve (1993) examined 270 patients and the WCST taps into specific executive skills, it
controls and found only two factors that ac- appears to not heavily load on those executive
counted for 91% of the total variability. They functioning tests that require speed (Boone
called factor 1 "Problem Solving" and factor 2 et al., 1993a; Welsh et al., 1991). Boone et al.
(1998) examined the factor structure of the ranging 2%-21%. They found a significant
WCST and three other tests of axecutive improvement in WCST scores from ages 6.5 to
function (Stroop, Verbal Fluency, .nd Au- approximately 19 years, with little change over
ditory Consonant Trigrams) in a woup of ages 20--50 years, and then a relatively sharper
250 patients and controls. They found: that the decline in performance after 60 years. Virtually
four outcome measures of the WCST (Cate- identical age effects were found when the
gories, Percent Conceptual Level Ref;ponses, same data were reanalyzed using responses
Errors, Percent Perseverative Refponses) from only the first deck of cards (in the manual
loaded robustly on their own factor (and infor the 64-card version; Kongs et al., 2000).
dependent of the other executive tef;ts), ac- These results are consistent with other findings
counting for 23% of the total varianct. using the 64-card version (Axelrod et al.,
A word of caution comes from Bow4en et al. 1993). The reports of Heaton et al. (1993) and
(1998), in a study examining the reliaWity and Kongs et al. (2000) of declining WCST per-
validity of the WCST in a group of Wcohol- formance after the sixth decade of life are
dependent individuals and a group ot college consistent with those of other investigators
students. They found only one u~erlying (Axelrod & Henry, 1992; Compton et al., 2000;
factor out of six WCST outcome measures ex- Craik et al., 1990). Axelrod and Henry (1992)
amined. The authors suggest that ~atients' compared the WCST performance of Heaton's
"pattern" of performance on the clifferent (1981) sample of 40-year-olds to their sample of
WCST outcome measures should be inter- individuals in their 50s, 60s, 70s, and 80s. They
preted with caution given that all of th. WCST found that only for Perseverative Responses
outcome measures were accounted fot by one did the age-related decline begin at age 50 and
factor. Further, based on their data, ~ey sug- that, for Categories Completed and Persever-
gest that test-retest assessment over tipne may ative Errors, differences were observed for
not produce accurate clinical interpre;ation. those 60 years and older. In a study examining
highly educated individuals (university fac-
ulty), Compton et al. (2000) found that there
were essentially no changes in any of the
RELATIONSHIP BETWEEN W~T WCST measures in those in their 30s, 40s, and
PERFORMANCE AND DEMOGRAJ-HIC 50s but that those over 60 years completed
FACTORS fewer categories, required more trials, had
fewer percent correct responses, committed
Effect of Age
more perseverative errors, and had lower con-
Age-related declines in WCST perfqrmance ceptual level responses than those in their 30s
on both the 128-card and the 64-card versions and40s.
have been widely and consistently ~ported In contrast, while Boone et al. (1993) also
(Anderson et al., 1991; Arbuckle & Gold, 1993; found age-related declines in WCST perfor-
Axelrod & Henry, 1992; Axelrod et 4. 1993, mance, they report that performance does not
1996; Crockett et al., 1986; Daigneault et al., deteriorate until after the age of 70 years. Fur-
1992; Davis et al., 1990; Head et all, 2002; ther, individuals who were 70 years or older also
Heaton et al., 1993; Kramer et al.;, 1994; only displayed deficits on Total Errors and
Laiacona et al., 2000; Mejia et al., 199B; Parkin Conceptual Level Responses relative to indi-
& Walter, 1991, 1992; Salthouse et al, 1996, viduals who were 60 years or younger. These
2003; Spencer & Raz, 1994), but a sltht dis- findings are more consistent with the results of
crepancy appears regarding the age ~ which Haaland et al. (1987), who found that in a sam-
WCST scores begin to decline and the ~pecific ple aged 17-87 years only those 80-87 years old
WCST measures that show age-relafed de- performed poorer than those 64-69 years
cline. Heaton et al. (1993), using 899 Jndivid- old. These authors also found that the SO-year-
uals aged 6.5-90 years, reported a qe.adratic old group committed a greater total number of
effect for age on all WCST measures, With the errors but that this group completed fewer cat-
proportion of variance accounted for; by age egories relative to the 60-year-old group.
Other studies which have not used contin- In sum, robust findings of age-related de-
uous age groupings but rather younger vs. cline in WCST performance have been re-
older groups have also found age-related dif- ported, and while there is some discrepancy
ferences on the WCST. Beatty (1993) found regarding the age at which WCST perfor-
that older participants (mean age 70 years) mance begins to deteriorate, studies tend to
performed poorer than middle-aged (mean agree that there is little change in performance
age 40) and young (mean age 20) participants between ages 20 and 50 years (Boone et al.,
on Categories Completed, Number of Per- 1993; Compton et al., 2000; Heaton et al.,
severative Errors, and Total Errors but not on 1993; Yeudall et al., 1986). The changes appear
Nonperseverative Errors, Trials to First Cat- to occur after the age of 60 (Heaton et al.,
egory, or Failure to Maintain Set. He also 1993), with sharper deterioration in perfor-
noted greater within-group variability for the mance occurring in the seventh and eighth
older group. Parkin and Walter (1991) ob- decades of life (Beatty, 1993; Boone et al.,
served a difference between their sample of 1993; Craik et al., 1990; Haaland et al., 1987;
"young" adults, who were an average of Uneweaver et al., 1999; Parin & Walter, 1991).
34 years, and their "old" adults, who were an Additionally, there appears to be relatively
average of 80 years, on all four of the WCST good agreement that the total number of er-
measures they examined (Categories Com- rors increases with increasing age and that
pleted, Total Errors, Perseverative Errors, number of categories completed decreases
and Nonperseverative Errors). These findings with age (Beatty, 1993; Boone et al., 1993;
were replicated in a second study (Parkin & Haaland et al., 1987; Heaton et al., 1993;
Walter, 1992). In a sample aged 60-80, Craik Parkin & Walter, 1991; Spencer & Raz, 1994).
et al. (1990) found a significant correlation Some investigators have reported increased
between age and Perseverative Errors. perseverative errors as a result of increasing
Spencer and Raz (1994) found an interest- age (Craik et al., 1990; Heaton et al., 1993;
ing age-by-education interaction effect; their Parkin & Walter, 1991; Spencer & Raz, 1994),
older group (mean age 69.5 years) was more while others have shown age-related increases
highly educated than their younger group of set failures (Beatty, 1993; Heaton et al.,
(mean age 23.8 years), but the younger group 1993). Positive findings (Uneweaver et al., 1999;
outperformed the older group on the three Heaton et al., 1993; Parkin & Walter, 1991)
WCST scores measured (Total Errors, Per- and negative findings (Beatty, 1993; Boone
severative Errors, and Categories Completed). et al., 1993) regarding the relationship be-
Using her version of the WCST, Nelson tween nonperseverative errors and age have
(1976) reported that age had a deleterious been reported. Conceptual level responses
effect on the number of categories achieved have also been shown to reduce with age
but not on the types of error. Similarly, Isin- (Boone et al., 1993; Compton et al., 2000;
grini and Vazou (1997), using the same ver- Heaton et al., 1993).
sion, found that individuals aged 25-46 years
achieved a greater number of categories and
Effect of Education
committed fewer total and perseverative er-
rors than a group of individuals aged 70- The effect of education on the WCST has been
79 years. Uneweaver et al. (1999) observed well documented in the literature (Boone et al.,
that age correlated with various measures of 1993a, 1998; Heaton, 1981; Heaton et al., 1993;
Nelson's MCST, including number of cat- Laiacona et al., 2000; Stratta et al., 1993).
egories achieved and number of nonper- In his first manual, Heaton (1981) reported
severative errors. However, they noted these that the effects of education on WCST mea-
age effects to be rather subtle until the eighth sures were most apparent after 15 years of
decade of life, when a sharper decline in formal education. For their large sample of
performance was detected. Caffarra et al. adults (aged 20 years and older), Heaton et al.
(2004) also found age effects for perseverative (1993) reported a steady, linear relationship
errors and number of categories achieved. between most age-adjusted WCST measures
and education, with poorer WCST perfor- MCST. Specifically, they found that individ-
mance for lower educational levels. l'he only uals with an elementary school education
two measures that did not exhibit a relation- committed more perseverative errors and com-
ship with education were Failure to Maintain pleted fewer categorical sorts relative to those
Set and Learning to Learn. Laiacor¥t et al. with greater education. Similarly, Caffarra
(2000) reported a similar improvethent in et al. (2004) found a relationship between
WCST scores with greater education in a education and MCST (Perseverative Errors
group of healthy Italian individuals. Education and Categories Completed).
particularly affected Perseverative Responses
and Nonperseverative Errors. Compt~n et al.
Effect of Gender
(1997) found that education was the most
predictive demographic factor of the ~umber Most studies have not found effects of gender
of set failures in a highly educated woup of on the WCST (Heaton et al., 1993; Kongs et al.,
healthy adults aged 25-72 years. H~ver, in 2000; Laiacona et al., 2000; Yendall et al.,
another sample of 102 highly educated adults, 1987). Heaton et al. (1993), in their normative
Compton et al. (2000) did not find a correla- sample of 899 children, adolescents, and ad-
tion between education and any of th~ WCST ults, found no gender-related differences on
measures. any of the WCST outcome measures.
Boone et al. (1993) found that :healthy In contrast, using an older sample of healthy
middle-aged and older adults with >l6 years participants (mean age approximately 62
of education outperformed those wlh :::;12 years), Boone et al. (1993) found that females
years of education on most of the ~ WCST outperformed males on the following six
measures (e.g., Total Errors, Perseyerative WCST measures: Categories Completed,
Responses, Percent Perseverative Errers, and Total Errors, Perseverative Responses, Per-
Percent Conceptual Level Response$). In a cent Perseverative Errors, Percent Concep-
later study, Boone et al. (1998), using •epwise tual Level Responses, and Trials to First
multiple regression analyses, found ~t edu- Category. Women and men performed simi-
cation emerged as a significant prediJ:tor for larly on Set Failure and "Other" Responses. It
Perseverative Responses, Percent Coifeptual is possible that the superior performance by
Level Responses, and Total Errors. ' women relative to men appears only in middle
One study reported an interaction ~tween to old age. Ferland et al. (1998) found an op-
gender, education, and WCST performance. posite gender effect in a group of young un-
In a group of healthy individuals aged 15-40 dergraduate college students. The authors
years, Yendall et al. (1986) found tqat only administered a modified version of the
"other" (unique) responses correlat~ with 128-card WCST (in which all 128 cards were
education in women but that in men Perse- administered regardless of whether or not six
verative Errors, Nonperseverative Err~rs. and categories were completed) to healthy and
Total Errors correlated significantt with brain-injured individuals. The test was ad-
education. ministered two different times (separated by
Interestingly, in a Hispanic, Spanish- 5 months), and the authors found that males
speaking sample, Mejia et al. (1998) fQUnd no outperformed females when the test scores
difference in any of the WCST rqeasures were collapsed over the two times. No gender
between adults with very little education effects, however, were present when only the
(2--5 years) and those with more education first (standard) test administration trial was
(6-11 years). However, performances ~ported examined. Thus, these findings may be a result
for both groups were strikingly poor. It is of differential practice effects for males vs.
possible that, as indicated by Heaton~(1981), females, which was not further examined in
effects of education emerge only in thqse with this study.
a college education. No gender differences have been reported
Lineweaver et al. (1999) demonstra4ed that for Nelson's MCST Caffarra et al., 2004;
education affects performance on ~elson's Lineweaver et al., 1999).
Effect of Intellectual Level similar to that of North American samples

when raw scores were converted into demo-
Very few studies have specifically examined
graphically corrected standard scores using the
the relationship between the WCST and in-
Heaton et al. (1993) norms. Rey et al. (1999)
tellectual functioning, although those that are
examined the performance of 75 Hispanic in-
available have observed a relationship be-
dividuals on the WCST and other neu-
tween WCST scores and IQ. Heaton (1981)
ropsychological tests. Participants' nationalities
reported a correlation between the majority of
included Cuba, Peru, Venezuela, Puerto Rico,
the WCST outcome measures and FSIQ.
Panama, Colombia, Honduras, and Nicaragua.
Similarly, Merriam et al. (1999) found that IQ
While no statistical analyses are reported, the
correlated with most WCST measures in a
authors note that means and SDs for all of the
group of healthy controls as well as in patients
WCST measures of their Hispanic sample ap-
with major depression or schizophrenia.
pear comparable to that of Heaton's (1981)
Boone et al. (1993a) did not find differences
normative sample.
on any of the WCST measures for four IQ
Artiola i Fortuny et al. (1999) have included
levels (90--109, 110-119, 120-129, and 130+).
the WCST in their standardized and validated
However, none of their participants had lower
battery of neuropsychological tests culturally
than average IQ; thus, the nonsignificant
adapted for Spanish-speaking individuals. Nor-
findings may have been due to a restricted
mative data based on 390 participants, which
range in IQ. In a later study, using multiple
were collected in Spain, Mexico, and the Uni-
regression analyses, Boone et al. (1998) found
ted States, are stratified by geographical
that FSIQ accounted for significant test score
area x age x education.
variability in all of the WCST measures ex-
amined (Categories Completed, Perseverative
Responses, Percent Conceptual Level Re-
sponses, Total Errors, and Trials to First
Category). METHOD FOR EVALUATING THE
Isingrini and Vazou (1997) using Nelson's NORMATIVE REPORTS
MCST found that better performance on the To adequately evaluate the WCST normative
WAIS Similarities subtest and Cattell's Ma- reports, six key criterion variables were deemed
trices (a measure of fluid intelligence) related critical. The first four of these relate to subject
to better performance on Categories Com- variables, and the last two refer to procedural
pleted, Total Errors, and Perseverative Er- issues.
rors. Yet, no relationship between these Minimal requirements for meeting the cri-
MCST scores and WAIS Vocabulary or In- terion variables were as follows.
formation subtests was found. The authors
interpreted these results to mean that the
MCST (and perhaps WCST) is correlated
with fluid, but not crystallized, intelligence. Subject Variables
Further studies are needed to better de- Sample Size
lineate the relationship between intelligence
As discussed in previous chapters, a mlm-
and WCST performance.
mum of 50 subjects per grouping interval is
optimal.
Effect of Ethnicity
To date, very few studies have examined the Sample Composition Description
effects of ethnicity on WCST performance. In As discussed earlier, information regarding
a study comparing manual and computer medical and psychiatric exclusion criteria is
administrations of the WCST, Artiola i Fortuny important; it is unclear if geographic recruit-
and Heaton (1996) found that the performance ment region, socioeconomic status or oc-
of healthy adults from Madrid, Spain, was quite cupation, ethnicity, gender, and recruitment
procedures are relevant, so until determined, Some studies present data divided into age
it is best that this information be provided. groups. Few studies classify participants into
education groups or present data for males
Age Group Interval and females separately; few studies report
Given the association between age and WCST data for males only or present data in age-by-
performance, information regarding the age of education cells. Data collected on individuals
the normative sample is critical and normative from Spain, South America (e.g., Peru, Ven-
data should be presented by age interval. ezuela, and Columbia), and Central America
(e.g., Panama, Honduras, and Nicaragua), and
Grouping by Educational Level Italy are presented in this chapter.
Given consistent evidence of effects of educa- Test-retest data are reported in some
tional level on WCST performance, normative studies, with typically 1-year intertrial inter-
data should be grouped by educational level. vals. Issues of reliability and/or practice effects
are discussed in these studies.
The studies vary in the WCST outcome
measures reported. The majority report at
Description of Administration Procedures least four or more measures. In the studies
reviewed below, all available WCST scores
Due to variability in administration procedures,
reported will be presented.
a detailed description, including identification
Among all the studies available in the lit-
of the version of the test administered, is de-
erature, we selected for review those based on
sirable. This would allow one to select the most
well-defined samples. Additionally, compre-
appropriate norms or to make corrections in
hensive normative data are available in the
interpretation of the data.
administration and scoring manuals of the
Data Reporting 128-card and 64-card versions of the WCST
(Heaton, 1981; Heaton et al., 1993; Kong et al.,
To facilitate interpretation of the data,
2000). Thus, among all of the available studies
group means and standard deviations should be
for these two versions of the WCST, only
presented at minimum for categories achieved
those that were published in 1993 and later
and one of the perseverative measures (e.g.,
and contain sample sizes of ~50 will be re-
errors or responses).
viewed in this chapter. We hope to accomplish
two goals with our reviews of these studies:
(1) to aid the reader to relatively quickly ex-
amine the most recent normative information
in order to supplement the Heaton et al.
OF THE NORMS
(1993) norms and (2) to cover the most recent
Data reporting for the WCST differs across literature in order to address any cohort ef-
studies. Some of these differences will be fects (e.g., changes in scores over the past
summarized below. decade) in the WCST. For Nelson's MCST,
Our review of the literature located WCST all of the studies containing sample sizes >50
normative reports for adults as well as ad- are reviewed.
ministration manuals containing comprehen- Summaries of the studies are presented in
sive normative data (Heaton, 1981; Heaton ascending chronological order for each ver-
et al., 1993; Kongs et al., 2000). Hundreds of sion of the test separately. Studies using the
other clinical studies have also reported con- original 128-card WCST administration pro-
trol subject data. The majority of studies cedure are presented first, followed by those
report the mean age, education, and gender using the 64-card version and then those using
distribution for the sample and/or for the Nelson's MCST. The text of study descrip-
age groups. Some studies report WAIS IQs tions contains references to the corresponding
or estimated intelligence levels and ethnic tables identified by number in Appendix 25.
composition. Table A25.1, the locator table, summarizes
information provided in the studies described for a study of pesticide poisoning conducted by
in this chapter. 1 Heaton et al. (1991) and was described in the
first WCST manual (Heaton, 1981). Exclusion
criteria are presented in the original Heaton et al.
SUMMARIES OF THE STUDIES (1991) study and include a history of learning
WCST 128-Card Administration Version disabilities, neurological illness, "significant"
head injury, "serious" psychiatric illness (e.g.,
WCST Manual schizophrenia), or substance abuse.
Sample 4: Fifty (34% male) healthy partici-
[WCST.1] Heaton, Chelune, Talley, Kay, and pants, recruited from Colorado, comprised this
Curtiss, 1993 (WCST 128-Card Version) sample. Participants were aged 58-84 years,
The Wisconsin Card Sorting Test Manual: with educational levels of 8-20 years. There is
Revised and Expanded is a modified edition of no mention of exclusion criteria for this sample.
Heaton's (1981) original WCST administra- Sample 5: This sample was collected on 124
tion and scoring manual. The goals of the (91% male, 9% female) commercial pilots
authors for this new edition were to include primarily recruited in Colorado (only five par-
wider normative age ranges (6.5--89 years), to ticipants were recruited in Washington DC).
further clarify scoring criteria with explicit Participants were aged 24--65 years and had
examples, and to revise scoring forms to fa- 14-20 years of education.
cilitate recording participants' responses and
Sample 6: This sample was based on a study
calculating outcome measures. The authors
by Axelrod and Henry (1992). Seventy-three
combined essentially six different samples to
healthy individuals (45% male, 55% female)
obtain normative data for 899 healthy indi-
were "recruited from a health promotion pro-
viduals. The six samples are described below.
ject, from independent living retirement resi-
Sample 1: This sample consisted of 453 (48% dences, and from the general community in the
Detroit metropolitan area." Ages ranged 51-
male) normal chiJdren and adolescents aged
89 years and years of education ranged ~20
6 years and 6 months to 17 years and 11 months
years. Exclusion criteria are presented in the
recruited from public schools. Data regarding
original study by Axelrod and Henry (1992) and
race were available only for 379 participants.
Whites comprised 78% of the sample, 11% include a history of psychiatric hospitalization or
use of psychotropic medication, history of sub-
were black, and 2% were classified as "other."
Exclusion criteria were a neurological dys- stance abuse, neurological disorder, head injury
function, learning disability, emotional disor- resulting in >5 minutes ofloss of consciousness,
der, or attention disorder. There is no mention significant illness such as diabetes or chronic
obstructive pulmonary disease (COPD) re-
of the geographic location from which this
sample was collected. quiring long-term medical treatment, or Mini-
Mental Status Exam (MMSE) scores of $24.
Sample 2: This sample consisted of "49 stu-
dents and friends of students who lived in the
community surrounding a large urban area in The manual provides regression-based raw to
T-score conversions for Total Errors, Percent
the southwestern United States." Participants
Errors, Perseverative Responses, Percent Per-
were 49% male, all 18 years of age, and had
severative Responses, Perseverative Errors,
12-15 years of education.
Percent Perseverative Errors, Nonperseverative
Sample 3: This sample was collected from Errors, Percent Nonperseverative Errors, and
Texas and Colorado and consisted of 150 (83% Percent Conceptual Level Responses. Addi-
male) healthy adult participants aged 15-- tionally, raw score to percentile conversions are
77 years, with educational levels ranging 7-20 provided for Categories Completed, Trials to
years. This sample originally served as controls Complete First Category, Failure to Maintain
'Nonns for children are available in Baron (2004) and Set, and Learning to Learn. The reader is
Spreen and Strauss (1998). referred to the manual, which stratifies the data
for these WCST outcome measures ~ed on on the WCST and the California Card Sorting
gender, 14 child and adolescent age groups Test. Participants were "young" (18-34 years;
(aged ~19 years), eight adult age groups (aged mean age= 25.5, SD = 5. 7), "middle-aged"
20--79 years), and six education groupsr(~8, 9- (35-49 years; mean age= 40.6, SO= 6.1), and
11, 12, 13-15, 1~17, ~18 years). Stan4ard test "old" (~60 years; mean age=70.9, SD=6.5)
administration and scoring criteria are well individuals who served as normal controls for
described. the author's previous studies. They were aged
Normative data based on this WCSTimanual 18-75 and had education of 8-20 years. In-
for Perseverative Responses are also kported dividuals with a history of medical illness such
in the Comprehensive Norms for an E!anded as diabetes, head injuries, neurological dis-
Halstead-Reitan Neuropsychological attery: ease, psychiatric illness, or substance abuse
Demographic, Corrections, Research F: ndings, "that could affect their performance" were
and Clinical Applications (Heaton et al1, 1991), excluded from the study. Standard procedures
and Perseverative Errors are reprod~ced in based on the Heaton (1981) manual were
the most updated edition (Heaton et al,, 2004). used. Means and SDs are reported.
The authors found that on the California
Study strengths Card Sorting Test the older subjects achieved
1. Sample composition is well desc~bed in fewer sorts but that they did not have increased
terms of age, gender, and education. verbal or nonverbal perseverative responses
2. Adequate exclusion criteria. and they were able to explain their correct
3. Test administration procedures are sorting strategies as well as the younger group.
specified. ; On the WCST, numbers of perseverative re-
4. Means and SDs for the entire saptple as sponses and perseverative errors were greater
well as T scores for the groups ±atified for the older relative to the younger group.
by age and education are presen ed.
5. Sample is stratified into numer s age- Study strengths
by-education groupings. 1. Sample composition is well described in
terms of age, education, and gender.
Considerations regarding use of the ~dy 2. Adequate exclusion criteria.
1. Overall sample size is adequate, but in- 3. Test administration procedures are
dividual cells for certain WCST ~tcome specified.
measures are relatively small. 4. Means and SDs for the test scores are
2. Recruitment procedures were ~t well reported.
described for some of the sub-samples. 5. Sample is stratified into three age groups.
3. Exclusion criteria are not speci,ed for
some of the subsamples. Considerations regarding use of the study
1. Overall sample size is adequate; however,
Other comments individual cells are relatively small, and
1. The interested reader is referred :to Fas- precise sizes are not provided (e.g., "age
tenau and Adams (1996) critiqu~ of the groups of 20-21 individuals" per cell).
Heaton et al. (1991) norms, and lleaton
) 2. Data are not stratified by education.
et al.'s 1996 response to this critique. 3. Educational levels are relatively high.
4. Sample composition is not well de-
Normative Studies and Control Gro&4PS scribed in terms of recruitment proce-
in Clinical Comparison Studies for dures and geographic location.
the WCST
[WCST.3] Boone, Ghaffarian, Lesser,
[WCST.2] Beatty, 1993 (WCST 128-Card Hiii-Gutierrez, and Berman, 1993 (WCST
Version) (Table A25.2) 128-Card Version) (Table A25.3)
The author examined the test performjl.nce of The purpose of this study was to provide
65 (31 male, 34 female) healthy p~ipants further information regarding the effects of
age, education, IQ, and gender on the WCST Considerations regarding use of the study
in older adults. The sample was recruited 1. Overall sample size is adequate, but
through newspaper advertisements, flyers, individual cells are relatively small.
and personal contact from the Los Angeles, 2. Educational levels are relatively high.
California, area. It consisted of 91 (35 males,
56 females) healthy adults who were fluent in [WCST.4] Stratta, Rossi, Mancini, Cupillari,
English and aged 45-83 years, with an average Matteri, and Casacchia, 1993 (WCST
education of 14.5 (2.5) years and an average 128-Card Version) (Table A25.4)
IQ of 115.89 (12.97). Exclusion criteria were a
The authors examined the effects of educa-
history of psychotic or major affective disor-
tion on WCST performance in a group of
der, current or past history of substance
patients with schizophrenia and healthy con-
abuse, documented neurological illness, and
trols. Sixty-one control participants were re-
significant medical illness that could affect the
cruited from among relatives and employees
central nervous system (e.g., diabetes). Indi-
of the S. Salvatore Hospital in L'Aquila, Italy.
viduals were also excluded from the study
Subjects were excluded from the study if they
based on abnormal neurological examination,
had a personal history of substance abuse,
significant metabolic abnormalities detected
head injury, "serious" medical illness, or
in blood tests, or abnormal MRI findings.
psychiatric disorder or a family history of
Seventy-one subjects were white, 10 were
psychiatric illness. Participants were right-
African American, five were Asian, and five
handed, with an average age of 31.93 (5.95)
were Hispanic. Standard procedures based on
years and 12.65 (4.3) years of education. The
the Heaton (1981) manual were used, and the
data are stratified into three education groups
protocols were computer-scored using the
(0--8, 9-13, and 2::14 years). Standard proce-
Harris (1988) software.
dures based on the Heaton (1981) manual
The results of the study indicated that
were used.
healthy middle-aged and older adults with
Differences between the schizophrenic and
> 16 years of education outperformed those
control groups attenuated when the data were
with $12 years of education on most of
stratified by educational level.
the WCST measures (e.g., Total Errors, Per-
severative Responses, Percent Perseverative
Errors, and Percent Conceptual Level Study strengths
Responses). Also, females scored higher 1. Sample composition is well described in
than males on almost all of the measures, terms of age, education, recruitment
and individuals older than 70 years per- procedures, and geographic location.
formed poorer than the younger subjects on 2. Adequate exclusion criteria.
Total Errors and Percent Conceptual Level 3. Test administration procedures are
Responses. specified.
1. Sample composition is well described in 5. Sample is stratified into two age
terms of age, education, ethnicity, gen- groups.
der, IQ, recruitment procedures, and
geographic location. Considerations regarding use of the study
2. Adequate exclusion criteria. 1. Overall sample size is adequate, but in-
3. Test administration procedures are dividual cells are relatively small.
specified. 2. Gender composition of the sample is not
4. Means and SDs for the test scores are reported.
reported. 3. Data were obtained on Italian subjects,
5. Sample is stratified first into three age which may limit their usefulness for
groups, then by gender, and finally into clinical interpretation in the United
three educational levels. States.
[WCST.S] Kramer, Humphrey, Larish, Logan, aging. A total of64 (23 male, 41 female) healthy
and Strayer, 1994 (WCST 128-Card Ver9ion) participants were divided into two age groups.
(Table A25.5) The "young adults" were aged 18-35 years, with
The authors collected WCST data on controls an average of 23.8, and the "older adults" were
as part of a study examining "whether inhibi- aged 65-80 years, with an average of 69.5. The
tory failures are general or specific in •ature." older group had an average of 15.3 years of
The sample included 62 (26 male, 36 female) education, which was significantly greater than
healthy individuals. The data were stratified the younger age group, who had an average of
into two age groups. The "young" in~viduals 13.5 years of education. No SDs were provided
were aged 18-28, with an average age; of 20.6 for age or education. Subjects were recruited
and an average of 16.4 (1.1) years of edocation. from the undergraduate psychology subject
The "old" individuals were aged 60-74, with an pool, advertisements, and personal invitations.
average age of 67.8 years and an avqrage of Participants were excluded if they had a history
16.3 (1.8) years of education. No SDs,for age of head injury, diabetes, epilepsy, "severe"
were reported. The younger group ha4 an av- substance abuse, neurological disease, use of
erage IQ of 117.8 (8.5) based on the K'Jlufman psychotropic medications, sleep deprivation, or
Brief Intelligence Test, and the oldef group color blindness. The test was administered us-
had an average IQ of 117.6 (8.4). S4rutdard ing a computer-administered version of the
procedures based on the Heaton (1981) man- WCST (Neurosoft Corp., McLean, VA), which
ual were used. . follows the standard testing procedures. With
With respect to the WCST, the authors this computerized version, the participant re-
found a significant age effect, with thf youn- ceives auditory (2,000 Hz tone for correct and
ger group outperforming the older gq:>up on 20 Hz tone for incorrect) and written ("right"
all but Trials to First Category and qoncep- or "wrong'') feedback from the computer after
tual Level Responses. ; each response.
The authors found age-related declines on
Study strengths all of the WCST scores they examined and
1. Sample composition is well described in further reported that Perseverative Errors was
terms of age, education, gender, and IQ. "inversely related to both factual and contex-
2. Test administration procedurcts are tual memory tests," with the relationship to
specified. contextual memory being stronger.
3. Means and SDs for the test sc~s are
4. Sample is stratified into three edpcation 1. Sample composition is well described in
groups. terms of age, education, gender, and
recruitment procedures.
Considerations regarding use of the stt~dy
1. Overall sample size is adequate, but in-
specified.
dividual cells are relatively small.
2. Exclusion criteria are not descrihl!d.
reported.
3. Recruitment procedures wer' not
5. Sample is stratified into three education
reported.
groups.
4. Educational levels and IQ are relatively
high. .
[WCST.6] Spencer and Raz, 1994 dividual cells are relatively small.
(WCST 128-Card Version) (Table A25.6) 2. Educational levels are relatively high for
The authors obtained WCST data on controls as the older group and significantly different
part of a study examining memory fot facts, for the two age groups (older group has
sources, and contextual detail as it re~es to higher education than younger group).
[WCST.7] Paolo, Troester, Axelrod, and Koller, monolingual Spanish-speaking. Participants

1995 (WCST 128-Card Version) (Table A25.7) were part of a larger normative study. Exclu-
These authors conducted principal compo- sion criteria were a history of learning dis-
nents analyses separately for patients with ability, "significant" head trauma, neurological
Parkinson's disease and normal controls. A illness, toxic exposure, major psychiatric ill-
total of 187 (69 male, 118 female) control ness, or substance abuse. For the total sample,
participants who were part of a longitudinal participants were aged 15-59 years, with an
study were recruited from the Kansas City, KS average of 27.32 (9.11), and had 11-18 years
community and retirement centers. Partici- of education, with an average of 14.35 (2.25).
pants were an average age of 69.74 (6.96) Sixty participants with an average age of 27.32
years, with an average of 14.91 (2.57) years of (10.82) and an average of 14.13 (2.33) years
education. Ninety-seven percent of partici- of education were administered the standard
pants were white, 1.1% were black, and 1.1% (manual) version of the WCST. Fifty-nine
were Hispanic. Participants were excluded if participants with an average age of 27.32 (7.06)
they had a history of stroke, psychiatric illness, years and an average of 14.58 (2.16) years of
"significant" head trauma, substance abuse, or education were administered the computer-
evidence of neurological disorder that may ized version of the WCST (i.e., WCST-CV2;
compromise cognition and scores <130 on the Heaton et al., 1993). Protocols administered
Dementia Rating Scale (DRS). Standard pro- in the standard format were computer-scored.
cedures based on the Heaton (1981) manual Testing was administered in Spanish to all par-
and computerized scoring based on the Hea- ticipants, using standard procedures (Heaton
ton (1993) scoring software were used. et al., 1993).
Principal components analysis revealed The study revealed that the demographi-
three WCST factors for the control partici- cally corrected WCST data for this Spanish-
pants and for patients with Parkinson's disease, speaking sample were quite similar to the
but the factor structures were not the same for published normative data derived from North
the two groups. American subjects.

1. Sample composition is well described 1. Sample composition is well described in
in terms of age, education, gender, eth- terms of age, education, gender, recruit-
nicity, recruitment procedures, and geo- ment procedures, and geographic location.
graphic location. 2. Relatively large sample size.
2. Relatively large sample size. 3. Test administration procedures are
3. Adequate exclusion criteria. specified.
4. Test administration procedures are 4. Means and SDs for the test scores are
specified. reported.
1. Sample is not stratified by age or edu-
Considerations regarding use of the study cation groups.
1. Sample is not stratified by age or edu- 2. Educational levels are relatively high.
cation groups. 3. Data were obtained on subjects from
2. Educational levels are relatively high. Spain, which may limit their usefulness for
clinical interpretation in the United States.
[WCST.8] Artiola i Fortuny and Heaton, 1996
(WCST 128-Card Version) (Table A25.8) [WCST.9] Hoff, Riordan, Morris, Cestaro,
The authors examined manual and computer Wieneke, Alpert, Wang, and Volkow, 1996
test administrations of the WCST in a group (WCST 128-Card Version) (Table A25.9)
of 119 (51 male, 68 female) healthy partici- The study examined the effects of crack co-
pants from Madrid, Spain, who were primarily caine use on cognitive functioning. A total of
54 male normal control participants ~ere re- had a history of stroke, "significant" head
cruited from the community. Participaqts were trauma, substance abuse, or neurological dis-
an average of 32.1 (9. 7) years of age ~d had orders. Additionally, participants had to have a
an average of 15.4 (2.4) years of edqcation. DRS score of >130 at each testing session and
Exclusion criteria were a history of rtedical, could not display a >10-point drop at the
neurological, or psychiatric illnesses; suJ>stance second testing probe. Standard procedures
abuse; or learning disability. Forty-eif~It par- based on the Heaton (1981) manual and
ticipants were White, four were Afri~ Amer- computerized scoring based on the Heaton
ican, and two were Hispanic. It appe~s that (1993) scoring software were used. Normal-
the Heaton et al. (1993) WCST adminiftration ized age- and education-corrected standard
was used, but the procedures are nof refer- scores (Heaton et al., 1993) were used for
enced nor are they well described. I Total Errors, Perseverative Errors, and Per-
Among other findings, the authors\ report cent Conceptual Level Responses.
that, surprisingly, crack cocaine use isiassoci- The study revealed practice effects for
ated with better performance on Ca~gories testing probes separated by approximately
Completed. I 1 year, with performance on most WCST
measures improving by 5-7 standard points.
Study strengths The authors provide a number of indices, in-
1
1. Sample composition is well desctlbed in cluding discrepancy scores, to assist clinicians
terms of age, education, ethnicifY. and in better interpreting test-retest change scores.
gender.
2. Adequate exclusion criteria. : Study strengths
3. Relatively large sample size. : 1. Sample composition is well described in
4. Means and SDs for the test scotes are terms of age, education, ethnicity, and
reported. t gender.
2. Relatively large sample size.
Considerations regarding use of the s~dy 3. Means and SDs for some test scores are
1. Sample is not stratified by age qr edu- reported.
cation groups.
2. Educational levels are relatively high. Considerations regarding use of the study
3. Test administration procedures ~e not 1. Sample is not stratified by age or edu-
specified. cation groups.
4. Recruitment procedures were not 2. Educational levels are relatively high.
reported. 3. Recruitment procedures were not
reported.
[WCST.10] Paolo, Axelrod, and Troester, 1 4. Adequate exclusion criteria.
1996a (WCST 128-Card Version)
(Table A25.10) [WCST.11] Rosselli and Ardila, 1996
The authors examined the test-retest reli- (WCST 128-Card Version) (Table A25.11)
ability and practice effects of the WCST in a The effects of substance abuse on cognitive
group of older adults. A total of 87 (2l'J male, functioning were assessed in this study. A total
62 female) participants, with an averag~ age of of 63 males aged 15-48 years, with an average
68.8 (6.21) years and an average of 14.a (2.42) age of25.61 (7.54) years and an average of 10.5
years of education who were part of ailongi- (4.58) years of education, participated in
tudinal study, were assessed at two ditferent this study. Participants were recruited from
times. Testing probes were separattd by Bogota, Colombia, and were Spanish-speaking.
approximately 1 year. Ninety-five perqent of Standard procedures based on the Heaton
participants were White, 2% were African (1981) manual were used.
American, and 2% were Hispanic. jartici- The authors report poorer performance on
pants were excluded from the study If they virtually all WCST measures for cocaine and
polysubstance abusers relative to normal 4. Means and SDs for the test scores are
controls. reported.

1. Sample composition is well described in 1. Sample is not stratified by age or edu-
terms of age, education, gender, re- cation groups.
cruitment procedures, and geographic 2. Educational levels are relatively high.
location. 3. Exclusion criteria are not described.
2. Relatively large sample size. 4. Recruitment procedures are not
3. Test administration procedures are reported.
specified.
4. Means and SDs for the test scores are [WCST.13] Compton, Bachman, and Logan,
reported. 1997 (WCST 128-Card Version) (Table A25.13)
These authors examined age-associated
Considerations regarding use of the study changes in intelligence and other cognitive
1. Sample is not stratified by age or edu- domains in a group of university faculty. Par-
cation groups. ticipants were 52 (30 male, 22 female) non-
2. Exclusion criteria are not described. psychology faculty members of the Georgia
3. All-male sample. College and State University. Participants
4. Data were obtained on subjects from were aged 25-72, with an average of 47.74
Colombia, which may limit their useful- (11.77) years, and ranged in educational level
ness for clinical interpretation in the 16-20 years, with an average of 18.44 (1.69).
United States. Participants were recruited by phone and due
to the "invasive nature of the protocol" and
[WCST.12] Salthouse, Fristoe, and Rhee, 1996 privacy issues given that psychology graduate
(WCST 128-Card Version) (Table A25.12) students conducted the testing, faculty from
The authors examined the effects of age on the Psychology Department were not included
various neuropsychological measures, includ- in the sample. The computerized administra-
ing the WCST. A total of 259 (approximately tion version of the standard procedures based
63% female) healthy participants aged 18--94, on the Heaton (1981) manual was used.
with an average age of 51.4 (18.4) years The authors found that age predicted the
and approximately 15 years of education (no number of categories completed and the av-
SD available), were recruited. All participants erage response time on the WCST, while ed-
were self-reported to be in good, very good, or ucation predicted the number of set failures in
excellent health. Standard procedures based this highly educated group.
on the Heaton et al. (1993) manual were used.
A significant age-related decline in most Study strengths
WCST measures was reported; however, the 1. Sample composition is well described in
strong intercorrelation among the WCST terms of age, education, gender, re-
scores and with other neuropsychological cruitment procedures, and geographic
test scores suggests "only a portion of the age- location.
related influences on many commonly used 2. Relatively large sample size.
neuropsychological measures is specific and 3. Test administration procedures are
potentially localized." specified.
terms of age, gender, and education. Considerations regarding use of the study
2. Relatively large sample size. 1. Sample is not stratified by age and edu-
3. Test administration procedures are cation groups.
specified. 2. Educational levels are relatively high.
3. Exclusion criteria are not described. [WCST.15] Artiola i Fortuny, Heaton, and
4. Precise computerized version ;of the Hermosillo, 1998 {WCST 128-Card Version)
WCST is not specified. (Table A25.15)
The authors examined differences in perfor-
[WCST.14] Fristoe, Salthouse, and Woodard,
mance on various neuropsychological tests,
1997 {WCST 128-Card Version) (Table Al5.14)
' including the WCST, in Spanish-speaking in-
The authors examined the processes mediating dividuals from the USA-Mexico border and
age-related differences in WCST ~·erfor- those from Spain. The study collected a total
mance. A total of 97 individuals parti "pated. of 390 participants aged 15-76 years, with 0--
Participants were divided into youn r and 20 years of education. Of these, 185 (47 male,
older groups. The younger group cons ted of 138 female) were from the USA-Mexico
48 (25% male) participants aged 18-3 years, border, were an average of 42.2 (13.5) years of
with an average of 26.7 (5.7), and 13~ (1.3) age, and had an average of 9.6 (6.1) yem of
years of education. The older group C<tlsisted education; 205 (91 male, 114 female) were
of 49 (35% male) participants ag~ 60- from Madrid, Spain, with an average age of
86 years, with an average of70.1 (7.2), aJ!,d 13.9 36.3 (16.1) years and an average education of
(2.0) years of education. Two comptferized 12.7 (4.4) years. The US-Mexico participants
versions were administered. The fi~t was were Mexicans who lived within Mexico or in
based on the standard version of Heat<t. et al. close proximity to the USA-Mexico border,
'
(1981), developed by Woodard (1994; ~rsonal Mexican Americans whose years of residency
communication). The second is a hypothesis- in the United States and years of education in
generated WCST version, in which partici- Mexico varied. Exclusion criteria for all par-
pants are required to indicate the prin~pal on ticipants were history of neurological illness,
which they plan to make their subsequ1nt sort use of psychoactive medication, chronic
(i.e., color, form, or number). The s~dard medical conditions (e.g., diabetes, hyperten-
version was always administered first. : sion), complaints of current cognitive or
The results indicated that feedback usage emotional problems, history of substance use,
(assessed by subject's verbalization bf the and learning problems or disability. Also,
hypothesis regarding sorting stratelf' and participants had to declare Spanish as their
the actual sorting choice), working n¢mory, primary language and to demonstrate native
and "perceptual-comparison" (e.g., digit fluency in Spanish.
symbol) performance accounted for a large Both the standard manual (Heaton et al.,
amount of the age-related deficits on WCST 1993) and the computerized WCST were ad-
performance. · ministered. Data for Categories completed
and Perseverative Responses are reported.
Study strengths
1. Sample composition is well desci1}>ed in Study strengths
terms of age, gender, and education. 1. Sample composition is well described
2. Relatively large sample size, with both in terms of age, gender, education, and
subgroups approaching 50. ethnicity.
3. Test administration procedure$ are 2. Large sample size.
specified. I 3. Test administration procedures are
5. Data are stratified into two age gro,pings. reported.
I 5. Data are stratified by USA-Mexico and
Considerations regarding use of the stftly Spain samples.
1. Recruitment procedures are llPt re-
ported. , Considerations regarding use of the study
2. Exclusion criteria are not cleady de- 1. Data are not stratified by age or educa-
scribed. tion groups.
2. Scores for only two WCST measures 3. Test administration procedures are
(Categories and Perseverative Re- specified.
sponses) are reported. 4. Means and SDs for the test scores are
reported.
[WCST.16] Boone, 1998 (WCST 728-Card 5. Data are stratified into two vascular sta-
Version) (Table A25.16) tus by two age groups and into two vas-
The effects of various demographic and cular status by three IQ groups.
health-risk factors, such as age, education, IQ,
and vascular status, were examined in a group Considerations regarding use of the study
of middle-aged and older individuals. Partic- 1. Overall sample size is adequate, but in-
ipants were 155 (53 male, 102 female) healthy dividual cells are relatively small.
adults aged 45-84, with a mean of 63.07 2. Educational levels are relatively high.
(9.29), who had an average of 14.57 (2.55)
years of education and an average IQ of [WCST.17] Mejia, Pineda, Alvarez, and Ardila,
115.41 (14.11). Exclusion criteria were a his- 1998 (WCST 728-Card Version)
tory of psychotic or major affective disorder, (Table A25.17)
current or past history of substance abuse, The authors examined the effects of age, gen-
documented neurological illness, or signifi- der, and education on a variety of memory and
cant medical illness that could affect the executive function tests, including the WCST.
central nervous system (CNS; e.g., diabetes). Participants were 60 (21 male, 39 female)
Individuals were also excluded based on ab- healthy Colombian adults aged 55-85 years,
normal neurological examination or significant with an average age of 69.66 years. Educational
metabolic abnormalities detected in blood level ranged 2-11 years. All participants were
tests. Individuals who reported or showed native Spanish speakers and were recruited
medical evidence of current or past hyper- from Medellin, Colombia. Exclusion criteria
tension, arrhythmia, large white-matter hy- were a history of psychiatric illness, neurolog-
perintensity on MRI (>10 cm2 ), coronary ical disorder, or psychotropic medication at the
artery bypass graft, angina, or old myocardial time of testing. The data are stratified first by
infarction were classified as having vascular two age groupings (55-70, 71-85 years) and
illness. Fifty-one participants were classified then by two education groupings (2-5, 6--11
as having vascular disease, and 104 partici- years). Standard procedures based on the
pants were classified as being "healthy." Heaton (1981) manual were used.
Given that age, IQ, and vascular status sig- These authors did not find significant dif-
nificantly predicted most WCST measures, ferences on the WCST between their samples
the data were stratified first by vascular status aged 55-77 and 71-85 years. Likewise, they
(vascular and healthy) by two age groups did not find differences between those with 2-
(<65, ;:::65) and then by vascular status (vas- 5 years and those with 6--11 years of educa-
cular and healthy) and three IQ groups (av- tion. They did, however, observe that those
erage, high average, and ;:::superior). Standard who attended rural schools committed fewer
procedures based on the Heaton (1981) perseverative errors than those who attended
manual were used, and data were computer- urban schools.
scored.
The results indicated that vascular status, Study strengths
age, IQ, education, and gender were signifi- 1. Sample composition is well described in
cant predictors of WCST performance. terms of age, gender, ethnicity, educa-
tion, and geographic location.
1. Sample composition is well described in 3. Test administration procedures are
terms of age, gender, education, IQ, and specified.
recruitment procedures. 4. Means and SDs for the test scores are
2. Adequate exclusion criteria. reported.
5. Sample is stratified into two age and two [WCST.19] Gooding, Kwapil, and Tallent, 1999
education groups. (WCST 128-Card Version) (Table A25.19)
The WCST performance of college students
Considerations regarding use of the study with schizotypal traits and normal controls
1. Overall sample size is adequate, but were examined. Control participants were
individual cells are relatively sm~. 104 (43 male, 61 female) college students
2. Recruitment procedures are not from the University of Wisconsin, Madison.
reported. Individuals with a history of psychotic illness
3. Data were obtained on subjects from and/or "psychoactive substance use disorder,
Colombia, which may limit their useful- family history of psychotic disorder, learning
ness for clinical interpretation · in the disability, epilepsy, TBI, or other medical ill-
United States. nesses were excluded from the study. Subjects
were an average of 18.72 (0.86) years of age,
[WCST.18] Basso, Bornstein, and Lang, 1999
with an average prorated WAIS-R IQ (using
(WCST 128-Card Version) (Table A25.18).
Vocabulary and Block Design subtests) of
This study examined practice effects f+r most 116.26 (12.56). A computerized version of the
common tests of executive functioning, ;nclud- WCST (Harris, 1988), based on standard
ing the WCST. Participants were 82 ~ealthy procedures described in the Heaton et al.
males recruited via community newspaper (1993) manual, was used.
advertisements. The authors note that no fe- Among other findings, the results revealed
males were recruited due to "logistic"\ limita- that college students with schizotypal traits
tions. Of the 82 participants, 50 were retested achieve fewer categorical sorts, commit more
in 1 year. Those participants were an av~rage of perseverative errors, and have more set fail-
32.50 (9.27) years of age and had an av~age of ures compared to controls.
14.98 (1.93) years of education. Therr were
48 Caucasians, one African American, ~d one
Study strengths
Hispanic. Exclusion criteria were psythiatric
disorder, neurological disease, head !injury,
terms of age, gender, IQ, education, and
learning disability, or other medical illness.
geographic location.
Standard procedures based on the Heaton et al.
(1993) manual were used.
Among other findings, the authors : report
significant practice effects for all WCST
specified.
measures with the exception of Categories
Completed and Failure to Maintain $et for
reported.
testing probes separated by 12 month~
Study strengths · Considerations regarding use of the study

1. Sample composition is well described in 1. Sample is not stratified by age and edu-
terms of age, gender, education, and cation groups, although it can be as-
recruitment procedures. sumed that it is a homogeneous sample
2. Relatively large sample size. with narrow ranges on these variables.
3. Test administration procedures are 2. Recruitment procedures and educa-
specified. tional levels are not reported.
4. Means and SDs for the test sCOfeS are 3. Relatively high IQ.
reported. ·
[WCST .20] Merriam, Thase, Haas, Keshavan,
Considerations regarding use of the st4£fy and Sweeney, 1999 (WCST 128-Card Version)
1. Sample is not stratified by age ar;J edu- (Table A25.20)
cation groups. The WCST performance of a group of pa-
2. Educational levels are relatively ~gh. tients with major depression was compared to
those with schizophrenia and normal controls. 1 was from Honduras, 8 were from Nicaragua,
Control participants were 61 healthy individ- and 19 "other." Data were collected in Dade
uals aged 18-50, with an average age of 26.08 County, Florida. All participants were pri-
(7.67) years, an average of 14.66 (2.39) years marily Spanish-speaking, and the test instruc-
of education, and an average IQ of 103.90 tions for the neuropsychological instruments
(9.22). Exclusion criteria were no history of were "adapted or translated from the original
electroconvulsive therapy, neurological illness, versions." The data are stratified into two ed-
head injury, or substance dependence within ucation groupings (12-15, >15 years). Stan-
6 months of testing. A computerized version dard procedures based on the Heaton (1981)
of the WCST was administered following the manual were used.
Heaton et al. (1993) standard procedures; The WCST scores of their sample were
however, there is no mention of the specific comparable to those of the age- and education-
computerized version used. matched sample reported by Heaton (1981).
Patients with depression performed poorer
than controls but not compared to patients Study strengths
with schizophrenia on virtually all WCST 1. Sample composition is well described in
measures, and their scores were related to the terms of age, education, gender, ethnic-
severity of their illness. ity, language, and geographic location.
1. Sample composition is well described in 3. Means and SDs for the test scores are
terms of age, education, and IQ. reported.
2. Relatively large sample size. 4. Data are stratified into two education
3. Adequate exclusion criteria. groupings.
specified.
reported.
dividual cells are relatively small.
2. Exclusion criteria are not described.
3. Recruitment procedures not reported.
1. Sample is not stratified by age and edu-
4. Educational levels for the overall sample
cation groups.
are relatively high.
2. Educational levels are relatively high.
3. Gender and recruitment procedures
were not reported. [WCST.22] Snitz, Curtis, Zald, Katsanis,
4. The exact computerized version of the and Iacono, 1999 (WCST 128-Card Version)
WCST used is not specified. (Table A25.22)
The study examined the relationship of spatial
[WCST.21] Rey, Feldman, Rivas-Vazquez, working memory to neuropsychological func-
Levin, and Benton, 1999 (WCST 128-Card tioning and oculomotor activity in patients
Version) (Table A25.21) with schizophrenia and healthy individuals.
The authors report normative data for a num- Control participants were 54 (19 male, 35
ber of neuropsychological tests, including the female) individuals who lived in the Minneap-
WCST, for Hispanics. Participants included olis, Minnesota, community. Control partici-
75 (56 male, 19 female) healthy individuals, pants were recruited via flyers placed in
with an average age of 33.45 (19.75) and an various clinics (e.g., general medical clinics,
average of 14.53 (3.25) years of education. dental clinics, dermatology clinics) in the same
Participants reflected the following nationali- hospital from which the schizophrenic pa-
ties: 53 were from Cuba, 3 were from Peru, 1 tients were recruited. Most of the sample
was from Venezuela, 6 were from Puerto Rico, were patients with nonneurological condi-
1 was from Panama, 6 were from Columbia, tions, and a smaller number were employees
of the hospital. Participants were also re- "serious" head injury, learning disability, epi-
cruited from vocationaVtechnical schools and lepsy, history of DSM Axis I disorder, psy-
medical clinics in a university hospital. Ex- chosis in the family, or medical condition that
clusion criteria were history of substance use would interfere with completing the tasks
disorder, diagnosis of major affective or psy- (e.g., color blindness) were excluded from the
chotic disorder, history of neurological disor- study. A computerized version of the WCST
der or any other medical illness affecting CNS (Harris, 1988), based on standard procedures
functioning, history of head injury, or mental described in the Heaton et al. (1993) manual,
retardation. Participants averaged 36.0 (13.4) was used.
years of age, 15.0 (1.7) years of education, and Individuals with schizotypal traits performed
IQ of 109.7 (13.4). The authors used the poorer than controls on Categories Completed
computerized version of the WCST developed and Failure to Maintain Set. Additionally, a
by Rezai (1988, personal communication) but negative relationship between working mem-
do not mention whether this version follows ory performance and number of perseverative
the standard testing procedures. errors (r = - 0.17) and Trials to Complete First
Among other findings, these authors report Category (r=- 0.15) was noted.
that the working memory impairment in pa-
tients with schizophrenia was related to fewer Study strengths
categories completed. 1. Sample composition is well described in
terms of age, gender, IQ, current edu-
Study strengths cation status, and geographic location.
1. Sample composition is well described in 2. Relatively large sample size.
terms of age, education, gender, IQ, re- 3. Adequate exclusion criteria.
cruitment procedures, and geographic 4. Test administration procedures are
location. specified.
2. Relatively large sample size. 5. Means and SDs for the test scores are
1. Data are not stratified by age and edu-
Considerations regarding use of the study cation groupings.
1. Data are not stratified by age or educa- 2. Recruitment procedures are not re-
tion groupings. ported.
2. Educational levels are relatively high. 3. Relatively high IQ.
3. Test administration procedures are not
specified. [WCST.24] Compton, Bachman, Brand, and
Avet, 2000 (WCST 128-Card Version)
[WCST.23] Tallent and Gooding, 1999 (Table A25.24)
{WCST 128-Card Version) (Table A25.23) The authors examined the relationship between
The authors examined the relationship be- age and WCST performance in a group of
tween working memory and WCST perfor- highly educated professionals. Participants
mance in individuals with schizotypal traits. were 102 (53 male, 49 female) healthy adults
This was a replication and extension of earlier >30 years of age, with an average of 19.5 years
work. Control participants were 63 (22 male, of education (SD for education is not reported).
41 female) undergraduate students at the Of these, 87 were current college professors
University of Wisconsin, Madison, who were and 17 were in professions which required a
an average age of 19.11 (1.03) and had an high level of cognitive skill (e.g., a significant
average prorated WAIS-R IQ (using Vocabu- amount of reading and writing). All participants
lary and Block Design subtests) of 114.63 were recruited from Atlanta, Georgia. English
(11.93). Participants whose primary language was their primary language, and 94 were
was not English, those with a history of Caucasian. The data were stratified into four
age groupings: :»-39 (mean=34.9, SD=3.71; personality disorder, head injury, neurological
mean education= 18.95), 40-49 (mean =46.19, disorder, or substance abuse or a famiJy history
SD = 2.80; mean education= 19.18), 50-59 of psychosis. The educational level of the
(mean= 54.03, SD = 3.20; mean education= controls was reported to be comparable to that
19.92), and 60 years and older (mean= 65.49, of the patient group, who had an average of
SD=5.72; mean education=19.47). A com- 13.2 (3.8) years of education; however, the
puterized version of the WCST (Loong, per- exact education of the controls was not re-
sonal communication, 1990), based on standard ported. Standard procedures based on the
procedures described in the Heaton (1981) Heaton (1981) manual were used.
manual, was used. Patients with schizophrenia committed more
The results revealed a significant relation- perseverative errors and achieved fewer cat-
ship between age and most WCST measures, egorical sorts than their unaffected siblings
including response time (r=0.49), Categories and controls, but no differences were found
Completed (r = -0.36), total trials (r = 0.32), between controls and unaffected siblings.
total correct responses (r= -0.36), and Per-
cent Conceptual Level Responses (r= -0.53). Study strengths
However, no relationship between education 1. Sample composition is well described in
and WCST scores was reported, which was terms of age and gender.
likely due to the restricted educational range. 2. Relatively large sample size.
1. Sample composition is well described in 4. Adequate exclusion criteria.
terms of age, education, gender, re- 5. Means and SDs for the test scores are
cruitment procedures, and geographic reported.
location.
2. Relatively large sample size, although Considerations regarding use of the study
individual cells are small. 1. Sample is not stratified by age or edu-
3. Test administration procedures are cation groups.
specified. 2. Actual educational level is not reported.
4. Means and SDs for the test scores are 3. Recruitment procedures are not re-
reported. ported.
5. Data are stratified into four age
[WCST.26] Laiacona, lnzaghi, De Tanti,
groupings.
and Capitani, 2000 (WCST 128-Card
The authors developed a new "global effi-
1. Data are not stratified by education, sam-
ciency" score for the WCST and normative
ple reflects a narrow educational range,
data for this new measure as well as other
and educational level is high.
WCST scores for a sample of Italian partici-
2. Exclusion criteria are not described.
pants. Participants were adults (100 male,
105 female) aged 15-85 years, with an average
[WCST.25] Ismail, Cantor-Graae, and McNeil, age of 46.5, and ranging in education 5-
2000 (WCST 128-Card Version) (Table A25.25) 17 years, with an average of 11.4 years. Nine-
The WCST performance of patients with teen were admitted to the Valduce Hospital
schizophrenia, their unaffected siblings, and located in Costa Masnaga, Italy. The illnesses
normal controls were compared. Control par- for which they were admitted to the hospital
ticipants were 75 (59 male, 16 female) healthy were not related to neurological or psychiatric
adults aged 20-54, with an average age disorders, and the average hospital stay was
of 35.9 years (no SD). Participants were ex- 20 (1.4) days. The remaining 186 participants
cluded if they had a personal history of psy- were healthy individuals recruited from
chosis, affective disorder, schizophrenia-related the same geographic area, and some were
relatives of patients attending the hospital. All bipolar disorder and healthy controls. Con-
participants were reported to be free from trols were 64 (30 male, 34 female) healthy
medical illnesses that affect cognitive perfor- individuals who were an average of 26.4 (5.44)
mance (e.g., substance use). However, the years of age and had 14.69 (2.99) years of
authors note that they did not want a '"hyper- education. Participants were employees and
normal" subject; thus, selection criteria were relatives or acquaintances of staff in the clin-
"not too selective." They admit that their ical and administrative areas of the S. Salva-
sample may have included participants with tore Hospital in L'Aguila, Italy. Exclusion
mild hypertension and diabetes who were re- criteria were a personal history of head injury,
ceiving medication. The data were part:ttioned substance abuse, "serious" neurological or
by gender, six age groups (15-29, 30-39,149-49, physical disease, or psychiatric disorder or a
50-59, 60--69, 70-85 years), and four education family history of psychosis or personality dis-
groups (5-6, 8-12, 13-16, 17-24 years). How- order. A computerized version of the WCST
ever, no data are reported for age groupa.15-29 (Schneider, 1989; personal communication),
and 30-39 with 5-6 years of education. based on standard procedures described in
Standard procedures based on the Heaton the Heaton et al. (1993) manual was used. The
(1981) manual were used. A global; score only exception to the standard scoring proce-
was created by multiplying the number of dure was "that the first ambiguous error re-
categories completed by 10 and subttacting peating the previously correct principle was
this value from the total number trials oi not scored as a perseveration."
administered: Among other findings, this study reports
that a discriminant function analysis was able
Global SC0!11 = [ ntrials administered to correctly classify 85.9% of controls but only
-(ncategories completed X 10)] · 48.5% of schizophrenic patients and 40% of
bipolar patients.
Study strengths .
1. Sample composition is well described in Study strengths
terms of age, gender, education, etlpllcity/ 1. Sample composition is well described in
language, recruitment procedur~, and terms of age, education, gender, recruit-
geographic location. ment procedures, ethnicity/language, and
2. Adequate exclusion criteria. geographic location.
3. Test administration procedures are 2. Relatively large sample size.
specified. . 3. Test administration procedures are
reported. 4. Adequate exclusion criteria.
5. Sample is stratified into gender, six age 5. Means and SDs for the test scores are
groups, and four education groupt. reported.
1. Overall sample size is adequate, ~ut in- 1. Sample is not stratified by age and edu-
dividual cells are relatively small. cation groups.
2. Data were obtained on subjects from 2. Data were obtained on subjects from
Italy, which may limit their usetplness Italy, which may limit their usefulness
for clinical interpretation in the lJnited for clinical interpretation in the United
States. States.
[WCST.27] Rossi, Arduini, Daneluzzo, Bustini, [WCST.28] Razani, Boone, Miller, Lee, and
Prosperini, and Stratta, 2000 (WCST 128-Card Sherman, 2001 (WCST 128-Card Version)
Version) (Table A25.27) (Table A25.28)
This study examined neuropsychological This study examined cognitive function in
functioning in patients with schizophrmia or patients with frontotemporal dementia or
Alzheimer's disease. Control participants were score= 12.2, SD = 2.7), and 60--84 (mean
104 (33 male, 71 female) healthy older adults age= 70.3, SD = 6.2; mean education= 16.4,
with an average of 60.36 (9.64) years of age, an SD = 2.9; mean Vocabulary scaled score= 12.8,
average of 14.82 (3.31) years of education, and SD = 2.4). A computerized version ofthe WCST
an average WAIS-R FSIQ of 116.81 (14.06). (Woodard, 1994), based on standard procedures
Data for control participants were selected described in the Heaton (1981) manual, was
from a larger pool of archival data. Exclusion used.
criteria were a history of head injury, major
affective or psychotic disorder, seizures, or Study strengths
substance abuse within the past 5 years. 1. Sample composition is well described in
Standard procedures based on the Heaton terms of age, education, gender, and
et al. (1993) manual were used. recruitment procedures.
Study strengths 3. Test administration procedures are
1. Sample composition is well described in specified.
terms of age, gender, education, and IQ. 4. Means and SDs for the test scores are
3. Test administration procedures are 5. Data are stratified into three groupings.
specified.
reported. 1. Exclusion criteria are not clearly de-
scribed.
Considerations regarding use of the study 2. Educational levels are high.
1. Recruitment procedures are not reported.
2. Educational and IQ levels are high.
3. Data are not stratified by age and WCST 64-Card Administration Version
education.
WCST-64 Manual
[WCST.29] Salthouse, Atkinson, and Berish,
2003 (WCST 128-Card Version) [WCST.30] Kongs, Thompson, Iverson, and
(Table A25.29) Heaton, 2000 (WCST 64-Card Version)
The authors examined age-related issues and The Wisconsin Card Sorting Test-64 Card
executive functioning in a group of 261 (35% Version professional manual is designed to
male, 65% female) healthy adults. Participants provide normative data on the shortened,
were aged 18-84 years and had an average of 64-card version of the WCST. This manual es-
approximately 16 years of education. They were sentially reanalyzes the first 64 responses (from
recruited via newspaper advertisements and the first deck of cards) in the sample reported in
Hyers to participate in a battery of neu- the WCST revised manual (Heaton et al., 1993).
ropsychological tests requiring three sessions of Data were not available for two participants;
approximately 2 hours' duration. No specific thus, this sample consisted of 897 children,
exclusion criteria are listed, but the authors adolescents, and adults. For a description of the
mention that six participants were excluded various samples, recruitment procedures, and
from the analysis for not completing the battery, demographic information, see WCST.1, above.
due to difficulty in understanding instructions, The manual provides regression-based raw to
and/or for obtaining WAIS-III Vocabulary T-score conversions for Total Errors, Percent
scaled scores of<4. The data were stratified into Errors, Perseverative Responses, Percent
three age grouping: 18--39 (mean= 27. 7, SD = Perseverative Responses, Perseverative Errors,
6.4; mean education= 15.5, SO= 3.3; mean Percent Perseverative Errors, N onperseverative
Vocabulary scaled score= 12.0, SO= 3.5), 40- Errors, Percent Nonperseverative Errors, and
59 (mean= 49.0, SD = 5.0; mean educa- Percent Conceptual Level Responses. Addi-
tion= 16.0, SD = 2.4; mean Vocabulary scaled tionally, raw score to percentile conversions are
provided for Categories Completed, Trials to into seven age groups by decade of life (i.e.,
Complete First Category, Failure to Maintain 20s, 30s, 40s, 50s, 60s, 70s, and 80s). Average
Set, and Learning to Learn. The reader is re- educational level ranged from 14.4 (3.0) to
ferred to the manual, which stratifies the data 15.6 (1.2). Subjects aged 20--49 were ei-
for these WCST outcome measures based on ther undergraduate students at Wayne State
gender, 14 child and adolescent age groups University or recruited via newspaper adver-
(ages ~19 years), eight adult age groups (ages tisements from the Detroit community. Par-
20--79 years), and six education groups ($8, 9- ticipants aged 50--89 were part of a previously
11, 12, 13-15, 1~17, 18> years). Standard test published study (Axelrod & Heruy, 1992). For
administration and scoring criteria are well these older participants, Axelrod and Heruy
described. (1992) list the following exclusion criteria:
history of psychiatric hospitalization or use of
Study strengths psychotropic medication, substance abuse,
1. Sample composition is well described in neurological disorder, head injury resulting in
terms of age, gender, and education. >5 minutes of loss of consciousness, signifi-
2. Adequate exclusion criteria. cant illness such as diabetes or COPD re-
3. Test administration procedures are quiring long-term medical treatment, or
specified. MMSE scores 2:::24. No exclusion criteria are
4. Means and SDs for the entire sample listed for participants younger than 50 years.
and T scores for the groups stratified by Standard procedures of the 128-card WCST
age and education are presented. based on the Heaton (1981) manual were
5. Sample is stratified into numerous age- used, and participants in the youngest three
by-education groupings. decade groups sorted all 128 cards even after
six categories were completed. For all partic-
Considerations regarding use of the study ipants, only data from the first 64 cards were
1. Overall sample size is adequate, but in- analyzed.
dividual cells for certain outcome mea- Results of trend analyses revealed a signifi-
sures are relatively small. cant age-related decline for Categories Com-
2. Recruitment procedures were not well pleted and an increase for Total Errors,
described for some subsamples. Perseverative Errors, and Perseverative Re-
3. Exclusion criteria are not specified for sponses for the 64-card version of the WCST.
some subsamples.
Study strengths
Other comments 1. Sample composition is well described in
1. The interested reader is referred to terms of age, gender, education, and
Fastenau and Adams (1996) critique of recruitment strategies.
the Heaton et al.'s normative approach, 2. Test administration procedures are
and Heaton et al.'s 1996 response to this specified.
critique. 3. Means and SDs for the entire sample are
reported.
Normative Studies and Control Groups 4. Sample is stratified into seven age
in Clinical Comparison Studies for groupings.
the WCST-64
[WCST.31] Axelrod, Jiron, and Henry, 1993 1. Overall sample size is adequate, but in-
(WCST 64-Card Version) (Table A25.30) dividual cells for certain measures are
The authors examined performance of healthy relatively small.
adults on the 64-card version of the WCST. 2. Recruitment procedures and exclusion
Participants were 140 (55 male, 85 female) were not well described for some
adults aged 20--90. The data were partitioned subsamples.
3. Educational levels are high; data are not 59 monolingual, Spanish-speaking Latino men
stratified by education. with $10 years of formal education in the Los
Angeles, California, community. Participants
[WCST.32] Paolo, Axelrod, Troester, Blackwell, were an average of 28.89 (8.37) years old and
and Koller, 1996b (64-Card Version) had an average of5.82 (2.49) years of education.
(Table A25.31) Exclusion criteria consisted of any self-report of
The authors examined performance of pa· head injury, neurological insults, prenatal or
tients with Parkinson's disease or Alzheimer's birth complications, learning disabilities, psy-
disease and normal controls on the 64-card chiatric problems, or substance abuse.
version of the WCST. Control participants Standard administration procedures were
were 35 (22 male, 13 female) older adults, used. Participants were tested in Spanish.
with an average age of71.34 (5.73), an average Selected subtests from the WAIS-111 (Mexi-
education of 13.11 (2.03) years, and an aver· can version) were included in the battery.
age DRS score of 137.37 (3.36). They were Mean performance on the Marin and Marin
part of a longitudinal study of neurodegener· (1991) acculturation scale for this sample was
ative disease. Participants were recruited via 17.61 (6.19). For the Los Angeles group,
advertisements, and all were interviewed, Picture Vocabulary subscale scores from the
completed a health questionnaire, and re· Woodcock-Johnson-III Tests of Achievement
ceived neurological examinations. Exclusion (M = 5.36, SD = 6.01) and the Bateria
criteria were history of stroke, psychiatric Woodcock-Muiioz-R, Pruebas de habilidad
disorder, "significant" head trauma, substance cognitiva-R (M = 29. 77, SD = 5.37) were used
abuse, or neurological disorders. Standard to assess level of English and Spanish word
procedures based on the Heaton et al. (1993) expressive abilities.
manual were used, but for the purposes of this The results are presented by education
study only the responses on the first 64 cards groupings (0-6, 7-10) and by age-and-
were scored and analyzed. The Heaton (1993) education groupings (18-29 years old, 0-6 or
computer scoring software was used. 7-10 years of education; 30-49 years old, 0-6
or 7-10 years of education). The authors
Study strengths found a significant difference in performance
1. Sample composition is well described in on the WCST (number correct and Categories
terms of age, education, gender, and Completed) between the two education
recruitment procedures. groups. However, the two age groups did not
2. Test administration procedures are differ significantly on any of the sections of the
specified. WCST.
3. Adequate exclusion criteria
reported. 1. Sample composition is well described
Considerations regarding use of the study geographic area.
1. Small sample size. 2. Data availability for a healthy, employable,
2. Sample is not stratified by age and edu· monolingual Spanish-speaking group with
cation groupings. low educational level.
3. Sample is stratified into two education
[WCST.33] Lopez-Carlos, Salazar, Villasenor, groups and two age-by· education groups.
Saucedo, and Peiia, 2003 (64-Card Version) 4. Adequate exclusion criteria.
(Tables A25.32, A25.33) 5. Means and SDs are reported.
The authors investigated the effects of demo-
graphic variables on cognitive abilities in Considerations regarding use of the study
Spanish-speaking individuals with low educa· 1. All-male sample.
tion. The WCST-64 was administered to 2. Small sample sizes.
MCST Administration Version: Normative WAIS IQ of 109.5 (13.2) from the United
Studies and Control Groups in Clinical Kingdom. Secretarial and support staff of a
Comparison Studies local hospital and university were recruited.
Exclusion criteria were no history of neu-
[WCST.34] Bondi, Monsch, Butters, Salmon, and rological or psychiatric disorder. Standard
Paulsen, 1993 (MCST Version) (Table A25.34) MCST administration and scoring procedures
11lis study examined the utility of the MCST in (Nelson, 1976) were used.
differentiating patients at various stages of Among other findings, the authors report
Alzheimer's disease and normal controls. that this version of the WCST did not dis-
Control participants were 75 (27 male, 48 fe- criminate well among frontal and non-frontal
male) older adults, with an average age of71.1 lobe lesion patients but that it did discriminate
(7.6) years, an average of 13.7 (2.6) years of between controls and patients (regardless of
education, and an average MMSE score of the site of lesion).
28.9 (1.2). Precise educational levels are not
reported. Participants were spouses of patients Study strengths
or recruited via newspaper advertisements 1. Sample composition is well described in
from the San Diego, California, community. terms of age, gender, IQ, and recruit-
Exclusion criteria were a history of substance ment procedures.
abuse, learning disability, or "serious" neu- 2. Relatively large sample size.
rological or psychiatric disorders. Standard 3. Test administration procedures are
MCST administration and scoring procedures specified.
(Nelson, 1976) were used. 4. Adequate exclusion criteria.
ROC curves found Categories Completed 5. Means and SDs for the test scores are
and Perseverative Errors to be more sensitive reported.
than Nonperseverative Errors at discriminat-
ing Alzheimer's patients from controls. Considerations regarding use of the study
1. Sample is not partitioned by age and
Study strengths education groups.
1. Sample composition is well described in 2. No information on education is reported.
terms of age, education, gender, re- 3. Data were obtained on subjects from the
cruitment procedures, and geographic United Kingdom, which may limit their
location. usefulness for clinical interpretation in
2. Relatively large sample size. the United States.
specified. [WCST.36] lsingrini and Vazou, 1997
4. Adequate exclusion criteria. (MCST Version) (Table A25.36)
5. Means and SDs for the test scores are Performance on tests of frontal lobe function
reported. and intelligence were assessed in a group of
healthy adults. Participants were 107 {52 male,
Considerations regarding use of the study 55 female) adults. 11le study was conducted
1. Data are not partitioned by age and in France, and the data were divided into two
education groups. age groups: 25-46 {mean= 35.5, SD = 7.58;
mean education= 12.5, SD = 3.03) and 70-99
[WCST.35] Van den Broek, Bradshaw, and (mean = 80.59, SD = 8.58; mean education=
Szabadi, 1993 (MCST Version) (Table A25.35) 8.54, SD = 1.18). Older participants who re-
11le sensitivity and specificity of the MCST sided independently or in senior-citizen resi-
were assessed in a group of patients with brain dential homes were recruited. There is
lesions and normal controls. 11le control no mention of the recruitment procedures
sample consisted of 77 (19 male, 58 female) for the younger group. Participants reported
Caucasian participants who were an average good health and were not on medications
of 35.2 (12.8) years old and had an average that affect cognitive functioning. However, no
other exclusion criteria were reported. Stan- The authors found significant effects of age
dard MCST administration and scoring pro- and education, but not gender, on the MCST.
cedures (Nelson, 1976) were used. Additionally, for testing probes separated by
The study found that MCST scores best 1 year, significant increases for Nonpersevera-
correlated with measures of Huid intelligence. tive Errors were found. Practice effects for
Categories Completed and Perseverative Er-
Study strengths rors were not observed.
terms of age, education, and gender. Study strengths
2. Test administration procedures are 1. Sample composition is well described in
specified. terms of age, education, gender, ethnic-
3. Means and SDs for the test scores are ity, and recruitment procedures.
reported. 2. Test administration procedures are
4. Sample is stratified into two age groupings. specified.
1. Overall sample is adequate, but individ- reported.
ual cells are relatively small. 5. Sample is stratified into four age and
2. Recruitment procedures not reported. four education groupings.
3. Data were obtained on French subjects,
which may limit their usefulness for Considerations regarding use of the study
clinical interpretation in the United 1. Overall sample is adequate, but individ-
States. ual cells are relatively small.
4. Low educational level for the older sub-
jects; data not stratified by education.
[WCST.37] Lineweaver, Bondi, Thomas, and CONCLUSIONS

Salmon, 1999 (MCST Version) (Table A25.37) A tremendous number of clinical studies, in-
The authors conducted a normative study cluding those on patients with brain lesions
on the MCST. Participants were 229 (97 male, and neurological and psychiatric patients, es-
132 female) healthy, community-dwelling, older pecially patients with schizophrenia and head
adults aged 45-91 years, with an average age of injury, as well as neuroimaging studies have
69.06 (8.58) years and an average ofl3.60 (4.57) established the WCST as a useful clinical tool
years of education. Seventy-eight percent were for detecting executive or frontal system
white, 21% were Mexican American or Spanish dysfunction.
American, 1% were African American, and the Up to three factor structures have been
remaining 1% were Cuban American. Partici- identified that best explain all the WCST
pants were part of a longitudinal study con- outcome measures, but when used individu-
ducted by the University of California, San ally, Perseverative Errors, Perseverative Re-
Diego Alzheimer's Disease Research Center. sponses, and Categories Completed appear to
Exclusion criteria were a history of substance be the most useful clinical measures (Heaton,
abuse, psychiatric illness, or neurological dis- 1981; Heaton et al., 2004; Milner, 1963).
order. The data were partitioned into four age Practice effects have been noted on most
groups (45-59, 60--69, 70-79, 80-91 years) and WCST and MCST measures in healthy adults
four education groups (1-6, 7-12, 13-16, 17- but need to be further investigated in clinical
20 years). Additionally, the authors provide populations.
demographically corrected norms and raw A review of the literature indicates a strong
score to scaled score conversions, which have effect of age, a moderate effect of education
not been reproduced in this chapter. and intelligence, and equivocal results for the
Standard MCST administration and scoring effect of gender on the various versions of
procedures (Nelson, 1976) were used. the WCST. Age-related decline in WCST
performance had been consistently ; docu- Very few studies have examined the rela-
mented, with virtually no changes between tionship between ethnicity and WCST scores.
the ages of 20 and 60 years but a relttively Of the few cross-cultural studies available,
steep decline during the sixth, seventJt, and primarily on Hispanic samples, there appear
eighth decades of life. Better performance on not to be large differences between these
the WCST appears to be related to ~gher samples and the nonnative samples developed
educational level, particularly when ~duca on North American populations. It is clear that
tionallevel is > 15 years. While relatively few additional research is needed to better under-
studies have explicitly examined the lation- stand the effects of factors such as culture,
ship between intelligence and WCS most ethnicity, and multilingualism on the WCST.
have found that higher intellectual fun oning Additionally, more normative information
leads to better WCST performance. Th+re are is needed for individuals with low educa-
mixed reports for the effects of gendet. with tional levels. Very few of the existing WCST
most studies reporting equivalent ferfor- studies included individuals with <12 years of
mance between males and females. education.2
2 Meta-analyses were not perfonned on the WCST data

as this chapter was not intended to summarize all of
the voluminous literature available on this test. Con-
versely, comprehensive sets of nonns are available in the
literature.
References
Abbruzzese, M., Ferri, S., & Scarone, S. (1996). Alcoholism: Clinical and Experimental Research,
Perfonnance on the Wisconsin Card Sorting 17(2), 205-210.
Test in schizophrenia: Perseveration in clinical Adams, K. M., Gilman, S., Koeppe, R., Kluin, K.,
subtypes. Psychiatry Research, 64(1), 27--33. Junek, L., Lohman, M., et al. (1995). Correlation
Abikoff, H., Alvir, J., Hong, G., Sukoff, R., Orazio, J., of neuropsychological function with cerebral
Solomon, S., et al. (1987). Logical Memory metabolic rate in subdivisions of frontal
Subtest of the Wechsler Memory Scale: Age lobes of older alcoholic patients measured with
and education norms and alternate-form reli- [-1-8F]fluorodeoxyglucose and positron emis-
ability of two scoring systems. Journal of Clinical sion tomography. Neuropsychology, 9(3),
and Experimental Neuropsychology, 9(4), 275-280.
435-448. Adams, R. L., & Trenton, S. L. (1981). Develop-
Abraham, E., Axelrod, B. N., & Ricker, J. H. ment of a paper-and-pen form of the Halstead
(1996). Application of the Oral Trail Making Test Category Test. Journal of Consulting and Clini-
to a mixed clinical sample. Archives of Clinical cal Psychology, 49(2), 298-299.
Neuropsychology, 11(8), 697-701. Akshoomoff, N. A., & Stiles, J. ( 1995a). Developmen-
Abrahams, S., Leigh, P. N., Harvey, A., Vythe- tal trends in visuospatial analysis and planning:
lingum, G. N., Grise, D., & Goldstein, L. H. I. Copying a complex figure. Neuropsychology,
(2000). Verbal fluency and executive dysfunction 9(3), 364-377.
in amyotrophic lateral sclerosis (ALS). Neu- Akshoomoff, N. A., & Stiles, J. (1995b ). Developmen-
ropsychologia, 38(6), 734-747. tal trends in visuospatial analysis and planning:
Abwender, D. A., Swan, J. G., Bowerman, J. T., & II. Memory for a complex figure. Neuropsy-
Connolly, S. W. (2001). Qualitative analysis chology, 9(3), 378-389.
of verbal fluency output: Review and compari- Akshoomoff, N. A., & Stiles, J. (2003). Children's
son of several scoring methods. Assessment, 8(3), performance on the ROCF and the development
323--336. of spatial analysis. In J. A. Knight (Ed.), The
Acevedo, A., Loewenstein, D. A., Barker, W. W., handbook of Rey-Osterrieth Complex Figure
HaiWOOd, D. G., Luis, C., Bravo, M., et al. usage: Clinical and research applications (pp.
(2000). Category Fluency Test: Normative data 393--409). Lutz, FL: Psychological Assessment
for English- and Spanish-speaking elderly. Jour- Resources.
nal of the International Neuropsychological Albert, M. S., Heller, H. S., & Milberg, W. (1988).
Society, 6(7), 760-769. Changes in naming ability with age. Psychology
Acker, M. B., & Davis, J. R. (1989). Psychology test and Aging, 3(2), 173-178.
scores associated with late outcome in head in- Alder, A. G., Adam, J., & Arenberg, D. (1990).
jury. Neuropsychology, 3(3), 123-133. Individual-differences assessment of the rela-
Adams, K. M., Gilman, S., Koeppe, R. A., Kluin, tionship between change in and initial level of
K. J., Brunberg, J. A., Dede, D., et al. (1993). adult cognitive functioning. Psychology and Ag-
Neuropsychological deficits are correlated with ing, 5(4), 560-568.
frontal hypometabolism in positron emission to- Alekoumbides, A., Charter, R. A., Adkins, T. G., &
mography studies of older alcoholic patients. Seacat, F. (1987). The diagnosis of brain damage
533
534 REFERENCES
by the WAIS, WMS, and Reitan battery utilizing Anderson, C. V., Bigler, E. D., & Blatter, D. D.
standardized scores corrected for age and edu- (1995). Frontal lobe lesions, diffuse damage, and
cation. International Journal of Clinical Neuro- neuropsychological functioning in traumatic
psychology, 9(1), 11-28. brain-injured patients. Journal of Clinical and
Alevriadou, A., Katsarou, Z., Bostantjopoulou, S., Experimental Neuropsychology, 17(6), 900-008.
Kiosseoglou, G., & Mentenopoulos, G. (1999). Anderson, S. W., Damasio, H., Jones, R. D., &
Wisconsin Card Sorting Test variables in relation Tranel, D. (1991). Wisconsin Card Sorting Test
to motor symptoms in Parkinson's disease. Per- performance as a measure of frontal lobe dam-
ceptual and Motor Skills, 89(3, Pt 1), 824--830. age. Journal of Clinical and Experimental Neu-
Allegri, R. F., Mangone, C. A., Villavicencio, A. F., ropsychology, 13, 909-922.
Rymberg, S., Taragano, F. E., & Baumann, D. Anger, W. K., Cassitto, M. G., Liang. Y-X, Amador,
(1997). Spanish Boston Naming Test norms. R., Hooisma, J., Chnislip, D. W., et al. (1993).
Clinical Neuropsychologist, 11(4), 416-420. Comparison of performance from three conti-
Allen, J., Blanton, P., Johnson-Greene, D., Murphy- nents on the WHO-recommended neurobeha-
Farmer, C., & Gross, A. (1992). Need for achieve- vioral core test battery. Environmental Research,
ment and performance on measures of behavioral 62, 1~147.
fluency. Psychological Reports, 71(2), 471-478. Anil, A. E., Kivircik, B. B., Batur, S., Kabakci, E.,
Allen, J. B., Gross, A.M., Aloia, M.S., & Billingsley, C. Kitis, A., Giiven, E., et al. (2003). The Turlcish
(1996). The effects of glucose on nonmemory version of the Auditory Consonant Trigram Test as
cognitive functioning in the elderly. Neurupsy- a measure of working memory: A normative study.
clwlogia, 34(5), 459-465. Clinical Neuropsychologist, 17(2), 159-lf;9.
American Academy of Neurology (1996). As- Annett, M. (1970). A classification of hand prefer-
sessment: Neuropsychological testing of adults. ence by association analysis. British Journal of
Considerations for neurologists. Report of the Psychology, 61, 303--321.
Therapeutics and Technology Assessment Sub- Anstey, K. J., & Smith, G. A. (1999). Interrelation-
committee. Neurology, 47(2), 592-599. ships among biological markers of aging. health,
American Psychological Association (1999). Stan- activity, acculturation, and cognitive performance
dards for Educational and Psychological Testing. in late adulthood. Psychology and Aging. 14(4),
American Psychological Association (2002). Ethical 605-618.
principles of psychologists and code of conduct. Anstey, K. J., Matters, B., Brown, A. K., &: Lord,
American Psychologist, 57(12), 1060-1073. S. R. (2000). Normative data on neuropsycholog-
Amieva, H., Lafont, S., Auriacombe, S., Rainville, C., ical tests for very old adults living in retirement
Orgogozo, J.-M., Dartigues, J.-F., et al. (1998). villages and hostels. Clinical Neuropsychologist.
Analysis of error types in the Trail Making Test 14(3), 309--317.
evidences inhibitory deficit in dementia of the Antes, G., & Oxman, A. D. (for the Cochrane Col-
Alzheimer type. Journal of Clinical and Experi- laboration) (2001). The Cochrane Collaboration
mental Neuropsychology. 20(2), 280-285. in the 20th century. In M. Egger, G. D. Smith, &
Amir, T. (2001). Benton Visual Retention Test: D. G. Altman (Eds.), Systematic rwlew8 in hetJth
Reliability, gender, and the effect of extended care: Meta-analysis in contert. London: BMJ.
practice on the performance of participants from Anthony, W. Z., Heaton, R. K., & Lehman, R. A. W.
UAE. Bulletin of the Faculty of Arts, Cairo (1980). An attempt to cross-validate two actuarial
University, 61(2), 7-17. systems for neuropsychological test interpreta-
Amir, T., & Bahri, T. (1994). Effect of substance tion. Journal of Consulting and Clinicol Psychol-
abuse on visuographic function. Perceptual and ogy, 48(3), 317-326.
Motor Skills, 78(1), 235-241. Arbuckle, T. Y., & Gold, D. P. (1993). Aging, in-
Anastasi, A. (1988). Norms and the interpretations hibition, and verbosity. Journals of Gerontology,
of test scores. In A. Anastasi (Ed.), Psychologi- 48(5), P~P232.
cal Testing (6th ed.) (pp. 71-108). New York: Arbuthnott, K., & Frank, J. (2000). Trail Making
MacMillan. Test, part B as a measure of executive control:
Andel, R., McCleary, C. A., Murdock, G. A., Fiske, A., Validation using a set-switching paradigm. Jour-
Wilcox, R. R., & Gatz, M. (2003). Performance nal of Clinical and Experimental Neuropsychol-
on the CERAD Word List Memory task: A ogy, 22(4), 518-528.
comparison of university-based and community- Ardila, A., & Rosselli, M. (1989). Neuropsycholog-
based groups. International Journal of Geriatric ical characteristics of normal aging. Develop-
Psychiatry, 18(8), 733--739. mental Neuropsychology, 5(4), 307-320.
REFERENCES 535
Ardila, A., & Rosselli, M. (2003). Educational ef- mance of Hispanics on selected Halstead-Reitan
fects on ROCF performance. In J. A. Knight neuropsychological tests. Assessment, 1(3),
(Ed.), The handbook of Rey-Osterrieth Complex 239-248.
Figure usage: Clinical and research applications. Arria, A.M., Tarter, R. E., Kabene, M.A., Laird,
Lutz, FL: Psychological Assessment Resources. S. B., Moss, H. & Van Thiel, D. M. (1991). The
Ardila, A., Rodriguez-Menendez, G., & Rosselli, M. role of cirrhosis in memory functioning of alco-
(2002). Current issues in neuropsychological holics. Alcolwlism: Clinical and Experimental
assessment with Hispanic/Latinos. In F.R. Fer- Research, 15(6), 932-937.
raro (Ed.), Minority and cross-cultural aspects of Artiola i Fortuny, L., & Heaton, R. K. (1996).
neuropsyclwlogical assessment. Studies on neu- Standard versus computerized administration of
ropsychology, development, and cognition. Lisse: the Wisconsin Card Sorting Test. Clinical Neu-
Swets & Zeitlinger. ropsyclwlogist, 10(4), 419-424.
Ardila, A., Rosselli, M., & Rosas, P. (1989). Neuro- Artiola i Fortuny, L., Heaton, R. K., & Hermosillo, D.
psychological assessment in illiterates: VJSuospa- (1998). Neuropsychological comparisons of
tial and memory abilities. Brain and Cognition, Spanish-speaking participants from the U.S.-
11(2), 147-166. Mexico border region versus Spain. Journal of the
Arena, R., & Gainotti, G. (1978). Constructional International Neuropsychological Society, 4(4),
apraxia and visuoperceptive disabilities in rela- 363--379.
tion to laterality of cerebral lesions. Cortex, Artiola i Fortuny, L., Hermosillo Romo, D. H.,
14(4), 463-473. Heaton, R. K., & Pardee, R. E. III (1999). Manual
Arenberg. D. (1978). Differences and changes with tk nonnas y procedimientos para Ia bateria neu-
age in the Benton Visual Retention Test. Journal ropsicologica en Espanal. Tucson, AZ: mPress.
of Gerontology, 33(4), 534-540. Ashendorf, L., O'Bryant, S. E., & McCaffrey, R. J.
Arenberg, D. (1982). Estimates of age changes on (2003). Specificity of malingering detection strat-
the Benton Visual Retention Test. Journal of egies in older adults using the CVLT and WCST.
Gerontology, 37(1), 87-90. Clinical Neuropsyclwlof!i.st, 17(2), 255-262.
Arias Bal, M.A., Vazquez-Barquero, J. L., Pena, C., Au, R., Joung, P., Nicholas, M., Ohler, L. K., et al.
Miro, J., & Berciano, J. A. (1991). Psychiatric (1995). Naming ability across the adult life span.
aspects of multiple sclerosis. Acta Psychiatrico Aging and Cognition, 2(4), 300--311.
Scandinavico, 83(4), 292-296. Audenaert, K., Brans, B., Van Laere, K., Lahorte, P.,
Arima, J. K. (1965). Performance of normal males · Versijpt, J., van Heeringen, K., Dierckx, R. (2000).
on the Halstead Tactual Performance Test under Verbal fluency as a prefrontal activation probe: a
severe environmental stress. Perceptual and Mo- validation study using 99mTc-ECD brain SPET.
tor Skills, 21, 83--90. European Journal of Nuclear Medicine, 27(12),
Armengol, C. G. (2002). Stroop test in Spanish: 1800-1808.
Children's norms. Clinical Neuropsychologist, Aupperle, R. L., Beatty, W. W., Shelton, F., &
16(1), 67--80. Gontkovsky, S. T. (2002). Three screening bat-
Armstrong. C., Onishi, K., Robinson, K., D'Espo- teries to detect cognitive impairment in multiple
sito, M., Thompson, H., Rostami, A., et al. (1996). sclerosis. Multiple Sclerosis, 8, 382-389.
Serial position and temporal cue effects in mul- Austin, M. P., Ross., M., O'Carroll, R. E., Ebmeier,
tiple sclerosis: Two subtypes of defective memory K. P., & Goodwin, G. M. (1992). Cognitive
mechanisms. Neuropsychologia, 34(9), 853--862. dysfunction in major depression. Journal of Af
Army Individual Test Battery (1994). Manual of fective Disorder, 25, 21-30.
directions and scoring. Washington, DC: War Axelrod, B. N. (2002). Are normative data from the64-
Department, Adjutant General's Office. card version of the WCST comparable to the full
Arnett, J. A., & Labovitz, S. S. (1995). Effect of WCST? Clinical Neuropsychologist,16(1), 7-11.
physical layout in performance of the Trail Making Axelrod, B. N., & Goldman, R. S. (1996). Use of
Test. Psychological Assessment, 7(2), 220-221. demographic corrections in neuropsychological
Arnett, P. A., Rao, S.M., Bernardin, L., Grafman, J., interpretation: How standard are standard scores?
Yetkin, F. Z., & Lobeck, L. (1994). Relationship Clinical Neuropsyclwlof!i.st, 10(2), 15~162.
between frontal lobe lesions and WCST perfor- Axelrod, B. N., & Henry, R. R. (1992). Age-related
mance by patients with multiple sclerosis. Neu- performance on the Wisconsin Card Sorting,
rology, 44, 420--424. Similarities, and Controlled Oral Word Associ-
Arnold, B. R., Montgomery, G. T., Castaneda, 1., & ation Tests. Clinical Neuropsyclwlogist, 6(1),
Longoria, R. (1994). Acculturation and perfor- 16-26.
536 REFERENCES
Axelrod, B. N., & Milner, I. B. (1997). Neuropsy- Baker, R., Donders, J., & Thompson, E. (2000).
chological findings in a sample of Operation Assessment of incomplete effort with the Cali-
Desert Storm veterans. Journal of Neuropsychi- fornia Verbal Learning Test. Applied Neuropsy-
atry and Clinical Neurosciences, 9(1), 23--38. chology, 7(2), 111-114.
Axelrod, B. N., Goldman, R. S., & Woodard, J. L. Baldo, J. V., Shimamura, A. P., Delis, D. C., Kra-
(1992a). Interrater reliability in scoring the Wis- mer, J., & Kaplan, E. (2001). Verbal and design
consin Card Sorting Test. Clinical Neuropsy- fluency in patients with frontal lobe lesions.
chologist, 6{2), 143-155. Journal of the International NeuropsychologU:td
Axelrod, B. N., Henry, R. R., & Woodard, J. L. Society, 7(5), 586-596.
(1992b). Analysis of an abbreviated form of the Baldo, J. V., Delis, D., Kramer, J., & Shimamura,
Wisconsin Card Sorting Test. Clinical Neuro- A. P. (2002). Memory performance on the Cali-
psychologist, 6(1), 27-31. fornia Verbal Learning Test-11: Findings from
Axelrod, B. N., Jiron, C. C., & Henry, R. R. (1993). patients with focal frontal lesions. Joumal of the
Performance of adults ages 20 to 90 on the Ab- International NeuropsychologU:td Society, 8(4),
breviated Wisconsin Card Sorting Test. Clinical 539-546.
Neuropsychologist, 7(2), ~209. Barbarotto, R., Laiacona, M., Frosio, R., Vecchio, M.,
Axelrod, B. N., Goldman, R. S., Tompkins, L. M., & Farinato, A., & Capitani, E. (1998). A normative
Jiron, C. C. (1994a). Poor differential perfor- study on visual reaction times and two Stroop
mance on the Wisconsin Card Sorting Test in colour-word tests. Italian Joumal of Neurolog#col
schizophrenia, mood disorder, and traumatic Science, 19(3), 161-170.
brain injury. Neuropsychiatry, Neuropsychology, Barcelo, F. (1999). Electrophysiological evidence of
and Behavioral Neurology. 7{1), 20-24. two different types of error in the Wisconsin Card
Axelrod, B. N., Greve, K. W., & Goldman, R. S. Sorting Test. Neuroreport, 10(6), 1~1303.
(1994b). Comparison of four Wisconsin Card Bardwell, W. A., Ancoli-Israel, S., Berry, C. C., &
Sorting Test scoring guides with novice raters. Dimsdale, J. E. (2001). Neuropsychological ef-
Assessment, 1{2), 115-121. fects of one-week continuous positive airway
Axelrod, B. N., Goldman, R. S., Heaton, R. K., pressure treatment in patients with obstructive
Curtiss, G., et al. (1996). Discriminability of the sleep apnea: A placebo-controlled study. Psy-
Wisconsin Card Sorting Test using the stan- chosomatic Medicine, 63(4), 579-584.
dardization sample. Journal of Clinical and Ex- Barker-Collo, S. L. (2001). The 60-item Boston
perimental Neuropsychology, 18(3), 338--342. Naming Test: Cultural bias and possible ad-
Axelrod, B. N ., Aharon-Peretz, J., Tomer, R., & aptations for New Zealand Aphasiology, 15(1),
Fisher, T. (2000a). Creating interpretation guide- 85-92.
lines for the Hebrew Trail Making Test. Applied Barker-Collo, S., Clarkson, A, Cribb, A, & Grogan. M.
Neuropsychology, 7(3), 186-188. (2002). The impact of American content on Ca-
Axelrod, B. N., Heilbronner, R., Barth, J., Larra- lifornia Verbal Learning Test performance: A
bee, G., Faust, D., Pliskin, N., et al. (2000b). New Zealand illustration. Clinical Neuropsy-
Test security: Official position statement of the chologist, 16(3), 290-299.
National Academy of Neuropsychology. Archives Barncord, S. W., & Wanlass, R. L. (1999). Paper or
of Clinical Neuropsychology. 15(5), 383-386. plastic: Another ecological consideration in neu-
Axelrod, B. N., Tomer, R., Fisher, T., & Aharon- ropsychological assessment. Applied Neuropsy-
Peretz, J. (2001). Preliminary analyses of Hebrew chology, 6(2), 121-122.
verbal fluency measures. Applied Neuropsychol- Barncord, S. W., & Wanlass, R. L. (2001). The
ogy, 8(4), 248-250. Symbol Trail Making Test: Test development
Baddeley, A. (1986). Working memory. Oxford and utility as a measure of cognitive impairment.
Psychology Series 11. New York: Oxford Uni- Applied Neuropsychology, 8(2), 99-103.
versity Press. Baron, I. S. (2004). NeuropsychologU:td evalua-
Baddeley, A. (1996). Exploring the central execution of the child. New York: Oxford University
tive. Quarterly Journal of Experimental Psychol- Press.
ogy: Human Experimental Psychology. Special Barr, A., & Brandt, J. (1996). Word-list generation
Issue: Working Memory, 49A(1), 5-28. deficits in dementia. Joumal of Clinical and Ex-
Bak, J. S., & Greene, R. L. (1980). Changes in perimental Neuropsychology, 18(6), 810-822.
neuropsychological functioning in an aging pop- Barr, W. B. (2003). Assessment of temporal lobe
ulation. Journal of Consulting and Clinical Psy- epilepsy using the ROCF. In J. A. Knight (Ed.),
chology, 48(3), 395-399. The handbook of Rey-Ostenieth Complex Figure
REFERENCES 537
usage: Clinical and research applications. Lutz, Bate, A. J., Mathias, J. L., & Crawford, J. R. (2001).
FL: Psychological Assessment Resources. Performance on the Test of Everyday Attention
Barrash, J., Suhr, J., & Manzel, K. (2004). Detecting and standard tests of attention following severe
poor effort and malingering with an expanded traumatic brain injury. Clinical Neuropsycholo-
version of the Auditory Verbal Learning Test gist, 15(3), 405--422.
(AVLTX): Validation with clinical samples. Jour- Battig, W. F., & Montague, W. E. (1969). Category
nal of Clinical and Experimental Neuropsychol- norms of verbal items in 56 categories: A repli-
ogy, 26(1), 125-140. cation and extension of the Connecticut category
Barreca, S. R., Finlayson, M.A. J., Gowland, C. A., & norms. Journal of Experimental Psychology.
Basmajian, J. V. (1999). Use of the Halstead 80(3), 1-46.
Category Test as a cognitive predictor of func- Bayles, K. A., & Tomoeda, C. K. (1983). Confron-
tional recovery in the hemiplegic upper limb: A tation naming impairment in dementia. Brain
cross-validation study. Clinical Neuropsychologist, and Language, 19, 98-112.
13(2), 171-181. Bayles, K. A., Salmon, D. P., Tomoeda, C. K., Ja-
Barresi, B. A., Nicholas, M., Tabor Connor, L., cobs, D., Caffrey, J. T., Kaszniak, A. W., et al.
Ohler, L. K., & Albert, M. L. (2000). Semantic (1989). Semantic and letter category naming in
degradation and lexical access in age-related Alzheimer's patients: A predictable difference.
naming failures. Aging. Neuropsychology, and Developmental Neuropsychology, 5(4), 335-347.
Cognition, 7(3), 16~178. Bayley, P. J., Salmon, D. P., Bondi, M. W., Bui,
Barrett, D. H., Morris, R. D., Akhtar, F. Z., & B. K., Olichney, J., Delis, D. C., et al. (2000).
Michalek, J. E. (2001). Serum dioxin and cog- Comparison of the serial position effect in very
nitive functioning among veterans of Operation mild Alzheimer's disease, mild Alzheimer's dis-
Ranch Hand. Neurotoxicology, 22, 491--502. ease, and amnesia associated with electro-
Barrett-Connor, E., & Goodman-Gruen, D. (1999). convulsive therapy. Journal of the International
Cognitive function and endogenous sex hor- Neuropsychological Society, 6(3), 290--298.
mones in older women. Journal of the American Beatty, W. W. (1993). Age differences on the Cali-
Geriatrics Society, 47(11), 128~1293. fornia Card Sorting Test: Implications for the
Bartels, M., & Themelis, J. (1983). Computerized assessment of problem solving by the elderly.
tomography in tardive dyskinesia: Evidence of BuUetin of the Psychonomic Society, 31(6), 511-
structural abnormalities in the basal ganglia sys- 514.
tem. Archiv foer Psychiatrie und Neroenkran- Beatty, W. W. & Monson, N. (1990). Problem
kheiten, 233(5), 371-379. solving in Parkinson's disease: Comparison of
Bartfai, A., Winborg, I. M., Nordstroem, P., & performance on the Wisconsin and California
Asberg, A. (1990). Suicidal behavior and cogni- Card Sorting Tests. Journal of Geriatric Psychi-
tive flexibility: Design and verbal fluency after atry and Neurology, 3, 163-171.
attempted suicide. Suicide and Life-Threatening Beatty, W. W., Goodkin, D. E., Monson, N., &
Behavior, 20(3), 254-266. Beatty, P. A. (1989). Cognitive disturbances in
Bartok, J. A., Wilson, C. S., Giordani, B., Keys, B. A., patients with relapsing remitting multiple scle-
Persad, C. C., Foster, N. L., et al. (1997). Varying rosis. Archives of Neurology, 46, 1113-1119.
patterns of verbal recall, recognition, and re- Beatty, W. W., Jocic, Z., Monson, N ., & Staton,
sponse bias with progression of Alzheimer's dis- R. D. (1993). Memory and frontal lobe dys-
ease. Aging. Neuropsychology, and Cognition, function in schizophrenia and schizoaffective
4(4), 266--272. disorder. Journal ofNeroous and Mental Disease,
Baser, C. A., & Ruff, R. M. (1987). Construct va- 181(1), 448-453.
lidity of the San Diego Neuropsychological Test Beatty, W. W., Jocic, Z., Monson, N ., & Katzung,
Battery. Archives of Clinical Neuropsychology, V. M. (1994). Problem solving by schizophrenic
2(1), 13--32. and schizoaffective patients on the Wisconsin
Basso, M. R., Bomstein, R. A., & Lang, J. M. (1999). and California Card Sorting Tests. Neuropsy-
Practice effects on commonly used measures of chology, 8(1), 4~4.
executive function across twelve months. Clinical Beatty, W. W., Hames, K. A., Blanco, C. R., Paul,
Neuropsychologist, 13(3), 283-292. R. H., et al. (1995). Verbal abstraction deficit
Basso, M. R., Harrington, K., Matson, M., & Low- in multiple sclerosis. Neuropsychology, 9(2),
ery, N. (2000). Sex differences on the WMS-III: 198-205.
Findings concerning verbal paired associates and Beatty, W. W., Krull, K. R., Wilbanks, S. L., Blanco,
faces. Clinical Neuropsychologist, 14(2), 231-235. C. R., Hames, K. A., & Paul, R. H. (1996a).
538 REFERENCES
Further validation of constructs from the Selec- Visuospatial Memory Test: Studies of normal
tive Reminding Test. Journal of Clinical and performance, reliability, and validity. Psycho-
Experimental Neuropsychology, 18(1), 52-55. logical Assessment, 8(2), 145-153.
Beatty, W. W., Wilbanks, S. L., Blanco, C. R., Benedict, R. H. B., Schretlen, D., Groninger, L., &
Hames, K. A., Tivis, R., & Paul, R. H. (1996b). Brandt, J. (1998). Hopkins Verbal Learning
Memory disturbance in multiple sclerosis: Re- Test-Revised: Normative data and analysis of
consideration of patterns of performance on the inter-form and test-retest reliability. Clinical
Selective Reminding Test. Journal ofClinical and Neuropsychologist, 12(1), 43--55.
Experimental Neuropsychology, 18(1), 56--62. Benedict, R. H. B., & Zgaljardic, D. J. (1998).
Beatty, W. W., Testa, J. A., English, S., & Winn, P. Practice effects during repeated administrations
(1997). Influences of clustering and switching on of memory tests with and without alternate
the verbal fluency performance of patients with forms. Journal of Clinical and Experimental
Alzheimer's disease. Aging, Neuropsychology, Neuropsychology, .20(3), 33~2.
and Cognition, 4(4), 273-279. Benito-Cuadrado, M. M., Esteba-CastiDo, S.,
Beatty, W. W., Salmon, D. P., Troester, A. 1., & Boehm, P., Cejudo-Bolivar, J., & Pena-Casanova,
Tivis, R. D. (2002). Do primary and supple- J. (2002). Semantic verbal fluency of animals: A
mentary measures of semantic memory predict normative and predictive study in a Spanish
cognitive decline by patients with Alzheimer's population. Journal of Clinical and Experimental
disease? Aging, Neuropsychology, and Cogni- NeuropsycholorJJ. 24(8), 1117-1122.
tion, 9(1), 1-10. Bennett-Levy, J. (1984). Determinants of perfor-
Bechtoldt, H. P., Benton, A. L., & Fogel, M. L. mance on the Rey-Osterrieth Complex Figure
(1962). An application of factor analysis in neu- test: An analysis, and a new technique for single-
ropsychology. Psychological Record, 12, 147-156. case assessment. British Journal of Clinical
Becker, J. T. (1988). Working memory and sec- Psychology, .23, 109-119.
ondary memory deficits in Alzheimer's Disease. Benton, A. (1945). A visual retention test for clini-
Journal of Clinical and Experimental Neuropsy- cal use. Archives of Neurolora and Psychiatry,
chology, 10(6), 73~753. 54, 212-216.
Becker, J. T., Huff, F. J., Nebes, R. D., Holland, A., & Benton, A. (1962). The VISual Retention Test as a
Boller, F. H. (1988). Neuropsychological function constructional praxis task. ConfinitJ Neurologica,
in Alzheimer's disease: Patterns ofimpainnent and 22, 141-155.
rate of progression. Archives of Neurology, 45, Benton, A. (1963). Revised Visual Retention Test:
263-268. Clinical and experimental applications (3rd ed.).
Beebe, D. W., Ris, M.D., & Dietrich, K. N. (2000). New York: The Psychological Corporation.
The relationship between CVLT-C process scores Benton, A. (1967). Problems of test construction in
and measures of executive functioning: Lack of the field of aphasia. Cortex, 3, 32-58.
support among community-dwelling adolescents. Benton, A. (1972). Abbreviated versions of the Vi-
Journal of Clinical and Experimental Neuropsy- sual Retention Test. Journal of Psycho/ora, 80,
chology, 22(6), 77~792. 18~192.
Bell, B. D., Davies, K. G., Hermann, B. P., & Benton, A. (1974). Revised Visual Retention Test:
Walters, G. (2000). Confrontation naming after Clinical and experimental applications (4th ed.).
anterior temporal lobectomy is related to age of San Antonio, TX: The Psychological Corporation.
acquisition of the object names. Neuropsycho- Benton, A., & Hamsher, K. (1978). Multilingual
logia, 38(1), 83-92. Aphasia Examination mtlnual. Iowa City: Uni-
Bell, B. D., Hermann, B. P., Woodard, A. R., Jones, versity of Iowa.
J. E., Rutecki, P. A., Sheth, R., et al. (2001). Benton, A., Hannay, H. J., & Varney, N. R. (1975).
Object naming and semantic knowledge in Visual perception of line direction in patients
temporal lobe epilepsy. Neuropsychology, 15(4), with unilateral brain disease. Neurolora• .25(10),
434-443. 907-910.
Bench, C. J., Frith, C. D., Grasby, P. M., Friston, Benton, A., Varney, N. R., & Hamsher, K. (1978).
K. J., Paulseu, E., Frackowiak, R. S. J., et al. Visuospatial judgment: A clinical test. Archives of
(1993). Investigations of the functional anatomy Neurology, 35(6), 364-367.
of attention using the Stroop Test. Neuropsy- Benton, A., Eslinger, P. J., & Damasio, A. R. (1981).
chologia, 31, 907-922. Normative observations on neuropsychological
Benedict, R. H. B., Schretlen, D., Groninger, L., test performances in old age. Journal of Clinical
Dobraski, M., et al. (1996). Revision of the Brief Neuropsycholora. 3(1), 33-42.
REFERENCES 539
Benton, A., Hamsher, K., Varney, N. R., & Spreen, hypothesis. Archives of Clinical Neuropsychol-
0. (1983a). Contributions to neuropsyclwlogical ogy, 11, 231-245.
assessment: A Clinical manual. New York: Ox- Berning, L. C., Weed, N.C., & Aloia, M. S. (1998).
ford University Press. Interrater reliability of the Ruff Figural Fluen(.y
Benton, A., Hamsher, K., Varney, N. R., & Sp- Test. Assessment, 5(2), 181-186.
reen, 0. (1983b). Visual Form Discrimination. Bernstein, J. H. (2003). Interpreting the ROCF
New York: Oxford University Press. productions of children. In J. A. Knight (Ed.),
Benton, A., Hamsher, K., & Sivan, A. B. (1994a). The handbook of Rey-Ostenieth Complex Figure
Multilingual Aphasia Examination. Iowa City: usage: Clinical and research applications. Lutz,
AJA Associates. FL: Psychological Assessment Resources.
Benton, A., Sivan, A. B., Hamsher, K., Varney, N. R., Bernstein, J. H., & Waber, D. P. (1996). Develop-
& Spreen, 0. (1994b). Contributions to neuro- mental Scoring System for the Rey-Osterrieth
psyclwlogical assessment-A clinical manual (2nd ComplexFigure(DSS-ROCF):Professionalmanual.
ed.). New York: Oxford University Press. Odessa, FL: Psychological Assessment Resources.
Benton-Sivan, A. (1992). Benton Visual Retention Berry, D. T. R., & Carpenter, G. C. (1992). Effect
Test (5th ed.). San Antonio: Psychological Cor- of four different delay periods on recall of the
poration. Rey-Osterrieth Complex Figure by older per-
Berg, E. A. (1948). A simple objective technique for sons. Clinical Neuropsychologist, 6( 1), 80-84.
measuring flexibility in thinking. Journal of Gen- Berry, D. T. R., Allen, R. S., & Schmitt, F. A.
eral Psychology, 39, 15--22. (1991). Rey-Osterrieth Complex Figure: Psycho-
Berman, K. F., Doran, A. R., Pickar, D., & Wein- metric characteristics in a geriatric sample. Clini-
berger, D. R. (1993). Is the mechanism of pre- cal Neuropsyclwlogist, 5(2), 143--153.
frontal hypofunction in depression the same as in Bertolucci, P. H. F., Okamoto, I. H., Brucki, S. M. D.,
schizophrenia? Regional cerebral blood flow Siviero, M. 0., Toniolo Neto, J., & Ramos, L. R.
during cognitive activation. British Journal of (2001). Applicability of the CERAD neuropsycho-
Psychiatry, 162, 183-192. logical battery to Brazilian elderly. Arquivos de
Berman, K. F., Ostrem, J. L., Randolph, C., Gold, J., Neuro-Psiquiatria, 59(3-A), 532-536.
Goldberg, T. E., Coppola, R. E., et al. (1995). Bherer, L., Belleville, S., & Peretz, I. (2001). Edu-
Physiological activation of a cortical network dur- cation, age, and the Brown-Peterson technique.
ing performance of the Wisconsin Card Sorting Developmental Neuropsyclwlogy, 19(3), 237-251.
Test: A positron emission tomography study. Neu- Bieliauskas, L. A., Adams, K. M., Fennell, E.,
ropsychologia, 33, 1027-1046. Hammeke, T., & Rourke, B. (1997a). Assessment
Bernard, L. C. (1989). Halstead-Reitan Neuropsy- of neuropsychological testing. Neurology, 49,
chological Test performance of black, Hispanic, 1182--1183.
and white young adult males from poor aca- Bieliauskas, L. A., Fastenau, P. S., Lacy, M. A., &
demic backgrounds. Archives of Clinical Neu- Roper, B. L. (1997b). Use of the odds ratio
ropsychology, 4, 267-274. to translate neuropsychological test scores into
Bernard, L. C. (1990). Prospects for faking believ- real-world outcomes: From statistical significance
able memory deficits on neuropsychological tests to clinical significance. Journal of Clinical and
and the use of incentives in simulation research. Experimental Neuropsychology, 19(6), 889-896.
Journal of Clinical and Experimental Neuropsy- Bigler, E. D. (2003). Neuroimagingand the ROCF. In
chology, 12, 715--728. J. A. Knight (Ed.), The handbookofRey-Osterrieth
Bernard, L. C. (1991). The detection of faked Complex Figure usage: Clinical and research
deficits on the Rey Auditory Verbal Learning applications. Lutz, FL: Psychological Assessment
Test: The effect of serial position. Archives of Resources.
Clinical Neuropsychology, 6, 81-88. Bigler, E. D., & Dodrill, C. B. (1997). Assessment
Bernard, L. C., Houston, W., & Natoli, L. (1993). of neuropsychological testing: Comment. Neu-
Malingering on neuropsychological memory tests: rology, 49(4), 1180-1181.
Potential objective indicators. Journal of Clinical Bigler, E. D., & Tucker, D. M. (1981). Comparison
Psyclwlogy, 49, 45--53. of Verbal IQ, Tactual Performance, Seashore
Bernard, L. C., McGrath, M. J., & Houston, W. Rhythm and Finger Oscillation tests in the blind
(1996). The differential effects of simulating and brain-damaged. Journal of Clinical Psy-
malingerers, closed head injury, and other CNS chology, 37(4), 849-851.
pathology on the Wisconsin Card Sorting Test: Bigler, E. D., Steinman, D. R., & Newton, J. S.
Support for the "pattern of performance" (1981a). Clinical assessment of cognitive deficit
540 REFERENCES
in neurologic disorder: I. Effects of age and cognitive impairment in the oldest old. Neurol-
degenerative disease. International journal of ogy, 60(3), 477-480.
Clinical Neuropsychology, 3(3), 5-13. Boll, T. J., & Reitan, R. M. (1973). Effect of age on
Bigler, E. D., Steinman, D. R., & Newton, J. S. performance on the Trail Making Test. Percep-
(1981b). Clinical assessment of cognitive deficit tual and Motor Skills, 36, 691--694.
in neurologic disorder. II: Cerebral trauma. Bolla, K. 1., Lindgren, K. N., Bonaccorsy, C., &
Clinical Neuropsychology, 3(3), 13--18. Bleecker, M. L. (1990). Predictors of verbal
Bigler, E. D., Rosa, L., Schultz, F., Hall, S., & fluency (FAS) in the healthy elderly. Journal of
Harris, J. (1989). Rey Auditory-Verbal Learning Clinical Psychology, 46(5), 623--628.
and Rey-Osterrieth Complex Figure Design Bolla-Wilson, K., & Bleecker, M. L. (1986). Influ-
performance in Alzheimer's disease and closed ence of verbal intelligence, sex, age, and educa-
head injury. Journal of Clinical Psyehology, tion on the Rey Auditory Verbal Learning Test.
45(2), 277-280. Developmental Neuropsychology, 2(3), 203--211.
Binder, L. M. (1982). Constructional strategies on Bondi, M. W., Kaszniak, A. W., Bayles, K. A., &
complex figure drawings after unilate!ll brain Vance, K. (1993). Contributions of frontal system
damage. Journal of Clinical Neuropsythology, dysfunction to memory and perceptual abilities in
4(1), 51-58. Parkinson's patients. Neuropsychology, 7, 89-102.
Binder, L. M., Villanueva, M. R., Howiesoo, D., & Bondi, M. W., Monsch, A. U., Butters, N., Salmon,
Moore, R. T. (1993). The Rey AVLT recpgnition D. P., & Paulsen (1993). Utility of a modified
memory task measures motivational im!*irment version of the Wisconsin Card Sorting Test in the
after mild head trauma. Archives of Clinical detection of dementia of the Alzheimer type.
Neuropsychology, 8, 137-147. Clinical Neuropsychologist, 7(2), 161-170.
Binder, E. F., Storandt, M., & Birge, S. J. (1999). Boone, K. B. (1999). Neuropsychological assess-
The relation between psychometric test. perfor- ment of executive functions: Impact of age, ed-
mance and physical performance in oldet adults. ucation, gender, intellectual level, and vascular
Journals of Gerontology: Series A: Biologfcal Sci- status on executive test scores. In B. L. Miller &
ences and Medical Sciences, 54(8), M~M432. J. L. Cummings (Eds.), The human frontal lobes:
Binder, L. M., Kelly, M. P., Villanueva, M. R., & Functions and disorders. The science and prac-
Winslow, M. M. (2003). Motivation ~d neu- tice of neuropsychology series (pp. 247-260).
ropsychological test performance following mild New York: Guilford Press.
head injury. Journal of Clinical and Experimen- Boone, K. B. (2000). The Boston Qualitative Scor-
tal Neuropsychology, 25(3), 420-430. ing System for the Rey-Osterrieth Complex
Binetti, G., Magni, E., Padovani, A., Cappa, S. F., Figure. Journal of Clinical and Experimental
et al. (1995). Release from proactive interference Neuropsychology, 22(3), 430-432.
in early Alzheimer's disease. Neuropsyd10logia, Boone, K. B., Miller, B. L., Rosenberg, L., Durazo, A.,
33(3), 379--384. Mcintyre, H., & Wei), M. (1988). Neuro-
Binetti, G., Magni, E., Padovani, A., Capp., S. F., psychological and behavioral abnormalities in an
Bianchetti, A., & Trabucchi, M. (1996). Execu- adolescent with frontal lobe seizures. Neurology,
tive dysfunction in early Alzheimer's disease. 38, 583--586.
Journal of Neurology, Neurosurgery, a'd Psy- Boone, K. B., Miller, B. L., Lesser, I. M., Hill, E., &
chiatry, 60, 91-93. D'Elia, L. (1990). Performance on frontal lobe
Blachstein, H., Vakil, E., & Hoffien, D. (19fl3). Im- tests in healthy, older individuals. Developmental
paired learning in patients with clo~d-head Neuropsychology. 6(3), 215-223.
injuries: An analysis of components of th, acqui- Boone, K. B., Ananth, J., Philpott, L., Kaur, A., &
sition process. Neuropsychology, 7(4), 53()....535. Djenderedjian, A. (1991). Neuropsycholog-
Bleecker, M. L., Bolla-Wilson, K., Agnew, J., & ical characteristics of nondepressed adults with
Meyers, D. A. (1988). Age-related se" differ- obsessive-compulsive disorder. Neuropsychia-
ences in verbal memory. Journal of Clinical try, Neuropsychology, and Behavioral Neurol-
Psychology, 44(3), 403--411. ogy, 4, 96-109.
Blusewicz, M. J., Kramer, J. H., & Delmonioo, R. L. Boone, K. B., Miller, B. L., Lesser, I. M.,
(1996). Interference effects in chronic alctholism. Mehringer, C. M., Hill-Gutierrez, E., Gold-
Journal of the International Neuropsychological berg, M. A., et al. (1992). Neuropsychological
Society, 2(2), 141-145. correlates of white-matter lesions in healthy
Boeve, B., McCormick, J., Smith, G., Fennan, T., elderly subjects: A threshold effect. Archives of
Rummans, T., Carpenter, T., et al. (200~). Mild Neurology, 49, 549--554.
REFERENCES 541
Boone, K. B., Ghaffarian, S., Lesser, I. M., Hill- sample. Journal of Clinical Psychology, 41(5),
Gutierrez, E., & Bennan, N. G. (1993a). 651-658.
Wisconsin Card Sorting Test perfonnance in Bornstein, R. A. (1986a). Classification rates obtained
healthy, older adults: Relationship to age, sex, with "standard" cut-off scores on selected neu-
education, and IQ. Journal of Clinical Psychol- ropsychological measures. Journal of Clinical and
ogy. 49(1), 54-60. Experimental Neuropsychology, 8(4), 413-420.
Boone, K. B., Lesser, I. M., Hill-Gutierrez, E. H., Bornstein, R. A. (1986b). Contribution of various
Bennan, N. G., & D'Elia, L. F. (1993b). Rey- neuropsychological measures to detection of
Osterrieth Complex Figure perfonnance in healthy, frontal lobe impainnent. International Journal of
older adults: Relationship to age, education, sex Clinical Neuropsychology, 8(1), 18-22.
and IQ. Clinical Neuropsychologist. 7(1), 22-28. Bornstein, R. A. (1986c). Normative data on in-
Boone, K. B., Miller, B. L., & Lesser, I. M. (1993c). tennanual differences on three tests of motor
Frontal lobe cognitive functions in aging: Meth- perfonnance. Journal of Clinical and Experi-
odologic considerations. Dementia, 4, 232-236. mental Neuropsychology, 8(1), 12-20.
Boone, K. B., Lesser, I. M., Miller, B. L., Wohl, M., Bornstein, R. A. (1986d). Consistency of inter-
Bennan, N., Lee, A., et al. (1995). Cognitive manual discrepancies in nonnal and unilateral
functioning in older depressed outpatients: Re- brain lesion patients. Journal of Consulting and
lationship of presence and severity of depression Clinical Psychology, 54(5), 719-723.
to neuropsychological test scores. Neuropsy- Bornstein, R. A. (1990). Neuropsychological test
chology, 9(3), 390--398. batteries in neuropsychological assessment. In
Boone, K. B., Ponton, M. 0., Gorsuch, R. L., A. A. Boulton, G. B. Baker, & M. Hiscock (Eds.),
Gonzalez, J. J., & Miller, B. L. (1998). Factor Neuromethods-17: Neuropsychology. Clifton, NJ:
analysis of four measures of prefrontal lobe Humana Press.
functioning. Archives of Clinical Neuropsychol- Bornstein, R. A., & Suga, L. J. (1988). Educational
ogy, 13(1), 585-595. level and neuropsychological perfonnance in
Boone, K. B., Miller, B. L., Lee, A., Bennan, N., healthy elderly subjects. Developmental Neuro-
Shennan, D., & Stuss, D. T. (1999). Neuropsy- psychology, 4(1), 17-22.
chological patterns in right versus left fronto- Bornstein, R. A., Baker, G. B., & Douglass, A. B.
temporal dementia. Journal of the International (1987a). Short-tenn retest reliability of the Halstead-
Neuropsychological Society, 5(7), 616-622. Reitan Battery in a normal sample. Journal of Ner-
Boone, K. B., Swerdloff, R. S., Miller, B. L., vous and Mental Disease,l75(4), 229-232.
Geschwind, D. H., Razani, J., Lee, A., et al. Bornstein, R. A., Paniak, C., & O'Brien, W. (1987b).
(2001). Neuropsychological profiles of adults with Preliminary data on classification of nonnal and
Klinefelter syndrome. Journal ofthe International brain-damaged elderly subjects. Clinical Neuro-
Neuropsychological Society, 7(4), 446-456. psychologist, 1(4), 315-323.
Borak, J., Sliwinski, P., Pobiasz, M., Gorecka, D., & Borod, J. C., Goodglass, H., & Kaplan, E. (1980).
Zielinski, J. (1996). Psychological status ofCOPD Nonnative data on the Boston Diagnostic
patients before and after one year of long-tenn Aphasia Examination, Parietal Lobe Battery, and
oxygen therapy. Monaldi Archives of Chest the Boston Naming Test. Journal of Clinical
Diseases, 51(1), 7-11. Neuropsychology, 2(3), 209-215.
Boringa, J. B., Lazeron, R., Reuling, 1., Ader, H., Borod, J. C., Caron, H. S., & Koff, E. (1984). Left-
PCennings, L., Linderboom, J., et al. (2001). banders and right-banders compared on perfor-
The Brief Repeatable Battery of Neuropsycho- mance and preference measures of lateral
logical Tests: Nonnative values allow application dominance. British Journal of Psychology. 75(2),
in multiple sclerosis clinical practice. Multiple 177-186.
Sclerosis, 7, 263-267. Botwinick, J. (1981). Neuropsychology of aging. In
Borkowski, J., Benton, A., & Spreen, 0. (1967). S. Filskov & T. Boll (Eds.), Handbook of Clinical
Word fluency and brain damage. Neuropsycho- Neuropsychology. New York: Wiley.
logia, 5, 135-140. Bowden, S., & Bell, R. (1992). Relative usefulness
Bornstein, M. H. (1973). Color vision and color of the WMS and WMS-R: A comment on D'Elia
naming: A psychological hypothesis of cul- et al. (1989). Journal of Clinical and Experi-
tural difference. Psychological Bulletin, 80, mental Neuropsychology, 14(2), 340-346.
257-285. Bowden, S., Fowler, K. S., Bell, R. C., Whelan, G.,
Bomstein, R. A. (1985). Nonnative data on selected Clifford, C. C., Ritter, A. J., et al. (1998). The
neuropsychological measures from a nonclinical reliability and internal validity of the Wisconsin
542 REFERENCES
Card Sorting Test. Neuropsychological Rehabil- Brandt,J., & Benedict, R. H. B. (2001). Hopkins
itation, 8(3), 243-254. Verbal Learning Test-Revised. Lutz, FL: Psy-
Bowles, N. L., & Poon, L. W. (1985). Aging and chological Assessment Resources.
retrieval of words in semantic memory. Journal Brebion, G., Smith, M. J., Connan, J. M., &
of Gerontology, 40(1), 71-77. Amador, X. (1996). Reality monitoring failure in
Boyd, J. L. (1981). A validity study of the Hooper schizophrenia: The role of selective attention.
Visual Organization Test. Journal of Consulting Schizophrenia Research, 22, 173-180.
and Clinical Psychology, 49,15-19. Breteler, M. M., van Amerongen, N. M., van
Boyd, J. L. (1982a). Reply to Rathbun and Smith: Swieten, J. C., Claus, J. J., Grobbee, D. E., van
Who made the Hooper blooper? Journal of Gijn, J., et al. (1994). Cognitive correlates of
Consulting and Clinical Psychology, 50, 284-285. ventricular enlargement and cerebral white mat-
Boyd, J. L. (1982b). Reply to Woodward. Journal ter lesions on magnetic resonance imaging. The
of Consulting and Clinical Psychology, 50(2), Rotterdam Study. Stroke, 25,1109-1115.
289-290. Brislin, R. W. (1983). Cross-cultural research in
Boyle, G. J. (1975). Shortened Halstead Category psychology. Annual Review of Psychology, 34,
Test. Australian Psychologist, 10(1), 81-84. 363-400.
Boyle, G. J. (1986). Clinical neuropsychological as- Brittain,J. L.,laMarche,J. A., Reeder, K. P., Roth, D. L,
sessment: Abbreviating the Halstead Category & Boll (1991). Effects of age and IQ on Paced
Test of brain dysfunction. Journal of Clinical Auditory Serial Addition Task (PASAT) perfor-
Psychology, 42(4), 615-625. mance. Clinical Neuropsychologjst, 5(2), 163-175.
Boyle, G. J. (1988). What does the neuropsycho- Brooke, M. M., Questad, K. A., Patterson, D. R., &
logical category test measure? Archives of Clin- Valois, T. A. (1992). Driving evaluation after trau-
ical Neuropsychology, 3, 69-76. matic brain injury. American Journal of Physical
Boyle, G. J., Ward, J., & Steindl, S. R. (1994). Medicine Rehabilitation, 71(3),177-182.
Psychometric properties of Russell's short fonn Brooks, D. N. (1972). Memory and head injury.
of the Booklet Category Test. Perceptual and Journal of Neroous and Mental Disease, 155(5),
Motor Skills, 79(1, Pt 1), 128-130. 350--355.
Bradford, D. T. (1992).Interpretive Reasoning and Brooks, J., Fos, L., Greve, K., & Hammond, J. S.
the Halstead-Reitan Tests. Brandon, Vf: Clinical (1999). Assessment of executive function in pa-
Psychology. tients with mild traumatic brain injury. The
Brady, C. B., Spiro, A., III, McGlinchey-Berroth, R., Journal of Trauma, 461, 159-163.
Milberg, W., & Gaziano, J. M. (2001). Stroke risk Brown, G. G., Kindennann, S. S., Siegle, G. J.,
predicts verbal fluency decline in healthy older Granholm, E., Wong. E. C., & Buxton, R. B.
men: Evidence from the nonnative aging study. (1999 ). Brain activation and pupil response during
Journals of Gerontology: Series B: Psychological covert perfonnance of the Stroop Color Word
Sciences and Social Sciences, 568(6), P340-P346. task. Journal ofthe International Neuropsycholog-
Braff, D. L. (1989). Sensory input deficits and neg- ical Society, 5(4), 308--319.
ative symptoms in schizophrenic patients. Amer- Brown, J. (1958). Some tests of the decay theory of
ican Journal of Psychiatry, 146(8), 1006-1011. immediate memory. Quarterly Journal of Ex-
Braff, D. L., Heaton, R. K., Kuck, J., Cullum, M., perimental Psychology, 10, 12-21.
Maranville, J., Grant, 1., et al. (1991). The gen- Buchanan, R. W., Strauss, M. E., Kirkpatrick, B.,
eralized pattern of neuropsychological deficits in Holstein, C., Breier, A., & Carpenter, W. T.
outpatients with chronic schizophrenia with (1994). Neuropsychological impairments in def-
heterogeneous Wisconsin Card Sorting Test re- icit vs. nondeficit fonns of schizophrenia. Ar-
sults. Archives of General Psychiatry, 48(10), chives of General Psychiatry, 51, 804-811.
891-898. Burgess, P. W. (2003). Assessment of executive
Brandon, A. D., & Chavez, E. L. ( 1985). Order and function. In P. W. Halligan, U. Kischka, & J. C.
delay effects on neuropsychological test presen- Marshall (Eds.), Handbook of clinical neuro-
tation: The Halstead Category and Wisconsin psychology. New York: Oxford University Press.
Card Sorting Tests. Clinical Neuropsychology, Burton, C. L., Hultsch, D. F., Strauss, E., &
7(3), 152-153. Hunter, M. A. (2002). Intraindividual variability
Brandt, J. (1991). The Hopkins Verbal Learning in physical and emotional functioning: Compar-
Test: Development of a new memory test with ison of adults with traumatic brain injuries and
six equivalent fonns. Clinical Neuropsychologist, healthy adults. Clinical Neuropsychologist, 16(3),
5(2), 125-142. 264--279.
REFERENCES 543
Buschke, H. (1973). Selective reminding for anal- Cahn, D. A., Marcotte, A. C., Stern, R. A.,
ysis of memory and learning. Journal of Verbal Arruda, J. E., Akshoomoff, N. A., & Leshko, I. C.
Learning and Verbal Behavior, 12(5), 543--550. (1996). The Boston Qualitative Scoring System
Buschke, H. (1984). Cued recall in amnesia. Jour- for the Rey-Osterrieth Complex Figure: A study
nal of Clinical Neuropsychology, 6(4), 433-440. of children with attention deficit hyperactiv-
Buschke, H., & Fuld, P. (1974). Evaluating storage, ity disorder. Clinical Neuropsychologist, 10(4),
retention, and retrieval in disordered memory 397-406.
and learning. Neurology, 24(1l), 1019-1025. Cahn-Weiner, D. A., Boyle, P. A., & Malloy, P. F.
Bustini, M., Stratta, P., Daneluzzo, E., Pollice, R., (2002). Tests of executive function predict
Prosperini, P., & Rossi, A. {1999). Tower of instrumental activities of daily living in community-
Hanoi and WCST performance in schizophrenia: dwelling older individuals. Applied Neuropsychol-
Problem-solving capacity and clinical correlates. ogy, 9(3), 187-191.
Journal of Psychiatric Research, 33, 285-290. Caine, E. D. (1986). The neuropsychology of de-
Butler, M., Retzlaff, P. D., & Vanderploeg, R. pression: The pseudodementia syndrome. In
(1991). Neuropsychological test usage. Profes- I. Grant & K. M. Adams (Eds.), Neuropsychologi-
sional Psychology: Research and Practice, 22(6), cal assessment of neuropsychiatric disorders
510-512. (pp. 221-243). New York: Oxford University Press.
Butler, R. W., Jenkins, M. A., Sprock, J., & Calero, M. D., Arnedo, M. L., Navarro, E., Ruiz-
Braff, D. L. (1992). Wisconsin Card Sorting Pedrosa, M., & Camero, C. (2002). Usefulness of a
Test deficits in chronic paranoid shcizophrenia: 15-item version of the Boston Naming Test in
Evidence for a relatively discrete subgroup? neuropsychological assessment of low-educational
Schizophrenia Research, 7, 169-176. elders with dementia Journal of Gerontology.
Butler, R. W., Horsman, 1., Hill, J. M., & Tuma, R. Series B: Psychological Sciences and Social
(1993). The effects of frontal brain impairment Sciences, 57(2), P187-P191.
on fluency: Simple and complex paradigms. Calsyn, D. A., O'Leary, M. R., & Chaney, E. F.
Neuropsychology, 7(4), 519-529. (1980). Shortening the Category Test. Journal of
Butman, T. J. (2001). Designing an instrument Consulting and Clinical Psychology, 48(6),
of early diagnosis of dementia in primary care/ 788-789.
Hacia un protocolo clinico de deteccion precoz Caltagirone C. Carlesimo, A., Nocentini, U. &
de demencia en asistencia primaria. Acta Psi- Vican, S. (1989). Defective concept formation in
quiatrica y Psicologica de America Latina, 47{1), parkinsonians is independent from mental de-
79-87. terioration. Journal of Neurology, Neurosurgery,
Butters, N., Granholm, E., Salmon, D.P., & Grant, I. and Psychiatry, 52, 334-337.
(1987). Episodic and semantic memory: A com- Camara, W. J., Nathan, J. S., & Puente, A. E.
parison of amnesic and demented patients. (2000). Psychological test usage: Implications in
Journal of Clinical and Experimental Neuropsy- professional psychology. Professional Psycholo-
chology, 9, 479-497. gy: Research and Practice, 31(2), 141-154.
Cabeza, R., & Nyberg. L. (2002). Imaging cogni- Campo, P., & Morales, M. (2003). Reliability and
tion: An empirical review of 275 PET and fMRI normative data for the Benton Visual Form
studies. Journal of Cognitive Science, 12, 1-47. Discrimination Test. Clinical Nettropsychologist,
Caffarra, P., Vezzadini, G., Dieci, F., Zonato, F., & 17(2), 220-225.
Venneri, A. (2002). Rey-Osterrieth Complex Campo, P., & Morales, M. (2004). Normative data
Figure: Normative values in an Italian population and reliability for a Spanish version of the verbal
sample. Neurological Sciences, 22(6), 443-447. Selective Reminding Test. Archives of Clinical
Caffarra, P., Vezzadini, G., Dieci, F., Zonato, F., & Neuropsychology, 19(3), 421-235.
Venneri, A. (2004). Modified card sorting test: Campo, P., Morales, M., & Juan-Malpartida, M.
Normative data. Journal of Clinical and Experi- (2000). Development of two Spanish versions of
mental Neuropsychology, 26(2), 246-250. the verbal Selective Reminding Test. Journal of
Cahn, D. A., Salmon, D. P., Butters, N., Wie- Clinical and Experimental Neuropsychology, 22,
derholt, W. C., Corey-Bloom, J., Edelstein, S. L., 279-285.
et al. (1995). Detection of dementia of the Alz- Campo, P., Morales, M., & Martinez-Castillo, E.
heimer type in a population-based sample: Neu- (2003). Discrimination of normal from demented
ropsychological test performance. Journal of the elderly on a Spanish version of the Verbal Selec-
International Neuropsychological Society, 1(3), tive Reminding Test. Journal of Clinical and Ex-
252-260. perimental Neuropsychology, 25(7), 991-999.
544 REFERENCES
Canadian Study of Health and Aging Working Carter, S. L., Shore, D., Harnadek, M. C. S., &
Group (1994). The Canadian Study of Health Kubu, C. S. (1998). Normative data and inter-
and Aging: Study methods and prevalence of rater reliability of the Design Fluency Test.
dementia. Canadian Medical Association Jour- Clinical Neuropsychologist, 12(4), 531--534.
nal, 150, 899-913. , Casey, M. B., Winner, E., Hurwitz, 1., & DaSilva,
Caplan, B. (1985). Stimulus effects in unilateral D. (1991). Does processing style affect recall of
neglect? Cortex, 21(1), 69-80. the Rey-Osterrieth or Taylor Complex Figures?
Caplan, B., & Caffery, D. (1996). Visual form dis- Journal of Clinical and Experimental Neuropsy-
crimination as a multiple-choice visual memory chology, 13, ~06.
test: Illustrative data. Clinical Neuropsycholo- Catafau, A. M., Parellada, E., Lomena, F. J., Ber-
gist, 10(2), 152-158. nardo, M., Pavia, J., Ros, D., et al. (1994). Pre-
Caplan, B., & Schultheis, M. (1998). An interpre- frontal and temporal blood flow in schizophrenia:
tative table for the Visual Form Discrittination Resting and activation technetium-99m-HMPAO
Test. Perceptual and Motor Skills, B?f, 1203- SPECT patterns in young neuroleptics-naive
1207. patients with acute disease. Journal of Nuclear
Caplan, B., & Shechter, J. (1995). The, role of Medicine, 35,935--941.
nonstandard neuropsychological assess~ent in Cattell, J. (1886). The time it takes to see and name
rehabilitation: History, rationale, and exaqtples. In objects. Mind, 11, 63-65.
L. A. Cushman & M. J. Scherer (Ed$.), Psy- Cauthen, N. (1977). Extension of the Wechsler
chological assessment in medical refuzbilita- Memory Scale norms to the older age groups.
tion. Washington DC: American Psyciological Journal of Clinical Psychology, 33, 208-212.
Association. Cauthen, N. (1978a). Normative data for the Tac-
Carew, T. G., Lamar, M., Cloud, B.S., Gross~. M., & tual Performance Test. Journal of Clinical Psy-
Libon, D. J. (1997). Impairment in category chology, 34(2), 456-460.
fluency in ischemic vascular dementia. Neuro- Cauthen, N. (1978b). Verbal fluency: Normative
psychology, 11(3), 400-412. data. Journal of Clinical Psychology, 32(1),
Carmelli, D., Swan, G. E., Reed, T., Schellen- 126-129.
berg, G. D., & Christian, J. C. (1999). The effect Cavalli, M., De Renzi, E., Faglioni, P., & Vitale, A.
of apolipoprotein E epsilon4 in the relationships (1981). Impairment of right brain-damaged
of smoking and drinking to cognitive fPnction. patients on a linguistic cognitive task. Cortex, 17,
Neuroepidemiology, 18(3), 125-133. 546-556.
Carmelli, D., DeCarli, C., Swan, G. E., Kelly- Cerhan, J. H., Ivnik, R. J., Smith, G. E.,
Hayes, M., Wolf, P. A., Reed, T., et al. (2000). Tangalos, E. C., Petersen, R. C., & Boeve, B. F.
The joint effect of apolipoprotein E epsilon4 and (2002). Diagnostic utility of letter fluency, cate-
MRI findings on lower-extremity function and gory fluency, and fluency difference scores in
decline in cognitive function. Journals of Ger- Alzheimer's disease. Clinical Neuropsychologist,
ontology: Series A: Biological Sciences and 16(1), 35-42.
Medical Sciences, 55A(2), M103-M109. Cermak, L. S., & Butters, N. (1972). The role of
Carr, E. K., & Lincoln, N. B. (1988). Inter-rater interference and encoding in the short-term
reliability of the Rey figure copying test, British memory deficits of Korsakoff patients. Neu-
Journal of Clinical Psychology, 27, 267-268. ropsychologia, 10, 8~95.
Carstairs, J. R., & Shores, E. A. (2000). The Mac- Chan, A. S., & Poon, M. W. (1999). Performance of
quarie University Neuropsychological Normative 7- to 95-year-old individuals in a Chinese version
Study (MUNNS): Rationale and methqdology. of the category fluency test. Journal of the In-
Australian Psychologist, 35, 36-40. · ternational Neuropsychological Society, 5(6),
Carter, C. S., Mintun, M., & Cohen, U. J.D. (1995). 525-533.
Interference and facilitation effects duri~ selec- Chan, R. C. K. (2001). Base rate of post-concussion
tive attention: An H 20 150 PET study of Stroop symptoms among normal people and its neu-
task performance. Neuroimage, 2, 264-272. ropsychological correlates. Clinical Rehabilita-
Carter, C. S., Mintun, M., Nichols, T., & Co~n. J.D. tion, 15(3), 266-273.
(1997). Anterior cingulate gyrus dyst.mction Channon, S. (1996). Executive dysfunction in de-
and selective attention deficits in schizophrenia: pression: The Wisconsin Card Sorting Test.
[-sup-1-sup-SO]H-sub-20 PET study: during Journal of Affective Disorders, 39, 107-114.
single-trial Stroop task performance. Arerican Charter, R. A. (1994). Determining random
Journal of Psychiatry, 154(12), 167~16'f5. responding for the Category, Speech-Sounds
REFERENCES 545
Perception, and Seashore Rhythm tests. Journal Chen, E. Y. H., Kwok, C. L., Chen, R. L., &
of Clinical and Experimental Neuropsychology, Kwong, P. P. K. (2001). Insight changes in acute
16(5), 7~748. psychotic episodes: A prospective study of Hong
Charter, R. A. (1999). Sample size requirements for Kong Chinese patients. Journal of Neroous and
precise estimates of reliability, generalizability, Mental Disease, 189(1), 24-30.
and validity coefficients. Journal of Clinical and Chen, H., & Ho, C. (1986). Developmental study of
Experimental Neuropsychology, 21(4), 559-566. the reversed Stroop effect in Chinese-English
Charter, R. A. (2000a). Internal consistency reli- bilinguals. Journal of Generol Psychology, 113,
ability of the Tactual Performance Test trials. 121-125.
Perceptual and Motor Skills, 91(2), 460-462. Chen, P., Ratcliff, G., Belle, S. H., Cauley, J. A.,
Charter, R. A. (2000b). Item difficulty analysis of DeKosky, S. T., & Ganguli, M. (2000). Cognitive
the tactual performance test trials. Perceptual tests that best discriminate between pre-
and Motor Skills, 91(3, Pt 1), 903-909. symptomatic AD and those who remain non-
Charter, R. A. (200la). Coefficients alpha for the demented. Neurology, 55(12), 1847-1853.
Tactual Performance Test trials. Perceptual and Chen, Y. L. R., Chen, Y. H. E., & Lieh, M. F.
Motor Skills, 92(3, Pt 1), 893. (2000). Semantic verbal fluency deficit as a fa-
Charter, R. A. (200lb). Difference score reliability milial trait marker in schizophrenia. Psychiatry
for Tactual Performance Test trials. Perceptual Research, 95(2), 133-148.
and Motor Skills, 92(3, Pt 1), 941-942. Chen, Z. Q., Yu, J. H., & Cao, S. H. (1990). Ref-
Charter, R. A. (200lc). Tactual performance test erence values of indicators for WHO neurobe-
trials: Internal consistency reliability using the havioral core test battery. Chinese Medical
Gilmer-Feldt coefficient. Perceptual and Motor Journal, 103(1), 61-65.
Skills, 93(2), 363--366. Cherrier, M. M., Mendez, M. F., Dave, M., &
Charter, R. A., & Dutra, R. L. (200la). Item diffi- Perryman, K. M. (1999). Performance on the
culty of the Tactual Performance Test Location Rey-Osterrieth Complex Figure test in Alzheimer
score. Perceptual and Motor Skills, 93(3), disease and vascular dementia. Neuropsychiatry,
899-900. Neuropsychology, and Behaviorol Neurology,
Charter, R. A., & Dutra, R. L. (200lb). Tactual 12(2), 95-101.
Performance Test: Item analysis of the Memory Chervinsky, A. B., Mitrushina, M., & Satz, P.
and Location scores. Perceptual and Motor (1992). Comparison of four methods of scoring
Skills, 92(3, Pt 1), 899-902. the Rey-Osterrieth Complex Figure Drawing
Charter, R. A., Adkins, T. G., Alekoumbides, A., & Test on four age groups of normal elderly. Broin
Seacat, G. F. (1987). Reliability of the WAIS, Dysfunction, 5, 267-287.
WMS, and Reitan Batteries: Raw scores and Chia, S. E., Jeyaratnam, J., Ong, C. N., Ng, T. P., &
standardized scores corrected for age and edu- Lee, H. S. (1994). Impairment of color vision
cation. International Journal of Clinical Neuro- among workers exposed to low concentrations of
psychology, 9(1), 2lh12. styrene. American Journal of Industrial Medi-
Charter, R. A., Walden, D. K., & Hoffman, C. cine, 26(4), 481-488.
(1998). Interscorer reliabilities for memory and Chiaravalloti, N.D., Demaree, H., Gaudino, E. A., &
localization scores of the Tactual Performance De Luca, J. (2003). Can the repetition effect
Test. Clinical Neuropsychologist, 12(2), 245-247. maximize learning in multiple sclerosis? Clinical
Charter, R. A., Dutra, R. L., & Rapport, L. J. (2000). Rehabilitation, 17(1), 58-68.
Tactual performance test: Internal consistency Chiu, H. F. K., Chan, C. K. Y., Lam, L. C. W., Ng, K.,
reliability of the memory and location scores. Li, S., Wong, M., et al. (1997). The modi6ed
Perceptual and Motor Skills, 91(1), 143-146. Fuld Verbal Fluency Test: A validation study in
Chavez, E. L., Schwartz, M. M., & Brandon, A. Hong Kong. Journals of Gerontology. Series B:
(1982). Effects of sex of subjects and method of Psychological Sciences and Social Sciences, 528(5),
block presentation on the Tactual Performance P247-P250.
Test. Journal of Consulting and Clinical Psy- Chiulli, S. J., Haaland, K. Y., Ellis, H. C., &
chology, 50(4), 600-601. Rhodes, J. M. (1985, February). Recall and
Chelune, G., Bomstein, R.L., & Prifitera, A. (1989). recognition memory with a variant of the Rey
The Wechsler Memory Scale-Revised: Current Auditory-Verbal Learning Test in a clinically
status and applications. In J. Rosen, P. McRey- depressed populotion. Paper presented at the
nolds, & G. Chelune (Eds.), Advances in psy- 13th Annual Convention of the International
chological assessment. New York: Plenum Press. Neuropsychological Society, San Diego, CA.
546 REFERENCES
Chiulli, S. J., Yeo, R., Haaland, K., & Garry, P. Test in an adult sample. Journal of Clinical and
(1989). Complex figure copy and recaD in the Experimental Neuropsychology, 10(2), 175-184.
elderly. Journal of Clinical and Experimental Clark, C., Jacova, C., Klonoff, H., Kremer, B., et al.
Neuropsychology, 11, 95. (1997). Pathological association and dissociation
Chiulli, S. J., Haaland, K. Y., LaRue, A., & Garry, P. J. of functional systems in multiple sclerosis and
(1995). Impact of age on drawing the Rey- Huntington's disease. Journal of Clinical and
Osterrieth Figure. Clinical Neuropsychologist, Experimental Neuropsychology, 19(1), 63-76.
9(3), 219-224. Coelho, C. A. (2002). Story narratives of adults with
Choca, J.P., Laatsch, L., Wetzel, L., & Aw-esti, A. closed head injury and non-brain-injured ad-
(1997). The Halstead Category Test: A fifty ults: Inftuence of socioeconomic status, elicita-
year perspective. Neuropsychology Review, 7(2), tion task, and executive functioning. Journal of
61-75. Speech, lAnguage, and Hearing Research, 45(6),
Christensen, A.-L. (1974). Luria's neuropsjcholog- 1232-1248.
ical investigation. Copenhagen: Munksgaard. Coello, E., Ardila, A., & Rosselli, M. (1990). Is there
Christensen, H., Mackinnon, A. J., Korten, A. E., a cognitive marker in major depression? Inter-
Jorm, A. F., Henderson, A. S., Jacomb, P., et al. national Journal of Neuroscience, 50, 137-145.
(1999). An analysis of diversity in the cognitive Coen, R. F., Maguire, C., Swanwick, G. R., Kirby, M.,
performance of elderly community dwellers: In- Burke, T., Lawlor, B. A., et al. (1996). Letter and
dividual differences in change scores as a function category fluency in Alzheimer's disease: A
of age. Psychology and Aging. 14(3), 365-379. prognostic indicator of progression. Dementia,
Christensen, H., Mackinnon, A. J., Korteq, A., & 7(5), 246-250.
Jorm, A. F. (2001). The "common cause hy- Coffey, C. E., Ratcliff, G., Saxton,J. A., Bryan, R.N.,
pothesis" of cognitive aging: Evidence for not Fried, L. P., & Lucke, J. F. (2001). Cognitive
only a common factor but also specific !lSSocia- correlates of human brain aging: A quantitative
tions of age with vision and grip strength in a magnetic resonance imaging investigation. Jour-
cross-sectional analysis. Psychology and Aging. nal of Neuropsychiatry and Clinical Neurosci-
16(4), 588-599. ences, 13(4), 471-485.
Christensen, K. J., Kim, S. W., Dysken, M. W., & Cohen, M. J., & Stanczak, D. E. (2000). On the
Hoover, K. M. (1992). Neuropsychologi~al per- reliability, validity, and cognitive structure of the
formance in obsessive-compulsive disorder. Thurstone Word Fluency Test. Archives of
Biological Psychiatry, 31, 4-18. Clinical Neuropsychology, 15(3), 267-279.
Christensen, K. J., Riley, B. E., Heffernan, K. A., Cohen, R. A., Kaplan, R. F., Zuffante, P., Moser, D. J.,
Love, S. B., & Sta Maria, M. E. M. (2002). Jenkins, M. A., Salloway, S., et al. (1999). Alter-
Neuropsychological tests in the elderly: Methods ation of intention and self-initiated action associ-
and sample characteristics of a GRECC study. ated with bilateral anterior cingulotomy. Journal
Clinical Neuropsychologist, 16(1), 43--50. of Neuropsychiatry and Clinical Neurosciences,
Chronicle, E. P., & MacGregor, N. A. (1998). 11(4), 444-453.
Are P ASAT scores related to mathematical Cohen, R. A., Brumm, V., Zawacki, T. M., Paul, R.,
ability? NeuropsychologiCill Rehabilitation, 8(3), Sweet, L., & Rosenbaum, A. (2003). Impulsivity
27~282. and verbal deficits associated with domestic
Cicchetti, D. V. (1999). Sample size requirements violence. Journal of the International Neu-
for increasing the precision of reliability esti- ropsychological Society, 9(5), 760-770.
mates: Problems and proposed solutions. Journal Cohn, N. B., Dustman, R. E., & Bradford, D. C.
of Clinical and Experimental Neuropsydtology, (1984). Age-related decrements in Stroop Color
21(4), 567--570. Test performance. Journal of Clinical Psychol-
Cicerone, K. D. (1997). Clinical sensitivity of ogy, 40, 1244-1250.
four measures of attention to mild traumatic Collie, A., Shafiz-Antonacci, R., Maruff, P., Tyler,
brain injury. Clinical Neuropsychologist. 11{3), P., & Currie, J. (1999). Norms and the effects of
266--272. demographic variables on a neuropsychological
Cicerone, K. D., & Azulay, J. (2002). Diagnostic battery for use in healthy ageing Australian
utility of attention measures in post.,ncus- populations. Australian and New Zealand Jour-
sion syndrome. Clinical Neuropsychologist 16(3), nal of Psychiatry, 33{4), 568-575.
280--289. Collier, A. C., Marra, C., Coombs, R. W., Claypoole, K.,
Clark, C., & Klonoff, H. (1988). Reliability and con- Cohen, W., et al. (1992). Central nervous sys-
struct validity of the six-block Tactual Performance tem manifestations in human immunodeficiency
REFERENCES 547
virus infection without AIDS. Journal of recorder. Clinical Neuropsychologist, 16(3),

Acquired Immune IHJiclency Syndrome, 5(3), 407-412.
229-241. Cooper, H., & Hedges, L. V. (1994). The handbook
Colombo, F., & Assai, G. (1992). Adaptation fran- of research synthesis. New York: Russell Sage
caise du test de denomination de Boston. Ver- Foundation.
sions abregees. European Review of Applied Cooper,J.A., Sagar, H.J.,Jordan, N., Harvey, N. S. &
Psyclwlogy, 42(1), 67-73. Sullivan, E. V. (1991). Cognitive impairment in
Comalli, P. E., Wapner, S., & Werner, H. (1962). early untreated Parkinson's disease and it's
Interference effects of Stroop Color-Word Test relationship to motor disability. Brain, 114,
in childhood, adulthood, and aging. Journal of ~2122.
Genetic Psyclwlogy, 100, 47-53. Copersino, M. L., Serper, M., & Allen, M. H.
Coman, E., Moses, J. A., Jr., Kraemer, H. C., Fried- (2003). Rapid screening for cognitive impair-
man, L., Benton, A. L., & Yesavage, J. (1999). Ge- ment in the psychiatric emergency service: II. A
riatric performance on the Benton Visual Retention flexible test strategy. Psychiatric Seroices, 54(3),
Test: Demographic and diagnostic considerations. 314-316.
Clinical Neuropsyclwlogist, 13(1), 66-77. Corey, D. M., Hurley, M. M., & Foundas, A. L.
Coman, E., Moses, J. A., Jr., Kraemer, H. C., (2001). Right and left handedness defined: A
Friedman, L., Benton, A. L., & Yesavage, J. multivariate approach using hand preference
(2002). Interactive inHuences on BVRT perfor- and hand performance measures. Neuropsychi-
mance level: Geriatric considerations. Archives atry, Neuropsychology, and Behavioral Neurol-
of Clinical Neuropsyclwlogy, 17(6), 595--610. ogy, 14(3), 144-152.
Compton, D. M., Bachman, L. D., & Logan, J. A. Corrigan, J. D., & Hinkeldey, N. S. (1987). Rela-
(1997). Aging and intellectual ability in young, tionships between parts A and B of the Trail
middle-aged, and older educated adults: Pre- Making Test. Journal of Clinical Psyclwlogy,
liminary results from a sample of college faculty. 43(4), 402-409.
Psychologiad Reports, 81(1), 79-90. Corrigan, J.D., Agresti, A. A., & Hinkeldey, N. S.
Compton, D. M., Bachman, L. D., Brand, D., & (1987). Psychometric characteristics of the Cat-
Avet, T. L. (2000). Age-associated changes in cog- egory Test: Replication and extension. Journal of
nitive function in highly educated adults: Emerg- Clinical Psyclwlogy, 43(3), 368--376.
ing myths and realities. International Journal of Corwin, J., & Bylsma, F. W. (1993). Psychological
Geriatric Psychiatry, 15(1), 75-85. examination of traumatic encephalopathy. The
Concha, M., Seines, 0. A., McArthur, J. C., & Complex Figure Copy Test. Clinical Neuropsy-
Nance-Sproson, T. (1995). Normative data for a chologist, 7, 4-21.
brief neuropsychological test battery in a cohort Craddick, R. A., & Stem, M. R. (1963). Practice
of injecting drug users. International Journal of effects on the Trail Making Test. Perceptual and
the Addictions, 30(7), 823--841. Motor Skills, 17(3), 651-653.
Conners, C. K., Epstein, J., Stem, R. A., March, J., Craik, F. 1., Byrd, M., & Swanson, J. M. (1987).
Sparrow, E., & Javorsky, D. J. (2003). Subtyping Patterns of memory loss in three eldery samples.
attention-deficit/hyperactivity disorder (ADHD): Psyclwlogy and Aging, 2, 79-86.
Use of the ROCF. In J. A. Knight (Ed.), The Craik, F. I., Morris, L. W., Morris, R. G., & Loe-
handbook of Rey-Osterrieth Complex Figure wen, E. R. (1990). Relations between source
usage: Clinical and research applications. Lutz, amnesia and frontal lobe functioning in older
FL: Psychological Assessment Resources. adults. Psyclwlogy and Aging, 5(1), 14S-151.
Connor, A., Franzen, M., & Sharp, B. (1988). Ef- Crawford, J. R. (1992). Current and premorbid in-
fects of practice and differential instructions on telligence measures in neuropsychological assess-
Stroop performance. International Journal of ment. In J. R. Crawford & D. M. Parker (Eds.),
Clinical Neuropsyclwlogy, 10, 1-4. A lwndbook of neuropsyclwlogical assessment
Connor, P. D., Sampson, P. D., Bookstein, F. L., (pp. 21-49). Hillsdale, NJ: Lawrence Erlbaum.
Barr, H. M., & Streissguth, A. P. (2001). Direct Crawford, J. R., & Garthwaite, P. H. (2002). In-
and indirect effects of prenatal alcohol damage vestigation of the single case in neuropsychology:
on executive function. Developmental Neuro- Confidence limits on the abnormality of test
psyclwlogy, 18(3), 331-354. scores and test score differences. Neuropsycho-
Constantinou, M., Ashendorf, L., & McCaffrey, logia, 40(8), 1196-1208.
R. J. (2002). When the third party observer Crawford, J. R., & Howell, D. C. (1998a). Com-
of a neuropsychological evaluation is an audio- paring an individual's test score against norms
548 REFERENCES
derived from small samples. Clinical Neuropsy- frontal lobe dysfunction in neuropsychiatry
chologist, 12(4), 482-486. samples. International Journal of Neurosciences,
Crawford, J. R., & Howell, D. C. (1998b). Re- 30, 241-248.
gression equations in clinical neuropsychology: Crockett, D. J., Hadjistavropoulos, T., & Hurwitz, T.
An evaluation of statistical methods for com- (1992). Primacy and recency effects in the
paring predicted and obtained scores. Journal of assessment of memory using the Rey Auditory
Clinical and Experimental Neuropsychology, Verbal Learning Test. Archives of Clinical Neu-
20(5), 755-762. ropsychology, 7, 97-107.
Crawford, J. R., Stewart, L. E., & Moore, J. W. Crookes, T. G., & McDonald, K. G. (1972). Benton's
(1989). Demonstration of savings on the AVLT Visual Retention Test in the differentiation of
and development of a parallel form. ]011mal of depression and early dementia. British Journal of
Clinical and Experimental Neuropsy~hology, Social and Clinical Psychology,ll(l), 66-69.
11(6), 975-981. Crossen, J. R., & Wiens, A. N. (1994). Comparison
Crawford, J. R., Howell, D. C., & Garthwaite, P. H. of Auditory-Verbal Learning Test (AVLT) and
(1998a). Payne and Jones revisited: Estimating California Verbal Learning Test (CVLT) in a
the abnormality of test score differences using a sample of normal subjects. Journal of Clinical
modified paired samples t test. journal of Clin- and Experimental Neuropsychology. 16(2),
ical and Experimental Neuropsychology. 20(6), 190-194.
898-905. Crossley, M., D'Arcy, C., & Rawson, N. S. B. (1997).
Crawford, J. R., Obonsawin, M. C., & Allan, K. M. Letter and category fluency in community-
(1998b). PASAT and components of WAIS-R dwelling Canadian seniors: A comparison of
performance: Convergent and discriminant val- normal participants to those with dementia of the
idity. Neuropsychological Rehabilitatioa, 8(3), Alzheimerorvasculartype.JoumalofClinicaland
255-272. Experimental Neuropsychology, 19(1), 52-62.
Crawford, J. R., Garthwaite, P. H., Howell, D. C., & Crosson, B., Hughes, C., Roth, D., & Monkowski, P.
Venneri, A. (2003). Intra-individual mea$ures of (1984a). Review of Russell's (1975) norms for
association in neuropsychology: Inferential the Logical Memory and Visual Reproduction
methods for comparing a single case with a subtests of the Wechsler Memory Scale. Journal
control or normative sample. Journal of the of Consulting and Clinical Psychology, 52(4),
International Neuropsychological Se>cietr, 9(7), 635--641.
989-1000. ' Crosson, B., Hughes, C., Roth, D., & Mankowski, P.
Crews, W. D., Jr., Harrison, D. W., & Rhodes, R. D. (1984b). Use of errors and correct ideas in
(1999). Neuropsychological test performances of scoring Wechsler Memory Scale stories. Paper
young depressed outpatient women: An· exami- presented at the meeting of the International
nation of executive functions. Archives of Clini- Neuropsychological Society, Houston, TX.
cal Neuropsychology, 14(6), 517-529. Crosson, B., Benefield, H., Cato, M.A., Sadek, J. R.,
Crews, W. D., Jefferson, A. L., Bolduc, T., Moore, A. B., Wierenga, C. E., et al. (2003). Left
Elliott, J. B., Ferro, N. M., Broshek, D. K., et al. and right basal ganglia and frontal activity during
(2001). Neuropsychological dysfunction in pa- language generation: Contributions to lexical,
tients suffering from end-stage chronic obstruc- semantic, and phonological processes. Journal of
tive pulmonary disease. Archives of Clinical the 1ntemational Neuropsychological Society,
Neuropsychology, 16(1), 643-652. 9(7), 1061-1077.
Crews, W. D., Jr., Jefferson, A. L., Broshek. D. K., Crovitz, H. F., & Zener, K. (1962). A group-test for
Rhodes, R. D., Williamson, J., Brazil, A. M., et al. assessing hand-and-eye dominance. American
(2003). Neuropsychological dysfunction. in pa- Journal of Psychology, 75, 271-276.
tients with end-stage pulmonary disease: Lung Crowe, S. F. (1998a). Decrease in performance on
transplant evaluation. Archives of Clinical Neu- the verbal fluency test as a function of time:
ropsychology, 18(4), 353-362. Evaluation in a young healthy sample. Journal of
Crockett, D., Clark, C. M., Browning, J., & Clinical and Experimental Neuropsychology,
MacDonald, J. (1983). An application of the 20(3), 391-401.
background interference procedure to the Ben- Crowe, S. F. (1998b). 11te differential contribution
ton Visual Retention Test. Journal of Clinical of mental tracking, cognitive flexibility, visual
Neuropsychology, 5(2), 181-185. search, and motor speed to performance on parts
Crockett, D. J., Blisker, D., Hurwitz, T., & Kozak, J. A and B of the Trail Making Test. Journal of
(1986). Clinical utility of three measures of Clinical Psychology, 54(5), 585-591.
REFERENCES 549
Cruice, M. N., Worrall, L. E., & Hickson, L. M. H. Davies, K., Bell, B. D., Bush, A. J., Hermann, B. P.,
(2000). Boston Naming Test results for healthy Dohan, F. C., Jr., & Jaap, A. S. (1998). Naming
olderAustralians:Alongitudinalandcross-sectional decline after left anterior temporal lobectomy
study. Aphasiology, 14(2), 143-155. correlates with pathological status of resected
Cullum, C. M., Steinman, D. R., & Bigler, E. D. hippocampus. Epilepsia, 39, 407-419.
(1984). Relationship between fluid and crystal- Davis, K. L., Price, C. C., Kaplan, E., & Libon, D.
lized cognitive functions using Category Test and J. (2002). Error analysis of the nine-word Cali-
W AIS scores. International Journal of Clinical fornia Verbal Learning Test (CVLT-9) among
Neuropsychology, 6(3), 172-174. older adults with and without dementia. Clinical
Cullum, C. M., Butters, N., Truster, A., & Salmon, D. Neuropsychologist, 16(1), 81-89.
(1990). Normal aging and forgetting rates on the Davis, R. D., Adams, R. E., Gates, D. 0., &
Wechsler Memory Scale-Revised. Archives of Cheramie, G. M. (1989). Screening for learning
Clinical Neuropsychology, 5, 23--30. disabilities: A neuropsychological approach. Jour-
Curtiss, G., Vanderploeg, R. D., Spencer, J., & nal of Clinical Psychology, 45(3), 423-429.
Salazar, A. M. (2001). Patterns of verbal learning Dawes, R. M., Faust, D., & Meehl, P. E. (1993).
and memory in traumatic brain injury. Journal of Statistical prediction versus clinical prediction:
the International Neuropsychological Society, lmprovingwhatworks. loG. L. Keren&C. Lewis
7(5), 574-585. (Eds.), A handbook for data analysis in the be-
Dahan, C., Amado, 1., Bayle, F., Gut, A., Willard, D., havioral sciences: Methodological issues (pp. 351-
Bourdel, M.-C., et al. (2002). Correlation be- 367). Hillsdale, NJ: Lawrence Erlbaum.
tween clinical syndromes and neuropsychological Dawes, R. M., Faust, D., & Meehl, P. E. (2002).
tasks in unmedicated patients with recent onset Clinical versus actuarial judgment. In T. Gilo-
schizophrenia. Psychiatry Research, 113(1-2), vicb, D. Griffin, et al. (Eds.), Heuristics and
83-92. biases: The psychology of intuitive judgment (pp.
Daigneault, S., Braun, C. M. J., & Whitaker, H. A. 716-729). New York: Cambridge University
(1992). Early effects of normal aging on per- Press.
severative and non-perseverative prefrontal Dawson, L. K., & Grant, I. (2000). Alcoholics'
measures. Developmental Neuropsychology, 8, initial organizational and problem-solving skills
99-114. predict learning and memory performance on
Dalrymple-Alford, J. C., Kalders, A. S., Jones, R. D., & the Rey-Osterrieth Complex Figure. Journal of
Watson, R. W. (1994). A central executive deficit the International Neuropsychological Society,
in patients with Parkinson's disease. Journal of 6(1), 12-19.
Neurology, Neurosurgery, and Psychilltry, 57(3), Dealberto, M.-J., Pajot, N., Courbon, D., & Al-
36<h'367. perovitch, A. (1996). Breathing disorders during
Dartigues, J. F., Gagnon, M., Mazaux, J. M., sleep and cognitive performance in an older
Barberger-Gateau, P., Commenges, D., Le- community sample: The EVA study. Journal
tenneur, L., et al. (1992). Occupation during life of the American Geriatrics Society, 44(11),
and memory performance in nondemented 1287-1294.
French elderly community residents. Neurology, Deary, I. J., Ebmeier, K. P., MacLeod, K. M.,
42(9), 1697-1701. Dougall, N., Hepburn, D. A., Frier, B. M., et al.
Dartigues, J. F., Commenges, D., Letenneur, D., (1994). PASAT performance and the pattern of
Barberfer-Gateau, P., Gillem, V., Fabrigoule, C., uptake of -super(99m)Tc-exametazime in brain
et al. (1997). Cognitive predictors of dementia in estimated with single photon emission tomogra-
elderly community residents. Neuroepidemiol- phy. Biological Psychology, 38(1), 1-18.
ogy, 16(1), 29-39. Deary, I. J., Langan, S. J. Hepburn, D. A., &
Datto,C.J., Levy,S., Miller, D. S.,&Katz, I. R. (2001). Frier, B. M. (1991). Which abilities does the
Impact of maintenance ECTon concentration and PASAT test? Personality and Individual Differ-
memory. Journal ofECT,17(3), 170-174. ences, 12, 983-987.
Davis, H. P., Cohen, A., Gandy, M., Colombo, P., DeckeL A. W. (1999). Tests of executive functioning
Van Dusseldorp, G., Simplke, N., et al. (1990). predict scores on the MacAndrew Alcoholism
Lexical priming deficits as a function of age. Scale. Progress in Neuro-Psychupharmacology
Behavioral Neuroscience, 104(2), 288-297. and Biological Psychiatry, 23(2), 209-223.
Davies, A. D. (1968). The influence of age on Trail Deckersbacb, T., Otto, M. W., Savage, C. R., Baer, L., &
Making Test performance. Journal of Clinical Jenike, M. A. (2000a). The relationship be-
Psychology, 24, 96-98. tween semantic organization and memory in
550 REFERENCES
obsessive-compulsive disorder. Psychotherapy Delis, D. C., Kaplan, E., & Kramer, J. (2001).
and Psychosomatics, 69(2), 101-107. Delis-Kilplan Executive Function System. San
Deckersbach, T., Savage, C. R., Henin, A., Mataix- Antonio, TX: Psychological Corporation.
Cols, D., Otto, M. W., Wilhelm, S., et al. (2000b). Dellatolas, G., Braga, L., Souza, L., Filho, G. N.,
Reliability and validity of a scoring system for Queiroz, E., & Deloche, G. (2003). Cognitive
measuring organizational approach in the Com- consequences of early phase of literacy. Journal
plex Figure Test. Journal of Clinical and Experi- of the 1ntemational Neuropsychological Society,
mental Neuropsychology, 22(5), 64()....64$. 9(5), 771-782.
DecouHe, P., Holmgreen, P., Calle, E., & ~eeks, M. Deloche, G., Hannequin, D., Dordain, M., Perrier, D.,
(1991). Nonresponse and intensityoffollow-up in et al. (1996). Picture confrontation oral naming:
an epidemiologic study of Vietnam-era veterans. Performance differences between aphasics and
American Journal of Epidemiology, 133,,83-93. normals. Brain and Language, 53(1), 105-120.
D'Eiia, L. F., Satz, P., & Schretlen, D. (1989). DeLuca, J., Gaudino, E., Diamond, B., Christo-
Wechsler Memory Scale: A critical appraisal of doulou, C., & Engel, R. (1998). Acquisition and
the normative studies. Journal of Clirdcal and storage deficits in multiple sclerosis. Journal of
Experimental Neuropsychology, 11(4), 551-568. Clinical and Experimental Neuropsychology,
D'Elia, L., Satz, P., Uchiyama, C., & White, T. 20(3), 376-390.
(1999). Color Trails Test, Professional manual. Demakis, G. J. (1999). Serial malingering on verbal
Odessa, FL: Psychological Assessment Re- and nonverbal fluency and memory measures:
sources. An analog investigation. Archives of Clinical
DeFilippis, N. A., McCampbell, E., & Rogers, P. Neuropsychology, 14(4), 401-410.
(1979). Development of a booklet forro of the Demakis, G. J. (2003). A meta-analytic review of
Category Test: Normative and validity run. Jour- the sensitivity of the Wisconsin Card Sorting
nal of Clinical Neuropsychology, 1(4), 339-342. Test to frontal and lateralized frontal brain
Degenszajn, J., Caramelli, P., Caixeta, L., & damage. Neuropsychology, 17(2), 255-264.
Nitrini, R. (2001). Encoding process in delayed Demakis, G. J., & Harrison, D. W. (1997). Rela-
recall impairment and rate of forgetting in Alz- tionships between verbal and nonverbal fluency
heimer's disease. Arquivos de Neuro-Psifuiatria, measures: Implications for assessment of execu-
59(2,A), 171-174. tive functioning. Psychological &ports, 81(2),
Dehaene, S., & Changeux, J. P. (1991). Ute Wis- 443-448.
consin Card Sorting Test: Theoretical analysis Demakis, G. J., Mercury, M. G., Sweet, J. J.,
and modeling in a neuronal network. Cerebral Rezak, M., Eller, T., & Vergenz, S. (2003).
Cortex, 1, 62-79. Qualitative analysis of verbal fluency before and
de Jager, C. A., Hogervorst, E., Combrinck, M., & after unilateral pallidotomy. Clinical Neuropsy-
Budge, M. M. (2003). Sensitivity and specificity chologist, 17(3), 322--330.
of neuropsychological tests for mild cognitive Demaree, H., Gaudino, E., & DeLuca, J. (2003).
impairment, vascular cognitive impairment and The relationship between depressive symptoms
Alzheimer's disease. Psychological Medicine, and cognitive dysfunction in multiple sclerosis.
33(6), 1039-1050. Cognitive Neuropsychiatry, 8(3), 161-171.
Delaney, R. C., Prevey, M. L., Cramer, J., & Demery, J. A., Pedraza, 0., & Hanlon, R. E. (2002).
Mattson, R. H. (1992). Test-retest comparability Differential profiles of verbal learning in traumatic
and control subject data for the Rey Auditory- brain injury. Journal of Clinical and Experimental
Verbal Learning Test and Rey-Osterrietbl Taylor Neuropsychalogy, 24(6), 818-827.
complex figures. Archives of Clinical Neuropsy- Demick, J., & Harkins, D. (1997). Role of cognitive
chology, 7(6), 523-528. style in the driving skills of young, middle-aged,
Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. A. and older adults (American Association of Re-
(1987). California Verbal Learning Test Manual. tired Persons Andrus Foundation Final Grant
San Antonio, TX: Harcourt Brace Jovanovich. Report). Washington, DC: AARP.
Delis, D. C., Kramer, J. H., Kaplan, E., &: Ober, Demick, J., & Wapner, S. (1985, August). Age dif-
B. A. (1994). California Verbal Learning Test- ferences in processes underlying sequential ac-
Children's Version. San Antonio, TX: Psycho- tivity (Stroop Color-Word Test). Paper presented
logical Corporation. at the Eastern Psychological Association annual
Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. A. meeting, Los Angeles, CA.
(2000). California Verbal Learning Test (~d ed.). Demick, J., Salas-Passeri, J., & Wapner, S. (1986,
San Antonio, TX: Psychological Corporation. August). Age differences among preschoalers in
REFERENCES 551
processes underlying sequential activity. Paper (PASAT): Norms for age, education, and eth-
presented at the Eastern Psychological Associa- nicity. Assessment, 5(4), 375-387.
tion annual meeting, New York, NY. Diehr, M. C., Chemer, M., Wolfson, T. J.,
Denman, S. (1984). Denman Neuropsychology Miller, S. W., Grant, 1., Heaton, R. K., et al.
Memory Scale. Charleston, SC: Author. (2003). The 50 and 100-item short forms of the
Desmond, D. W., Glenwick, D. S., Stem, Y., & Paced Auditory Serial Addition Task (PASAT):
Tatemichi, T. K. (1994). Sex differences in the Demographically corrected norms and compar-
representation of visuospatial functions in the isons with the full PASAT in normal and clinical
human brain. Rehabilitation Psychology, 39(1), samples. Journal of Clinical and Experimental
3-14. Neuropsychology, 25(4), 571-585.
D'Esposito, M., Onishi, K., Thompson, H., Dikmen, S. S., Machamer, J. E., Winn, H. R., &
Robinson, K., Armstrong, C., & Grossman, M. Temkin, N. R. (1995). Neuropsychological out-
(1996). Working memory impairments in multi- come at 1-year post head injury. Neuropsychol-
ple sclerosis: Evidence from a dual-task para- ogy. 9(1), 80-90.
digm. Neuropsychology, 10(1), 51-56. Dikmen, S. S., Heaton, R. K., Grant, 1., &
DesRosiers, G., & Ivison, D. (1986). Paired asso- Temkin, N. R. (1999). Test-retest reliability and
ciate learning: Normative data for differences practice effects of Expanded Halstead-Reitan
between high and low associate word pairs. Neuropsychological Test Battery. Journal of the
Journal of Clinical and Experimental Neuropsy- International Neuropsychological Society, 5(4),
chology, 8(6), 637-642. 346--356.
DesRosiers, G., & Kavanagh, D. (1987). Cognitive Diniz, L. F. M., da Cruz, M.d. F., Torres, V. d. M., &
assessment in closed head injury: Stability, va- Cosenza, R. M. (2000). 0 teste de aprendizagem
lidity and parallel forms for two neuropsycholog- auditive-verbal de Rey: Normas para uma po-
ical measures of recovery. International Journal pulacao brasileira. The Rey Auditory-Verbal
of Clinical Neuropsychology, 9(4), 162-173. Learning Test: Norms for a Brazilian sample.
de Zubicaray, G., & Ashton, R. (1996). Nelson's &vista Brasileira de Neurologia, 36(3), 79-83.
(1976) Modified Card Sorting Test: A review. Doan, Q. T., & Swerdlow, N. R. (1999). Prelimi-
Clinical Neuropsychologist, 10(3), 245-254. nary findings with a new Vietnamese Stroop test.
Diamond, B. J., & DeLuca, J. (1996). Rey-Osterrieth Perceptual and Motor Skills, 89(1), 173-182.
Complex Figure Test performance following an- Dodrill, C. B. (1978a). A neuropsychological bat-
terior communicating artery aneurysm. Archives tery for epilepsy. Epilepsia, 19, 611--623.
of Clinical Neuropsychology, 11(1), 21-28. Dodrill, C. B. (1978b). The Hand Dynamometer
Diamond, B. J., DeLuca, J., Kim, H., & Kelley, S. M. as a neuropsychological measure. Journal of
(1997). The question of disproportionate impair- Consulting and Clinical Psychology, 46(6),
ments in visual and auditory information process- 1432-1435.
ing in multiple sclerosis. Journal of Clinical and Dodrill, C. B. (1979). Sex differences on the
Experimental Neuropsychology, 19(1), 34-42. Halstead-Reitan Neuropsychological Battery and
Diamond, R., Krengel, M., White, R. F., & Ja- on other neuropsychological measures. Journal
vorsky, D. J. (2003). The ROCF in assessment of of Clinical Psychology, 35(2), 236-241.
individuals exposed to toxicants. In J. A. Knight Dodrill, C. B. (1987). What's normal: Presidential
(Ed.), The handbook of Rey-Osterrieth Complex address. Paper presented at the first annual
Figure usage: Clinical and research applications. meeting of the Pacific Northwest Neurological
Lutz, FL: Psychological Assessment Resources. Association, Seattle, WA.
DiCarlo, M.A., Gfeller, J.D., & Oliveri, M. V. (2000). Dodrill, C. B., & Troupin, A. S. (1975). Effects of
Effects of coaching on detecting feigned cognitive repeated administrations of a comprehensive
impairment with the Category Test. Archives of neuropsychological battery among chronic epi-
Clinical Neuropsychology, 15(5), 399-413. leptics. Journal of Nervous and Mental Disease,
Dick, F., Semple, S., Soutar, A., Osborne, A., 161(3), 185-190.
Cherrie, J. W., & Seaton, A. (2004). Is colour Danders, J. (1999a). Cluster subtypes in the stan-
vision impairment associated with cognitive dardization sample of the California Verbal
impairment in solvent exposed workers? Occu- Learning Test-Children's Version. Developmen-
pational and Environmental Medicine, 61(1), tal Neuropsychology, 16(2), 163-175.
76-78. Donders, J. (1999b). Specificity of a malingering
Diehr, M. C., Heaton, R. K., Miller, W., & Grant, I. formula for the Wisconsin Card Sorting Test.
(1998). The Paced Auditory Serial Addition Task Journal of Forensic Neuropsychology, 1, 35-42.
552 REFERENCES
Donders, J. (2001). A suiVey of report writing by Dugbartey, A. T., Townes, B. D., & Mahurin, R. K.
neuropsychologists. II: Test data, report format, (2000). Equivalence of the Color Trails Test
and document length. Clinical Neuropsycholo- and Trail Making Test in nonnative English-
gist, I5(2), 150-161. speakers. Archives of Clinical Neuropsychology,
Donders, J., & Kirsch, N. (1991). Nature and im- I5(5), 425-431.
plications of selective impairment on the Booklet Duley, J. F., Wilkins, J. W., Hamby, S. L., Hopkins,
Category Test and the Wisconsin Card Sorting D. G., Burwell, R. D., & Barry, N. S. (1993).
Test. Clinical Neuropsychologist, 5, 78-82. Explicit scoring criteria for the Rey-Osterrieth
Downey, J., Elkin, E. J., Ehrhardt, A. A., Meyer- and Taylor complex figures. Clinical Neuropsy-
Bahlburg, H. F., Bell, J. J., & Morishima, A. chologist, 7(1), 29-38.
(1991). Cognitive ability and everyday function- Dunn, E. J., Margolis, R. B., & Taylor, J. M. (1985).
ing in women with Turner syndrome. Journal of Short forms of the Category Test: Applications
Learning Disabilities, 24(1), 32-39. for geriatric patients. International Journal of
Drane, D. L., Loring, D. W., Lee, G. P., & Clinical Neuropsychology, 7(1), 29-31.
Meador, K. J. (1998). Trial-length sensitivity of Dwyer, C. A. (1996). Cut scores and testing: Sta-
the Verbal Selective Reminding Test to later- tistics, judgment, truth, and error. Psychological
alized temporal lobe impairment. Clinical Neu- Assessment, 8(4), 360-362.
ropsychologist, I2(1), 68-73. Dye, 0. A. (1979). Effects of practice on Trail
Drane, D. L., Yuspeh, R. L., Huthwaite, J. S., & Making Test performance. Perceptual and Motor
Klingler, L. K. (2002). Demographic character- Skills. 48(1), 296.
istics and normative obseiVations for derived Dyer, F. N. (1973). The Stroop phenomenon and
Trail Making Test indices. Neuropsychiatry, Neu- its use in the study of perceptual, cognitive, and
ropsychology, and Behavioral Neurology, I5(1), response processes. Memory and Cognition, I,
39-43. 106-120.
Drebing, C. E., Van Gorp, W. G., Stuck, A. E., Dywan, J., Segalowitz, S. J., & Unsal, A. (1992).
Mitrushina, M., & Beck, J. (1994). Early detec- Speed of information processing, health, and
tion of cognitive decline in higher cognitively cognitive performance in older adults. Develop-
functioning older adults: Sensitivity and speci- mental Neuropsychology, 8(4), 473-490.
ficity of a neuropsychological screening battery. Dywan, J., Segalowitz, S. J., Henderson, D., & Ja-
Neuropsychology, 8(1), 31-38. coby, L. L. (1993). Memory for source after
Drewe, E. A. (1974). The effect of type and area traumatic brain injury. Brain and Cognition,
of brain lesion on Wisconsin Card Sorting Test 2I(l), 20-43.
performance. Cortex, IO, 159-170. Echternacht, R. (1981). Neuropsychological assess-
Duara, R., Grady, C., Haxby, J., Ingvar, D., So- ment of motor functioning. The Finger Tapping
koloff, L., Margolin, R. A., et al. (1984). Human Test: Data on an adult female inpatient psychiatric
brain glucose utilization and cognitive runction population. Clinical Neuropsychology, 3(2), 8-9.
in relation to age. Annals of Neurology, I6(6), Eckardt, M. J., & Matarazzo, J. D. (1981). Test-
703-713. retest reliability of the Halstead impairment in-
Duchnick, J. J., Vanderploeg, R. D., & Curtiss, G. dex in hospitalized alcoholic and nonalcoholic
(2002). Identifying retrieval problems using males with mild to moderate neuropsychological
the California Verbal Learning Test. Journal impairment. Journal of Clinical Neuropsychol-
of Clinical and Experimental Neuropsychology, ogy, 3(3), 257-269.
24(6), 840--852. Eckman, P. S., & Shean, G. D. (2000). Impairment in
Duff, K., WesteiVelt, H. J., McCaffrey, R. J., & test performance and symptom dimensions of
Haase, R. F. (2001). Practice effects, test-retest schizophrenia. Jotimal of Psychiatric Research,
stability, and dual baseline assessments with the 34(2), 147-153.
California Verbal Learning Test in an HIV sam- Egan, V. (1988). PASAT: ObseiVed correlations
ple. Archives of Clinical Neuropsychology,I6(5), with IQ. Personality and Individual Differences,
461-476. 9(1), 179-180.
Duff, K., Pattern, D., Schoenberg, M. R., Mold, J., Egger M., Smith, G. D., & Sterne, A. C. (2001). Uses
Scott, J. G., & Adams, R. L. (2003). Age- and abuses of meta-analysis. Clinical Medicine,
and education-corrected independent normative I(6), 478-484.
data for the RBANS in a community dwelling Egger, M., Ebrahim, S., & Smith, G. D. (2002).
elderly sample. Clinical Neuropsychologist, Where now for meta-analysis? International Jour-
17(3), 351-366. nal of Epidemiology, 3I, 1-5.
REFERENCES 553
Elfgren, C. 1., & Risberg, J. (1998). Lateralized Elwood, R. W. (1995). The California Verbal
frontal blood flow increases during fluency taslcs: Learning Test: Psychometric characteristics and
Influence of cognitive strategy. Neuropsycholo- clinical application. Neuropsychology Review,
gia. 36(6), 505--512. 5(3), 173-201.
Elfgren, C. 1., Ryding, E., & Passant, U. (1996). Embretson, S. E. (1996). The new rules of measure-
Performance on neuropsychological tests related ment. Psychalogical Assessment, 8(4), 341-349.
to single photon emission computerised tomog- Engelsmann, F., Katz, J., Ghadirian, A. M., &
raphy findings in frontotemporal dementia. Brit- Schachter, D. (1988). Lithium and memory:
ish journal of Psychiatry, 169, 416-422. A long-term follow-up study. Journal of Clinical
Elias, M. F., Podraza, A. M., Pierce, T. W., & Psychuphamwcology, 8(3), 207-212.
Robbins, M. A. (1990). Determining neu- Epker, M. 0., Lacritz, L. H., & Cullum, C. M.
ropsychological cut scores for older, healthy (1999). Comparative analysis of qualitative ver-
adults. Experimental Aging Research, 16(4), bal fluency performance in normal elderly and
209-220. demented populations. Journal of Clinical and
Elias, M. F., Robbins, M. A., Walter, L. J., & Experimental Neuropsychology, 21(4), 425-434.
Schultz, N. R. (1993). The influence of gender Epperson, R. C., & Cripe, L. (1985). Relationship
and age on Halstead-Reitan Neuropsycholog- of PASAT performance and IQ. Unpublished
ical Test performance. Journals of Gerontology, manuscript.
48(6), P278-P281. Ernst, J. (1987). Neuropsychological problem-
Ellingsen, D. G., Bast-Pettersen, R., Efskind, J., & solving skills in the elderly. Psychology and Ag-
Thomassen, Y. (2001 ). Neuropsychological effects ing. 2(4), 363-365.
of low mercmy vapor exposure in chloralkali Ernst, J. (1988). Language, grip strength, sensory-
workers. Neurotoxicology, 22(2), 249-258. perceptual, and receptive skills in a normal elderly
El-Sheikh, M., El-Nagdy, S., Townes, B. D., & sample. Clinical Neuropsychologist, 2(1), 30--40.
Kennedy, M. C. (1987). The Luria-Nebraska and Ernst, J., Warner, M. H., Townes, B. D., Peel, J. H., &
Halstead-Reitan neuropsychological test batteries: Preston, M. (1987). Age group differences on
A cross-cultural study in English and Arabic. In- neuropsychological battery performance in a
ternational Journal of Neuroscience, 32, 757-764. neuropsychiatric population: An international
Elvevag, B., Weinstock, D. M., Akil, M., Kleinman, descriptive study with replications. Archives of
J. E., & Goldberg, T. E. (2001). A comparison of Clinical Neuropsychology, 2, 1-12.
verbal fluency taslcs in schizophrenic patients Escalona, E., Yanes, L., Feo, 0., & Maizlish, N.
and normal controls. Schizophrenia Research, (1995). Neurobehavioral evaluation of Venezue-
51(2--3), 119-126. lan workers exposed to organic solvent mixtures.
Elvevag,B.,Fisher,J.E.,Gunl,J.M.,&Goldberg, T.E. American Journal of Industrial Medicine, 27,
(2002). Semantic clustering in verbal fluency: 15-27.
Schizophrenic patients versus control partici- Escandell, V. A. (2002). Cross-cultural neuropsy-
pants. Psychalogical Medicine, 32(5), 909-917. chology in Saudi Arabia. In F.R. Ferraro (Ed.),
Elwan, 0., Hassan, A., Naseer, M., Fahmy, M., Minority and cross-cultural aspects of neuro-
Elwan, F., Kader, A., et al. (1996). Brain aging in psychological assessment. Studies on neuropsy-
normal Egyptians: Neuropsychological, electro- chology, development, and cognition. (pp. 299-
physiological and cranial tomographic assessment. 325) Lisse: Swets & Zeitlinger.
Journal of Neurological Sciences, 136, 7~. Eslinger, P. J., & Benton, A. L. (1983). Visuo-
Elwan, 0., Hassan, A., Naseer, M., Elwan, F., Deif, R., perceptual performances in aging and dementia:
Serafy, 0., et al. (1997). Brain aging in a sample of Clinical and theoretical implications. Journal of
normal Egyptians cognition: education, addiction Cltnical Neuropsychology, 5(3), 213-220.
and smoking. Journal of Neurological Sciences, Eslinger, P. J., & Grattan, L. M. (1993). Frontal lobe
148,79-86. and frontal-striatal substrates for different forms
Elwood, R. W. (1991). Factor structure of the of human cognitive flexibility. Neuropsychologia,
Wechsler Memory Scale-Revised (WMS-R) in 31(1), 17-28.
a clinical sample: A methodological reappraisal. Eslinger, P. J., Damasio, H., Graff-Radford, N. R., &
Clinical Neuropsychalogist, 5, 329-337. Damasio, A. R. ( 1984). Examining the relationship
Elwood, R. W. (1993). Psychological tests and between computed tomography and neuro-
clinical discriminations: Beginning to address the psychological measures in normal and demented
base rate problem. Clinical Psychalogy Review, elderly. Journal of Neurology, Neurosurgery, and
13(5), 409--419. Psychiatry, 47(12), 1319-1325.
554 REFERENCES
Eslinger, P. J., Damasio, A. R., Benton, A. L., & Van measure parietal lobe functions. American Jour-
Allen, M. (1985). Neuropsychological detEction of nal of OcCI.lpational Therapy, 36(7), 444-449.
abnormal mental decline in older persons. Journal Fastenau, P. S. (1996a). An elaborated administra-
of American Medical Association, 253, 670--674. tion of the Wechsler Memory Scale-Revised.
Eson, M. E., Yen, J. K., & Bourke, R. S. (1978). Clinical Neuropsychologist, 10(4), 425-434.
Assessment of recovery from serious head injury. Fastenau, P. S. (1996b). Development and prelimi-
Journal of Neurology, Neurosurgery, and Psy- nary standardization of the "Extended Complex
chiatry, 41(11), 1036-1042. Figure Test" (ECFT). Journal of Clinical and
Esposito, G., Kirkby, B. S., Van Hom, J. D., Experimental Neuropsychology, 18(1), 63-76.
Ellmore, T. M., & Berman, K. F. (1999). Con- Fastenau, P. S. (1998). Validity of regression-based
text-dependent, neural system-specific neuro- norms: An empirical test of the comprehensive
physiological concomitants of ageing: Mapping norms with older adults. Journal of Clinical and
PET correlates during cognitive activation. Brain, Experimental Neuropsychology, 20(6), 906-916.
122(5), ~79. Fastenau, P. S. (2002a). Examination of the ap-
Estes, W. K. (1974). Learning theory and intelli- propriateness of 30-50-year-old ECFT norms
gence. American Psychologist, 29, 740-749. for younger adults: Supporting evidence. Ar-
Evans, R. W., Ruff, R. M., & Gualtieri, C. T. (1985). chives of Clinical Neuropsychology, 17(8), 835.
Verbal fluency and figural fluency in bright chil- Fastenau, P. S. (2002b). The Extended Complex
dren. Perceptual and Motor Skills, 61(3, Pt 1), Figure Test (ECFT). Los Angeles: Western Psy-
699-709. chological Services.
Fabian, M. S., Jenkins, R. L., & Parsons, 0. A. Fastenau, P. S. (2003). Extended Complex Figure
(1981). Gender, alcoholism and neuropsycholog- Test (ECFT): Rationale and empirical support
ical functioning. Journal of Consulting and for recognition and matching. In J. A. Knight
Clinical Psychology. 49(1), 138-140. : (Ed.), The handbook of Rey-Osterrieth Complex
Fabrigoule, C., Rouch, 1., Taberly, A., Letenaeur, L., Figure usage: Clinical and research applications.
Commenges, D., Mazaux, J.-M., et al. (1998). Lutz, FL: Psychological Assessment Resources.
Cognitive process in preclinical phase of demen- Fastenau, P. S., & Adams, K. M. (1996). Heaton,
tia. Brain, 121(1), 135-141. Grant and Matthews' comprehensive norms: An
Faglioni, P., Bertolani, L., Botti, C., & Merelli, E. overzealous attempt [Book review]. Journal of
(2000a). Verbal learning strategies in patients Clinical and Experimental Neuropsychology,
with multiple sclerosis. Cortex, 36(2), 243-263. 18(3), 444-448.
Faglioni, P., Saetti, M. C., & Botti, C. (2000b). Fastenau, P. S., Denburg, N. L., & Mauer, B. A.
Verbal learning strategies in Parkinson's disease. (1998). Parallel short forms for the Boston Nam-
Neuropsychology, 14(3), 456-470. ing Test: Psychometric properties and norms for
Fama, R., Sullivan, E. V., Shear, P. K., Cabo- older adults. Journal ofClinical and Experimental
Weiner, D. A., Yesavage, J. A., Tinklenberg, J. R., Neuropsychology, 20(6), 828-834.
et al. (1998). Fluency performance patterns in Fastenau, P. S., Denburg, N. L., & Hufford, B. J.
Alzheimer's disease and Parkinson's disease. Clin- (1999). Adult norms for the Rey-Osterrieth
ical Neuropstjchologist, 12(4), 487-499. Complex Figure Test and for supplemental
Fama, R., Sullivan. E. V., Shear, P. K., Cabo- recognition and matching trials from the Ex-
Weiner, D. A., Marsh, L., Lim, K. 0., et al. tended Complex Figure Test. Clinical Neuro-
(2000). Structural brain correlates of verbal and psychologist, 13(1), 30-47.
nonverbal fluency measures in Alzheimer's dis- Fastenau, P. S., Evans, J. D., Johnson, K. E., &
ease. Neuropsychology. 14(1), 29-41. . Bond, G. R. (2002). Multicultural training in
Farabat, T. M., Abdelrasoul, G. M., Amr, M. M., clinical neuropsychology. In F. R. Ferraro (Ed.),
Shebl, M. M., Farabat, F. M., & Anger, W. K. Minority and cross-cultural aspects of neu-
(2003). Neurobehavioural effects among workers ropsychological asessment. Studies on neuro-
occupationally exposed to organophosphorous psychology, development, and cognition. Lisse:
pesticides. OcCI.lpational and Environmental Swets & Zeitlinger.
Medicine, 60, 279-286. Faust, D. (1991). Forensic neuropsychology: The art
Farmer, A. (1990). Performance of normal males of practicing a science that does not yet exist
on the Boston Naming Test and the Word Test. [Special Issue: Forensic clinical neuropsychology].
Aphasiology, 4(3), 293-296. Neuropsychology Review, 2(3), 205-231.
Farver, P. F., & Farver, T. B. (1982). Performance Faust, D., Ziskin, J., & Hiers, J. (1991). Brain
of normal older adults on tests desi~ed to damage claims: Coping with neuropsychological
REFERENCES 555
evidence (Vols. 1, 2). Los Angeles: Law and Clinical and Experimental Neuropsychology,
Psychology Press. 19(2), 204-210.
Feinstein, A., Brown, R., & Ron, M. (1994). Effects Fillenbaum, G. G., Peterson, B., Welsh-Bohmer, K. A.,
of practice of serial tests of attention in healthy Kukull, W. A., & Heyman, A. (1998). Progression
subjects. Journal of Clinical and Experimental of Alzheimer's disease in black and white patients:
Neuropsychology, 16, 436--447. The CERAD experience, part XVI. Neurology.
Feldstein, S. N., Keller, F. R., Portman, R. E., 51(1), 154-158.
Durham, R. L., Klebe, K. J., & Davis, H. P. (1999). Fillenbaum, G. G., Heyman, A., Huber, M. S.,
A comparison of computerized and standard Ganguli, M., & Unverzagt, F. W. (2001). Per-
versions of the Wisconsin Card Sorting Test. formance of elderly African American and white
Clinical Neuropsychologist, 13(3), 303--313. community residents on the CERAD Neu-
Fenwick, P., Galliano, S., Coate, M.A., Rippere, V., & ropsychological Battery. Journal of the Inter-
Brown, D. (1985). "Psychic sensitivity," mystical national Neuropsychological Society, 7(4),
experience, head injwy and brain pathology. 502-509.
British Journal of Medical Psychology, 58(1), Fillenbaum, G. G., Unverzagt, F. W., Ganguli, M.,
35-44. Welsh-Bohmer, K. A., & Heyman, A. (2002).
Ferland, M. B., Ramsay, J., Engeland, C., & The CERAD Neuropsychology Battery: Perfor-
O'Hara, P. (1998). Comparison of the performance of representative community and tertiary
mance of normal individuals and survivors of care samples of African-American and Europe-
traumatic brain injwy on repeat administrations an-American elderly. In F. R. Ferraro (Ed.),
of the Wisconsin Card Sorting Test. Journal of Minority and cross-cultural aspects of neu-
Clinical and Experimental Neuropsychology, ropsychological assesment. Studies on neuro-
20(4), 473-482. psychology, development, and cognition. Lisse:
Ferman, T. J., Ivnik, R. J., & Lucas, J. A. (1998). Swets & Zeitlinger.
Boston Naming Test discontinuation rule: Rig- Filskov, S. B., & Catanese, R. A. (1986). Effects of
orous versus lenient interpretations. Assessment, sex and handedness on neuropsychological test-
5(1), 13-18. ing. In S. B. Filskov & T. J. Boll (Eds.), Hand-
Ferraro, F. R., & Barth, J. (2003). Speeded lexical book of Clinical Neuropsychology (Vol. 2). New
decision performing on 15-item forms of the York: Wiley.
Boston Naming Test. Psychology and Education: Fimm, B., Bartl, G., Zimmermann, P., & Wallesch, C.
An Interdisciplinary Journal, 40(2), 38-40. (1994). Different mechanisms underlie shifting
Ferraro, F. R., & Bercier, B. (1996). Boston Nam- set on external and internal cues in Parkinson's
ing Test performance in a sample of Native disease. Brain and Cognition, 25, 287-304.
American elderly adults. Clinical Gerontologist, Finlayson, M.A. J., Johnson, K. A., & Reitan, R. M.
17(1), 58--60. (1977). Relationship of level of education to
Ferraro, F. R., Blaine, T., Flaig, S., & Bradford, S. neuropsychological measures in brain-damaged
(1998). Familiarity norms for the Boston Naming and non-brain-damaged adults. Journal of Con-
Test stimuli. Applied Neuropsychology, 5(1), sulting and Clinical Psychology. 45(4), 536--542.
43-47. Finlayson, M. A., Sullivan, J. F., & Alfano, D. P.
Ferraro, F. R., Bercier, B. J., Holm, J., & McDo- (1986). Halstead's Category Test: Withstanding
nald, J. D. (2002). Preliminary normative the test of time. Journal of Clinical and Experi-
data from a brief neuropsychological test battery mental Neuropsychology, 8(6), 706-709.
in a sample of Native American elderly. In Fisher, L. M., Freed, D. M., & Corkin, S. (1990).
F. R. Ferraro (Ed.), Minority and cross-cultural Stroop Color-Word Test performance in patients
aspects of neuropsychological assessment. Lisse: with Alzheimer's disease. Journal of Clinical and
Swets & Zeitlinger. Experimental Neuropsychology, 12, 745-758.
Feyereisen, P. (1997). A meta-analytic procedure Fisher,N.J.,Tiemey, M.C.,Snow, W. G., &Szalai,J. P.
shows an age-related decline in picture naming: (1999). Odd/even short forms of the Boston
Comments on Goulet, Sica, and Kahn (1994). Naming Test: Preliminary geriatric norms. Clini-
journal of Speech, Language, and Hearing Re- cal Neuropsychologist, 13(3), 359--364.
search, 40, 1328-1333. Fisk, J. D., & Archibald, C. J. (2001). Limitations of
Fillenbaum, G. G., Huber, M., & Taussig, I. M. the Paced Auditory Serial Addition Test as a
(1997). Performance of elderly white and African measure of working memory in patients with
American community residents on the abbrevi- multiple sclerosis. Journal of the International
ated CERAD Boston Naming Test. Journal of Neuropsychological Society, 7(3), 363-372.
556 REFERENCES
Fitz, A. G., Conrad, P. M., Hom, D. L., Sarff, P., & Flynn, J. R. (1998). WAIS-III and WISC-III gains
Majovski, L. V. (1992). Hooper Visual Organi- in the United States from 1972 to 1995: How to
zation Test performance in lateralized brain in- compensate for obsolete norms. Perceptual and
jury. Archives of Clinical Neuropsychology, 7, Motor Skills, 86(3, Pt 2), 1231-1239.
243-250. Folbrecht, J. R., Charter, R. A., Walden, D. K., &
Fitzhugh, K. B., Fitzhugh, L. C., & Reitan, R. M. Dobbs, S.M. (1999). Psychometric properties of
(1964). Influence of age upon measures of the Boston Qualitative Scoring System for the
problem solving and experimental background in Rey-Osterrietb Complex Figure. Clinical Neu-
subjects with longstanding cerebral dysfunction. ropsychologist, 13(4), 442-449.
Journal of Gerontology, 19, 132-134. Folstein, M., Folstein, S., & McHugh, P. (1975).
Fix, A. J., Daughton, D., Kass, 1., Bell, C. W., & Mini-Mental State: A practical method for grading
Golden, C. J. (1985). Cognitive functioning and the cognitive state of patients for the clinician.
survival among patients with chronic obstructive Journal of Psychiatric Research,12, 189-198.
pulmonary disease. International Journal of Neu- Fontenot, D. J., & Benton, A. L. (1972). Perception
roscience, 27(1-2), 13-17. of direction in the right and left visual fields.
Flanagan, J. L., & Jackson, S. T. (1997). Test-retest Neuropsychologia, 10(4), 447-452.
reliability of three aphasia tests: Performance of Foo, S. C., Jeyaratnam, J., & Koh, D. (1990).
non-brain damaged older adults. Journal of Chronic neurobehavioural effects of toluene.
Communication Disorders, 30, 33-43. British Journal of Industrial Medicine, 47(1),
Fleming, K., Goldberg, T. E., Gold, J. M., & 480-484.
Weinberger, D. R. (1995). Verbal working Foo, S. C., Lwin, S., Chia, S. E., & Jeyaratnam, J.
memory dysfunction in schizophrenia: Use of a (1994). Chronic neurobehavioural effects in
Brown-Peterson paradigm. Psychiatry Research, paint formulators exposed to solvents and noise.
56(2), 155-161. Annals of the Academy of Medicine, Singapore,
Fleming, K., Goldberg, T. E., Binks, S., Randolph, 23(5), 650--654.
C., et al. (1997). Visuospatial working memory in Fos, L.A., Greve, K. W., South, M. B., Mathias, C., &
patients with schizophrenia. Biological Psychia- Benefield, H. (2000). Paced VISual Serial Addition
try, 41(1), 43-49. Test: An alternative measure of information pro-
Fletcher, R. H., Fletcher, S. W., & Wagner, E. H. cessing speed. Applied Neuropsychology, 7(3),
(1996). Clinical epidemiology: The essentials (3rd 140-146.
ed.). Philadelphia: Williams & Wilkins. Fossati, P., Amar, G., Raoux, N., Ergis, A. M., &
Fletcher-Janzen, E., Strickland, T. L., & Reynolds, Allilaire, J. F. (1999). Executive functioning
C. R. (2000). Handbook of cross-cultural neu- and verbal memory in young patients with uni-
ropsychology. Netherlands: Kluwer Academic polar depression and schizophrenia. Psychiatry
Publishers. Research, 89(3), 171-187.
Flicker, C., Ferris, S., Crook, T., & Bartus, R. Fossati, P., Ergis, A.-M., & Allilaire, J.-F. (2001).
(1987). Implications of memory and language Problem-solving abilities in unipolar depressed
dysfunction in the naming deficit of senile de- patients:Comparisonofperformanceonthemodi-
mentia. Brain and Language, 31, 187-200. fied version of the Wisconsin and the California
Flinton, M. J., Lucas, J. A., Graff-Radford, N. R., & sorting tests. Psychiatry Research, 104(2),
Uitti, R. J. (1998). Analysis of visuospatial errors 1~156.
in patients with Alzheimer's disease or Parkin- Fossum, B., Holmberg, H., & Reinvang, I. (1992).
son's disease. Journal of Clinical and Experi- Spatial and symbolic factors in performance on
mental Neuropsychology, 20(2), 186-193. the Trail Making Test. Neuropsychology, 6(1),
Flowers, K. A., & Robertson, C. (1985). The effect 71-75.
of Parkinson's disease on the ability to maintain a Foster, H. G., Hillbrand, M., & Silverstein, M.
mental set. Journal of Neurology, Neurosurgery, (1993). Neuropsychological deficit and aggres-
and Psychiatry, 48, 517-529. sive behavior: A prospective study. Progress in
Fluck, E., Fernandes, C., & File, S. E. (2001). Are Neuro-PsychopharmllCology and Biological Psy-
lorazepam-induced deficits in attention similar to chiatry, 17(6), 939-946.
those resulting from aging? Journal of Clinical Francis, W. N., & Kueera, H. (1982). Frequency
Psychopharmacology, 21(2), 126-130. analysis of English usage: Lexicon and grammar.
Flynn, J. R. (1984). The mean IQ of Americans: Boston: Houghton Miffiin.
Massive gains 1932 to 1978. Psychological Bul- Frank, E. M., McDade, H. L., & Scott, W. K. (1996).
letin, 95(1), 29-51. Naming in dementia secondary to Parkinson's,
REFERENCES 557
Huntington's, and Alzheimer's diseases. Journal brain activity in chronic schizophrenic patients
of Communication Disorders, 29, 183-197. during the performance of a verbal fluency task.
Frank, R. M., & Bryne, G. J. (2000). The clinical British Journal of Psychiatry, 167, 343-349.
utility of the Hopkins Verbal Learning Test as a Fromm-Auch, D., & Yeudall, L. T. (1983). Normative
screening test for mild dementia. International data for the Halstead-Reitan neuropsychological
Journal of Geriatric Psychiatry,15(4), 317-324. tests. Journal of Clinical Neuropsychology, 5(3),
Franklin, R. D. (2003). Predktion in forensic and 221-238.
neuropsychology. Hillsdale, NJ: Lawrence Erlbaum. Fujii, D. E., Uoyd, H. A., & Miyamoto, K. (2000).
Franzen, M. D. (2000). &liability and validity in The salience of visuospatial and organizational
neuropsychological assessment (2nd ed.). New skills in reproducing the Rey-Osterrieth Com-
York: Kluwer Academic/Plenum. plex Figure in subjects with high and low IQs.
Franzen, M. D., Smith, S. S., Paul, D. S., & Mac- Clinical Neuropsychologist, 14(4), 551-554.
Innes, W. D. (1993). Order effects in the ad- Fukui, T., Sugita, K., Sato, Y., Takeuchi, T., & Tsu-
ministration of the Booklet Category Test and kagoshi, H. (1994). Cognitive functions in sub-
Wisconsin Card Sorting Test. Archives of Clini- jects with incidental cerebral hyperintensities.
cal Neuropsychology, 8(2), 105-110. European Neurology, 34, 272--276.
Franzen, M. D., Haut, M. W., Rankin, E., & Fuld, P. A. (1981). The Fuld Object Memory Test.
Keefover, R. (1995). Empirical comparison of Chicago: Stoelting Instrument.
alternate forms of the Boston Naming Test. Fuller, K. H., Gouvier, W. D., & Savage, R. M.
Clinical Neuropsychologist, 9(3), 225-229. (1997). Comparison oflist Band list C of the Rey
Franzen, M. D., Paul, D., & Iverson, G. L. (1996). Auditory Verbal Learning Test. Clinical Neuro-
Reliability of alternate forms of the Trail Making psychologist, 11(2), 201-204.
Test. Clinical Neuropsychologist, 10{2), 125-129. Furry, C. A., & Baltes, P. B. (1973). The effect of age
Frazier, T. W., Adams, N. L., Stauss, M. E., & differences in ability-extraneous performance
Redline, S. (2001). Comparability of the Rey and variables on the assessment of intelligence in
Mack forms of the Complex Figure Test. Clini- children, adults, and the elderly. Journal of Ger-
cal Neuropsychologist, 15(3), 337-344. ontology, 28, 73-80.
Freides, D. (1993). Proposed standard of profes- Gaddes, W. H., & Crockett, D. J. (1975). Spreen-
sional practice: Neuropsychological reportsdisplay Benton aphasia tests: Normative data as a mea-
all quantitative data. Clinical Neuropsychologist, sure of normal language development. Brain and
7(2).~235. Language, 2(3), 257-280.
Freides, D. (1995). Interpretations are more benign Gaillard, W. D., Hertz-Pannier, L., Mott, S. H.,
than data? Clinical Neuropsychologist, 9, 248. Barnett, A. S., LeBihan, D., & Theodore, W. H.
Friedman, L., Jesberger, J. A., & Meltzer, H. Y. (2000). Functional anatomy of cognitive devel-
(1991). A model of smooth pursuit performance opment: fMRI of verbal fluency in children and
illustrates the relationship between gain, catch-up adults. Neurology, 54(1), 180-185.
saccade rate, and catch-up saccade amplitude in Galasko, D., Edland, S.D., Morris, J. C., Clark, C.,
normal controls and patients with schizophrenia. Mohs, R., & Koss, E. (1995). The Consortium to
Biological Psychiatry, 30(6), 537-556. Establish a Registry for Alzheimer's Disease
Friedman, L., Kenny, J. T., Jesberger, J. A., (CERAD): XI. Clinical milestones in patients
Choy, M. M., & Meltzer, H. Y. (1995). Rela- with Alzheimer's disease followed over 3 yrs.
tionship between smooth pursuit eye-tracking Neurology, 45(8), 1451-1455.
and cognitive performance in schizophrenia. Galindo, G. & Cortes, J. F. (2003). The ROCF and
Bwlogtcal Psychiatry, 37(4), 265-272. the Complex Figure for Children in Spanish-
Friedman, M.A., Schinka, J. A., Mortimer, J. A., & speaking populations. In J. A. Knight (Ed.), The
Borenstein Graves, A. (2002). Hopkins Verbal handbook of &y-Osterrieth Complex Figure
Learning Test-Revised: Norms for elderly African usage: Clinical and research applications. Lutz,
Americans. Clinical Neuropsychologist, 16(3), FL: Psychological Assessment Resources.
356-372. Gambini, 0., Macciardi, F., Abbruzzese, M., &
Fristoe, N. M., Salthouse, T. A., & Woodard, J. L. Scarone, S. (1992). Influence of education on
(1997). Examination of age-related deficits on WCST performances in schizophrenic patients.
the Wisconsin Card Sorting Test. Neuropsy- International Journal of Neuroscience, 67(1-4),
chology, 11(3), 428-436. 105-109.
Frith, C. D., Friston, K. J., Herold, S., Silbersweig, D., Ganguli, M., Ratcliff, G., Huff, F., Belle, S., Kano, M.,
Fletcher, P., Cahill, C., et al. (1995). Regional Fischer, L., et al. (1991). Effects of age, gender,
558 REFERENCES
and education on cognitive tests in a rural elderly Gerson, A. (1974). Validity and reliability of the
community sample: Norms from the Mono- Hooper Visual Organization Test. Perceptual and
ngahela Valley Independent Elders Survey. Motor Skills, 39, 95-100.
Neuroepidemiology. 10, 42-52. Giambra, L. M., Arenberg, D., Kawas, C.,
Ganguli, M., Seaberg, E., Belle, S., Fischer, L., & Zonderman, A. B., & Costa, P. T. (1995). Adult
Kuller, L. H. (1993). Cognitive impainnent and life span changes in immediate visual memory
the use of health services in an elderly rural and verbal intelligence. Psychology and Aging.
population: The MoVIES project. Journal of the 10(1), 123-139.
American Geriatrics Society, 41, 1065-1070. Gigi, A., Schnaider-Beeri, M., Davidson, M., &:
Ganguli, M., Seaburg, E. C., Ratcliff, G. G., & Prohovnik, I. (1999). Validation of a Hebrew
Belle, S. H. (1996). Cognitive stability over selective reminding test. Israel Journal of Psy-
2 years in a rural elderly population: 'Ihe MoV- chiatry and Reloted Sciences, 36(1), 11-17.
IES project. Neuroepidemiology, 15(1), 42-50. Gilandas, A. J., Touyz, S., Beumont, P. J. V., &:
Gansler, D. A., Fucetola, R., Krengel, M., Stetson, S., Greenberg, H. P. (1984). Handbook of neu-
Zimering, R., & Makaly, C. (1998). Are there ropsychological assessment. Sydney: Grone &:
cognitive subtypes in adult attention deficit/ Stratton.
hyperactivity disorder? Journal of NerlJous and Gilleard, E., & Gilleard, C. (1989). A comparison of
Mental Disease, 186(12), 776-781. Turkish and Anglo-American normative data on
Garb, H. N., & Schramke, C. J. (1996). Judgment the Wechsler Memory Scale. Journal of Clinical
research and neuropsychological assessment: A Psychology, 45(1), 114-117.
narrative review and meta-analyses. Psgchologi- Giovagnoli, A. R. (1996). Trail Making Test: Nor-
cal Bulletin, 1.20(1), 140-153. mative values from 287 normal adult controls.
Gasquoine, P. G. (2001). Research in clinical Italian Journal of Neurological Sciences, 17(4),
neuropsychology with Hispanic American par- 305-310.
ticipants: A review. Clinical Neuropsydwlogist, Giovagnoli, A. R., &: Avanzini, G. (1996). Forgetting
15(1), 2-12. rate and interference effects on a verbal memory
Gaudino, E. A., Geisler, M. W., & Squires, N. K. distractor task in patients with temporal lobe
(1995). Construct validity in the Trail Making epilepsy. Journal of Clinical and Experimental
Test: What makes part B harder? journal of Neuropsychology, 18, 259-264.
Clinical and Experimental Neuropsychology, Giovannetti, T., Goldstein, R. Z., Schullery, M.,
17(4), 529-535. Barr, W. B., & Bilder, R. M. (2003). Category
Geffen, G. M., Moar, K. J., O'Hanlon, A. P., fluency in first-episode schizophrenia. Journal of
Clark, C. R., & Geffen, L. B. (1990). Perfor- the International Neuropsychological Society,
mance measures of 16- to 86-year-old males and 9(3), 384-393.
females on the Auditory Verbal Learning Test. Gladsjo, J. A., Schuman, C. C., Evans, J. D., Peavy, G. M.,
Clinical Neuropsychologist, 4(1), 45--63. Miller, S. W., &: Heaton, R. K. (1999). Norms for
Geffen, G. M., Bate, A., Wright, M., Rozenbilds, U., letter and category fluency: Demographic cor-
& Geffen, L. (1993). A comparison of cognitive rections for age, education, and ethnicity. As-
impairments in dementia of the Alzheimer type sessment, 6(2), 147-178.
and depression in the elderly. Demenla, 4(5), Glass, G. V. (1976). Primary, secondary, and meta-
294--300. analysis. Educational Researcher, 5, 3--5.
Geffen, G. M., ButteiWorth, P., & Geffen, L. B. Gleissner, U., & Elger, C. E. (2001). The hippo-
(1994). Test-retest reliability of a new fonn of the campal contribution to verbal fluency in patients
Auditory Verbal Learning Test (AVLT). Archives with temporal lobe epilepsy. Cortex, 37(1), 55--63.
of Clinical Neuropsychology, 9(4), 303-316. Glennerster, A., Palace, J., Warburton, D., & Ox-
George, M. S., Ketter, T. A., Parekh, P. I., Ro- bury, S. (1996). Memory in myasthenia gravis:
sinksy, N., Ring, H., Casey, B. J., et al. (1994). Neuropsychological tests of central cholinergic
Regional brain activity when selecting a response function before and after effective immunologic
despite interference: An H 20 150 PET study of treatment. Neurology, 46(4), 1138-1142.
the Stroop and an emotional Stroop. Human Goethe, K. E., Mitchell, J. E., Marshall, D. W.,
Brain Mapping. 1, 194-209. Brey, R. L., Cahill, W. T., Leger, G. D., et al.
Gershberg, F. B., &: Shimamura, A. P. (1995). Im- (1989). Neuropsychological and neurological
paired use of organizational strategies in free recall function of human immunodeficiency virus se-
following frontal lobe damage. Neuropsychologia, ropositive asymptomatic individuals. Archives of
33(10), 1305-1333. Neurology, 46(2), 129-133.
REFERENCES 559
Gold, A. E., Macl..axl, K. M., Deary, I. J., & Frier, B. M. Goldstein, G., & Shelly, C. H. (1972). Statistical
(1995). Hypoglycemia-induced cognitive dys- and nonnative studies of the Halstead neu-
function in diabetes mellitus: Effect of hypogly- ropsychological test battery relevant to a neu-
cemia unawareness. Physiology and Behavior, ropsychiatric hospital setting. Perceptual and
58(3), 501-511. Motor Skills, 34, 603-620.
Goldberg, T. E., Bennan, K. F., Mohr, E., & Goldstein, G., & Shelly, C. H. (1975). Similarities
Weinberger, D. R. (1990). Regional cerebral and differences between psychological deficit in
blood How and cognitive function in Huntington's aging and brain damage. Journal of Gerontology,
disease and schizophrenia: A comparison of pa- 30(4), 448-455.
tients matched for perfonnance on a prefronatal- Goldstein, S. G., & Braun, L. S. (1974). Reversal of
type task. Archives of Neurology, 47, 418-422. expected transfer as a function ofincreased age.
Golden, C. (1978). Stroop experimental uses. Chi- Perceptual and Motor SkiUs, 38, 1139-1145.
cago: Stoelting. Gollan, T. H., Montoya, R.I., & Werner, G. A. (2002).
Golden, C. (1981). A standardized version of Lur- Semantic and letter Huency in Spanish-English
ia's neuropsychological tests. In S. Filskov & bilinguals. Neuropsychology, 16(4), 562-576.
T. Boll (Eds.), Handbook of clinical neuropsy- Gontkovsky, S. T., & Souheaver, G. T. (2002). T-
chology. New York: Wiley-Interscience. score and raw-score comparisons in detecting
Golden, C. J., Kupennan, S. K, Macinnes, W. D., & brain dysfunction using the Booklet Category
Moses, J. A. (1981a). Cross-validation of an ab- Test and the Short Category Test. Perceptual
breviated fonn of the Halstead Category Test. and Motor Skills, 94(1), 319-322.
Journal of Consulting and Clinical Psychology, Gonzalez, E. A., Dieter, J. N. 1., Natale, R. A., &
49(4), 606--607. Tanner, S. L. (2001). Neuropsychological eval-
Golden, C. J., Osmon, D. C., Moses, J. A., Jr., & uation ofhigher functioning homeless persons: A
Berg. R. A. (1981b). Interpretation ofthe Halstead- comparison of an abbreviated test battery to the
Reitan Neuropsychological Test Battery: A Case- Mini-Mental State Exam. Journal of Neroous
book Approach. New York: Grune & Stratton. and Mental Disease, 189(3), 176-181.
Goldman, R. S., Axelrod, B. N., Heaton, R. K., Goodglass, H. (1980 ). Disorders of naming following
Chelune, G. J., et al. (1996). Latent structure of brain injury. American Scientist, 68(6), 647--655.
the WCST with the standardization samples. Goodglass, H., Kaplan, E., & Barressi, B. (2001).
Assessment, 3(1), 7:>---78. The Boston Diagnostic Aphasia Examination
Goldman, W. P., Baty, J. D., Buckles, V. D., (BDAE) (3rd ed.). Odessa, FL: Psychological
Sahnnann, S., & Morris, J. C. (1998). Cognitive Assessment Resources.
and motor functioning in Parkinson disease: Gooding, D. C., & Tallent, K. A. (2002). Spatial
Subjects with and without questionable demen- working memory perfonnance in patients with
tia. Archives of Neurology, 55(5), 674--680. schizoaffective psychosis versus schizophrenia: A
Goldsmith, R. W., & Brengelmann, J. C. (1971). tale of two disorders? Schizophrenia Research,
Interactions between personality, abnonnality 53(3), 209-218.
and test conditions in a battery of tests. Archives Gooding. D. C., Kwapil, T. R., & Tallent, K. A.
of Psychology, 123(3), 217-224. (1999). Wisconsin Card Sorting Test deficits in
Goldstein, D., Mercury, M., Azrin, R., Millsapa. C., schizotypic individuals. Schizophrenia Research,
Ventura, T., & Pliskin, N. (2000). Cultural con- 40(3), 201-209.
siderations on the Boston Naming Test: The ef- Goodman, A. M., Delis, D. C., & Mattson, S. N.
fects of race and geographic region. Journal of the (1999). Nonnative data for 4-year-old children
International Neuropsychological Society, 6, 143. on the California Verbal Learning Test-
Goldstein, G. (1997). The clinical utility of stan- Children's Version. Clinical Neuropsychologist,
dardized or flexible battery approaches to neuro- 13(3), 274-282.
psychological assessment. In G. I. Goldstein & Gordon, H. W., & Lee, P. A. (1986). A relationship
T. M. Incagnoli (Eds.), Contemporary approaches between gonadotropins and visuospatial func-
to neuropsychological assessment. New York: tion. Neuropsychologia, 24(4), 56:>-576.
Plenum. Gordon, N. G. (1972). The Trail Making Test in
Goldstein, G., & Zubin, J. (1990). Neuropsycho- neuropsychological diagnosis. Journal of Clinical
logical differences between young and old Psychology, 28, 167-169.
schizophrenics with and without associated neu- Gordon, N. G., & O'Dell, J. W. (1983). Sex dif-
rological dysfunction. Schizophrenia Research, ferences in neuropsychological perfonnance.
3(2), 117-126. Perceptual and Motor Skills, 56, 126.
560 REFERENCES
Gorsuch, R. L. (1983). The theory of continuous changes in old age. journal of Clinical and Ex-
nanning. In R. L. Gorsuch (chair), Continuous perimental Neuropsychology, 17, 390-415.
nonning: An alternative to tabled nonns? Sym- Grafman, J., Jonas, B., & Salazar, A. (1990). Wis-
posium conducted at the 9lst Annual Conven- consin Card Sorting Test performance based
tion of the American Psychological Association, on location and size of neuroanatomical lesion
Anaheim, August 26--30. in Vietnam veterans with penetrating head in-
Gottschalk, L.A., & Selin, C. (1991). Comparative jury. Perceptual and Motor Skills, 71(3, Pt 2),
neurobiological and neuropsychological deficits 1120-1122.
in adolescent and adult schizophrenic and non- Grant, D. A., & Berg, E. A. (1948). Behavioral
schizophrenic patients. Psychotherapy and Psy- analysis of degree of reinforcement and ease of
chosomatics, 55(1), 32-41. shifting to new responses in a Weigl-type card-
Goul, W. R., & Brown, M. (1970). Effects of age sorting problem. Journal of Experimental Psy-
and intelligence on Trail Making Test perfor- chology, 38, 404-411.
mance and validity. Perceptual and Motor Skills, Grant, 1., Prigatano, G. P., Heaton, R. K.,
30, 319-326. McSweeny, A. J., Wright, E. C., & Adams, K. M.
Goulet, P., Ska, B., & Kahn, H. J. (1994). Is there a (1987). Progressive neuropsychologic impair-
decline in picture naming with advancing age? ment and hypoxemia: Relationship in chronic
Journal of Speech and Hearing Research, 37(3), obstructive pulmonary disease. Archives of
629--644. General Psychiatry, 44(11), 999-1006.
Gouvier, W. D. (1999). Base rates and clinical de- Green, B., & Hall, J. (1984). Quantitative methods
cision making in neuropsychology. In J. J. Sweet for literature review. Annual Review of Psy-
(Ed.), Forensic neuropsychology: Fundamentals chology, 35, 31-53.
and practice. Studies on neuropsychology, de- Greene, R. L., & Farr, S. P. (1985, August). Mul-
velopment, and cognition. Lisse: Swets & tiple regression of moderator variables on Trail
Zeitlinger. Making Test peifonnance. Paper presented at
Gouvier, W. D. (2001). Are you sure you're really the annual meeting of the American Psycholog-
telling the truth? [Special Issue: Controversies in ical Association, Los Angeles, CA.
neuropsychology] NeuroRehabilitation, 16(4), Gregg, E. W., Yaffe, K., Cauley, J. A., Rolka, D. B.,
215-219. Blackwell, T. L., Narayan, K. M., & Cummings,
Gouvier, W. D., Hayes, J. S., & Smiroldo, B. B. S. R. (2000). Is diabetes associated with cognitive
(1998). The significance of base rates, test sen- impairment and cognitive decline among older
sitivity, test specificity, and subjects' knowledge women? Study of Osteoporotic Fractures Re-
of symptoms in assessing TBI sequelae and search Group. Archives of Internal Medicine,
malingering. In C. R. Reynolds (Ed.), Detection 160(2), 17~180.
of malingering during head injury litigation. Gregory, R., & Paul, J. (1980). The effects of hand-
Critical issues in neuropsychology, (pp. 55-79). edness and writing posture on neuropsycho-
NY: Plenum Press. logical test results. Neuropsychologia, 18,
Gouvier, W. D., Pinkston, J. B., Santa Maria, M.P., & 231-235.
Cherry, K. E. (2002). Base rate analysis in cross- Gregory, R. J., Paul, J. J., & Morrison, M. W. (1979).
cultural clinical psychology-Diagnostic accuracy A short form of the Category Test for adults.
in the balance. In F. R. Ferraro (Ed.), Minority journal of Clinical Psychology, 35(4), 795-798.
and cross-cultural aspects of neuropsychological Greiffenstein, M. F., Baker, W. J., & Gola, T.
assessment. Studies on neuropsychology. devel- (1994). Validation of malingered amnesia mea-
opment, and cognition. Lisse: Swets & Zeitlinger. sures with a large clinical sample. Psychological
Grady, D., Yaffe, K., Kristof, M., Lin, F., Richards, Assessment, 6(3), 218-224.
C., & Barrett-Connor, E. (2002). Effect of Greve, K. W. (1993). Can perseverative responses
postmenopausal hormone therapy on cognitive on the Wisconsin Card Sorting Test be scored
function: The Heart and Estrogen/Progestin accurately? Archives of Clinical Neuropsychol-
Replacement Study. American Journal of Medi- ogy, 8, 511-517.
cine, 113(7), 543-548. Greve, K. W. & Bianchini, K. J. (2002). Using the
Graf, P., & Uttl, B. (1995). Component processes of Wisconsin Card Sorting Test to detect malin-
memory: Changes across the adult lifespan. gering: An analysis of the specificity of two
Swiss Journal of Psychology, 54(2), 11~130. methods in nonmalingering normal and patient
Graf, P., Uttl, B., & Tuokko, H. (1995). Color- samples. Journal of Clinical and Experimental
and Picture-Word Stroop Tests: Performance Neuropsychology, 24, 48-54.
REFERENCES 561
Greve, K. W., Williams, M. C., Haas, W. G., elderly. Journal of Clinical and Experimental
Littell, R. R., & Reinoso, C. (1996). The role of Neuropsychology, 19(5), 643-654.
attention in Wisconsin Card Sorting Test per- Grober, E., Lipton, R., Katz, M., & Sliwinski, M.
formance. Archives of Clinical Neuropsychology, (1998). Demographic inHuences on free and
11(3), 215-222. cued selective reminding performance in older
Greve, K. W., Brooks, J., Crouch, J. A., persons. Journal of Clinical and Experimental
Williams, M. C., et al. (1997). Factorial structure Neuropsychology, 20(2), 221-226.
of the Wisconsin Card Sorting Test. British Grober, E., Lipton, R., Hall, C., & Crystal, H.
Journal of Clinical Psychology, 36(2), 283-285. (2000). Memory impairment on free and cued
Greve, K. W., Ingram, F., & Bianchini, K. J. (1998). selective reminding predicts dementia. Neurol-
Latent structure of the Wisconsin Card Sorting ogy, 54(4), 827--832.
Test in a clinical sample. Archives of Clinical Groff, M. G., & Hubble, L. M. (1981). A factor
Neuropsychology, 13(7), 597-609. analytic investigation of the Trail Making Test.
Greve, K. W., Bianchini, K. J., Hartley, S. M., & International Journal of Clinical Neuropsychol-
Adams, D. (1999). The Wisconsin Card Sorting ogy, 3(4), 11-13.
Test in stroke rehabilitation: Factor structure Gronwall, D. (1977a). Paced Auditory Serial-
and relationship to outcome. Archives of Clinical Addition task: A measure of recovery from concus-
Neuropsychology, 14, 497-509. sion. Perceptual and Motor SkiUs, 44(2), 367-373.
Greve, K. W., Lindberg, R. F., Bianchini, K. J., & Gronwall, D. (1977b). PASAT (Paced Auditory
Adams, D. (2000). Construct validity and pre- Serial Addition Test): Manual of instructions and
dictive value of the Hooper Visual Organization norms. Victoria: University of Victoria.
Test in stroke rehabilitation. Applied Neuropsy- Gronwall, D., & Sampson, H. (1974). The psycho-
chology, 7(4), 215-222. logical effects of concussion. Auckland, New
Greve, K. W., Love, J. M., Sherwin, E., Mathias, C. Zealand: Auckland University Press-Oxford Uni-
W., Ramzinski, P., & Levy, J. (2002). Wisconsin versity Press.
Card Sorting Test in chronic severe traumatic Gronwall, D., & Wrightson, P. (1974). Delayed
brain injury: Factor structure and performance recovery of intellectual function after minor
subgroups. Brain Injury, 16(1), 29-40. head injury. lAncet, 2, 605--609.
Greve, K. W., Bianchini, K. J., & Adams, D. Gronwall, D., & Wrightson, P. (1981). Memory and
(2003a). The ROCF in stroke rehabilitation and information processing capacity after closed
recovery. In J. A. Knight (Ed.), The handbook of head injury. Journal of Neurology, Neurosurgery,
Rey-Osterrieth Complex Figure usage: Clinical and Psychiatry, 44, 88~95.
and research applications (pp. 525-541). Lutz, Gross-Isseroff, R., Sasson, Y., Voet, H., Hendler,
FL: Psychological Assessment Resources. T., Luca-Haimovici, K., Kandel-Sussman, et al.
Greve, K. W., Hartley, S. M., Houston, R. J., (1996). Alternation learning in obsessive-
Bianchini, K. J., Adams, D., & Stanford, M. S. compulsive disorder. Biological Psychiatry,
(2003b). The ROCF in patients with vascular 39(8), 733-738.
lesions of the cerebellum. In J. A. Knight (Ed.), Groth-Marnat, G. (2000). Introduction to neu-
Rey-Osterrieth Complex Figure usage: Clinical ropsychological assessment. In G. Groth-Mamat
and research applications (pp. 611-624). Lutz, (Ed.), Neuropsychological assessment in clinical
FL: Psychological Assessment Resources. proctice. New York: Wiley.
Griffiths, P. (1991). Word-finding ability and design Groth-Mamat, G. (2003). Handbook of psycholog-
Huency in developmental dyslexia. British Jour- ical assessment (4th ed.). New York: Wiley.
nal of Clinical Psychology, 30(1), 47-60. Gruenewald, P. J., & Lockhead, G. R. (1980). The
Grigsby, J., & Kaye, K. (1995). Alphanumeric se- free recall of category examples. Journal of Ex-
quencing and cognitive impairment among perimental Psychology: Human Learning and
elderly persons. Perceptual and Motor Skills, Memory, 6(3), 225-240.
80(3, Pt 1), 732-734. Gruzelier, J., & Warren, K. (1993). Neuropsycho-
Grigsby, J., Kaye, K., & Busenbark, D. (1994). Al- logical evidence of reductions on left frontal tests
phanumeric sequencing: A report on a brief with hypnosis. Psychological Medicine, 23(1),
measure of information processing used among 93-101.
persons with multiple sclerosis. Perceptual and Guilford, J.P. (1965). Fundamental statistics in psy-
Motor SkiUs, 78(3, Pt 1), 883--887. chology and education. New York: McGraw-Hill.
Grober, E., Merling, A., Heimlich, T., & Lipton, R. Guilmette, T. J., & Rasile, D. (1995). Sensitivity,
(1997). Free and cued selective reminding in the specificity, and diagnostic accuracy of three
562 REFERENCES
verbal memory measures in the assessment and change in multiple sclerosis. Journal of
of mild brain injury. Neuropsychology, 9(3), Clinical Psychology, 44(4), 540--548.
338--344. Halligan, P. W., Cockburn, J., & Wilson, B. A.
Guo, Q., Lu, C., & Hong, Z. (2000). Application of (1991). The behavioural assessment of visual
Rey-Osterrieth Complex Figure Test in Chinese neglect. Neuropsychological Rehabilitation, 1(1),
normal older people. Chinese Journal of Clinical 5-32.
Psychology, 8(4), 205-207. Halstead, W. C. (1947). Brain and intelligence.
Gur, R. C., Alsop, D., Glahn, D., Petty, R., Swan- Chicago: University of Chicago Press.
son, C. L., Maldjian, J. A., et al. (2000). An fMRI Hamby, S. L., Wilkins, J. W., & Barry, N. S. (1993).
study of sex differences in regional activation to a Organizational quality on the Rey-Osterrieth and
verbal and a spatial task. Brain and Lan{!}Jage, Taylor complex figure tests: A new scoring sys-
74(2), 157-170. tem. Psychological Assessment, 5(1), 27--33.
Gurd, J. M. (2000). Verbal Huency deficits in Hamby, S. L., Bardi, C. A., & Wilkins, J. W. (1997).
Parkinson's disease: Individual differences in Neuropsychological assessment of relatively in-
underlying cognitive mechanisms. Journal of tact individuals: Psychometric lessons from an
NeurolinfYJisHcs, 13(1), 47-55. Hw+ sample. Archives of Clinical Neuropsy-
Curling, H. M., Curtis, D., & Murray, R. M. chology, 12(6), 545-556.
(1991). Psychological deficit from excessive al- Hameleers, P. A. H. M., Van Boxtel, M. P. J.,
cohol consumption: Evidence from a co-twin Hogervorst, E., Riedel, W. J., Houx, P. J., Bun-
control study. British Journal ofAddiction, 86(2), tinx, F., et al. (2000). Habitual caffeine con-
151-155. sumption and its relation to memory, attention,
Guruje, 0., Unverzargt, F., Osuntokun, B., Hen- planning capacity and psychomotor performance
drie, H., Baiyewu, 0., Ogunniyi, A., et al. (1995). across multiple age groups. Human Psycho-
The CERAD Neuropsychological Test Battery: pharmacology: Clinical and Erperimental, 15(8),
Norms from a Yoruba-speaking Nigerian sample. 573-581.
West African Journal of Medicine, 14, 29-33. Hamilton, M. (1960). A rating scale for depression.
Guskiewicz, K. M., Ross, S. E., & Marshall, S. W. Journal of Neurology, Neurosurgery, and Psy-
(2001). Postural stability and neuropsychological chiatry, 23, 56-62.
deficits after concussion in collegiate athletes Hanks, R. A., Allen, J. B., Ricker, J. H., & Desh-
[Special Issue: Concussion in athletes]. Journal pande, S. A. (1996). Normative data on a mea-
of Athletic Training, 36(3), 263-273. sure of design Huency: The Make A Figure Test.
Haaland, K. Y., Cleeland, C. S., & Carr, D. (1977). Assessment, 3(4), 459-466.
Motor performance after unilateral hemisphere Hannay, H. J., & Levin, H. S. (1985). Selective
damage in patients with tumor. Archives of Reminding Test: An examination of the equiva-
Neurology, 34, 556-559. lence of four forms. Journal of Clinical and Ex-
Haaland, K. Y., Linn, R., Hunt, W., & Goodwin, J. perimental Neuropsychology, 7(3), 251-263.
(1983). A normative study of Russel's variant of Hannay, H. J., Falgout, J. C., Leli, D. A.,
the Wechsler Memory Scale in a healthy elderly Katholi, C. R., et al. (1987). Focal right temporo-
population. Journal of Consulting and Clinical occipital blood How changes associated with
Psychology, 51(6), 87~1. judgment of line orientation. Neuropsychologia,
Haaland, K. Y., Vranes, L. F., Goodwin, J. S., & 25(5), 755-763.
Garry, P. J. (1987). Wisconsin Card Sort Test Hanninen, T., Hallikainen, M., Koivisto, K., Paar-
performance in a healthy elderly population. tanen, K., Laakso, M. P., Riekkinen, P. J., et al.
Journal of Gerontology, 42(3), 345-346. (1997). Decline of frontal lobe functions in
Haaland, K. Y., Price, L., & LaRue, A. (2003). What subjects with age-associated memory impair-
does the WMS-111 tell us about memory changes ment. Neurology, 48, 148-153.
with normal aging? Journal of the International Harnadek, M. C. S., & Rourke, B. P. (1994).
Neuropsychological Society, 9, 89-96. Principal identifying features of the syndrome of
Haddock, C. K., Rindskopf, D., & Shadish, W. R. nonverbal learning disabilities in children. Jour-
(1998). Using odds ratios as effect sizes for meta- nal of Learning Disabilities, 27(3), 144-154.
analysis of dichotomous data: A primer on Harris, J. G., Cullum, C. M., & Puente, A. E.
methods and issues. Psychological Methods, 3(3), (1995). Effects of bilingualism on verbal learning
339-353. and memory in Hispanic adults. Journal of
Halligan, F. R., Reznikoff, M., Friedman, H. P., & the International Neuropsychological Society,
LaRocca, N. G. (1988). Cognitive dysfunction 1(1), 10--16.
REFERENCES 563
Harris, M., Cross, H., & VanNieuwkerk, R. (1981). Hays, J. R. (1995). Trail Making Test norms for
The effects of state depression, induced de- psychiatric patients. Perceptual and Motor Skills,
pression and sex on the Finger Tapping and 80(1), 187-194.
Tactual Performance Tests. Clinical Neuropsy- Head, D., Bolton, D., & Hymas, N. (1989). Deficits
chology, 3(4), 28-34. in cognitive shifting ability in patients with
Harris, M. E. (1988). Wisconsin Card Sorting Test obsessive-compulsive disorder. Biological Psy-
computer version. Odessa, FL: Psychological chiatry, 25, 929-937.
Assessment Resources. Head, D., Raz, N., Gunning-Dixon, F., Williamson,
Harris, M. J., & Rosenthal, R. (1985). Mediation A., & Acker, J. D. (2002). Age-related differ-
of interpersonal expectancy effects: 31 meta- ences in the course of cognitive skill acquisition:
analyses. Psychological BuUetin, 97, 363-386. The role of regional cortical shrinkage and cog-
Hart, R. P., Kwentus, J. A., Wade, J. B., & Taylor, J. nitive resources. Psychology and Aging, 17(1),
R. (1988). Modified Wisconsin Sorting Test in 72-84.
elderly normal, depressed and demented pa- Heaton, R. K. (1981). Wisconsin Card Sorting Test
tients. Clinical Neuropsychologist, 2(1), 49-56. manual. Odessa, FL: Psychological Assessment
Harter, S., Hart, C., & Harter, G. (1999). Expanded Resources.
scoring criteria for the Design Fluency Test: Heaton, R. K. (1985). Importance of demographic
Reliability and validity in neuropsychological and variables in interpreting scores on the Halstead-
college samples. Archives of Clinical Neuropsy- Reitan Battery. Paper presented at the 13th annual
chology, 14(5), 419-432. meeting of the International Neuropsychological
Hartlage, L. C. (2001). Neuropsychological testing Society, San Diego, CA.
of adults: Further considerations for neurolo- Heaton, R. K. (1992). Comprehensive nonns for an
gists. Archives of Clinical Neuropsychology, expanded Halstead-Reitan Battery: A supple-
16(3), 201-213. ment for the WAIS-R Odessa, FL: Psychological
Hartman, M., & Potter, G. (1998). Sources of age Assessment Resources.
differences on the Rey-Osterrieth Complex Heaton, R. K. (1993). Wisconsin Card Sorting Test
Figure Test. Clinical Neuropsychologist, 12(4), Computer version 2.0. Odessa, FL: Psychological
513-524. Assessment Resources.
Harvey, N. S. (1986). Impaired cognitive set- Heaton, R. K. (2003a). Wisconsin Card Sorting
shifting in obsessive-compulsive neurosis. IRCS Test computer version 4.0. Odessa, FL: Psycho-
Medical Science, 14, 936-937. logical Assessment Resources.
Hasselblad, V., & Hedges, L. V. (1995). Meta- Heaton, R. K. (2003b). Wisconsin Card Sorting
analysis of screening and diagnostic tests. Psy- Test-64 research edition computer version
chological BuUetin, 117(1), 167-178. 2.0. Odessa, FL: Psychological Assessment
Haut, M. W., Cahill, J., Cutlip, W. D., Resources.
Stevenson, J. M., Makela, E. H., Bloomfield, Heaton, R. K., Vogt, A. T., Hoehn, M. M., Lewis, J. A.,
S. M. (1996). On the nature of Wisconsin Card Crowley, T. J., & Stallings, M. A. (1979). Neu-
Sorting Test performance in schizophrenia. ropsychological impairment with schizophrenia
Psychiatry Research, 65(1), 15-22. vs. acute and chronic cerebral lesions. Journal of
Hawkins, K. A., & Bender, S. (2002). Norms and Clinical Psychology, 35(1), 46-53.
the relationship of Boston Naming Test perfor- Heaton, R. K., Nelson, L. M., Thompson, D. S.,
mance to vocabulary and education: A review. Burks, J. S., & Franklin, G. M. (1985). Neu-
Aphasiology, 16(12), 1143-1153. ropsychological findings in relapsing-remitting
Hawkins, K. A., Sledge, W. H., Orleans, J. F., Quin- and chronic-progressive multiple sclerosis. Jour-
lan, D. M., et al. (1993). Nonnative implications of nal of Consulting and Clinical Psychology, 53(8),
the relationship between reading vocabulary and 103-110.
Boston Naming Test performance. Archives of Heaton, R. K., Grant, 1., & Matthews, C. G. (1986).
Clinical Neuropsychology, 8(6), 525-537. Differences in neuropsychological test perfor-
Haxby, J. V., Grady, C. L., Duara, R., Robertson- mance associated with age, education, and sex.
Tehabo, E. A., Koziarz, B., Cutler, N. R., et al. In I. Grant & K. Adams (Eds.), Neuropsycholog-
(1986). Relations among age, visual memory, and ical assessment of neuropsychiatric disorders.
resting cerebral metabolism in 40 healthy men. New York: Oxford University Press.
Brain and Cognition, 5(4), 412-427. Heaton, R. K., Grant, 1., & Matthews, C. (1991).
Hayes, W. L. (1963). Statistics. New York: Rinehart Comprehensive nonns for an expanded Halstead-
&Winston. Reitan Neuropsychological Battery: Demographic
564 REFERENCES
corrections, research findings, and clinical appli- Henderson, L. W., Frank, E. M., Pigatt, T.,
cations. Odessa. FL: Psychological Assessment Abramson, R. K., & Houston, M. (1998). Race,
Resources. gender, and educational level effects on Boston
Heaton, R. K., Chelune, G. J., Talley, J. L., Kay, Naming Test scores. Aphasialogy, 12(10),
G. G., & Curtiss, G. (1993). Wiscon$n Card 901-911.
Sorting Test manual: revised and etpanded. Henderson, V. W., Mack, W., Freed, D. M.,
Odessa, FL: Psychological Assessment R•sources. Kempler, D., & Andersen, E. S. (1990). Naming
Heaton, R. K., Matthews, C., Grant, 1., & ~vitable, consistency in Alzheimer's disease. Brain and
N. (1996a). Demographic corrections with Language,39, 530-538.
comprehensive norms: an overzealous attempt or Hermann, B. P., Wyler, A. R., & Richey, E. T.
a good start? Journal of Clinical and1 Experi- (1988). Wisconsin Card Sorting Test perfor-
mental Neuropsychology, 18(3), 449--458. mance in patients with complex partial seizure of
Heaton, R. K., Ryan, L., Grant, 1., & Matthews, temporal-lobe origin. Journal of Clinical and
C. G. (1996b). Demographic influences: on neu- Experimental Neuropsychology, 10, 467-476.
ropsychological test performance. In I. Grant & Hesselbrock, M. N., Weidenman, M.A., & Reed,
K. M. Adams (Eds.), Neuropsychological as- H. B. (1985). Effect of age, sex, drinking history
sessment of neuropsychiatric disorders ($nd ed.). and antisocial personality on neuropsychology of
New York: Oxford University Press. alcoholics. Journal of Studies on Alcohol, 46(4),
Heaton, R. K., Avitable, N., Grant, 1., & Matthews, 313-320.
C. G. (1999). Further crossvalidation of regres- Heubrock, D. (1995). Error analysis in neu-
sion-based neuropsychological norms with an ropsychological assessment of verbal memory
update for the Boston Naming Test. ]~mal of and learning. European Journal of Psychological
Clinical and Experimental Neurops!Jfhology, Assessment, 11(1). 21-28.
21(4), 571-582. Hilgert, L. D., & Treloar, J. H. (1985). The rela-
Heaton, R. K., Temkin, N., Dikmen, S., Avitable, tionship of Hooper Visual Organization Test to
N., Taylor, M. J., Marcotte, T. D., et al, (2001). sex, age and intelligence of elementary school
Detecting change: A comparison of th~e neu- children. Measurement and Evaluation in
ropsychological methods, using nofllal and Counseling and Development, 17(4), 203-206.
clinical samples. Archives of Clinical Niuropsy- Hinton, V. J., Dobkin, C. S., Halperin, J. M.,
chology, 16(1), 75-91. Jenkins, E. C., Brown, W. T., Ding, X. H., et al.
Heaton, R. K.• Miller, S. W., Taylor, M. J., &Grant, I. (1992). Mode of inheritance influences behav-
(2004). Revised comprehensive nonns for an ex- ioral expression and molecular control of cogni-
panded Holstead-Reitan Battery: Demllgraphi- tive deficits in female carriers of the fragile X
caUy adjusted neuropsychological no~ for syndrome. American Journal of Medical Genet-
African American and Caucasian adulfiJ. Lutz, ics, 43(1-2), 87-95.
FL: Psychological Assessment Resources. Ho, A. K., Sahakian, B. J., Robbins, T. W., Barker,
Hedges, L. (1982). Estimation of effect size from a R. A., Rosser, A. E., & Hodges, J. R. (2002).
series of independent experiments. Psyc~logical Verbal Huency in Huntington's disease: A longi-
BuUetin, 92, 490-499. tudinal analysis of phonemic and semantic clus-
Hedges, L., & Olkin, I. (1985). Statistical tpethock tering and switching. Neuropsychologia, 40(8),
for meta-analysis. Orlando, FL: Academ~ Press. 1277-1284.
Heilbronner, R. L., Henry, G. K., Buck, P.;Adams, Hochberg, F. H., & Slotnick, B. (1980). Neu-
R. L., & Fogle, T. (1991). Lateralized brain ropsychologic impairment in astrocytoma survi-
damage and performance on Trail Maki~ A and vors. Neurology, 30(2), 172-177.
B, Digit Span forward and backward, and TPT Hochla, N. N., Fabian, M. S., & Parsons, 0. A.
memory and location. Archives of Cliniapl Neu- (1982). Brain-age quotients in recently detoxified
ropsychology, 6(4), 251-258. alcoholic, recovered alcoholic and nonalcoholic
Helmstaedter, C., Pohl, C., & Elger, C. E.,(1995). women. Journal of Clinical Psychology, 38(1),
Relations between verbal and nonverbal ..emory 207-212.
performance: Evidence of confounding; effects Hodges, J. R., Salmon, D. P., & Butters, N. (1991).
particularly in patients with right tempotal lobe The nature of the naming deficit in Alzhei-
epilepsy. Cortex, 31(2), 345-355. mer's and Huntington's disease. Brain, 114,
Hemsley, D. (1974). Relationship between two 1547-1558.
tests of visual retention. Perceptual an4 Motor Hoff, A. L., Riordan, H., Morris, L., Cestaro, V.,
Skills, 39, 1132-1134. · Wieneke, M., Alpert, R., et al. (1996). Effects
REFERENCES 565
of crack cocaine on neurocognitive function. and norms. In J. A. Knight (Ed.), The handbook
Psychiatry Research, 60(2-3), 167-176. of Rey-Osterrieth Complex Figure usage: Clini-
Hoffman, D. T. (1969). Sex differences in preferred cal and research applications. Lutz, FL: Psy-
finger tapping rates. Perceptual and Motor Skills, chological Assessment Resources.
29,676. Hubley, A. M., & Tremblay, D. (2002). Compara-
Hogervorst, E., Combrinck, M., Lapuerta, P., bility of total score performance on the Rey-
Rue, J., Swales, K., & Budge, M. (2001). The Osterrieth Complex Figure and a modified
Hopkins Verbal Learning Test and screening for Taylor Complex Figure. Journal of Clinical and
dementia. Dementia and Geriatric Cognitive Experimental Neuropsychology, 24(3), 370-382.
Disorders, 13(1), 13-20. Hubley, A.M., Tombaugh, T. N., & Hemingway,
Holdwick, D. J., Jr., & Wingenfeld, S. A. (1999). D. (2003). A modification of the Taylor Figure
The subjective experience of PASAT testing: and the development of new figures for older
Does the PASAT induce negative mood? Ar- adults. In J. A. Knight (Ed.), The handbook of
chives of Clinical Neuropsychology, 14(3), Rey-Osterrieth Complex Figure usage: Clinical
273-284. and research applications. Lutz, FL: Psycholog-
Hom, J. (2003). Forensic neuropsychology: Are we ical Assessment Resources.
there yet? Archives of Clinical Neuropsychology, Huff, F. J., Collins, C., Corkin, S., & Rosen, T. J.
18, 827-845. (1986a). Equivalent forms of the Boston Naming
Hom, J., & Reitan, R. M. (1990). Generalized Test. Journal of Clinical and Experimental
cognitive function after stroke. Journal of Clin- Neuropsychology, 8(5), 556-562.
ical and Experimental Neuropsychology, 12, Huff, F. J., Corkin, S., & Growdon, J. H. (1986b).
644-654. Semantic impairment and anomia in Alzheimer's
Honn, V. J., Para, M. F., Whitacre, C. C., & disease. Brain and Language, 28, 235-249.
Bomstein, R. A. (1999). Effect of exercise on Hugdahl, K., & Franzon, M. (1985). Visual half-field
neuropsychological performance in asymptom- presentations of incongruent color-words reveal
atic HIV infection. AlDS and Behavior, 3(1), mirror-reversal of language lateralization in dex-
67-74. tral and sinistral subjects. Cortex, 21, 359-374.
Hooper, H. E. (1958, 1983, 1997). Hooper Visual Hughes, D. L., & Bryan, J. (2002). Adult age dif-
Organization Test (VOT). Los Angeles: Western ferences in strategy use during verbal fluency
Psychological Services. performance. Journal of Clinical and Experi-
Homer, M.D., Flashman, L.A., Freides, D., Ep- mental Neuropsychology, 24(5), 642-654.
stein, C. M., & Bakay, R. A. E. (1996). Temporal Huhtaniemi, P., Haier, R. J., Fedio, P., & Buchs-
lobe epilepsy and performance on the Wisconsin baum, M. S. (1983). Neuropsychological char-
Card Sorting Test. Journal of Clinical and Ex- acteristic of college males who show attention
perimental Neuropsychology, 18, 310-313. dysfunction. Perceptual and Motor Skills, 57,
Horton, A.M., & Roberts, C. (2003). Demographic 399-406.
effects on the Trail Making Test in a drug abuse Hulicka, I. M. (1966). Age differences in Wechsler
treatment sample. Archives of Clinical Neu- Memory Scale scores. Journal of Genetic Psy-
ropsychology, 18(1), 49-56. chology, 190, 135-145.
Houx, P. J., Jolles, J., & Vreeling, F. (1993). Stroop Hultsch, D. F., Hammer, M., & Small, B. J. (1993).
interference: Aging effects assessed with the Age differences in cognitive performance in later
Stroop Color-Word Test. Experimental Aging life: Relationships to self-reported health and
Research, 19, 209-224. activity life style. Journals of Gerontology, 48(1),
Hsieh, S., & Riley, N. (1997, November). Neu- Pl-Pll.
ropsychological peiformtlnce in the People's Re- Hunter, J. E., Schmidt, F. L., & Jackson, G. B.
public of China: Age and educational nonns for (1982). Meta-analysis: Cumulating research find-
four attention tasks. Paper presented at the ings across studies. Beverly Hills: Sage.
National Academy of Neuropsychology, Las Iijima, M., Osawa, M., Iwata, M., Miyazaki, A., &
Vegas, NV. Tei, H. (2000). Topographic mapping of P300
Hsieh, S., Lee, C. Y., & Tai, C. T. (1995). Set- and frontal cognitive function in Parkinson's
shifting aptitude in Parkinson's disease: External disease. Behavioural Neurology, 12(3), 143-148.
versus internal cues. Psychological Reports, 77, Ilonen, T., Taiminen, T., Lauerma, H., Karlsson, H.,
339--349. Helenius, H. Y. M., Tuimala, P., et al. (2000).
Hubley, A. M., & Tombaugh, T. N. (2003). Taylor Impaired Wisconsin Card Sorting Test perfor-
Complex Figure: Comparability to the ROCF mance in first-episode schizophrenia: Resource or
566 REFERENCES
motivation deficit? Comprehensive Psychiatry, Iverson, G. L., Franzen, M. D., & Lovell, M. R.
41(5), 385-391. (1999). Normative comparisons for the Con-
Ingraham, L. J., & Aiken, C. B. (1996). An empir- trolled Oral Word Association Test following
ical approach to determining criteria for abnor- acute traumatic brain injury. Clinical Neuro-
mality in test batteries with multiple measures. psychologist, 13(4), 437-441.
Neuropsychology, 10(1), 120-124. Iverson, G. L., Woodward, T. S., & Smith-
Ingraham, L. J., Chard, F., Wood, M., & Mirsky, Seemiller, L. (2000). Internal consistency and
A. F. (1988). A Hebrew language version of the concurrent validity of two short forms of the
Stroop test. Perceptual and Motor Skills, 67(1), Visual Form Discrimination Test. Applied Neu-
187-192. ropsychology, 7, 108-110.
Ingram, F., Greve, K. W., Ingram, P., & Soukup, V. Iverson, G. L., Lange, R. T., Green, P., & Franzen, M.
M. (1999). Temporal stability of the Wisconsin (2002). Detecting exaggeration and malingering
Card Sorting Test in an untreated patient sam- with the Trail Making Test. Clinical Neuropsy-
ple. British Journal of Clinical Psychology, 38, chologist, 16(3), 398-406.
209-211. lvinskis, A., Allen, S., & Shaw, E. (1971). An ex-
Inman, V. W., & Parkinson, S. R. (1983). Differ- tension of Wechsler Memory Scale norms to
ences in Brown-Peterson recall as a function of lower age groups. Journal of Clinical Psychology,
age and retention interval. Journal of Gerontol- 27, 354--357.
ogy, 38, 58-64. lvison, D. (1977). The Wechsler Memory Scale:
Insel, T. R., Donnelly, E. F., Lalakea, M. L., Preliminary findings toward an Australian stan-
Alterman, I. S., & Murphy, D. L. (1983). Neu- dardization. Australian Psychologist, 1.2, 303-
rological and neuropsychological studies of 312.
patients with obsessive-compulsive disorder. lvison, D. (1986). Anna Thompson and the Amer-
Biological Psychiatry, 18(7), 741-751. ican Liner New York: Some normative data.
Isaacs, B., & Kennie, A. T. (1973). The Set Test as Journal of Clinical and Experimental Neuropsy-
an aid to the detection of dementia in old people. chology, 8(3), 317-320.
British Journal of Psychiatry, 123, 467-470. lvison, D. (1993). Logical memory in the Wechsler
Ishikawa, S. S., Raine, A., Lencz, T., Bihrle, S., & Memory Scales: Does the order of passages af-
Lacasse, L. (2001). Autonomic stress reactivity fect difficulty in an university sample? Clinical
and executive functions in successful and un- Neuropsychologist, 7(2), 215-218.
successful criminal psychopaths from the com- lvnik, R. J., Sharbrough, F. W., & Laws, E. R.
munity. Journal of Abnon11al Psychology, 110(3), (1987). Effects of anterior temporal lobectomy
423-432. on cognitive function. Journal of Clinical Psy-
lsingrini, M., & Vazou, F. (1997). Relation between chology, 43, 128-137.
fluid intelligence and frontal lobe functioning in lvnik, R. J., Malec, J. F., Tangalos, E. G.,
older adults. International Journal of Aging and Petersen, R. C., Kokmen, E., & Kurland, L. T.
Human Development, 45(2), 99-109. (1990). The Auditory-Verbal Learning Test
Ismail, B., Cantor-Graae, E., & McNeil, T. F. (AVLT): Norms for ages 55 and older. Psycho-
(2000). Minor physical anomalies in schizo- logical Assessment: A Journal of Consulting and
phrenia: Cognitive, neurological and other clin- Clinical Psychology, 2, 304-312.
ical correlates. Journal of Psychiatric Research, lvnik, R. J., Smith, G., Tangalos, E., Petersen, R.,
34(1), 45-56. Kokmen, E., & Kurland, L. (1991). Wechsler
Iverson, G. L. (2001). Can malingering be identi- Memory Scale: IQ-dependent norms for persons
fied with the Judgment of Line Orientation Test? ages 65 to 97 years. Psychological Assessment,
Applied Neuropsychology. 8(3), 167-173. 3(2), 156-161.
Iverson, G. L., Sherman, E. M. S., & Smith- lvnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. G.,
Seemiller, L. ( 1997a). Evaluation of a short form of Petersen, R. C., Kokmen, E., et al. (1992a).
the Visual Form Discrimination Test for assessing Mayo's Older Americans Normative Studies:
cognitive decline associated with dementia. Jour- WAIS-R norms for ages 56 to 97. Clinical Neu-
nal of Cognitive Rehabilitation, 15, 20-21. ropsychologist, 6(Suppl.), 1-30.
Iverson, G. L., Slick, D., & Smith-Seemiller, L. Ivnik, R., Malec, J., Smith, G., Tangalos, E.,
(1997b). Screening for visual-perceptual deficits Petersen, R., Kokman, E., et al. (1992b). Mayo's
following closed head injury: A short form of the Older Americans Normative Studies: WMS-R
Visual Form Discrimination Test. Brain Injury, norms for ages 56 to 94. Clinical Neuropsy-
11, 125-128. chologist, 6(Suppl.), 49-82.
REFERENCES 567
Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. Janowsky, J. S., Shimamura, A. P., Kritchevsky, M.,
G., Petersen, R. C., Kokmen, E., et al. (1992c). & Squire, L. R. (1989). Cognitive impairment
Mayo's Older Americans Normative Studies: following frontal lobe damage and its relevance
Updated AVLT norms for ages 56 to 97. Clinical to human amnesia. Behavioral Neuroscience,
Neuropsychologist, 6, 83--104. 103, 548--560.
lvnik, R., Smith, G., Malec, J., Tangalos, E., & Jarvis, P. E., & Barth, J. T. (1984). Holstead-Reitan
Parisi, J. (1993). Comparison of Wechsler vs. Test Battery: An interpretive guide. Odessa, FL:
Mayo summary scores in a clinical sample. Psychological Assessment Resources.
Journal of Clinical Psychology, 49(4), 534-542. Javorsky, D., & Stem, R. A. (1999). Validity of
Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. the Boston Qualitative Scoring System (BQSS)
G., & Petersen, R. C. (1996). Neuropsychologi- for the Rey-Osterrieth Complex Figure in dis-
cal tests' norms above age 55: COWAT, BNT, criminating between Alzheimer's and vascular
MAE Token, WRAT-R Reading, AMNART, dementia. Journal of the International Neu-
STROOP, TMT, and JLO. Clinical Neuropsy- ropsychological Society, 5, 120.
chologist, 10(3), 262-278. Jenkins, R. L., & Parsons, 0. A. ( 1978). Cognitive
lvnik, R., Smith, G., Lucas, J., Tangalos, E., Kok- deficits in male alcoholics as measured by a
men, E., & Petersen, R. (1997). Free and cued modified Wisconsin Card Sorting Test. Alcohol
selective reminding test: MOANS norms. Jour- Technical Reports, 7, 76--83.
nal of Clinical and Experimental Neuropsychol- Jenkins, R. L., & Parsons, 0. A. (1981). Neu-
ogy, 19(5), 676--691. ropsychological effect of chronic alcoholism on
lvnik, R., Smith, G., Petersen, R., Boeve, B., Kok- tactual-spatial performance and memory in
men, E., & Tangalos, E. G. (2000). Diagnostic males. Alcoholism: Clinical and Experimental
accuracy of four approaches to interpreting Research, 5(1), 26--33.
neuropsychological test data. Neuropsychology, Jenkins, R. L., & Parsons, 0. A. (1989). Hemi-
14, 163--177. spheric asymmetry in the processing of tactual-
lvnik, R. J., Smith, G. E., Cerhan, J. H., Boeve, spatial material of low verbal codability in
B. F., Tangalos, E. G., & Petersen, R. C. (2001). normal subjects. Archives of Clinical Neuropsy-
Understanding the diagnostic capabilities of chology, 4(4), 311-321.
cognitive tests. Clinical Neuropsychologist, 15(1), Jensen, A. (1965). Scoring the Stroop Test. Acta
114-124. Psychologica, 24, 398-408.
Jackson, S. T., & Tompkins, C. A. (1991). Supple- Jensen, A., & Rohwer, W. (1966). The Stroop
mental aphasia tests: Frequency of use and Color-Word Test: A review. Acta Psychologica,
psychometric properties. Clinical Aphasiology, 25,36-93.
20, 91-99. Jeste, D. V., Harris, M. J., Krull, A., Kuck, J.,
Jacobs, D. M., Sano, M., Dooneief, G., Marder, K., McAdams, L. A., & Heaton, R. (1995). Clinical
Bell, K., & Stem, Y. (1995). Neuropsychological and neuropsychological characteristics of pa-
detection and characterization of preclinical tients with late-onset schizophrenia. American
Alzheimer's disease. Neurology, 45, 957-962. Journal of Psychiatry, 152(5), 722-730.
Jacobs, D. M., Sano, M., Albert, S., Schofield, P., Johnson, D. A., Roethig-Johnson, K., & Middleton,
Dooneief, G., & Stem, Y. (1997). Cross-cultural J. (1988). Development and evaluation of an at-
neuropsychological assessment: A comparison of tentional test for processing capacity in a normal
randomly selected, demographically matched sample. Journal of Child Psychology and Psy-
cohorts of English- and Spanish-speaking older chiatry, 2, 199-208.
adults. Journal of Clinical and Experimental Neu- Johnson, S. C., Saykin, A. J., Flashman, L. A.,
ropsychology, 19(3), 331-339. McAllister, T. W., & Sparling, M. B. (2001).
Jacqmin-Gadda, H., Fabrigoule, C., Commenges, Brain activation on fMRI and verbal memory
D., Letenneur, L., & Dartigue, J. F. (2000). A ability: Functional neuroanatomic correlates of
cognitive screening battery for dementia in the CVLT performance. Journal of the International
elderly. Journal of Clinical Epidemiology, 53(10), Neuropsychological Society, 7(1), 55-62.
980--987. Johnson, S. K., DeLuca, J., Diamond, B. J., &
Janowsky, J. S., & Thomas-Thrapp, L. J. (1993). Natelson, B. H. (1996). Selective impairment of
Complex figure recall in the elderly: A deficit in auditory processing in chronic fatigue syndrome:
memory or constructional strategy? Journal of A comparison with multiple sclerosis and healthy
Clinical and Experimental Neuropsychology, controls. Perceptual and Motor Skills, 83(1),
15(2), 159-169. 51--62.
568 REFERENCES
Johnson-Greene, D., Adams, K. M., Gilman, S., & Kalechstein, A. D., van Gorp, W. G., & Rapport,
Junek, L. (2002). Relationship between neu- L. J. (1998). Variability in clinical classification of
ropsychological and emotional functioning in raw test scores across normative data sets.
severe chronic alcoholism. Clinical Neuropsy- Clinical Neuropsychologist, 12(3), 339--347.
chologist, 16(3), 300--309. Kalinowski, A. G., Weinstein, C. S., & Seidman,
Johnson-Selfridge, M., Zalewski, C., & Aboudar- L. J. (2003). Organizational and retrieval deficits
ham, J. (1998). The relationship between eth- on the ROCF in schizophrenia. In J. A. Knight
nicity and word fluency. Archives of Clinical (Ed.), The handbook of Rey-Osterrieth Complex
Neuropsychology, 13(3), 319--325. Figure usage: Clinical and research applications.
Johnstone, B., & Wilhelm, K. L. (1997). The con- Lutz, FL: Psychological Assessment Resources.
struct validity of the Hooper Visual Organization Kaltreider, L. B., Cicerello, A. R., Lacritz, L. H.,
Test. Assessment, 4(3), 243-248. Weiner, M. F., Honig, L. S., Rosenberg, R.N.,
Johnstone, B., Holland, D., & Hewett, J. E. (1997). et al. (2000). Comparison of the Cerad and
The construct validity of the Category Test: Is it CVLT list-learning tasks in Alzheimer's disease.
a measure of reasoning or intelligence? Psycho- Clinical Neuropsychologist, 14(3), 269-274.
logical Assessment, 9(1), 28--33. Kanaya, T., Scullin, M. H., & Ceci, S. J. (2003). The
Jones, B., Teng, E., Folstein, M., & Harrison, K. Flynn effect and U.S. policies: The impact of
(1993). A new bedside test of cognition for pa- rising IQ scores on american society via mental
tients with HIV infection. Annals of Internal retardation diagnoses. American Psychologist,
Medicine, 119, 1001-1004. 58(10), 778-790.
Jones, B. P., Mirsky, A., & Duncan, C. C. (2003). Kane, R. L., Parsons, 0. A., & Goldstein, G. (1985).
ROCF performance, attention disorders, and Statistical relationships and discriminative accu-
neuropsychiatric disorders. In J. A. Knight (Ed.), racy of the Halstead-Reitan, Luria-Nebraska,
The handbook of Rey-Osterrieth Complex Figure and Wechsler IQ scores in the identification of
usage: Clinical and research applications. Lutz, brain damage. Journal of Clinical and &peri-
FL: Psychological Assessment Resources. mental Neuropsychology, 7(3), 211-223.
Jones-Cotman, M. (1991a). Localization of lesions by Kang, S. K. (2000). TheapplicabilityofWHO-NCTB
neuropsychological testing. Epilepsia, 32(Suppl. 5), in Korea. Neurotoxicology, 21(5), 697-701.
S41-S52. Kanter, G. (1984). PASAT performance and intel-
Jones-Cotman, M. (199lb). Presurgical psycholog- ligence: A relationship? The International Jour-
ical assessment in children: Special tests. Journal nal of Cltnical Neuropsychology, 6, 84.
of Epilepsy, 3(Suppl. 1), 93-102. Kaplan, E. (1988). A process approach to neu-
Jones-Cotman, M., & Milner, B. (1977). Design ropsychological assessment. In T. Boll & B. K.
fluency: The invention of nonsense drawings Bryant (Eds.), Clinical neuropsychology and
after focal cortical lesions. Neuropsychologia, brain fonction: Research, measurement and prac-
15(Suppl. 5), 653-674. tice (Vol. 7). Washington, DC: American Psy-
Joyce, E., Blumenthal, S., & Wessely, S. (1996). chological Association.
Memory, attention, and executive function Kaplan, E., Goodglass, H., & Weintraub, S. (1978).
in chronic fatigue syndrome. Journal of Neurol- The Boston Naming Test. Experimental edition.
ogy. Neurosurgery, and Psychiatry, 60(5), Boston: Kaplan and Goodglass.
495--503. Kaplan, E., Goodglass, H., & Weintraub, S. (1983).
Judd, P. H., & Ruff, R. M. (1993). Neuropsycho- The Boston Naming Test. Philadelphia: Lea and
logical dysfunction in borderline personality Febiger.
disorder. Journal of Personality Disorders, 7(4), Kaplan, E., Fein, D., Morris, R., & Delis, D. C.
275-284. (1991). WAIS-R as a neuropsychological instru-
Jung, R. E., Yeo, R. A., Chiulli, S. J., Sibbitt, W. L., ment: WAIS-R-Nl manual. New York: Psycho-
Jr., Weers, D. C., Hart, B. L., et al. (1999). logical Corporation.
Biochemical markers of cognition: A proton MR Kaplan, E., Goodglass, H., & Weintraub, S. (2000).
spectroscopy study of normal human brain. Boston Naming Test, Second edition. Philadel-
Neuroreporll0(16), 3327-3331. phia: Lippincott Williams & Wilkins.
Kaasa, S., Olsnes, B. T., & Mastekaasa, A. (1988). Kareken, D. A., Moberg, P. J., & Gur, R. C. (1996).
Neuropsychological evaluation of patients with Proactive inhibition and semantic organization:
inoperable non-small cell lung cancer treated Relationship with verbal memory in patients
with combination chemotherapy and radiother- with schizophrenia. Journal of the International
apy. Acta Oncologica, 27(3), 241-246. Neuropsychological Society, 2(6), 486-493.
REFERENCES 569
Kasahara, H., Tanno, M., Yamada, H., Endoh, K., Kellner, C. H., Rubinow, D. R., & Post, R. M.
Kobayashi, M., et al. (1993). MRI study of the (1986). Cerebral ventricular size and cognitive
brain in aged volunteers: T 2 high signal intensity impairment in depression. Journal of Affective
lesions and high cortical function. Nippon Ronen Disorders, 10(3), 215-219.
Igakkai Zasshi, 30(10), 892-900. Kelly, M. D., Kundert, D. K., & Dean, R. S. (1992).
Kasahara, H., Yamada, H., Tanno, M., Kobayashi, M., Factor analysis and matrix invariance of the
Karasawa, A., Endo, K., et al. (1995). Magnetic HRNB-C Category Test. Archives of Clinical
resonance imaging study of the brain in aged vol- Neuropsychology, 7, 411>-418.
unteers: T 2 high intensity lesions and higher order Kempen, J. H., Kritchevsky, M., & Feldman, S. T.
cortical function. European Archives ofPsychiatry (1994). Effect of visual impairment on neu-
and Clinical Neuroscience, 49(5-6), 273-279. ropsychological test performance. Journal of
Kaskie, B., & Storandt, M. (1995). Visuospatial Clinical and Experimental Neuropsychology,
deficit in dementia of the Alzheimer type. Ar- 16(2), 223-231.
chives of Neurology, 52, 422--425. Kempler, D., Teng, E. L., Dick, M., Taussig, I. M.,
Katz, L. J., Wood, D. S., Goldstein, G., Auchenbach, & Davis, D. S. (1998). The effects of age, edu-
R. C., & Geckle, M. (1998). The utility of neu- cation, and ethnicity on verbal Ruency. Journal
ropsychological tests in evaluation of attention- of the International Neuropsychological Society,
deficit/hyperactivity disorder (ADHD) versus 4(6), 531-538.
depression in adults. Assessment, 5(1), 45-51. Kennedy, K. J. (1981). Age effects on Trail Making
Kawas, C. H., Corrada, M. M., Brookmeyer, R., Test performance. Perceptual and Motor Skills,
Morrison, A., Resnick, S. M., Zonderman, A. B., 52(2), 671-675.
et al. (2003). Visual memory predicts Alzheimer's Kibby, M. Y., Schmitter-Edgecombe, M., & Long,
disease more than a decade before diagnosis. C. J. (1998). Ecological validity of neu-
Neurology, 60(7), 1089-1093. ropsychological tests: Focus on the California
Kawasaki, Y., Maeda, Y., Suzuki, M., Urata, K., Hi- Verbal Learning Test and the Wisconsin Card
gashima, M., Kiba, K., etal. (1993). SPECTanalysis Sorting Test. Archives of Clinical Neuropsy-
of regional cerebral blood Row changes in patients chology, 13(6), 523-534.
with schizophrenia during Wisconsin Card Sorting Kilander, L., Nyman, H., Boberg, M., & Lithell, H.
Tests. Schizophrenia Research, 10, 109-116. (2000). The association between low diastolic
Kay, G. G. (2002). Guidelines for the psychological blood pressure in middle age and cognitive
evaluation of air crew personnel. Occupational function in old age. A population-based study.
Medicine, 17(2), 227-245. Age and Ageing, 29(3), 243-248.
Kear-Colwell, J. J., & Heller, M. (1978). A nor- Killgore, W. D. S., & Adams, R. L. (1999). Pre-
mative study of the Wechsler Memory Scale. diction of Boston Naming Test performance
Journal of Clinical Psychology, 34(2), 437-442. from vocabulary scores: Preliminary guidelines
Keenan, P. A., Ricker, J. H., Lindamer, L. A., for interpretation. Perceptual and Motor Skills,
Jiron, C. C., & Jacobson, M. W. (1996). Rela- 89(1), 327-337.
tionship between WAIS-R Vocabulary and Per- Kilpatrick, D. G. (1970). The Halstead Category
formance on the California Verbal Learning Test. Test of brain dysfunction: Feasibility of a short
Clinical Neuropsychologist, 10(4), 455-458. form. Perceptual and Motor Skills, 30, 577-578.
Kehrer, C. A., Sanchez,P. N.,Habif, U.,Rosenbaum,J. Kim, H., & Na, D. L. (1999). Normative data on the
G., & Townes, B. D. (2000). Effects of a significant- Korean version of the Boston Naming Test.
other observer on neuropsychological test perfor- Journal of Clinical and Experimental Neuropsy-
mance. Clinical Neuropsychologist, 14(1), 67-71. cholor!J, 21(1), 127-133.
Kelland, D. Z., & Lewis, R. F. (1994). Evaluation of Kim, J. K., & Kang, Y. (1999). Nonnative study of
the reliability and validity of the Repeatable the Korean-California Verbal Learning Test
Cognitive-Perceptual-Motor Battery. Clinical (K-CVLT). Clinical Neuropsychologist, 13(3),
Neuropsychologist, 8(3), 295-308. 365-369.
Kelland, D. Z., & Lewis, R. F. (1996). The Digit Kimbarow, M. L., Vangel, S. J., Jr., & Lichtenberg. P. A.
Vigilance Test: Reliability, validity, and sensitiv- (1996). The influence of demographic variables on
ity to diazepam. Archives of Clinical Neuropsy- normal elderly subjects' performance on the
chology, 11(4), 339-344. Boston Naming Test. Clinical Aphasiology, 24,
Keller, F. R., & Davis, H. P. (1998). Colorado as- 135-144.
sessment tests (version 1.0) [Computer software]. Kimura, S. D. (1981). A card form of the Reitan-
Colorado Springs: Colorado Assessment Tests. modified Halstead Category Test. Journal of
570 REFERENCES
Consulting and Clinical Psychology, 49(1), Klein, M., Ponds, R. W. H. M., Houx, P. J., &
1~146. Jolles, J. (1997). Effect of test duration on
Kindennann, S. S., Kalayam, B., Brown, G. G., age-related differences in Stroop interference.
Burdick, K. E., & Alexopoulos, G. S. (2000). Journal of Clinical and Experimental Neuropsy-
Executive functions and P300 latency in elderly chology, 19, 77-8.2.
depressed patients and control subjects. Ameri- Klicpera, C. (1983). Poor planning as a character-
can Journal of Geriatric Psychiatry, 8(1), 57-65. istic of problem-solving behavior in dyslexic
King, G. D., Hannay, H. J., Masek, B. J., &: Burns, children: A study with the Rey-Osterrieth Com-
J. W. (1978). Effects of anxiety and sex on neu- plex Figure test. Acta Paedopsychiatrica, 49,
ropsychological tests. Journal of Consulting and 7~2.
Clinical Psychology, 46(2), 375-376. · Klimczak, N. J., Donovick, P. J., & Burright, R.
King, J. H., Gfeller, J. D., & Davis, H. P. (1998). (1997). The malingering of multiple sclerosis and
Detecting simulated memory impairment with mild traumatic brain injury. Brain Injury, 11(5),
the Rey Auditory Verbal Learning Test: Impli- 343-35.2.
cations of base rates and study generalizability. Klonoff, H., & Kennedy, M. (1965). Memory and
Journal of Clinical and Experimental NtiUropsy- perceptual functioning in octogenarians and
chology, 20(5), 603-612. nonagenarians in the community. Journal of
King, M. C. (1981). Effects of non-focal brain Gerontology, 20, 328--333.
dysfunction on visual memory. Journal of Clini- Klonoff, H., & Kennedy, M. (1966). A comparative
cal Psychology, 37(3), 638-643. study of cognitive functioning in old age. Journal
Kirk, U. (1992a). Confrontation naming in normally of Gerontology, 21, 239--243.
developing children: Word-retrieval or word Klove, H. (1974). Validation studies in adult clinical
knowledge? Clinical Neuropsychologist, 6(2), neuropsychology. In R. M. Reitan & L. A. Da-
156-170. vison (Eds.), Clinical Neuropsychology: Current
Kirk, U. (1992b). Evidence for early acquisition Status and Applications (pp. 211-236). Wa-
of visual organization ability: A developmental shington, DC: Winston.
study. Clinical Neuropsychologist, 6(2), 171-177. Klusman, L. E., Cripe, L.l., & Dodrill, C. B. (1989).
Kirk, U., & Kelly, M.S. (1986). Scoringscaleforthe Analysis of errors on the Trail Making Test. Per-
Rey-Osterrieth Complex Figure. Paper pre- ceptual and Motor SkiUs, 68, 1199-1204.
sented at the meeting of the Internatio~ Neu- Knesevich, J. W., LaBarge, E., & Edwards, D.
ropsychological Society, Denver, CO. (1986). Predictive value of the Boston Naming
Kirshner, H. S., Webb, W. G., & Kelly, M. P. Test in mild senile dementia of the Alzheimer
(1984). The naming disorder of dementJa. Neu- type. Psychiatry Research, 19, 155-161.
ropsychologia, 2.2, 23--30. Knight, J. A. (2003). The handbook of Rey-
Kishi, R., Harabuchi, 1., Katakura, Y., Ikeda, T., & Osterrieth Complex Figure usage: Clinical and
Miyake, H. (1993). Neurobehavioral effects of research applications: Lutz, FL: Psychological
chronic occupational exposure to organic sol- Assessment Resources.
vents among Japanese industrial painters. Envi- Knight, J. A., Kaplan, E., & Ireland, L. (2003).
ronmental Research, 62(2), 303-313. Survey findings of Rey-Osterrieth Complex
Kivircik, B. B., Yener, G. G., Alptekin, K., & A)'din, H. Figure usage. In J. A. Knight (Ed.), The hand-
(2003). Event-related potentials and neuro- book of Rey-Osterrieth Complex Figure usage:
psychological tests in obsessive-compulsive dis- Clinical and research applications. Lutz, FL: Psy-
order. Progress In Neuro-Psychophormacology chological Assessment Resources.
and Biological Psychiatry, 27(4), 601-606. Knights, R. M., & Moule, A. D. (1967). Nonnative
Kivling-Boden, G., & Sundbom, E. (2003). Cogni- and reliability data on finger and foot tapping
tive abilities related to post-traumatic syinptoms in children. Perceptual and Motor Skills, 25,
among refugees from the fonner Yugoslavia in 717-720.
psychiatric treatment. Nordic Journal of Psychi- Koffier, S. P., & Zehler, D. (1985). Nonnative data
atry, 57(3), 191-198. for the hand dynamometer. Perceptual and Mo-
Kixmiller, J. S., Verfaellie, M., Mather, M. M., & tor Skills, 61, 589-590.
Cennak, L. S. (.2000). Role of perceptual and Kohn, S. E., & Goodglass, H. (1985). Picture-
organizational factors in amnesics' recall of the naming in aphasia. Brain and Language, 24(2),
Rey-Osterrieth Complex Figure: A cont>arison 266--283.
of three amnesic groups. Journal of Clinical and Kohnert, K. J., Hernandez, A. E., & Bates, E.
Experimental Neuropsychology, 22(2), 11}8-207. (1998). Bilingual performance on the Boston
REFERENCES 571
Naming Test: Preliminary norms in Spanish and Kujala, P., Portin, R., Revonsuo, A., & Ruutiainen,
English. Brain and Lan(YJage, 65(3), 422-440. J. (1995). Attention related performance in two
Kongs, S. K., Thompson, L. L., Iverson, G. L., & cognitively different subgroups of patients with
Heaton, R. K. (2000). Wisconsin Card Sorting multiple sclerosis. journal of Neurology, Neuro-
Test-64 Card Version. Lutz, FL: Psychological surgery, and Psychiatry, 59(1), 77-82.
Assessment Resources. Kulik, J. (1983). Review of G. V. Glass et al., Meta-
Konishi, S., Nakajima, K., Uchida, 1., Kameyama, M., analysis in social research [Book review].
Nakahara, K., Sekihara, K., et al. (1998). Tran- Evaluation News, 4, 101-105.
sient activation of inferior prefrontal cortex Kumar, P., Gupta, B. N., Pandya, K. P., & Clerk, S.
during cognitive set shilling. National Neurosci- H. (1988). Behavioral studies in petrol pump
ence, 1, 80--84. workers. International Archives of Occupational
Koren, D., Seidman, L. J., Harrison, R. H., and Environmental Health, 61(1-2), 35--38.
Lyons, M. J., Kremem, W. S., Caplan, B., et al. Kupke, T. (1983). Effect of subject sex, examiner
(1998). Factor structure of the Wisconsin Card sex, and test apparatus on Halstead Category and
Sorting Test: Dimensions of deficit in schizo- Tactual Performance tests. Journal of Consulting
phrenia. Neuropsychology, 12(2), 289-302. and Clinical Psychology, 51(4), 624--626.
Kortte, K. B., Horner, M. D., & Windham, W. K. Kurylo, M., Temple, R. 0., Elliott, T. R., &
(2002). The Trail Making Test, part B: Cognitive Crawford, D. (2001). Rey Auditory Verbal
flexibility or ability to maintain set? Applied Learning Test (AVLT) performance in individ-
Neuropsychology, 9(2), 106-109. uals with recent-onset spinal cord injury. Reha-
Koss, E., Ober, B. A., Delis, D. C., & Friedland, bilitation Psychology, 46(3), 247-261.
R. P. (1984). The Stroop Color-Word Test: In- Kuslansky, G., Katz, M., Verghese, J., Hall, C. B.,
dicator of dementia severity. International Lapuerta, P., LaRuffa, G., et al. (2004). Detect-
Journal of Neuroscience, 24, 53--61. ing dementia with the Hopkins Verbal Learning
Kozel, J., & Meyers, J. E. (1998). A cross-validation Test and the Mini-Mental State Examination.
study of the Victoria Revision of the Category Archives of Clinical Neuropsychology, 19(1),
Test. Archives of Clinical Neuropsychology, 89-104.
13(3), 327-332. Kuzis, G., Sabe, L., Tiberti, C., Merello, M., Lei-
Kozora, E., & Cullum, C. M. (1995). Generative guarda, R., & Starkstein, S. E. (1999). Explicit
naming in normal aging: Total output and qualita- and implicit learning in patients with Alzheimer
tive changes using phonemic and semantic con- disease and Parkinson disease with dementia.
straints. Clinical Neuropsychologist, 9(4), 313--320. Neuropsychiatry, Neuropsychology, and Behav-
Kramer, A. F., Humphrey, D. G., Larish, J. F., ioral Neurology, 12(4), 265-269.
Logan, G. D., & Strayer, D. L. (1994). Aging and Laatsch, L., & Choca, J. (1991). Understanding the
inhibition: Beyond a unitary view of inhibitory Halstead Category Test by using item analysis.
processing in attention. Psychology and Aging, Psychological Assessment: A Journal of Consul-
9(4), 491-512. ting and Clinical Psychology, 3, 701-704.
Kramer, J. H., Delis, D. C., & Daniel, M. H. Labarge, A. S., McCaffrey, R. J., & Brown, T. A.
(1988). Sex differences in verbal learning. Jour- (2003). Neuropsychologists' abilities to deter-
nal of Clinical Psychology, 44(6), 907-915. mine the predictive value of diagnostic tests.
Krebs, R. (1994). The Hopkins Verbal Learning Archives of Clinical Neuropsychology, 18(2),
Test: An alternative to the MMSE? Gerontolo- 165-175.
gist, 34(5), 692. LaBarge, E., Edwards, D., & Knesevich, J. W.
Kritz-Silverstein, D., & Barrett-Connor, E. (2002). (1986). Performance of normal elderly on the
Hysterectomy, oophorectomy, and cognitive Boston Naming Test. Brain and lAnguage, 27,
function in older women. Journal of the Ameri- 380--384.
can Geriatrics Society, 50(1), 55-61. LaBarge, E., Balota, D. A., Storandt, M., &
Krupp, L. B., Sliwinski, M., Masur, D. M., Fried- Smith, D. S. (1992). An analysis of confrontation
berg, F., & Coyle, P. K. (1994). Cognitive naming errors in senile dementia of the Alzhei-
functioning and depression in patients with mer type. Neuropsychology, 6(1), 77-95.
chronic fatigue syndrome and multiple sclerosis. Labreche, T. M. (1983). The Victoria Revision of
Archives of Neurology, 51(7), 7~710. the Halstead Category Test. Unpublished doc-
Kuehn, S. M., & Snow, W. G. (1992). Are the Rey toral dissertation, University of Victoria, Canada.
and Taylor figures equivalent? Archives of Clin- Lacritz, L. H., & Cullum, C. M. (1998). The
ical Neuropsychology, 7, 445-448. Hopkins Verbal Learning Test and CVLT:
572 REFERENCES
A preliminary comparison. Archives of Clinical Larrabee, G. J. (2003). Detection of malingering

Neuropsychology, 13(1), 623--628. . using atypical performance patterns on standard
Lacritz, L. H, & Cullum, M. (2003). The WAIS-III neuropsychological tests. Clinical Neuropsy-
and WMS-III: Practical issues and frequently chologist, 17(3), 410-425.
asked questions. In D. Tulsky, D. Sakl~fske, G. Larrabee, G. J., & Curtiss, G. (1995). Construct
Chelune, R. Heaton, R. Ivnik, R. Bornstein, et al. validity of various verbal and visual memory
W
(Eds.), Clinical interpretation of the AlS-lil tests. Journal of Clinical and Experimental
and WMS-111. San Diego: Academic Press. Neuropsychology, 17(4), 536-547.
Lacritz, L. H., Cullum, C. M., Frol, A. B., Larrabee, G. J., & Levin, H. S. (1986). Memory
Dewey, R. B., Jr., & Giller, C. A. (2o00). Neu- self-ratings and objective test performance in a
ropsychological outcome following unilateral ste- normal elderly sample. Journal of Clinical and
reotactic pallidotomy in intractable P~ldnson's Experimental Neuropsychology, 8(3), 27~284.
disease [Special Issue: Neurobehavioral4Bsues in Larrabee, G. J., Largen, J. W., & Levin, H. S.
the neurosurgical treatment of move111e1Jt disor- (1985). Sensitivity of age-decline resistant
ders, Part II: Pallidotomy and paUidal ;timula- ("hold") WAIS subtests to Alzheimer's disease.
tion]. Brain and Cognition, 42(3), 364-3T8. Journal of Clinical and Experimental Neuropsy-
Lacritz, L. H., Cullum, C. M., Weiner, ~· F., & chology, 7(5), 497-504.
Rosenberg, R. N. (2001). Compa1son of Larrabee, G. J., Levin, H. S., & High, W. M.
the Hopkins Verbal Learning Test-Re~ to the (1986). Senescent forgetfulness: A quantitative
California Verbal Learning Test in ~eimer's study. Developmental Neuropsychology, 2(4),
disease. Applied Neuropsychology, 8(3), 180-184. 373--385.
Lacy, M. A., Gore, P. A., Jr., Pliskin, IN. H., Larrabee, G. J., Trahan, D. E., Curtiss, G., & Le-
Henry, G. K., Heilbronner, R. L., & Hamf::r, D.P. vin, H. S. (1988). Normative data for the Verbal
(1996). Verbal Huency task equivalence. 'Clinical Selective Reminding Test. Neuropsychology,
Neuropsychologist, 10(3), 305-308. . 2(3-4), 173-182.
LaHeche, G., & Albert, M. S. (1995). Ejecutive Larrabee, G. J., Trahan, D. E., & Levin, H. S.
function deficits in mild Alzheimer's idisease. (2000). Normative data for a six-trial adminis-
Neuropsychology, 9(3), 313--320. 1 tration of the Verbal Selective Reminding Test.
Laiacona, M., Inzaghi, M.G., De Tanti, A~ & Ca- Clinical Neuropsychologist, 14(1), ll0-118.
pitani, E. (2000). Wisconsin Card Sorting Test: A Larrain, C. M., & Cimino, C. R. (1998). Alternate
new global score, with Italian norms, i and its forms of the Boston Naming Test in Alzheimer's
relationship with the Weigl sorting test.! Neuro- disease. Clinical Neuropsychologist, 12(4),
logical Sciences, 21(5), 279-291. 5~.
Lamar, M., Zonderman, A. B., & Res{rlck, S. La Rue, A., D'Elia, L., Clark, E., Spar, J., & Jarvik,
(2002). Contribution of specific cogni~e pro- L. (1986). Clinical tests of memory in dementia,
cesses to executive functioning in an a~g pop- depression, and healthy aging. Psychology and
ulation. NeuropSfJchology, 16(2), 156-1~. Aging, 1(1), 69-77.
Lamberty, G. J., Putnam, S. H., Chatel, :D. M., LaRue, A., Romero, L., Ortiz, I., Liang, H. C., &
Bieliauskas, L. A., et al. (1994). Deriv~ Trail Lindeman, R. D. (1999). Neuropsychological
Making Test indices: A preliminary · report. performance of Hispanic and non-Hispanic
Neuropsychiatry, Neuropsychology, andl Behav- older adults: An epidemiologic survey. Clinical
ioral Neurology, 7(3), 230-234. Neuropsychologist, 13(4), 474--486.
Lannoo, E., & Vingerhoets, G. (1997). Flemish Lawton, M. P., & Brody, E. M., (1969). Assessment
normative data on common neuropsychplogical of older people: Self-maintaining and instru-
tests: InHuence of age, education, and gender. mental activities of daily living. Gerontologist, 9,
Psychologica Belgica, 37(3), 141-155. 179-188.
Lansdell, H., & Donnelly, E. F. (1977). · Factor Le Carret, N., Rainville, C., Lechevallier, N., La-
analysis of the Wechsler Adult Intelligence Scale font, S., Letenneur, L., & Fabrigoule, C. (2003).
subtests and the Halstead-Reitan Categpry and InHuence of education on the Benton Visual
Tapping tests. Journal of Consulting and f:linical Retention Test performance as mediated by a
Psychology, 45, 412--416. strategic search component. Brain and Cogni-
Lansing, A. E., Ivnik, R. J., Cullum, C., M., & tion, 53(2), 408-411.
Randolph, C. (1999). An empirically derived Leckliter, I. N., & Matarazzo, J. D. (1989). The in-
short form of the Boston Naming Test. t1rchives Huence of age, education, IQ, gender, and alcohol
of Clinical Neuropsychology, 14(6), 481~7. abuse on Halstead-Reitan Neuropsychological
REFERENCES 573
Test Battery performance. Journal of Clinical Halstead-Reitan Neuropsychological Battery.

Psychology, 45(4), 485-512. Clinical Neuropsychologist, 5, 83--88.
LeDorze, G., & Durocher, J. (1992). The effects of Levin, B. E., Uabre, M. M., Reisman, S., Weiner,
age, educational level, and stimulus length on W. J., Sahchez-Ramos, J., Singer, C., et al.
naming in normal subjects. Journal of Speech (1991). Visuospatial impairment in Parkinson's
Lan(!;Jage Pathology and Audiology, 16, 21-29. disease. Neurology, 41(3), 365-369.
Lee, G. P., Strauss, E., Loring, D. W., McCloskey, L., Levin, H. S., Benton, A. L., & Grossman, R. G.
et al. (1997). Sensitivity of figural fluency on the (1982). Neurobehavioml consequences of closed
Five-Point Test to focal neurological dysfunction. head injury. New York: Oxford University Press.
Clinical Neuropsychologist,11(1), 59-68. Levin, H. S., Mattis, S., Ruff, R. M., Eisenberg,
Lee, S.-H., & Lee, S. H. (1993). A study on the H. M., Marshal, L. F., Tabaddor, K., High, W.
neurobehavioral effects of occupational exposure M., & Frankowski, R. F. (1987). Neurobeha-
to organic solvents in Korean workers. Envi- vioral outcome following minor head injury: A
ronmental Research, 60, 227-232. three-center study. Journal of Neurosurgery, 66,
Lee, T. M. C., & Chan, C. C. H. (2000a). Are Trail 234-243.
Making and Color Trails tests of equivalent Levin, H. S., Song, J., Ewing-Cobbs, L., Chapman,
constructs? Journal of Clinical and Experimental S. B., & Mendelsohn, D. (2001). Word fluency in
Neuropsychology, 2.2(4), 529--534. relation to severity of closed head injury, asso-
Lee, T. M. C., & Chan, C. C. H. (2000b). Com- ciated frontal brain lesions, and age at injury in
parison of the Trail Making and Color Trails children. Neuropsychologia, 39(2), 122-131.
tests in a Chinese context: A preliminary report. Lewandowski, L., Kobus, D. A., Church, K. L., & Van
Perceptual and Motor Skills, 90(1), 187-190. Orden,K. (1982). Neuropsychological implications
Lee, T. M. C., Cheung, C. C. Y., Chan, J. K. P., & of hand preference versus hand grip performance.
Chan, C. C. H. (2000). Trail making across lan- Perceptual and Motor Skills, 55(1), 311-314.
guages. Journal of Clinical and Experimental Lewis, F. C., & Soares, L. (2000). Relationship
Neuropsychology, 22(6), 772-778. between semantic paraphasias and related non-
Lee, T. M. C., Yuen, K. S. L., & Chan, C. C. H. verbal factors. Perceptual and Motor Skills,
(2002). Normative data for neuropsychological 91(2), 366-372.
measures of fluency, attention, and memory Lewis, M. B., & Howdle, P. D. (2003). Cognitive
measures for Hong Kong Chinese. Journal of dysfunction and health-related quality of life
Clinical and Experimental Neuropsychology, in long-term liver transplant survivors. Liver
24(5), 615--632. Tronsplantotion and Surgery, 9(11), 1145-1148.
Lees, A. J. & Smith, E. (1983). Cognitive deficits in Lewis, R. (1995). Digit Vigilance Test: Professional
the early stages of Parkinson's disease. Broin, User's Guide. Odessa, FL: Psychological As-
106, 257-270. sessment Resources.
Lees-Haley, P., Smith, H., Williams, C., & Dunn, J. Lewis, R., & Rennick, P. (1979). Manual for the
(1996). Forensic neuropsychological test usage: Repeatable Cognitive-Perceptual-Motor Battery.
An empirical survey. Archives of Clinical Neu- Grosse Pointe Park, Ml: Axon.
ropsychology, 11(1), 45-51. Lewis, S., Campbell, A., Takushi-Chinen, R.,
Leggio, M. G., Silveri, M. C., Petrosini, L., & Brown, A., Dennis, G., Wood, D., et al. (1997).
Molinari, M. (2000). Phonological grouping is Visual Organization Test performance in an Af-
specifically affected in cerebellar patients: A rican American population with acute unilateral
verbal fluency study. Journal of Neurology, cerebral lesions. International Journal of Neu-
Neurosurgery, and Psychiatry, 69(1), 102-106. roscience, 91(3-4), 295-302.
Leng, N. R. C., & Parkin, A. J. (1989). Aetiological Lezak, M. D. (1976). Neuropsychological assess-
variation in the amnesic syndrome: Compari- ment. New York: Oxford University Press.
sons using the Brown-Peterson task. Cortex, 25, Lezak, M. D. (1982). The test-retest stability and
251-259. reliability of some tests commonly used in neu-
Lenzenweger, M. F., & Korfine, L. (1994). Per- ropsychological assessment. Paper presented at
ceptual aberrations, schizotypy, and the Wisco- the meeting of the International Neuropsycho-
nsin Card Sorting Test. Schizophrenia Bulletin, logical Society, Deauville, France.
20(2), 345--357. Lezak, M.D. (1983). Neuropsychological assessment
Leonberger, F. T., Nicks, S. D., Goldfader, P. R., & (2nd ed.). New York: Oxford University Press.
Munz, D. C. (1991). Factor analysis of the Lezak, M.D. (1995). Neuropsychological assessment
Wechsler Memory Scale-Revised and the (3rd ed.). New York: Oxford University Press.
574 REFERENCES
Lezak, M. D., Howieson, D. B., & Loring. D. W. Manual. Lutz, FL: Psychological Assessment
(2004). Neuropsychological assessment (4th ed.). Resources.
New York: Oxford University Press. Loberg, T. (1980). Alcohol misuse and neuro-
Liberman, J. N., Stewart, W., Seines, 0., & psychological deficits in men. Journal of Studies
Gordon, B. (1994). Rater agreement for the Rey- on Alcohol, 41(1), 119-128.
Osterrieth Complex Figure test. Jor,rnal of Locascio, J. L., Growdon, J. H., & Corkin, S.
Clinical Psychology, 50(4), 615-624. (1995). Cognitive test performance in detecting.
Libon, D. J., Glosser, G., Malamut, B. L., staging, and tracking Alzheimer's disease. Ar-
Kaplan, E., Goldberg, E., Swenson, R., et al. chives of Neurology, 52, 1087-1099.
(1994). Age, executive functions, and visuospatial Loewenstein, D. A., Rubert, M. P., Argueelles, T., &
functioning in healthy older adults. Neuropsy- Duara, R. (1995). Neuropsychological test per-
chology, 8, 38-43. formance and prediction of functional capacities
Libon, D. J., Freeman, R. Q., Giovannetti, T., among Spanish-speaking and English-speaking
Lamar, M., Cloud, B. S., Stem, R. A., et al. (2003). patients with dementia. Archives of Clinical
The ROCF and visuoconstructional impainnent in Neuropsychology,10(2), 75-88.
cortical and subcortical dementia. In J. A; Knight Loewenstein, D. A., Barker, W. W., Harwood, D. G.,
(Ed.), The handbook of Rey-Osterrieth Complex Luis, C., Acevedo, A., Rodriguez, 1., et al. (2000).
Figure usage: Clinical and research applications. Utility of a modified Mini-Mental State Exami-
Lutz, FL: Psychological Assessment Resources. nation with extended delayed recall in screening
Lichtenberg, P., & Christensen, B. (1992). Ex- for mild cognitive impairment and dementia
tended normative data for the Logical Memory among community dwelling elders. Interna-
subtest of the Wechsler Memory Scale-Revised: tional Journal of Geriatric Psychiatry, 15(5),
Responses from a sample of cognitively intact 434-440.
elderly medical patients. Psychological Beports, Logue, P. E., & Allen, K. (1971). WAIS-predicted
71, 745-746. Category Test scores with the Halstead Neu-
Lichtenberg, P. A., Ross, T., & Christell$en, B. ropsychological Battery. Perceptual and Motor
(1994). Preliminary normative data on the Bos- Skills, 33, 1~1096.
ton Naming Test for an older uroan population. Lombardi, W. J., Andreason, P. J., Sirocco, K. Y.,
Clinical Neuropsychologist, 8(1), 109-111. Rio, D. E., Gross, R. E., Umhau, J. C., et al.
Lichtenberg, P. A., Ross, T. P., Youngblade, L., & (1999). Wisconsin Card Sorting Test perfor-
Vangel, S. J. (1998). Normative studies research mance following head injury: Dorsolateral fronto-
project test battery: Detection of dementia in striatal circuit activity predicts perseveration.
African American and European American ur- Journal of Clinical and Experimental Neuropsy-
ban elderly patients. Clinical Neuropsychologist, chology, 21(1), 2-16.
12(2), 146-154. Loa, H., Bonne), J., Etevenon, P., Benyacoub, J., &
Light, R. J., & Pillemer, D. B. (1984). Sumrmng up: Slowen, P. (1981). Intellectual efficiency in
The science of reviewing research. Cambridge, manic-depressive patients treated with lithium:
MA: Harvard University Press. A control study. Acta Psychiatrica Scandinavica,
Lin, Y. G., & Rennick, P. M. (1974). Correlations 64(5), 423--430.
between performance on the Category Test and Loong, J. W. K. (1990). The Wisconsin Card Sort-
the Wechsler Adult Intelligence Scale in an ep- ing Test [Computer software]. San Luis Obispo,
ileptic sample. Journal of Clinical Psychology, CA: Loong.
31(1), 62-65. Loonstra, A. S., Tarlow, A. R., & Sellers, A. H.
Lineweaver, T. T., Bondi, M. W., Thomas, R. G., & (2001). COWAT metanorms across age, educa-
Salmon, D. P. (1999). A normative study of tion, and gender. Applied Neuropsychology, 8(3),
Nelson's (1976) modified version of the Wiscon- 161-166.
sin Card Sorting Test in healthy older adults. Lopez, M. N., Arias, G. P., Hunter, M. A, Charter, R. A.,
Clinical Neuropsyclwlogist, 13(3), 328-347. & Scott, R. R. (2003). Boston Naming Test:
Little, A. J., Templer, D. 1., Persel, C. S., & Problems with administration and scoring. Psy-
Ashley, M. J. (1996). Feasibility of the neu- clwlogj.cal Reports, 92(2), 468-472.
ropsychological spectrum in prediction of out- Lopez, M. N., Lazar, M. D., & Oh, S. (2003).
come following head injury. Journal of Clinical Psychometric properties of the Hooper Visual
Psychology, 52, 455--460. Organization Test. Assessment, 10(1), 66-70.
Uorente, A.M., Williams, J., Satz, P., & D'Elia, L. F. Lopez-Carlos, E. (1999). Validity study of the World
(2003). Children's Color Trails Test-Profe~sional Health Organization/University of California at
REFERENCES 575
Los Angeles Auditory Verbal Learning Test. (1998). Mayo's Older Americans Normative
Dissertation Abstracts International. Section B: Studies: Category 6uency norms. Journal of
The Sciences and Engineering, 59(9-B), 5095. Clinical and Experimental Neuropsychology,
Lopez-Carlos, E., Salazar, X. F., Villasenor, T., 20(2), 194-200.
Saucedo, C., & Peiia, R. (2003). Validez y datos Lucas, M. D., & Sonnenberg, B. R. (1996). Neuro-
nonnativos de la pruebas de nomtnacion en psychological trends in the Parkinsonism-plus
personas con educacton limttada. Poster pre- syndrome: A pilot study. Journal of Clinical and
sented at the Neuropsicologia-Congreso Lati- Experimental Neuropsychology, 18(1), 88-97.
noamericano por Ia Sociedad Latinoamericana Ludgate, J., Keating, J., O'Dwyer, R., & Callaghan,
de Neuropsicologia. Toronto, Canada. N. (1985). An improvement in cognitive function
Lorig, T. S., Gehring, W. J., & Hym, D. L. (1986). following polypharmacy reduction in a group
Period analysis of the EEG during performance of epileptic patients. Acta Neurologica Scandi-
of the Trail Making Test. International Journal navica, 71(6), 448-452.
of Clinical Neuropsyclwlogy, 8(3), 97-99. Luria, A. R. (1980). Higher cortical fUnctions in
Loring, D. W. (1989). The Wechsler Memory man. New York: Basic Books.
Scale-Revised, or the Wechsler Memory Lynch, W. J. (2002). Assessment in traumatic brain
Scale-Revisited? Clinical Neuropsyclwlogist, injury: Update on recent developments. Journal
3(1), 59--69. of Head Trauma Rehabilitation, 17(1), 66--70.
Loring, D. W., & Meador, K. J. (2003). The Med- Lyness, S. A., Eaton, E. M., & Schneider, L. S.
ical College of Georgia (MCG) Complex Fig- (1994). Cognitive performance in older and
ures: Four forms for follow-up. In J. A. Knight middle-aged depressed outpatients and controls.
(Ed.), The handbook of Rey-Osterrieth Complex Journal of Gerontology: Psyclwlogical Sciences,
Figure usage: Clinical and research applications. 49, P129-P136.
Lutz, FL: Psychological Assessment Resources. Lysaker, P., & Bell, M. (1994). Insight and cogni-
Loring, D. W., Lee, G. P., & Meador, K. J. (1988). tive impairment in schizophrenia: Performance
Revising the Rey-Osterrieth: Rating right hemi- on repeated administration of the Wisconsin
sphere recall. Archives of Clinical Neuropsy- Card Sorting Test Journal of Nervous and
clwlogy, 3, 239-247. Mental Disorders, 182, 656--660.
Loring, D. W., Martin, R. L., Meador, K. J., & Lysaker, P., Bell, M., & Beam-Goulet, J. (1995).
Lee, G. P. (1990). Psychometric construction Wisconsin Card Sorting Test and performance in
of the Rey-Osterrieth Complex Figure: Meth- schizophrenia. Psychiatry Research, 56, 45-51.
odological considerations and interrater reli- Lyvers, M., & Maltzman, I. (1991). Selective effects
ability. Archives of Clinical Neuropsyclwlogy, 5, of alcohol on Wisconsin Card Sorting Test per-
1-14. formance. British Journal of Addiction, 86(4),
LoSasso, G. L., Rapport, L. J., Axelrod, B. N., & 399-407.
Reeder, K. P. (1998). Intermanual and alternate- Macinnes, W. D., Golden, C. J., McFadden, J., &
form equivalence on the Trail Making Tests. Wilkening, G. N. (1983a). Relationships between
Journal of Clinical and Experimental Neuropsy- the Booklet Category Test and the Wisconsin
clwlogy, 20(1), 107-110. Card Sorting Test. International Journal of
Lu, L., & Bigler, E. D. (2000). Performance on Neuroscience, 21(34), 257-264.
original and a Chinese version of Trail Making Macinnes, W. D., McFadden, J. M., & Golden, C. J.
Test part B: A normative bilingual sample. Ap- (1983b). A short-portable version of the Cate-
plied Neuropsyclwlogy, 7(4), 243-246. gory Test. International Journal of Neuroscience,
Lu, L., & Bigler, E. D. (2002). Normative data on 18,41-44.
Trail Making Test for neurologically normal, Mack, J. L., &: Carlson, N. J. (1978). Conceptual
Chinese-speaking adults. Applied Neuropsy- deficits and aging: The Category Test. Perceptual
clwlogy, 9(4), 219-225. and Motor SlciUs, 46, 1~128.
Lu, P. H., Boone, K. B., Cozolino, L., & Mitchell, C. Mack, W. J., Freed, D. M., Williams, B. W., &
(2003). Effectiveness of the Rey-Osterrieth Henderson, V. W. (1992). Boston Naming Test:
Complex Figure Test and the Meyers and Shortened versions for use in Alzheimer's dis-
Meyers Recognition Trial in the detection of ease. Journals of Gerontology, 47(3), P154-P158.
suspect effort. Clinical Neuropsyclwlogist, 17(3), MacKay, A. J., Connor, L. T., Albert, M. L., &
426-440. Ohler, L. K. (2002). Noun and verb retrieval in
Lucas, J. A., lvnik, R. J., Smith, G. E., Bohac, D. L., healthy aging. Journal of the International Neu-
Tangalos, E. G., Graff-Radford, N. R., et al. ropsyclwlogical Society, 8(6), 764-770.
576 REFERENCES
MacLeod, C. (1991). Half century of research on Manly, J. J., Jacobs, D. M., Touradji, P., Small, S.
the Stroop effect: An integrative review. Psy- A., & Stem, Y. (2002). Reading level attenuates
chological Bullettn, 109, 163-203. differences in neuropsychological test perfor-
Maddocks, D., & Saling, M. (1996). Neurqpsycho- mance between African American and white
logical deficits following concussion. Brain In- elders. Journal of the International Neuropsy-
jury, 10(2), 99-103. chological Society, 8(3), 341--348.
Madison, L. S., George, C., & Moeschler, J. B. Marcopulos, B. A., McLain, C. A., & Giuliano, A. J.
(1986). Cognitive functioning in the hgile-X (1997). Cognitive impairment or inadequate
syndrome: A study of intellectual, mempry and norms? A study of healthy, rural, older adults
communication skills. Journal of Men,.U Defi- with limited education. Clinical Neuropsycholo-
ciency Research, 30, 12~148. gist, 11(2), 111-131.
Maj, M., Janssen. R., Satz, P., Zaudig, M., Mares, M. (2002). Demographic predictors of ver-
Starace, F., Boor, D., et al. (1991). The World bal learning and memory indices on the World
Health Organization's cross-cultural sflldy on Health Organization-University of California at
neuropsychiatric aspects of infection With the Los Angeles Auditory Verbal Learning Test in a
human immunodeficiency virus (HIV-1)1 Prepa- Hispanic sample. Dissertation Abstracts Inter-
ration and pilot phase. British Journal !of Psy- national. Section B: The Sciences and Engineer-
chunry,159,351-356. : ing, 62(10-B), 4793.
Maj, M., D'Eiia, L., Satz, P., Jansses, R., Za~g, M., Margolin, D., Pate, D. S., Friedrich, F. J., & Elia,
Uchiyama, C., et al. (1993). Evaluation oft-Yo new E. (1990). Dysnomia in dementia and in stroke
neuropsychological tests designed to niinimize patients: Different underlying cognitive deficits.
cultural bias in the assessment ofHIV-1 Seropos- Journal of Clinical and Experimental Neuropsy-
itive persons: A WHO study. Archives of Clinical chology, 12(4), 597-612.
Neuropsychology, 8, 123-135. Marie, R. M., Rioux, P., Eustache, F., Travere,
Majdan, A., Sziklas, V., & Jones-Go~. M. J. M., et al. (1995). Clues and functional neuro-
(1996). Performance of healthy subj~ and anatomy of verbal working memory: A study
patients with resection from the anteri.,r tem- about the resting brain glucose metabolism in
poral lobe on matched tests of verbal tand vi- Parkinson's disease. European Journal of Neu-
suoperceptual learning. Journal of Clin~ and rology, 2, 83-94.
Experimental Neuropsychology, 18(3), 4J6-430. Marien, P., Mampaey, E., Vervaet, A., Saerens, J.,
Malec, J., Ivnik, R., Smith, G., Tangalos, E., ~tersen, & De Deyn, P. P. (1998). Normative data for the
R., Kokmen, E., et al. (1992). Mayo's Older Boston Naming Test in native Dutch-speaking
Americans Normative Studies: Utility of· correc- Belgian elderly. Brain and Language, 65(3),
tions for age and education for the WAIS-R. 447-467.
Clinical Neuropsycholo,gst, 6(Suppl.), 31--47. Marin, G., & Marin, B. V. (1991). Research with
Malina, A., Regan, T., Bowers, D., & Millis, S. Hispanic populations. Newbury Park, CA: Sage
(2001). Psychometric analysis of the Visll41 Form Publications, Inc.
Discrimination Test. Perceptual and: Motor Martin, A., & Fedio, P. (1983). Word production
Skills, 92(2), 449-455. and comprehension in Alzheimer's disease: The
Malloy, P. (1987). Frontal lobe dysfunction in breakdown of semantic knowledge. Brain and
obsessive-compulsive disorder. In Perecman, E. Language, 19, 124-141.
(Ed.), The frontal lobes revisited (pp. rol-223). Martin, D. J., Oren, Z., & Boone, K. (1991). Major
Hillsdale, NJ: Lawrence Erlbaum. · depressive's and dysthymic's performance on the
Manly, J. J., Miller, S. W., Heaton, R. K., Btro, D., Wisconsin Card Sorting Test. Journal of Clinical
Reilly, J., Velasquez, R. J., et al. (1998). The effect Psychology, 47, 685-690.
of African-American acculturation on neu- Martin, N. J., & Franzen, M. D. (1989). The effect
ropsychological test performance in normal and of anxiety on neuropsychological function. In-
HIV-positive individuals. Journal of the .lPtema- ternational Journal of Neuropsychology, 11, 1-8.
tional Neuropsychological Society, 4(3), 2t1--302. Martin, P. W., & Greene, R. L. (1978, April). Inter-
Manly, J. J., Jacobs, D. M., Sano, M., B):lll, K., judge reliability ofmemory and location scores on
Merchant, C. A., Small, S. A., et al. (1Wg). Ef- the Halstead-Reitan Tactual Performance Test.
fect of literacy on neuropsychological ±t per- Paper presented at the meeting of the South-
formance in nondemented, education- atched western Psychological Association, New Orleans.
elders. Journal of the International N ropsy- Martin, R. C., Sawrie, S., Hugg, J., Gilliam, F.,
chological Society, 5(3), 191-202. Faught, E., & Kuzniecky, R. (1999). Cognitive
REFERENCES 577
correlates of 1H MRSI-detected hippocampal Matarazzo, J. D. (1990). Psychological assessment

abnormalities in temporal lobe epilepsy. Neu- versus psychological testing. American Psychol-
rology, 53(9), 2052-2058. ogist, 45(9), 999-1017.
Martin, R. C., Sawrie, S. M., Edwards, R., Roth, D. Matarazzo, J.D., Wiens, A. N., Matarazzo, R. G., &
L., Faught, E., Kuzniecky, R. 1., et al. (2000). Goldstein, S. G. (1974). Psychometric and clin-
Investigation of executive function change fol- ical test-retest reliability of the Halstead im-
lowing anterior temporal lobectomy: Selective pairment index in a sample of healthy, young.
normalization of verbal fluency. Neuropsychol- normal men. Journal of Neroous and Mental
ogy, 14(4), 501-508. Disease, 158(1), 37--49.
Martin, S. E., Engleman, H. M., Deary, I. H., & Matarazzo, R. G. (1995). Psychological report
Douglas, N. J. (1996). The effect of sleep frag- standards in neuropsychology. Clinical Neuro-
mentation on daytime function. American Jour- psychologist, 9(3), 249-250.
nal of Respiratory and Critical Care Medicine, Mathiesen, T., Ellingsen, D. G., & Kjuus, H.
153, 1328-1332. (1999). Neuropsychological effects associated
Martin, T. A., Hoffman, N. M., & Donders, J. with exposure to mercury vapor among former
(2003). Clinical utility of the Trail Making Test chloralkali workers. Scandinavian Journal of
ratio score. Applied Neuropsychology, 10(3), Work, Environment and Health, 25(4), 342-350.
163-169. Matthews, C. G. (1974). Application of neu-
Martinez, A. A., Penades, R., Vieta, E., Colom, F., ropsychological test methods in mentally re-
Reinares, M., Benabarre, A., Salamero, M., & tarded subjects. In R. M. Reitan & L. A. Davison
Gastro, C. (2002). Executive function in patients (Eds.), Clinical neuropsychology: Current status
with remitted bipolar and schizophrenia and its and applications. Washington, DC: Hemisphere.
relationship with functional outcome. Psycho- Matthews, K. A., Cauley, J., Yaffe, K., & Zmuda,
therapy and Psychosomatics, 71, 39-46. J. M. (1999). Estrogen replacement therapy and
Mason, C. F., & Ganzler, H. (1964). Adult norms cognitive decline in older community women.
for the Shipley Institute of living Scale and Journal of the American Geriatrics Society,
Hooper Visual Organization Test based on 47(5), 518--523.
age and education. Journal of Gerontology, 19, Mattis, S. (1976). Mental status examination for
419-424. organic mental syndrome in the elderly patient.
Massman, P. J., & Doody, R. S. (1996). Hemi- In L. Bellak & T. Karasu (Eds.), Geriatric psy-
spheric asymmetry in Alzheimer's disease is ap- chiatry. New York: Grune & Stratton.
parent in motor functioning. Journal of Clinical Mattis, S. (1988). Dementia Rating Scale. Odessa,
and Experimental Neuropsychology, 18(1), FL: Psychological Assessment Resources.
110-121. Mayeux, R., Brandt, J., Rosen, J., & Benson, F.
Mast, B. T., MacNeill, S. E., & Lichtenberg. P. A. (1980). Interictal memoty and language impair-
(2000). Clinical utility of the normative studies ment in temporal lobe epilepsy. Neurology, 30,
research project test battety among vascular 120-125.
dementia patients. Clinical Neuropsychologist, Mayr, U. (2002). On the dissociation between
10, 173-180. clustering and switching in verbal fluency:
Masur, D. M., Fuld, P. A., Blau, A. D., Thai, L. J., Comment on Troyer, Moscovitch, Winocur, Al-
Levin, H. S., & Aronson, M. K. (1989). Distin- exander and Stuss. Neuropsychologia, 40(5),
guishing normal and demented elderly with the 562-566.
Selective Reminding Test. Journal of Clinical McCaffrey, R. J., Krahula, M. M., & Heimberg,
and Experimental Neuropsychology, 11(5), R. G. (1989). An analysis of the significance of
615-630. performance errors on the Trail Making Test in
Masur, D. M., Fuld, P. A., Blau, A. D., CtyStal, H., & polysubstance users. Archives of Clinical Neu-
Aronson, M. K. (1990). Predicting development ropsychology, 4(4), 393-398.
of dementia in the elderly with the Selective McCaffrey, R. J., Ortega, A., Orsillo, S. M., Nelles,
Reminding Test. Journal of Clinical and Exper- W. B., et al. (1992). Practice effects in repeated
imental Neuropsychology, 12(4), 529--538. neuropsychological assessments. Clinical Neu-
Mataix-Cols, D., Barrios, M., Sanchez-Turet, M., ropsychologist, 6(1), 32-42.
Vallejo, J., & Junque, C. (1999). Reduced design McCaffrey, R. J., Ortega, A., & Haase, R. F. (1993).
fluency in subclinical obsessive-compulsive Effects of repeated neuropsychological assess-
subjects. Journal of Neuropsychiatry and Clini- ments. Archives of Clinical Neuropsychology,
cal Neurosciences, 11(3), 395-397. 8(6), 519-524.
578 REFERENCES
McCaffrey, R. J., Cousins, J. P., Westervelt, H. J., Mejia, S., Pineda, D., Alvarez, L. M., & Ardila. A.
Martynowicz, M., et al. (1995). Practice effects (1998). Individual differences in memory and
with the NIMH AIDS Abbreviated Neu- executive function abilities during normal aging.
ropsychological Battery. Archives of Clinical International Journal of Neuroscience, 95(3-4),
Neuropsychology, 10(3), 241-250. 271-284.
McCaffrey, R. J., Duff, K., & Westervelt, H. J. Merriam, E. P., Thase, M. E., Haas, G. L., Ke-
(2000). Practitioner's guide to evaluating change shavan, M. S., & Sweeney, J. A. (1999). Prefrontal
with neuropsychological assessment instruments. cortical dysfunction in depression determined by
New York: Kluwer Academic/Plenum. Wisconsin Card Sorting Test performance.
McCaffrey, R. J., Westervelt, H., & Haase, R. F. American Journal ofPsychiat1lJ,I56(5), 780-782.
(2001). Serial neuropsychological assessment with Merrick, E. E., Donders, J., & Wiersum, M. (2003).
the National Institute of Mental Health (NIMH) Validity of the WCST-64 after traumatic
AIDS Abbreviated Neuropsychological Battery. brain injury. Clinical Neuropsychologist, 17(2),
Archives ofClinical Neuropsychology, I6(1), 9-18. 153-158.
McCarthy, D. (1972). Manual for the McCarthy Merten, T. (2002). A short version of the Hooper Vi-
Scales for Children's Abilities. New York: Psy- sual Organization Test: Development and valida-
chological Corporation. tion. Clinical Neuropsychologist,I6(2), 13&-144.
McCracken, L. M., & Franzen, M. D. (1992). Merten, T., & Beal, C. (2000). An analysis of the
Principal-components analysis of the equiva- Hooper Visual Organization Test with neurolog-
lence of alternate forms of the Trail Making Test. ical patients. Clinical Neuropsychologist, 14(4),
Psychological Assessment, 4(2), 235-238. 521-529.
McCurry, S. M., Gibbons, L. E., Uomoto, J. M., Mesulam, M. M. (1985). Principles of behavioral
Thompson, M. L., Graves, A. B., Edland, S. D., and cognitive neurology. New York: Oxford
et al. (2001). Neuropsychological test perfor- University Press.
mance in a cognitively intact sample of older Mesulam, M. M. (2000). Principles of behavioral
Japanese American adults. Archives of Clinical and cognitive neurology (2nd ed.). New York:
Neuropsychology, I6, 447-459. Oxford University Press.
McFie, J. & Piercy, M. F. (1952). Intellectual im- Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G.,
pairment with localized cerebral lesions. Brain, Moreland, K. L., Dies, R. R., et al. (2001). Psy-
75, 292-311. chological testing and psychological assessment:
McKeever, W. F., & Abramson, M. (1991). Hal- A review of evidence and issues. American Psy-
stead and Halstead-Reitan norms for Finger chologist, 56(2), 128-165.
Tapping Test are severely biased against females Meyers, J. E., & Lange, D. (1994). Recognition
and left-banders. Journal of Clinical and Exper- subtest for the Complex Figure. Clinical Neu-
imental Neuropsychology, I3(1), 91. ropsychologist, 8(2), 153-166.
McKhann, G., Drachman, D., Folstein, M., Katz- Meyers, J. E., & Meyers, K. R. (1992). A training
man, R., Price, D., & Stadlan, E. M. (1984). manual for the clinical scoring of the Rey-
Clinical diagnosis of Alzheimer's disease: Report Osterrieth Complex Figure and the recognition
of the NINCDS-ADRDA Work Group. Neurol- subtest. Sioux City, lA: Author.
ogy, 34, 939-944. Meyers, J. E., & Meyers, K. R. (1995a). Rey
Meador, K. J., Loring, D. W., Allen, M. E., Complex Figure Test under four different ad-
Zamrini, E. Y., Moore, E. E., Abney, 0. L., et al. ministration procedures. Clinical Neuropsychol-
(1991). Comparative cognitive effects of carba- ogist, 9(1), 63-67.
mazepine and phenytoin in healthy adults. Meyers, J. E., & Meyers, K. R. (1995b). Hey
Neurology, 4I(l0), 1537-1540. Complex Figure Test and Recognition trial:
Meador, K. J., Moore, E. E., Nichols, M. E., Professional manual. Lutz, FL: Psychological
Abney, 0. L., Taylor, H. S., Zamrini, E. Z., et al. Assessment Resources.
(1993). The role of cholinergic systems in vi- Meyers, J. E., & Meyers, K. R. (1996). Hey Com-
suospatial processing in memory. Journal of plex Figure Test and Recognition trial: Supple-
Clinical and Experimental Neuropsychology, mental norms for children and adolescents. Lutz,
I5(5), 832-842. FL: Psychological Assessment Resources.
Mehta, Z. & Newcombe, F. (1996). Dissociable Meyers, J. E., Galinsky, A. M., & Volbrecht, M.
contributions of the two cerebral hemispheres to (1999). Malingering and mild brain injury: How
judgments of line orientation. Journal of Inter- low is too low? Applied Neuropsychology, 6(4),
national Neuropsychological Society, 2, 335--339. 208-216.
REFERENCES 579
Miceli, G., Caltagirone, C., Gainotti, G., Masullo, (Eds.), The frontal granular cortex and behavior
C., & Siweri, M. C. (1981). Neuropsychological (pp. 331-334). New York: McGraw-Hill.
correlates of localized cerebral lesions in non- Milner, B. (1970). Memory and the medial tem-
aphasic brain-damaged patients. Journal of poral regions of the brain. In K. H. Pribram &
Clinical Neuropsychology, 3, 53--63. D. E. Broadbent (Eds.), Biology of 1nenwry
Mickanin, J., Grossman, M., Onishi, K., Aur- (pp. 29--50). New York: Academic Press.
iacombe, S., & Clark, C. (1994). Verbal and Milner, B. (1972). Disorders of learning and
nonverbal fluency in patients with probable Alz- memory after temporal lobe lesions in man.
heimer's disease. Neuropsychology, 8, 385--394. Clinical Neurosurgery, 19, 421-446.
Miezejeski, C. M., Jenkins, E. C., Hill, A. L., Wis- Milner, B. (1975). Psychological aspects of focal
niewski, K., et al. (1986). A profile of cognitive epilepsy and its neurosurgical management.
deficit in females from fragile X families. Neu- Advances in Neurology, 8, 299--321.
ropsychologia, 24(3), 405-409. Miner, T., & Ferraro, F. R. (1998). The role of
Miller, B. L., Lesser, I. M., Boone, K. B., Hill, E., speed of processing, inhibitory mechanisms, and
Mehringer, C. M., & Wong, K. (1991). Brain presentation order in Trail-Making Test perfor-
lesions and cognitive function in late-life psy- mance. Brain and Cognition, 38(2), 246-253.
chosis. British Journal of Psychiatry, 158, 76-82. Misra, U. K., Prasad, M., & Pandey, C. M. (1994).
Miller, E. (1984). Verbal fluency as a function of a A study of cognitive functions and event related
measure of verbal intelligence and in relation to potentials following organophosphate exposure.
different types of cerebral pathology. British Electromyography and Clinical Neurophysiol-
Journal of Clinical Psychology, 23, 53-57. ogy, 34(4), 197-203.
Miller, E. (1985). Possible frontal impairments: A Mitropoulou, V., Harvey, P. D., Maldari, L. A.,
test using a measure of verbal fluency. British Moriarty, P. J., New, A. S., Silverman, J. M.,
Journal of Clinical Psychology, 24, 211-212. et al. (2002). Neuropsychological performance in
Miller, E. N. (2003). An update on the 1991 article schizotypal personality disorder: Evidence re-
by Seines, Jacobson, Machado, Becker, Wesch, garding diagnostic specificity. Biological Psychi-
Miller, Visscher and McArthur. atry, 52(12), 1175-1182.
Miller, E. N., Seines, 0. A., McArthur, J. C., Satz, P., Mitrushina, M., & Satz, P. (1991a). Effect of re-
Becker, J. T., Cohen, B. A., et al. (1990). Neu- peated administration of a neuropsychological
ropsychological performance in HIV-1-infected battery in the elderly. Journal of Clinical Psy-
homosexual men: The Multicenter AIDS Co- chology, 47(6), 790-801.
hort Study (MACS). Neurology, 40(2), 197-204. Mitrushina, M., & Satz, P. (1991b). Changes in
Miller, L. S., & Rohling, M. L. (2001). A statisti- cognitive functioning associated with normal
cal interpretive method for neuropsycholog- aging. Archives of Clinical Neuropsychology, 6,
ical test data. Neuropsychology Review, 11(3), 49-60.
143-169. Mitrushina, M., & Satz, P. (1995). Repeated testing
Millis, S. R., Rosenthal, M., & Lourie, I. F. (1994). of normal elderly with the Boston Naming Test.
Predicting community integration after trau- Aging: Clinical and Experimental Research, 7,
matic brain injury with neuropsychological 123-127.
measures. lntematWnal Journal of Neuroscience, Mitrushina, M., Satz, P., & Van Gorp, W. (1989).
79(3-4), 165-167. Some putative cognitive precursors in subjects
Millis, S. R., Rosenthal, M., Novack, T. A., hypothesized to be at-risk for dementia. Archives
Sherer, M., Nick, T. G., Kreutzer, J. S., et al. of Clinical Neuropsychology, 4, 323-333.
(2001). Long-term neuropsychological outcome Mitrushina, M., Satz, P., & Chervinsky, A. B.
after traumatic brain injury. Journal of Head (1990). Efficiency of recall on the Rey-Osterrieth
Trauma Rehabilitation, 16(4), 343-355. Complex Figure in normal aging. Brain Dtjs-
Milner, B. (1962). Laterality effects in audition. In fonction, 3, 148-150.
V. B. Mountcastle (Ed.), Interhemispheric rela- Mitrushina, M., Satz, P., Chervinsky, A., & D'Elia, L.
tions and cerebral dominance. Baltimore: Johns (1991). Performance of four age groups of
Hopkins University Press. normal elderly on the Rey Auditory-Verbal
Milner, B. (1963). Effects of different brain lesions Learning Test. Journal of Clinical Psychology,
on card sorting test. Archives of Neurology, 9, 47(3), 351-357.
90-100. Mitrushina, M., D'Elia, L., Satz, P., Uchiyama, C.,
Milner, B. (1964). Some effects of frontal lobec- Mathews, A., & Harker, J. (1993). A comparison
tomy in man. In J. M. Warren & K. A. Akert of selective attention deficits in normal elderly
580 REFERENCES
and AIDS patients. Developmental Brain Dys- quality of reporting of meta-analysis of random-
function, 6, 324--328. ' ized controlled trials: The QUOROM statement.
Mitrushina, M., Drebing, C., Uchiyama, C., Satz, P., Laneet, 354, 1896-1900.
Van Gorp, W., & Chervinsky, A. (1994). The Monsch, A. U., Bondi, M. W., Butters, N., Salmon,
pattern of deficit in different memory compo- D. P., Katzman, R., & Thai, L. J. (1992). Com-
nents in normal aging and dementia of Alzhei- parisons of verbal fluency tasks in the detection
mer's type. Journal ofClinical Psychology, 50(4), of dementia of the Alzheimer type. Archives of
591-596. Neurology, 49(12), 1253-1258.
Mitrushina, M., Fogel, T., D'Eiia, L., Uchiyama, C., Monsch, A. U., Bondi, M. W., & Butters, N. (1994).
& Satz, P. (1995a). Performance on motor tasks A comparison of category and letter fluency in
as an indication of increased behavioral asym- Alzheimer's disease. Neuropsychology, 8, 25-30.
metry with advancing age. Neuropsychologia, Montgomery, P., Silverstein, R., Wichmann, R., &
33(3), 359-364. Fleischaker, K. (1993). Spatial updating in
Mitrushina, M., Uchiyama, C., & Satz, P. (l995b). Parkinson's Disease. Brain and Cognition, 23,
Heterogeneity of cognitive profiles in normal 113-126.
aging: Implications for early manifestations of Montse, A., Pere, V., Carme, J., Francese, V., &
Alzheimer's disease. Journal of Clinical and Ex- Eduardo, T. (2001). Visuospatial deficits in Par-
perimental Neuropsychology, 17(3), 374-382. kinson's disease assessed by Judgment of Line
Mittenberg, W., Seidenberg, M., O'Leary, D. S., & Orientation Test: Error analyses and practice
DiGiulio, D. V. (1989). Changes in cerebral effects. Journal of Clinical and Experimental
functioning associated with normal aging. Jour- Neuropsychology, 23(5), 592-598.
nal of Clinical and Experimental Neuroplychol- Moore, T. E., Richards, B., & Hood, J. (1984).
ogy, 11, 91~932. Aging and the coding of spatial information.
Mittenberg, W., Burton, D., Darrow, E., & Journal of Gerontology, 39(2), 210-212.
Thompson, G. (1992). Normative data i>r the Morehouse, S. A., Szeliga, F., & DiTommaso, E.
WMS-R: 25 to 34 year olds. Psychological As- (2000). Characteristics of the bimanual deficit
sessment, 4(3), 363--368. using grip strength. Laterality: Asymmetries of
Moberg, M., Ferraro, F. R., & Petros, T. V. (2000). Body, Brain and Cognition, 5(2), 167-185.
Lexical properties of the Boston Naming Test Morey, C. E., Cilo, M., Berry, J., & Cusick, C.
stimuli: Age differences in word naming and (2003). The effect of Aricept in persons with
lexical decision latency. Applied Neuropllljclwl- persistent memory disorder following traumatic
ogy, 7(3), 147-153. brain injury: A pilot study. Brain Injury, 17(9),
Moehle, K. A., Fitzhugh-Bell, K. B., Engleman, E., & 809-815.
Hennon, D. (1988). Statistical and diagnostic ade- Morgan, J. E., & Caccappolo-van Vliet, E. (2001).
quacy of a short form of the Halstead Category Advanced years and low education: The case
Test. International Journal of Neuroscienee, 42, against the comprehensive norms. Journal of
107-112. Forensic Neuropsychology, 2(1), 53-69.
Moehle, K. A., Rasmussen, J. L., & Fitzhugh-Bell, Morice, R. (1990). Cognitive inflexibility and pre-
K. B. (1990). Factor analysis of neuropsyt:holo- frontal dysfunction in schizophrenia and mania.
gical tests in an adult sample. InterntJtional British Journal of Psychiatry, 157, 50-54.
Journal of Clinical Neuropsychology, 1.2(3-4), Mormont, C. (1984). The influence of age and de-
107-115. pression on intellectual and memory perfor-
Moering, R. G., Schinka, J. A., Mortimer, J. A., & mances. Acta Psychiatrica Belgica, 84(2), 127-134.
Graves, A. B. (2004). Normative data for elderly Morris, J. C., Heyman, A., Mohs, R. C., Hughes,
African Americans for the Stroop Color and S. P., van Belle, G., Fullenbaum, G., et al.
Word Test. Archives of Clinical Neuropsychol- (1989). The Consortium to Establish a Registry
ogy, 19(1), 61-71. for Alzheimer's Disease (CERAD). Part 1.
Moffoot, A. P. R., O'Carroll, R. E., Bennie, J., Clinical and neuropsychological assessment of
Carroll, S., Dick, H., Ebmeier, K. P., et al. Alzheimer's disease. Neurology, 39, 1159-1165.
(1994). Diurnal variation of mood and neuro Morris, J. C., Edland, S., Clark, C., Galasko, D.,
psychological function in major depression with Koss, E., Mohs, R., et al. (1993). The Consor-
melancholia. Journal of Affective Disorders, 32, tium to Establish a Registry for Alzheimer's
257-269. Disease (CERAD): N. Rates of cognitive change
Moher, D., Cook, D. J., Eastwood, S., Olkin, 1., in the longitudinal assessment of probable Alz-
Rennie, D., Stroup, D. (1999). Improving the heimer's disease. Neurology, 43(12), 2457-2465.
REFERENCES 581
Morrison, M. W., Gregory, R. J., & Paul, J. J. elderly medical inpatients. Clinical Gerontolo-
(1979). Reliability on the Finger Tapping Test gist, 17(1), 43--53.
and a note on sex differences. Perceptual and Nabors, N. A., Vangel, S. J., Lichtenberg, P. A., &
Motor Skills, 48, 13~142. Walsh, P. (1997). Normative and clinical utility
Morrow, L. A., Muldoon, S. B., & Sandstrom, D. J. of the Hooper Visual Organization Test with
(2001). Neuropsychological sequelae associated geriatric medical inpatients. Journal of Clinical
with occupational and environmental exposure Geropsyclwlogy, 3(3), 191-198.
to chemicals. In R. E. Tarter, M. Butters, & S. R. Nadler, J. D., Grace, J., White, D. A., Butters,
Beers, (Eds.), Medical neuropsyclwlogy (2nd M. A., & Malloy, P. F. (1996). Laterality differ-
ed.). New York: Kluwer Academic/Plenum. ences in quantitative and qualitative Hooper
Moses, J. A. (1986). Factor structure of Benton's performance. Archives of Clinical Neuropsy-
Tests of Vtsual Retention, Visual Construction, clwlogy, 11(3), 223-229.
and Visual Form Discrimination. Archives of Nagahama, Y., Fuyama, H., Yamauchi, H., Mat-
Clinical Neuropsyclwlogy, 1(2), 147-156. suszaki, S., Konishi, H., Shibasaki, H., et al.
Moses, J. A., Jr., Pritchard, D. A., & Adams, R. L. (1996). Cerebral activation during performance
(1999). Normative corrections for the Halstead of a card sorting test. Brain, 119, 1667-1675.
Reitan Neuropsychological Battery. Archives of Nagahama, Y., Sadoto, N., Yamauchi, H., Katsumi Y.,
Clinical Neuropsychology, 14(5), 445-454. Hayashi, T., Fukuyama, H., et al. (1998). Neural
Mount, D. L., Hogg, L., & Johnstone, B. (2002). activity during attention shifts between object
Applicability of the 15-item versions of the features. Neuroreport, 9, 2633-2638.
Judgment of Line Orientation Test for individ- Nagahama, Y., Okina, T., Suzuki, N., Matsuzaki, S.,
uals with traumatic brain injury. Brain Injury, Yamauchi, H., Nabatame, H., et al. (2003).
16(12), 1051-1055. Factor structure of a modified version of the
Mountain, M. A. & Snow, W. G. (1993). Wisconsin Wisconsin Card Sorting Test: An analysis of ex-
Card Sorting Test as a measure of frontal ecutive deficit in Alzheimer's disease and mild
pathology: A review. Clinical Neuropsyclwlogist, cognitive impairment. Dementia and Geriatric
7, 108-118. Cognitive Disorders, 16(2), 103-112.
Mungas, D. (1983). Differential clinical sensitivity Nalcaci, E., Kalaycioglu, C., Cicek, M., & Gene, Y.
of specific parameters of the Rey Auditory-Ver- (2001). The relationships between handedness
bal Learning Test. Journal of Consulting and and fine motor performance. Cortex, 37(4),
Clinical Psyclwlogy, 51(6), 848-855. 493-500.
Mungas, D., Marshall, S.C., Weldon, M., Haan, M. Naugle, R. 1., & McSweeny, A. J. (1995). On the
& Reed, B. R. (1996). Age and education cor- practice of routinely appending neuropsycholog-
rection of the Mini-Mental State Examination ical data to reports. Clinical Neuropsyclwlogist,
for English and Spanish-speaking older adults. 9(3), 245-247.
Psychology and Aging, 12, 718-725. Naugle, R. 1., & McSweeny, A. J. (1996). More
Mungas, D., Reed, B. R., & Kramer, J. H. (2003). thoughts on the practice of routinely appending
Psychometrically matched measures of global raw data to reports: Response to Freides and
cognition, memory, and executive function for Matarazzo. Clinical Neuropsychologist, 10(3),
assesment of cognitive decline in older persons. 313--314.
Neuropsyclwlogy, 17(3), 380--392. Nebes, R. D. (1989). Semantic memory in Alzhei-
Murkin, J., Newman, S., Stump, D., & Blumenthal, mer's disease. Psychological Bulletin, 106,
J. (1995). Statement of consensus on assessment 377-394.
of neurobehavioral outcomes after cardiac sur- Nebes, R. D., & Brady, C. B. (1990). Preserved
gery. Annals of Thoracic Surgery, 59, 1289-1295. organization of semantic attributes in Alzhei-
Murphy, C., Nordin, S., & Acosta, L. (1997). Odor mer's disease. Psychology and Aging, 5,
learning, recall, and recognition memory in 574--579.
young and elderly adults. Neuropsyclwlogy, Nebes, R. D., Martin, D. C., & Hom, L. C. (1984).
11(1), 126-137. Sparing of semantic memory in Alzheimer's
Mutchnick, M. G., Ross, L. K., & Long, C. J. (1991). disease. Journal of Ahnonnal Psyclwlogy, 93,
Decision strategies for cerebral dysfunction: IV. 321-330.
Determination of cerebral dysfunction. Archives Nehemkis, A. M., & Lewinsohn, P. M. (1972). Ef-
of Clinical Neuropsychology, 6(4), 25~270. fects of left and right cerebral lesions on the
Nabors, N. A., Vangel, S. J., & Lichtenberg, P. A. naming process. Perceptual and Motor Skills, 35,
(1996). Visual Form Discrimination test with 787-798.
582 REFERENCES
Neils, J., Brennan, M. M., Cole, M., Boller, F., & and dementia: A study of semantic relatedness.
Gerdeman, B. (1988). The use of phonemic Brain and Language, 54, 184-195.
cueing with Alzheimer's disease patients. Neu- Nielsen, H., Knudsen, L., & Daugbjerg, 0. (1989).
ropsychologia, 26, 351--354. Normative data for eight neuropsychological
Neils, J., Baris, J. M., Carter, C., Dell'aira, A. L., tests based on a Danish sample. Scandinavian
Nordloh, S. J., Weiler, E., et al. (1995). Effects of Journal of Psychology, 30(1), 37-45.
age, education, and living environment on Bos- Nielsen, H., Lolk, A., & Kragh-Sorensen, P. (1995).
ton Naming Test performance. Journal of Speech Normative data for eight neuropsychological
and Hearing Research, 38, 1143-1149. tests, gathered from a random sample of Danes
Nell, V. (2000). Cross-cultural neuropsychological aged 64 to 83 years. Nordisk Psykologi, 47(4),
assessment: Theory and practice. Hillsdale, NJ: 241-255.
Lawrence Erlbaum. Nishiwaki, Y., Maekawa, K., Ogawa, Y., Asukai, N.,
Nelson, H. E. (1976). A modified card sorting test Minami, M., et al. (2001). Effects of sarin on the
sensitive to frontal lobe defects. Cortex, 12(4), nervous system in rescue team staff members
313-324. and police officers 3 years after the Tokyo sub-
Nelson, H. E. (1982). National Adult Reading Test way sarin attack. Environmental Health Per-
(NART): Test manual. Windsor, Ontario: NFER spectives, 109(11), 1169--1173.
Nelson. N'Kaoua, B., Lespinet, V., Barsse, A., Rougier, A.,
Nelson, N. W., Boone, K., Dueck, A., Wagener, L., & Claverie, B. (2001). Exploration of hemi-
Lu, P., & Grills, C. (2003). Relationships be- spheric specialization and lexico-semantic pro-
tween eight measures of suspect effort. ~linical cessing in unilateral temporal lobe epilepsy with
Neuropsychologist, 17(2), 263-272. · verbal fluency tasks. Neuropsychologia, 39(6),
Netherton, S. D., Elias, J. W., Albrecht, ;N. N., 635--642.
Acosta, C., et al. (1989). Changes in the perfor- Nolin, P. (1999). Analyses psychometriques de
mance of parkinsonian patients and nomlfl} aged I'adaptation francaise du California Verbal Learn-
on the Benton VISual Retention Test. Experi- ing Test (CVLT). [Psychometric analyses of the
mental Aging Research, 15(1-2), 13-18. French version of the California Verbal Learning
Neuger, G. J., O'Leai)', D. S., Berent, S., Fish- Test (CVLT)]. Revue Quebecoise de Psychologie,
burne, F. J., Giordani, B., Boll, T. J., et al. 20(1), 39--55.
(1981). Order effects on the Halstead.:.Reitan Norman, M. A., Evans, J. D., Miller, S. W., &
Neuropsychological Test Battery and allied pro- Heaton, R. K. (2000). Demographically cor-
cedures. Journal of Consulting and Clinical rected norms for the California Verbal Learning
Psychology, 49, 722-730. Test. Journal of Clinical and Experimental
Newcombe, F. (1969). Missile wounds of the brain. Neuropsychology, 22(1), 80-94.
London: Oxford University Press. Norris, M. P., Blankenship-Reuter, L., Snow-
Ng, T. P., Lim, L. C., & Win, K. K. (1992). An Turek, A. L., & Finch, J. (1995). InHuence of
investigation of solvent-induced neuro-psychiat- depression on verbal fluency performance. Aging
ric disorders in spray painters. Annals of the and Cognition, 2(3), 206-215.
AcademyofMedicine, Singapore,21(6), 79'7-803. Numan, B., Sweet, J. J., & Ranganath, C. (2000).
Ng, V. W. K., Eslinger, P. J., Williams, S. C. R., Use of the California Verbal Learning Test to
Brammer, M. J., Bullmore, E. T., Andrew, C. M., detect proactive interference in the traumatically
et al. (2000). Hemispheric preference in visuos- brain injured. Journal of Clinical Psychology,
patial processing: A complementaJ}' approach 56(4), 553-562.
with fMRI and lesion studies. Human Brain Nyberg, L., Winocur, G., & Moscovitch, M. (1997).
Mapping, 10(2), 80--86. Correlation between frontal lobe functions and
Nicholas, L. E., Brookshire, R. H., MacLennan, D. L., explicit and implicit stem completion in health
Schumacher, J. G., & Porrazzo, S. A. (1989). Re- elderly. Neuropsychology, 11(1), 70-76.
vised administration and scoring procedures for Obayashi, S., Matsushima, E., Ando, H., Ando, K.,
the Boston Naming Test and norms for non-brain- & Kojima, T. (2003). Exploratory eye move-
damaged adults. Aphasia, 5(6), 569-580. ments during the Benton Visual Retention Test:
Nicholas, M., Ohler, L., Albert, M., & Goodglass, Characteristics of visual behavior in schizophre-
H. (1985). Lexical retrieval in healthy aging. nia. Psychiatry and Clinical Neurosciences,
Cortex, 21, 595-606. 57(4), 409--415.
Nicholas, M., Ohler, L., Au, R., & Albert, M. L. Ober, B. A., Dronkers, N. F., Koss, E., Delis, D. C., &
(1996). On the nature of naming errors in aging Friedland, R. P. (1986). Retrieval from semantic
REFERENCES 583
memory in Alzheimer-type dementia. Journal of Ostrosky-Solis, F., Jaime, R. M., & Ardila, A.
Clinical and Experimental Neuropsychology, 8, (1998). Memory abilities during normal aging.
75-92. International Journal of Neuroscience, 93(1-2),
Obonsawin, M. C., Robertson, A., Crawford, J. R., 151-162.
Perera, C., Walker, S., Blackmore, L., et al. Ostrosky-Solis, F., Ardila, A., & Rosselli, M. (1999).
(1998). Non-mnestic cognitive function in the NEUROPSI: A brief neuropsychological test
scopolamine model of Alzheimer's disease. Hu- battery in Spanish with norms by age and edu-
man PsychopharmtJCOlogy Clinical and Experi- cational level. Journal of the International Neu-
mental, 13(6), 439--450. ropsychological Society, 5(5), 413-433.
O'Connell, M. E., & Tuokko, H. (2002). The 12- Oswald, W., & Roth, E. (1978). Der Zahlen-
item Buschke Memory Test: Appropriate for use Verbindungs Test (ZVT). Gottingen: Hogrefe.
across levels of impairment. Applied Neuropsy- Pachana, N. A., Boone, K. B., Miller, B. L., Cum-
chology, 9(4), 226-233. mings, J. L., & Berman, N. (1996). Comparison
O'Connor, M. K. (2002). The predictive utility of of neuropsychological functioning in Alzheimer's
the Hopkins Verbal Learning Test-Revised in disease and frontotemporal dementia. Journal of
older adults with depression versus dementia of the International Neuropsychological Society, 2,
the Alzheimer's type. Dissertation Abstracts In- 505-510.
ternational. Section B: The Sciences and Engi- Paivio, A., Yuille, J. C., & Madigan, S. A. (1968).
neering, 63(1-B), 543. Concreteness, imagery, and meaningfulness val-
O'Donnell, J.P., MacGregor, L.A., Dabrowski, J. J., ues for 925 nouns. Journal of Experimental
Oestreicher, J. M., et al. (1994). Construct va- Psychology 76(Monogr. Suppl.), 1-25.
lidity of neuropsychological tests of conceptual Palmer, B. W., Boone, K. B., Chang, L., Lee, A., &
and attentional abilities. Journal of Clinical Black, S. (1994). Cognitive deficits and person-
Psychology, 50(4), 596-600. ality patterns in maternally versus paternally in-
O'Donnell, M. P., & Webb, M. G. (1986). Post- herited myotonic dystrophy. Journal of Clinical
ECT blood pressure rise and its relationship to and Experimental Neuropsychology, 16(5), 784-
cognitive and affective change. British Journal of 795.
Psychiatry, 149, 494-497. Palmer, C., Wolkenstein, B., LaRue, A., Swan, G.,
O'Donnell, P. 0., Kurtz, J., & Ramanaiah, N. V. & Smalley, S. (1994). Commingling analysis of
(1983). Neuropsychological test findings for memory performance in elderly men. Genetic
normal, learning-disabled and brain-damaged Epidemiology, 11, 443-449.
young adults. Journal of Consulting and Clinical Pan, J. W., Krupp. L. B., Elkins, L. E., & Coyle,
Psychology, 51(5), 726-729. P. K. (2001). Cognitive dysfunction lateralizes
Ojemann, G. A., Sutherling. W. W., Lesser, R. P., with NAA in multiple sclerosis. Applied Neuro-
Dinner, D. S., Jayakar, P., & Saint-Hilaire, J. M. psychology, 8(3), 155-160.
(1993). Cortical stimulation. In J. Engel (Ed.), Panek, P. E., Rush, M. C., & Slade, A. L. (1984).
Surgical treatment of the epilepsies (2nd ed., Locus of the age-Stroop interference relationship.
pp. 399--414). New York: Raven. Journal of Genetic Psychology,145(2), 209-216.
O'Leary, M. R., Donovan, D. M., & Chaney, E. F. Paniak, C. E., Shore, D. L., & Rourke, B. P. (1989).
(1977). The relationship of perceptual field ori- Recovery of memory after severe closed-head
entation to measures of cognitive functioning injury: Dissociations in recovery of memory pa-
and current adaptive abilities of alcoholics and rameters and predictors of outcome. Journal of
nonalcoholics. Journal of Neroous and Mental Clinical and Experimental Neuropsychology,
Disease, 165(4), 275-282. 11(5), 631-644.
Osterrieth, P. A. (1944). Le test de copie d'une figure Pantelis, C., Egan, G., Pipingas, A., Maruff, P.,
complexe. Archives de Psychologie, 30, 206-356. O'Keefe, G., Velakoulis, D., et al. (1996). Prac-
Osterrieth, P. A. (1993). The complex figure copy tice dependent alterations in activation of the
test. Clinical Neuropsychologist, 7(1), 3-21. anterior cingulate cortex during the Stroop task:
Ostrosky-Solis, F., Ardila, A., & Rosselli, M. (1997). A positron emission tomography study. Neuro-
NEUROPSI: Evaluaci6n Neuropsicologica Breve image, 3, S193.
en Espaiiol. Manual, Instructivo y Protocolo de Paolo, A. M., Troester, A. 1., Axelrod, B. N., &
Aplicaci6n [NEUROPSI: A brief neuropsycho- Koller, W. C. (1995). Construct validity of the
logical evaluation in Spanish. Manual, instruc- WCST in normal elderly and persons with Par-
tions, and application protocol]. Mexico City: kinson's disease. Archives of Clinical Neuropsy-
Bayer de Mexico. chology, 10(5), 463-473.
584 REFERENCES
Paolo, A. M., Axelrod, B. N., & Troester, A. I. Parallel distributed processing and neuropsy-
(1996a). Test-retest stability of the Wisconsin chology: A neural network model of Wisconsin
Card Sorting Test. Assessment, 3(2), 137-143. Card Sorting and Verbal Fluency. Neuropsy-
Paolo, A. M., Axelrod, B. N., Troester, A. 1., chology Review, 3(2), 213-233.
Blackwell, K. T., & Koller, W. C. (1996b). Utility Parsons, 0. A. (1975). Brain damage in alcoholics:
of a Wisconsin Card Sorting Test Short Fonn in Altered states of unconsciousness. In M. M.
persons with Alzheimer's and Parkinsoo's dis- Gross (Ed.), Alcohol intoxication and with-
ease. Journal uf Clinical and Expe1'imental drawal. Experimental studies (No. 2). New York:
Neuropsychology, 18(6), 892-897. Plenum.
Paolo, A. M., Cluff, B. R., & Ryan, J. J. (1996c). Parsons, 0. A., Maslow, H. 1., Morris, F., &
Influence of perceptual organization and naming Denny, J. P. (1964). Trail Making Test perfor-
abilities on the Hooper Visual Organization Test. mance in relation to certain experimenter, test
Neuropsychiatry, Neuropsychology, and Behav- and subject variables. Perceptual and Motor
ioral Neurology, 9(4), 254--257. Skills, 19, 199--206.
Pardo,J. V.,Pardo,P.J.,Janer, K. W., &Raichle, M. E. Pauker, J.D. (1980). Nonnsforthe Halstead-Reitan
(1990). The anterior cingulate cortex mediates Neuropsychological Test Battery based on a
processing selection in the Stroop attentional nonclinical adult sample. Address presented at
conflict paradigm. Proceedings of the National the meeting of the Canadian Psychological As-
Actukmy of Science, 87,256-259. sociation, Calgary, Alberta, Canada.
Parellada, E., Cataqfau, A. M., Bernardo, M., Lo- Pauker, J.D. (1988). Constructing overlapping cell
mena, F., Gonzalex-Monclus, E., & Setnain, J. tables to maximize the clinical usefulness of
(1994). Prefrontal dysfunction in young acute nonnative test data: Rationale and an example
neuroleptic-naive schizophrenic patients: A from neuropsychology. Journal of Clinical Psy-
resting and activation SPECT study. Psychiatry chology, 44(6), 930-933.
Research: Neuroimaging, 55, 131-139. Paul, R. H., Cohen, R., Moser, D., Ott, B.,
Parellada, E., Catarineu, S., Catafau, A., Zawacki, T., & Gordon, N. (2001). Perfonnance
Bernardo, M., & Lomena, F. (2000). Psychopa- on the Hooper Visual Organizational Test in
thology and Wisconsin Card Sorting Test per- patients diagnosed with subcortical vascular de-
fonnance in young unmedicated schizophrenic mentia: Relation to naming perfonnance. Neu-
patients. Psychopathology, 33(1), 14-18. ropsychiatry, Neuropsychology, and Behavioral
Parkin, A. J., & Java, R. I. (1999). Deterioration of Neurology, 14(2), 93-97.
frontal lobe function in nonnal aging: Influences Paul,R. H.,Cohen,R.A.,Moser, D.J.,Zawacki, T. M.,
of fluid intelligence versus perceptual speed. & Gordon, N. (2002). The serial position effect
Neuropsychology, 13(4), 539--545. in mild and moderately severe vascular demen-
Parkin, A. J., & Lawrence, A. (1994). A dissociation tia. Journal of the International Neuropsycholo-
in the relation between memory tasks and frontal gical Society, 8(4), 584-587.
lobe tests in the normal elderly. Neuropsycho- Paulsen, J. S., Heaton, R. K., Sadek, J. R., Perry, W.,
logia, 32(12), 1523--1532. et al. (1995a). The nature of learning and
Parkin, A. J., & Walter, B. M. (1991). Aging, short- memory impainnents in schizophrenia. Journal
tenn memory, and frontal dysfunction. Psycho- of the International Neuropsychological Society,
biology, 19(2), 175-179. 1(1), 88--99.
Parkin, A. J., & Walter, B. M. (1992). Recollective Paulsen, J. S., Salmon, D. P., Monsch, A. U.,
experience, nonnal aging, and frontal dysfunc- Butters, N., et al. (1995b). Discrimination of
tion. Psychology and Aging, 7(2), 290-298. cortical from subcortical dementias on the basis
Parkinson, S. R., Inman, V. W., & Dannenbaum, S. E. of memory and problem-solving tests. Journal of
(1985). Adult age differences in short-tenn for- Clinical Psychology, 51(1), 48--58.
getting. Acta Psychologica, 60, 83-101. Peaker, A., & Stewart, L. E. (1989). Rey's Auditory
Parks, R. W., Loewenstein, D. A., Dodrill, K. L., Verbal Learning Test-a review. In J. R. Craw-
Barker, W. W., Yoshii, F., Chang, J. Y., et al. ford & D. M. Parker (Eds.), Developments in
(1988). Cerebral metabolic effects of a verbal clinical and experimental Neuropsychology. New
fluency test: A PET scan study. Journal of York: Plenum.
Clinical and Experimental Neuropsychology, 10, Peirson, A. R., & Jansen, P. (1997). Comparability
565-575. of the Rey-Osterrieth and Taylor fonns of the
Parks, R. W., Levine, D. S., Long, D. L., Crockett, D. J., Complex Figure Test. Clinical Neuropsycholo-
Dalton, I. E., Weingartner, H., et al. (1992). gist, 11(3), 244-248.
REFERENCES 585
Pendleton, M. G., & Heaton, R. K. (1982). A blood flow at rest and during a verbal fluency
comparison of the Wisconsin Card Sorting Test task. Journal of Affective Disorders, 28(4),
and the Category Test. Journal of Clinical Psy- 233-240.
chology, 38, 392-396. Piatt, A. L., Fields, J. A., Paolo, A. M., Koller, W. C., &
Pendleton, M. G., Heaton, R. K., Lehman, R. A. Troester, A. I. (1999a). Lexical, semantic, and
W., & Hulihan, D. (1982). Diagnostic utility of action verbal fluency in Parkinson's disease
the Thurstone Word Fluency Test in neu- with and without dementia. Journal of Clinical
ropsychological evaluations. Journal of Clinical and Experimental Neuropsychology, 21(4),
Neuropsychology, 4, 307-317. 435-443.
Pentz, C. A., Elias, M. F., Wood, W. G., Piatt, A. L., Fields,J. A., Paolo, A. M.,&Troester,A. I.
Schultz, N. A., & Dineen, J. (1979). Relationship (1999b). Action (verb naming) fluency as an ex-
of age and hypertension to neuropsychological ecutive function measure: Convergent and di-
test performance. Experimental Aging Research, vergent evidence of validity. Neuropsychologia,
5(4), 351-372. 37(13), 1499-1503.
Perlmuter, L. C., Tun, P., Sizer, N., McGlinchey, Pieniadz, J., & Kelland, D. (2001). Reporting scores
R. E., & Nathan, E. M. (1987). Age and diabetes in neuropsychological assessment: Ethnicality,
related changes in verbal fluency. Experimental validity, practicality, and more. In C. G. Armen-
Aging Research, 13, 9-14. gol, E. Kaplan, & E. Moes (Ed.), The consumer-
Perret, E. (1974). The left frontal lobe of man and oriented neuropsycholo,g.cal report. Lutz, FL:
the suppression of habitual responses in verbal Psychological Assessment Resources.
categorical behaviour. Neuropsychologia, 12, Pierce, T. W., Elias, M. F., Keohane, P. J.,
323-330. Podraza, A.M., Robbins, M.A., & Schultz, N. R.
Perrine, K. (1993). Differential aspects of concep- (1989). Validity of a short form of the Category
tual processing in the Category Test and Test in relation to age, education and gender.
Wisconsin Card Sorting Test. Journal of Clini- Experimental Aging Research, 15(3), 137-141.
cal and Experimental Neuropsychology, 15, Pihlajamaeki, M., Tanila, H., Hanninen, T., Koe-
461-473. noenen, M., Laakso, M., Partanen, K., etal. (2000).
Petersen, R. C., Smith, G., Kokmen, E., lvnik, R. J., Verbal fluency activates the left medial temporal
et al. (1992). Memory function in normal aging. lobe: A functional magnetic resonance imaging
Neurology, 42(2), 396-401. study. Annals of Neurology, 47(4), 470-476.
Petersen, R. C., Smith, G. E., Waring, S.C., Ivnik, R. J., Pimental, P. A. & Ross, C. (2003). ROCF produc-
Tangalos, E. G., & Kokmen, E. (1999). Mild tions in right- and left-hemisphere lesion pa-
cognitive impairment: Clinical characterization tients. In J. A. Knight (Ed.), The handbook of
and outcome. Archives of Neurology, 56(3), Rey-Osterrieth Complex Figure usage: Clinical
303-308. and research applications. Lutz, FL: Psycholog-
Peterson, B., Anderson, A., Skudlarski, P., Zhang, H., ical Assessment Resources.
& Gore, J. (1996). An fMRI study of the Stroop Pineda, D. A., & Merchan, V. (2003). Executive
effect. Neuroimllge, 3, S195. function in young Colombian adults. Interna-
Peterson, L. R., & Peterson, M. J. (1959). Short- tional Journal of Neuroscience, 113(3), 397-410.
term retention of individual verbal items. Journal Pirozzolo, F. J., Hansch, E. C., Mortimer, J. A.,
of Experimental Psychology, 58, 193-198. Webster, D. D., & Kuskowski, M. A. (1982).
Peynirciglu, Z. F., Thompson, J. L. W., &Tanielian, T. B. Dementia in Parkinson disease: A neuropsycho-
(2000). Improvement strategies in free-throw logical analysis. Brain and Cognition, 1, 71-83.
shooting and grip-strength tasks. Journal of Gen- Polubinski, J. P., & Melamed, L. E. (1986). Ex-
eral Psychology, 127(2), 145-156. amination of the sex difference on a Symbol
Phelps, E. A., Hyder, F., Blamire, A. M., & Shul- Digit Substitution Test. Perceptual and Motor
man, R. G. (1997). fMRI of the prefrontal cortex Skills, 62, 975-982.
during overt verbal fluency. Neuroreport: An Ponsford, J., & Kinsella, G. (1992). Attentional
International Journal for the Rapid Communi- deficits following closed-head injury. Journal of
cation of Research in Neuroscience, 8(2), Clinical and Experimental Neuropsychology.
561-565. 14(5), 822-838.
Philpot, M. P., Banerjee, S., Needham-Bennett, H., Pontius, A., & Yudowitz, B. (1980). Frontal lobe
Costa, D. C., & Ell, P. J. (1993). 99mTc-HMPAO system dysfunction in some criminal actions as
single photon emission tomography in late life shown in the narratives test. Journal of Nervous
depression: A pilot study of regional cerebral and Mental Disease, 168, 111-117.
586 REFERENCES
Ponton, M. 0., & Ardila, A. (1999). The future of Price, L. J., Fein, G., & Feinberg, I. (1980). Neu-
neuropsychology with Hispanic populations in ropsychological assessment of cognitive JUnction
the United States. Archives of Clinical Neuro- in the elderly. In L. W. Poon (Ed.), Aging in the
psychology, 14(1), 565--580. 1980's. Washington, DC: American Psychologi-
Ponton, M. 0., Satz, P., Herrera, L., Ortiz, F., cal Association Press.
Urrutia, C. P., Young, R., et al. (1996). Nonna- Prigatano, G. P., & Borgaro, S. R. (2003). Quali-
tive data stratified by age and education for the tative features of finger movement during the
Neuropsychological Screening Battery for His- Halstead finger oscillation test following trau-
panics (NeSBHIS): Initial report. Journal of matic brain injury. Journal of the International
International Neuropsychological Society, 2(2), Neuropsychological Society, 9(1), 128-133.
96-104. Prigatano, G. P., & Parsons, 0. A. (1976). Rela-
Ponton, M. 0., Gonzalez, J. J., Hernandez, 1., tionship of age and education to Halstead test
Herrera, L., & Higareda, I. (2000). Factor perfonnance in different patient populations.
analysis of the Neuropsychological Sc~;eening Journal of Consulting and Clinical Psychology,
Battery for Hispanics (NeSBHIS). [Special 44(4}, 527-533.
Issue: Assessment of Spanish-speaking popula- Prigatano, G. P., Parsons, 0. A., Levin, D. C.,
tions}. Applied Neuropsychology, 7(1), 3i-39. Wright, E., & Hawryluk, G. (1983). Neuro-
Poreh, A.M., & Shye, S. (1998). Examination of the psychological test perfonnance in mildly hypox-
global and local features of the Rey-Osrerrieth emic patients with chronic obstructive pulmonary
Complex Figure using faceted smallest space disease. Journal of Consulting and Clinical Psy-
analysis. Clinical Neuropsychologist, 12(4}, chology, 51(1), 108-116.
453-467. Psychological Assessment Resources (1990). Wis-
Poreh, A. M., Ross, T. P., & Whitman, R. D. consin Card Sorting Test: Scoring program
(1995). Reexamination of executive functions in (Version 3.0). Odessa, FL: Author.
psychosis-prone college students. Personality Puckett, J. M:, & Lawson, W. M. (1989). Absence
and Individual Differences, 18(4), 535-539. of adult age differences in forgetting in Brown-
Portin, R., Saarijarvi, S., Joukamaa, M., & Salo- Peterson task. Acta Psychologica, 72, 159-175.
kangas, R. K. R. (1995). Education, gender and Puente, A. E., & Ardila, A. (2000). Neuropsycho-
cognitive perfonnance in a 62-year-old nonnal logical assessment of Hispanics. In E. Fletcher-
population: Results from the Turva Project. Janzen, T. L. Strickland, et al. (Eds.), Handbook
Psychological Medicine, 25, 1295-1298. of cross-cultural neuropsychology. Critical issues
Portin, R., Polo-Kantola, P., Polo, 0., Koskinen, T., in neuropsychology. Amsterdam: Kluwer
Revonsuo, A., Irjala, K., et al. (1999). Serum Academic.
estrogen level, attention, memory and other Pujol, J., Vendrell, P., Dues, J., Kulisevsky, J.,
cognitive functions in middle-aged women. Cli- Marti-Valalta, J. L., Garcia, C., et al. (1996).
macteric, 2(2), 115-123. Frontal lobe activation during word generation
Powell, G. E. (1979). The relationship between studied by functional MRI. Acta Neurologica
intelligence and verbal and spatial memory. Scandinavica, 93, 403-410.
Journal of Clinical Psychology, 35(2}, 335--340. Qualls, C. E., Bliwise, N. G., & Stringer, A. Y.
Power, D. G., Logue, P. E., McCarty, S. M., Ro- (2000). Short fonns of the Benton Judgment of
senstiel, A. K., & Ziezat, H. A. (1979). Inter-rater Line Orientation Test: Development and psy-
reliability of the Russell revision of the Wechsler chometric properties. Archives of Clinical Neu-
Memory Scale: An attempt to clarify some am- ropsychology, 15(2), 159-163.
biguities in scoring. Journal of Clinical Neuro- Query, W. T. (1979). Category Test score as related
psychology, 1, ~- to age in two brain-damaged populations. Jour-
Prakash, I. J., & Bhogle, S. (1992). Benton's Visual nal of Clinical Psychology, 35(4), 802--804.
Retention Test: Nonns for different age goups. Query, W. T., & Berger, R. A. (1980). AVLT memory
Journal of the Indian Academy of Applied Psy- scores as a function of age among general medical,
chology. 18(1-2), 33--36. neurologic and alcoholic patients. Journal of
Prevey, M., Delaney, R., Cramer, J., Mattson, R., & Clinical Psychology, 36(4), 1009-1012.
VA Epilepsy Cooperative Study 264 Group Query, W. T., & Megran, J. (1983). Age-related
(1998). Complex partial and secondarily gener- nonns for AVLT in a male patient population.
alized seizure patients: Cognitive functioning Journal of Clinical Psychology, 39(1}, 136-138.
prior to treatment with antiepileptic medication. Query, W. T., & Megran, J. (1984). Influence of
Epilepsy Research, 30(1-9}. depression and alcoholism on learning, recall
REFERENCES 587
and recognition. Journal of Clinical Psychology, Rao, S. M., Leo, G. J., Ellington; L., Nauertz, T.,
40(4), 1097-1100. Bernardin, L., & Unverzagt, F. (1991b). Cogni-
Radanovic, M., Azambuja, M., Mansur, L. L., tive dysfunction in multiple sclerosis: II. Impact
Porto, C. S., & Scaff, M. (2003). 1balamus and on employment and social functioning. Neurol-
language: interface with attention, memory and ogy, 41(5), 692-696.
executive functions. Arquivos de Neuro-psi- Rapport, L. J., Dutra, R. L., Webster, J. S., Charter, R., &
quiatria, 61(1), 34-42. Morrill, B. (1995). Hemispatial deficits on the
Rahman, Q., & Wilson, G. D. (2003). Large sexual- Rey-Osterrieth Complex Figure drawing. Clini-
orientation-related differences in performance cal Neuropsychologist, 9(2), 169-179.
on mental rotation and judgement of line ori- Rapport, L. J., Charter, R. A., Dutra, R. L.,
entation tasks. Neuropsychology, 17(1), 25--31. Farchione, T. J., & Kingsley, J. J. (1997). Psy-
Raine, A., Lencz, T., Reynolds, G. P., Harrison, G., chometric properties of the Rey-Osterrieth
Sheard, C., Medley, I., et al. (1992). An evaluation Complex Figure: Lezak-Osterrieth versus Den-
of structural and functional prefrontal deficits in man scoring systems. Clinical Neuropsycholo-
schizophrenia: MRI and neuropsychological gist, 11(1), 46-53.
measures. Psychiatry Research: Neuroimaging, Rapport, L. J., Van Voorhis, A., Tzelepis, A., &
45(2), 123-137. Friedman, S. R. (2001). Executive functioning in
Randall, C. M., Dickson, A. L., & Plasay, M. T. adult attention-deficit hyperactivity disorder.
(1988). 1be relationship between intellectual Clinical Neuropsychologist, 15(4), 479-491.
function and adult performance on the Benton Rapport, L. J ., & Webster, J. S. (2003). Assessment of
Visual Retention Test. Cortex, 24(2), 277-289. unilateral neglect using the ROCF. In J. A. Knight
Randolph, C. (1998). Repeatable Battery for the (Ed.), The hondbook of Rey-Osterrieth Complex
Assessment of Neuropsychological Status: Man- Figure usage: Clinical and research applications.
ual. San Antonio, TX: Psychological Corporation. Lutz, FL: Psychological Assessment Resources.
Randolph, C., Braun, A. R., Goldberg, T. E., & Raskin, S. A., Borod, J. C., Wasserstein, J., Bodis-
Chase, T. N. (1993). Semantic fluency in Alz- Wollner, 1., Coscia, L. & Yahr, M.D. (1990).
heimer's, Parkinson's and Huntington's disease: VtSuospatial orientation in Parkinson's disease. In-
Dissociation of storage and retrieval failures. ternotional Journal ofNeuroscience, 51(1-2), 9-18.
Neuropsychology, 7, 82--88. Raskin, S. A., Sliwinski, M., & Borod, J. C. (1992).
Randolph, C., Tierney, M. C., Mohr, E., & Chase, Clustering strategies on tasks of verbal fluency in
T. N. (1998). 1be Repeatable Battery for the As- Parkinson's disease. Neuropsychologia, 30(1),
sessment of Neuropsychological Status (RBANS): 95-99.
Preliminary clinical validity. Journal ofClinical and Rasmussen, K., Jeppesen, H. J., & Sabroe, S.
Experimental Neuropsychology, 20(3), 310--319. (1993). Psychometric tests for assessment of
Randolph, C., Lansing, A. E., Ivnik, R. J., Cullum, brain function after solvent exposure. American
C. M., & Hermann, B. P. (1999). Determinants Journal of Industrial Medicine, 24(5), 553-565.
of confrontation naming performance. Archives Rasmusson, D. X, Bylsma, F. W., & Brandt, J.
of Clinical Neuropsychology, 14(6), 489-496. (1995). Stability of performance on the Hopkins
Rao, S. L., & Andrade, C. (1998). Selective Re- Verbal Learning Test. Archives of Clinical Neu-
minding Test to measure verbal and visual ropsychology, 10(1), 21-26.
memory. Indian Journal of Clinical Psychology, Rasmusson, D. X., Carson, K. A., Brookmeyer, R.,
25(2), 149-153. Kawas, C., & Brandt, J. (1996). Predicting rate of
Rao, S. M., Mittenberg, W., Bernardin, L., cognitive decline in probable Alzheimer's dis-
Haughton, V., Leo, G. J. (1989). Neuropsycho- ease. Brain and Cognition. Special Issue: The
logical test findings in subjects with Leukoaraisis. dementias, 31(2), 133-147.
Archives of Neurology, 46, 40-44. Rasmusson, D. X., Zonderman, A. B., Kawas, C., &
Rao, S. M., & Cognitive Function Study Group of Resnick, S. M. (1998). Effects of age and de-
the National Multiple Sclerosis Society (1990). A mentia on the Trail Making Test. Clinical Neu-
manual for the Brief Battery of Neuropsycholog- ropsychologist, 12(2), 169-178.
ical Tests in multiple sclerosis. Milwaukee: Rathbun, J., & Smith, A. (1982). Comment on the
Medical College of Wisconsin. validity of Boyd's validation study of the Hooper
Rao, S.M., Leo, G. J., Bernardin, L., & Unverzagt, F. Visual Organization Test. Journal of Consulting
(1991a). Cognitive dysfunction in multiple scle- and Clinical Psychology, 50, 281-283.
rosis: I. Frequency, patterns, and prediction. Rattan, G., Dean, R. S., & Fischer, W. E. (1986).
Neurology, 41(5), 685-691. Response time as a dependent measure on the
588 REFERENCES
Category Test of the Halstead-Reitan Neu- Reinprecht, F., Elmstahl, S., Janzon, L., & Andre-
ropsychological Test Battery. Archives of Clini- Petersson, L. (2003). Hypertension and changes
cal Neuropsychology, 1(2), 17~182. of cognitive function in 81-year-old men: A
Ravdin, L. D., Katzen, H. L., Agrawal, P., & Relkin, N. R. 13-year follow-up of the population study "Men
(2003). Letter and semantic fluency in older hom in 1914," Sweden. Journal of Hypertension,
adults: Effects of mild depressive symptoms and 21,57-66.
age-stratified normative data. Clinical Neuro- Reitan, R. M. (1955a). Certain differential effects of
psychologist, 17(2), 19~202. left and right cerebral lesions in human adults.
Ravnkilde, B., Videbech, P., Rosenberg, R., Journal of Comparative and Physiological Psy-
Gjedde, A., & Gade, A. (2002). Putative tests of chology, 48, 474-477.
frontal lobe function: A PET-study of brain ac- Reitan, R. M. (1955b). Investigation of the validity
tivation during Stroop's Test and verbal fluency. of Halstead's measures of biological intelligence.
Journal of Clinical and Experimental N~ropsy Archives of Neurology and Psychiatry, 73, 28-35.
chology, 24(4), 534--547. Reitan, R. M. (1955c). The relation of the Trail
Raz, N., Gunning-Dixon, F., Head, D., Dupuls,J. H., Making Test to organic brain damage. Journal of
& Acher, J. D. (1998). Neuroanatomical corre- Consulting Psychology, 195, 393-394.
lates of cognitive aging: Evidence fro structural Reitan, R. M. (1955d). The distribution according
magnetic resonance imaging. Neuropsychology, to age of a psychologic measure dependent upon
12, 9~112. organic brain functions. Journal of Gerontology,
Razarli, J., Boone, K., Miller, B. L., Lee, A., & 10, 338--340.
Sherman, D. (2001). Neuropsychological per- Reitan, R. M. (1958). Validity of the Trail Making
formance of right- and left-frontotemporal de- Test as an indicator of organic brain damage.
mentia compared to Alzheimer's disease. Journal Perceptual and Motor Skills, 8, 271-276.
of the International Neuropsychological Society, Reitan, R. M. (1959). The comparative effects of
7(4), 468-480. brain damage on the Halstead impairment index
Reader, M. J., Harris, E. L., Schuerholz, L. J., & and the Wechsler-Bellevue scale. Journal of
Denckla, M. B. (1994). Attention deficit hyper- Clinical Psychology, 15, 281-285.
activity disorder and executive dysfunction. De- Reitan, R. M. (1964). Psychological deficits result-
velopmental Neuropsychology, 10(4), 493-512. ing from cerebral lesions in man. In J. M. War-
Rebok, G., Brandt, J., & Folstein, M. (1990). ren & K. A. Akert (Eds.), The frontal granular
Longitudinal cognitive decline in patients with corlex and behavior. New York: McGraw-Hill.
Alzheimer's disease. Journal of Geriatric Psy- Reitan, R. M. (1971). Trail Making Test results for
chiatry and Neurology, 3(2), 91-97. normal and brain-damaged children. Perceptual
Reed, H. B. C., & Reitan, R. M. (1962). The sig- and Motor Skills, 33(2), 57~81.
nificance of age in the performance of a complex Reitan, R. M. (1979). Manual for administration of
psychomotor task by brain-damaged and non- neuropsychological test batteries for adults and
brain-damaged subjects. Journal of Gerontology, children. Tucson, AZ: Neuropsychology Press.
17, 193-196. Reitan, R. M. (1985). The Holstead-Reitan Neuro-
Reed, H. B. C., & Reitan, R. M. (1963a). A com- psychological Test Battery. Tucson, AZ: Neuro-
parison of the effects of the nonnal aging process psychology Press.
with the effects of organic brain damage on Reitan, R. M., & Wolfson, D. (1985). The Halstead-
adaptive abilities. Journal of Gerontology, 18, Reitan Neuropsychological Test Battery. Theory
177-179. and clinical interpretation. Tucson, AZ: Neuro-
Reed, B. C., & Reitan, R. M. (1963b). Changes in psychology Press.
psychological test performance associated with Reitan, R. M., & Wolfson, D. (1988). Traumatic
the normal aging process. Journal of Gerontol- brain injury. Volume II: Recovery and rehabili-
ogy, 18, 271-274. tation. Tucson, AZ: Neuropsychology Press.
Regard, M., & Landis, T. (1994). The "smiley:" A Reitan, R. M., & Wolfson, D. (1993). The Holstead-
graphical expression of mood in right anterior ce- Reitan Neuropsychological Test Battery: Theory
rebral lesions. Neuropsychiatry, Neuropsychology, and clinical interpretation (2nd ed). Tucson, AZ:
and Behavioral Neurology, 7(4), 303-307. Neuropsychology Press.
Regard, M., Strauss, E., & Knapp, P. (1982). Reitan, R. M., & Wolfson, D. (1995a). Category
Children's production on verbal and non~verbal Test and Trail Making Test as measures of
fluency tasks. Perceptual and Motor Skills, 55, frontal lobe functions. Clinical Neuropsycholo-
839-844. gist, 9(1), 50--56.
REFERENCES 589
Reitan, R. M., & Wolfson, D. (1995b). Influence of Rezai, K. (1988). Wisconsin Card Sorting Test.
age and education on neuropsychological test Version 1.1. Iowa City: University of Iowa.
results. Clinical Neuropsychologist, 9(2), 151-158. Rich, J. B.• Troyer, A. K., Bylsma, F. W., & Brandt, J.
Reitan, R. M., & Wolfson, D. (2001). Critical (1999). Longitudinal analysis of phonemic clus-
evaluation of "Assessment: Neuropsycholgical tering and switching during word-list generation in
testing of adults." Archives of Clinical Neuro- HWltington's disease. Neuropsychology, 13(4),
psychology, 16(3), 215-226. 525-531.
Reiter, J. C. (2000). Measuring cognitive processes Richardson, E. D., & Marottoli, R. A. (1996).
Wlderlying picture naming in Alzheimer's and Education-specific normative data on common
cerebrovascular dementia: A general processing neuro-psychological indices for individuals older
tree approach. Journal of Clinical and Experi- than 75 years. Clinical Neuropsychologist. 10(4),
mental Neuropsychology, 22(3), 351-369. 375-381.
Rempfer, M. V., Hamera, E. K., Brown, C. E., & Ricker, J. H., & Axelrod, B. N. (1994). Analysis of
Cromwell, R. L. (2003). The relations between an oral paradigm for the Trail Making Test. As-
cognition and the independent living skill of sessment, 1(1). 47-51.
shopping in people with schizophrenia. Psychi- Ricker, J. H., & Axelrod, B. N. (1995). Hooper
atry Research, 117(2), 103-112. Visual Organization Test: Effects of object
Rende, B., Ramsberger, G., & Miyake, A. (2002). naming ability. Clinical Neuropsychologist, 9(1),
Commonalities and differences in the working 57-62.
memory components Wlderlying letter and cat- Ricker, J. H., Axelrod, B. N., & Houtler, B. D.
egory fluency tasks: A dual-task investigation. (1996). Clinical validation of the oral Trail
Neuropsychology, 16(3), 309--321. Making Test. Neuropsychiatry, Neuropsychol-
Resnick, S. M., Trotman, K. M., Kawas, C., & ogy. and Behavioral Neurology, 9(1), 50-53.
Zonderman, A. B. (1995). Age-associated chan- Rimel, R. W., Giordani, B., Barth, J. T., & Jane, J. A.
ges in specific errors on the Benton VISual Re- (1982). Moderate head injury: Completing the
tention Test. Journals of Gerontology. Series B: clinical spectrum of brain trauma. Neurosurgery,
Psychological Sciences and Social Sciences, 11(3), 344-351.
508(3), P171-P178. Ripich, D. N., Petrill, S. A., Whitehouse, P. J., &
Resnick, S. M., Metter, E. J., & Zonderman, A. B. Ziol, E. W. (1995). Gender differences in lan-
(1997). Estrogen replacement therapy and lon- guage of AD patients: A longitudinal study.
gitudinal decline in visual memory: A possible Neurology, 45,299--302.
protective effect? Neurology, 49(6), 1491-1497. Risser, M. G., & Bowers, T. G. (1993). Cognitive
Retzlaff, P., Butler, M., & Vanderploeg, R. D. and neuropsychological characteristics of atten-
(1992). Neuropsychological battery choice and tion deficit hyperactivity disorder children re-
theoretical orientation: A multivariate analysis. ceiving stimulant medications. Perceptual and
Journal of Clinical Psychology, 48(5), 666--672. Motor Skills, 77(3, Pt 1), 1023-1031.
Rey, A. (1941). L'examen psychologique dans les Ritchie, K., & Hallerman, E. (1989). Cross-valida-
cas d'encephalopathie traumatique. Archives de tion of a dementia screening test in a heteroge-
Psychologie, 28, 286--340. neous population. International Journal of
Rey, A. (1964). L'examen clinique en psychologie. Epidemiology, 18(3), 717-719.
Paris: Presses Universitaires de France. Robert, P. H., Lafont, V., Medecin, 1., Berthet, L.,
Rey, A., & Osterrieth, P. A. (1993). Translations of Thauby, S., Baudu, C., et al. (1998). Clustering
excerpts from Andre Rey's "Psychological ex- and switching strategies in verbal fluency tasks:
amination of traumatic encephalopathy" and Comparison between schizophrenics and healthy
P. A. Osterrieth's ''The Complex Figure Copy adults. Journal of the International Neu-
Test." Clinical Neuropsychologist, 7(1), 4-21. ropsychological Society, 4(6), 539-546.
Rey, G. J., & Benton, A. L. (1991). Examen de Roberts, P. M., Garcia, L. J., Desrochers, A., &
afasia multilingue. Iowa City: AJA Associates. Hernandez, D. (2002). English performance of
Rey, G. J., Feldman, E., Rivas-Vazquez, R., Levin, proficient bilingual adults on the Boston Naming
B. E., & Benton, A. (1999). Neuropsychological test Test. Aphasiology, 16(4-6), 635-645.
development and normative data on Hispanics. Ar- Robertson-Tchabo, E. A., & Arenberg, D. (1989).
chives ofClinical Neuropsychology, 14(7), 593-601. Assessment of memory in older adults. In
Reynolds, C. R. (2002). Comprehensive Trail Mak- T. HWlt & J. Clyde (Eds.), Testing older adults:
ing Test. Lutz, FL: Psychological Assessment A reference guide for geropsychological assess-
Resources. ments. Austin, TX: Pro-Ed.
590 REFERENCES
Robertson-Tchabo, E. A., Arenberg, D., Tobin, J.D., & Roman, G. C., Tatemichi, T. K., Erkinjuntti, T.,
Plotz, J. B. (1986). A longitudinal study of cog- Cummings, J. L., Masden, J. C., Garcia, J. H.,
nitive performance in noninsulin dependent et al. (1993). Vascular dementia: Diagnostic cri-
(type II) diabetic men. Experimental Gerontol- teria for research studies. Report of the NINDS-
ogy, 21(4-5), 459-467. AIREN International Workshop. Neurology,
Robinson, A. L., Heaton, R. K., Lehman, R. A., & 43(2), 250-260.
Stilson, D. (1980). The utility of the Wisconsin Rosen, W. G. (1980). Verbal fluency in aging and
Card Sorting Test in detecting and localizing dementia. Journal of Clinical Neuropsychology,
frontal brain lesions. Journal of Consul~ng and 2(2), 135-146.
Clinical Psychology, 48, 605-614. Rosenberg, S., Ryan, J. J., & Prifitera, A. (1984).
Robinson, L. J., Kester, D. B., Saykin~ A. J., Rey Auditory-Verbal Learning Test performance
Kaplan, E. F., et al. (1991). Comparison of two of patients with and without memory im-
short forms of the Wisconsin Card Sorting Test. pairment. Journal of Clinical Psychology, 40(3),
Archives of Clinical Neuropsychology, 6(1-2), 785-787.
27-33. Rosenfeld, B., Sands, S. A., & Van Gorp, W. G.
Robinson-Whelen, S. (1992). Benton Visual Re- (2000). Have we forgotten the base rate prob-
tention Test performance among normal and lem? Methodological issues in the detection of
demented older adults. Neuropsycholo~, 6(3), distortion. Archives of Clinical Neuropsychology,
261-269. 15(4), 349--359.
Rochford, J., Grant, I., & LaVigne, G. (1977). Rosenthal, R. (1979). The "file drawer" problem
Medical students and drugs: Further . neuro- and tolerance for null results. Psychological
psychological and use pattern considerations. Bulletin, 86, 638--641.
International Journal of the Addictions, 12(8), Rosenthal, R. (1983). Assessing the statistical and
1057-1065. social importance of the effects of psychother-
Rodriguez-Aranda, C. (2003). Reduced writing and apy. Journal of Consulting and Clinical Psy-
reading speed and age-related changes in verbal chology, 51, 4-13.
fluency tasks. Clinical Neuropsychologist, 17(2), Rosenthal, R. (1984). Meta-analytic procedures for
203-215. social research. Beverly Hills, CA: Sage.
Rogers, R. D., Andrews, T. C., Grasby, P. M., Rosenthal, R. (1995). Writing meta-analytic re-
Brooks, D. J., & Robbins, T. W. (2000). Con- views. Psychological Bulletin, 118(2), 183-192.
trasting cortical and subcortical activation pro- Ross, S. R., Millis, S. R., & Rosenthal, M. (1997).
duced by attentional-set shifting and reversal Neuropsychological prediction of psychosocial
learning in humans. Journal of Cognitive Neu- outcome after traumatic brain injury. Applied
roscience, 12, 142--162. Neuropsychology, 4(3), 165-170.
Rohling, M. L., Langhinrichsen-Rohling, J., & Ross, T. P. (2003). The reliability of cluster and
Miller, L. S. (2003a). Actuarial assesslllent of switch scores for the Controlled Oral Word As-
malingering: Rohling's interpretative method. In sociation Test. Archives of Clinical Neuropsy-
R. D. Franklin (Ed.), Prediction in forensic and chology, 18(2), 153-164.
neuropsychology: Sound statistical practices. Ross, T. P., & Lichtenberg, P. A. (1998). Expanded
Hillsdale, NJ: Lawrence Erlbaum. normative data for the Boston Naming Test for
Rohling, M. L., Williamson, D. J., Miller, L. S., & use with urban, elderly medical patients. Clinical
Adams, R. L. (2003b). Using the Halstead- Neuropsychologist, 12(4), 475-481.
Reitan Battery to diagnose brain damage: A Ross, T. P., Lichtenberg, P. A., & Christensen, B. K.
comparison of the predictive power of traditional (1995). Normative data on the Boston Naming
techniques to Rohling's interpretive Jllethod. Test for elderly adults in a demographically di-
Clinical Neuropsychologist, 17(4), 531-543. verse medical sample. Clinical Neuropsycholo-
Rollnik,J. D., Borsutzky, M., Huber, T. J., Mogk, H., gist, 9(4), 321-325.
Seifert, J., Emrich, H. M., et al. (2002). Short- Ross, T. P., Foard, E. L., Hiott, F. B., & Vincent, A.
term cognitive improvement in schizophrenics (2003). The reliability of production strategy
treated with typical and atypical neuroleptics. scores for the Ruff Figural Fluency Test. Archives
Neuropsychobiology, 45(2), 74--80. of Clinical Neuropsychology, 18(8), 879-891.
Roman, D. D., Edwall, G. E., Buchanan, R. J., & Rosselli, M., & Ardila, A. (1991). Effects of age,
Patton, J. H. (1991). Extended norms for the education, and gender on the Rey-Osterrieth
Paced Auditory Serial Addition Task. Clinical Complex Figure. Clinical Neuropsychologist, 5,
Neuropsychologist, 5(1), 33-40. 370-376.
REFERENCES 591
Rosselli, M., & Ardila, A. (1996). Cognitive effects Test (Adjusting-PSAT): Thresholds for speed of
of cocaine and polydrug abuse. Journal of Clin- information processing as a function of stimulus
ical and Experimental Neuropsychology, 18(1), modality and problem complexity. Archives of
122-135. Clinical Neuropsychology, 19(1), 131-143.
Rosselli, M., Ardila, A., Araujo, K., Weekes, V. A., Ruchinskas, R. (2003). Limitations of the Oral Trail
Caracciolo, V., Padilla, M., et al. (2000). Verbal Making Test in a mixed sample of older individ-
fluency and repetition skills in healthy older uals. Clinical Neuropsychologist, 17(2), 137-142.
Spanish-English bilinguals. [Special Issue: As- Ruchinskas, R., Broshek, D., Crews, W. D., Barth,
sessment of Spanish-speaking populations]. Ap- J., Francis, J., & Robbins, M. (2000). A neu-
plied Neuropsychology, 7(1), 17-24. ropsychological normative database for lung
Rosselli, M., Ardila, A., Bateman, J. R., & Guzman, M. transplant candidates. Journal of Clinical Psy-
(2001a). Neuropsychological test scores, academic chology in Medical Settings, 7(2), 107-112.
performance, and developmental disorders in Ruff, R. (1988). Ruff Figural Fluency Test. San
Spanish-speaking children. Developmental Neu- Diego: Neuropsychological Resources.
ropsychology, 20(1), 355-373. Ruff, R. M. (1994). What role does depression play
Rosselli, M., Ardila, A., Lubomski, M., Murray, S., & on the performance of the Ruff 2 & 7 Selective
King, K. (2001b). Personality profile and neu- Attention Test? Perceptual and Motor Skills,
ropsychological test performance in chronic 78(1), 63--66.
cocaine-abusers. International Journal of Neu- Ruff, R. M. (1996). Ruff Figural Fluency Test:
roscience, 110(1-2), 55-72. Professional manual. Odessa, FL: Psychological
Rosselli, M., Ardila, A., Salvatierra, J., Marquez, Assessment Resources.
M., Matos, L., & Weekes, V. A. (2002a). A cross- Ruff, M., & Allen, C. C. (1996). Ruff2 & 7 Selective
linguistic comparison of verbal fluency tests. Attention Test: Professional manual. Odessa, FL:
International Journal of Neuroscience, 112(6), Psychological Assessment Resources.
759-776. Ruff, R. M., & Crouch, J. A. (1991) Neuropsycho-
Rosselli, M., Ardila, A., Santisi, M. N., logical test instruments in clinical trials. In
Arecco, M.d. R., Salvatierra, J., Conde, A., et al. E. Mohr & P. Brouwers (Eds.), Handbook of
(2002b). Stroop effect in Spanish-English bilin- clinical trials: The neurobehavioral approach.
guals. Journal of the International Neu- Lisse: Swets & Zeitlinger.
ropsychological Society, 8(6), 819-827. Ruff, R. M., & Jurica, P. J. (2003). The ROCF and
Rosser, A., & Hodges, J. R. (1994). Initial letter and frontal lobe damage. In J. A. Knight (Ed.), The
semantic category fluency in Alzheimer's disease, handbook of Rey-Osterrieth Complex Figure
Huntington's disease, and progressive supra- usage: Clinical and research applications. Lutz,
nuclear palsy. Journal of Neurology, Neurosur- FL: Psychological Assessment Resources.
gery, and Psychiatry, 57(11), 1389-1394. Ruff, R. M., & Parker, S. B. (1993). Gender- and age-
Rossi, A., Arduini, L., Daneluzzo, E., Bustini, M., specific changes in motor speed and eye-hand
Prosperini, P., & Stratta, P. (2000). Cognitive coordination in adults: Normative values for the
function in euthymic bipolar patients, stabilized Finger Tapping and Grooved Pegboard Tests.
schizophrenic patients, and healthy controls. Perceptual and Motor Skills, 76, 1219-1230.
Journal of Psychiatric Research, 34, 333--339. Ruff, R. M., Evans, R. W., & Light, R. H. (1986a).
Rossini, E. D., & Karl, M. A. (1994). The Trail Automatic detection vs. controlled search: A
Making Test A and B: A technical note on paper-and-pencil approach. Perceptual and Mo-
structural nonequivalence. Perceptual and Motor tor Skills, 62(2), 407-416.
Skills, 78(2), 625-626. Ruff, R. M., Evans, R., & Marshall, L. F. (1986b).
Roth, E., Davidoff, G., Thomas, P., Doljanac, R., Impaired verbal and figural fluency after head
Dijkers, M., Berent S., et al. (1989). A controlled injury. Archives of Clinical Neuropsychology,
study of neuropsychological deficits in acute 1(2), 87-101.
spinal cord injury patients. Paraplegia, 27(6), Ruff, R. M., Levin, H. S., & Marshall, L. F. (1986c).
480-489. Neurobehavioral methods of assessment and the
Rounsaville, B. J., Jones, C., Novelly, R. A., & study of outcome in minor head injury. Journal
Kleber, H. (1982). Neuropsychological func- of Head Trauma Rehabilitation, 1(2), 43-52.
tioning in opiate addicts. Journal of Neroous and Ruff, R. M., Light, R. H., & Evans, R. W. (1987).
Mental Disease, 170(4), 209-216. The Ruff Figural Fluency Test: A normative
Royan, J., Tombaugh, T. N., Rees, L., & Francis, study with adults. Developmental Neuropsy-
M. (2004). The Adjusting-Paced Serial Addition chology, 3(1), 37-51.
592 REFERENCES
Ruff, R. M., Baser, C. A., Johnston, J. W., Halstead-Reitan Battery. In P. E. Logue & J. M.
Marshall, L. F., et al. (1989a). Neuropr;ycholo- Schear (Eds.), Clinical neuropsychology, a mul-
gical rehabilitation: An experimental stqdy with tidisciplinary approach (pp. 50-98). Springfield,
head-injured patients. Journal of Head Trauma IL: Thomas.
Rehabilitation, 4(3), 20-36. Russell, E. W. (1985). Comparison of the TPT 10
Ruff, R. M., Light, R. H., & Quayhagen, M. and 6 hole form board. Journal of Clinical Psy-
(1989b). Selective Reminding Tests: A normative chology, 41(1), 68-81.
study of verbal learning in adults. Joyrnal of Russell, E. W. (1987). A reference scale method for
Clinical and Experimental Neuropsy~hology, constructing neuropsychological test batteries.
11(4), 539-550. I Journal of Clinical and Experimental Neuropsy-
Ruff, R. M., Niemann, H., Allen, C. C., F~, C. E., chology, 9(4), 376--392.
& Wylie, T. (1992). The Ruff 2 & 7 ~elective Russell, E. W. (1988). Renorming Russell's version
Attention Test: !i neuropsychological apJ}Iication. of the Wechsler Memory Scale. Journal of
Perceptual and Motor Skills, 75(3, : Pt 2), Clinical and Experimental Neuropsychology,
1311-1319. . 10(2), 235-249.
Ruff, R. M., Marshall, L. F., Crouch, J.. Klaubt;r, M. R., Russell, E. W., & Levy, M. (1987). Revision of the
Levin H. S., Barth, J., et al. (1993). Pred,tctors of Halstead Category Test. Journal of Consulting
outcome following severe head trauma: :Follow- and Clinical Psychology, 55(6), 898-901.
up data from the Traumatic Coma Data Bank. Russell, E., & Starkey, R. (1993). Halstead-Russell
Brain Injury, 7(2), 101-111. ' neuropsychological evaluation system (HRNES).
Ruff, R. M., Allen, C. C., Farrow, C. E., Niemann, H., Los Angeles: Western Psychological Services.
& Wylie, T. (1994). Figural fluency: Differential Russell, E. W., & Starkey, R. I. (2001). Halstead-
impairment in patients with left versus right frontal Russell neuropsychological evaluation system-
lobe lesions. Archives ofClinical Neuropstjchology, revised (HRNES-R). Los Angeles: Western
9{1), 41-55. Psychological Services.
Ruff, R. M., Light, R. H., Parker, S. B., & Le~n. H. S. Russell, E. W., Neuringer, C., & Goldstein, G.
(1996). Benton Controlled Oral Word As$leiation (1970). Assessment of brain damage: A neu-
Test: Reliability and updated norms. Artihives of ropsychological key approach. New York: Wiley.
Clinical Neuropsychology, 11(4), 329-338. Rutschmann, J., Comblatt, B., & Erlenmeyer-
Ruff, R. M., Light, R. H., Parker, S. B., & Le'Jin, H. S. Kimling, L. (1980). Auditory recognition mem-
(1997). The psychological construct of word flu- ory in adolescents at risk for schizophrenia:
ency. Brain and Language, 57, 394-405. Report on a verbal continuous recognition task.
Ruffolo, L. F., Guilmette, T. J., & Willis; W. G. Psychiatry Research, 3(2), 151-161.
(2000). Comparison of time and error rates on Ryan, C. M., Morrow, L. A., Bromet, E. J., &
the Trail Making Test among patients with head Parkinson, D. K. (1987). Assessment of neu-
injuries, experimental malingerers, patients with ropsychological dysfunction in the workplace:
suspect effort on testing, and normal eontrols. Normative data from the Pittsburgh Occupa-
Clinical Neuropsychologist, 14(2), 223-230. tional Exposures Test Battery. Journal of Clini-
Ruffolo, J. S., Javorsky, D. J., Tremont, G., cal and Experimental Neuropsychology, 9(6),
Westervelt, H. J., & Stem, R. A. (2001). A com- 665-679.
parison of administration procedures for the Rey- Ryan, J. J., & Geisser, M. E. (1986). Validity and
Osterrieth Complex Figure: Flowcharts versus diagnostic accuracy of an alternate form of the
pen switching. Psychological Assessment, 13(3), Rey Auditory Verbal Learning Test. Archives of
299-305. Clinical Neuropsychology, 1, 209-217.
Ruggieri, R. M., Palermo, R., Vitello, G., Gennuso, M., Ryan, J. J., Rosenberg, S. J .. & Mittenberg, W.
Settipani, N., & Piccoli, F. (2003). Co~tive im- (1984). Factor analysis of the Rey Auditory-
pairment in patients suffering from relapsing- Verbal Learning Test. International Journal of
remitting multiple sclerosis with EDSS ~8.5. Acta Clinical Neuropsychology, 6(4), 239-241.
Neurologica Scandinavica, 108, 323-326; Ryan, J. J., Geisser, M. E., Randall, D. M., &
Russell, E. W. (1975). A multiple scoring method Georgemiller, R. J. (1986). Alternate form reli-
for the assessment of complex memory func- ability and equivalency of the Rey Auditory
tions. Journal of Consulting and Clinical Psy- Verbal Learning Test. Journal of Clinical and
chology, 43, 800-809. Experimental Neuropsychology. 8(5), 611-616.
Russell, E. W. (1984). Theory and development Ryan, J. J., Paolo, A. M., & Brungardt, T. M.
of pattern analysis methods related , to the (1990). Standardization of the Wechsler Adult
REFERENCES 593
Intelligence Scale-Revised for persons 75 years Salthouse, T. A., Atkinson, T. M., & Berish, D. E.
and older. Psychological Assessment, 2, 404-411. (2003). Executive functioning as a potential me-
Ryan, J. J., Paolo, A. M., & Skrade, M. (1992). Rey diator of age-related cognitive decline in nonnal
Auditory Verbal Learning Test performance of a adults. Journal of Experimental Psychology:
federal corrections sample with acquired im- General,132(4), 566-594.
munodeficiency syndrome. International Journal Sampson, H. (1956). Pacing and performance on a
of Neuroscience, 64, 177-181. serial addition task. Canadian Journal of Psy-
Rybakowski, J. & Borkowska, A. (2001). The effect chology, 10, 219-225.
of treatment with risperidone, olanzapine or Samuels, I., Butters, N., & Fedio, P. (1972). Short
phenothiazine on cognitive functions in patients tenn memory disorders following temporal lobe
with schizophrenia. International Journal of removals in humans. Cortex, 8, 283-298.
Psychiatry in Clinical Practice, 5, 249-256. Samuels, I., Butters, N., Fedio, P., & Cox, C.
Sackellares, D. K., & Sackellares, J. C. (2001). Im- (1980). Deficits in short-tenn auditory memory
paired motor function in patients with psychogenic for verbal material following right temporal re-
pseudoseizures. Eptlepsia, 42(12), 1600-1606. movals in humans. International Journal of Neu-
Sackett, D. L., Straus, S. E., Richardson, W. S., roscience, 11(2), 101-107.
Rosenberg, W., & Haynes, R. B. (2000). Evi- Sasher, T. M., & Fastenau, P. S. (2001). Prelimi-
dence-based medicine. New York: Churchill nary child nonnative data for the Extended
Livingstone. Complex Figure Test (ECFT). Clinical Neuro-
Sacks, T. L., Clark, C. R., Pols, R. G., & Geffen, L. B. psychologist, 15, 258.
(1991). Comparability and stability of perfor- Saoud, M., d'Amato, T., Gutknecht, C., Triboulet, P.,
mance of six alternate forms of the Dodrill-Stroop Bertaud, J.-P., Marie-Cardine, M., et al. (2000).
Color-Word Test. Clinical Neuropsychologist, 5, Neuropsychological deficit in siblings discordant
220-225. for schizophrenia. Schizophrenia Bulletin, 26(4),
Sagawa, K., Kawakatsu, S., Komatani, A., & 893-902.
Totsuka, S. (1990a). Frontality, laterality, and Satz, P. (1988). Neuropsychological testimony:
cortical-subcortical gradient of cerebral blood Some emerging concerns. Clinical Neuropsy-
How in schizophrenia: Relationship to symptoms chologist, 2, 89-100.
and neuropsychological functions. Neuropsy- Satz, P. (1993). Brain reserve capacity on symptom
chobiology, 24(1), 1-7. onset after brain injury: A formulation and re-
Sagawa, K., Kawakatsu, S., Shibuya, I., Oiji, A., view of evidence for threshold theory. Neuro-
Morinobu, S., Komatani, A., et al. (1990b). psychology, 7, 273-295.
Correlation of regional cerebral blood How with Satz, P., & Mogel, S. (1962). An abbreviation of the
performance on neuropsychological tests in WAIS for clinical use. Journal of Clinical Psy-
schizophrenic patients. Schizophrenia Research, chology, 18, 77-79.
3(4), 241-246. Savage, C. R., & Otto, M. W. (2003). Evaluating
Salthouse, T. A., & Fristoe, N. (1995). A process nonverbal memory in obsessive-<:ompulsive
analysis of adult age effects on a computer- disorder with the ROCF. In J. A. Knight (Ed.),
administered Trail Making Test. Neuropsychol- The handbook of Rey-Ostenieth Complex Figure
ogy, 9, 518--528. usage: Clinical and research applications. Lutz,
Salthouse, T. A., Fristoe, N., & Rhee, S. H. (1996). FL: Psychological Assessment Resources.
How localized are age-related effects on neu- Savage, C. R., Baer, L., Keuthen, N.J., Brown, H. D.,
ropsychological measures? Neuropsychology, Rauch, S. L., & Jenike, M. A. (1999). Organiza-
10(2), 272-285. tional strategies mediate nonverbal memory im-
Salthouse, T. A., Toth, J. P., Hancock, H. E., & pairment in obsessiv~ompulsive disorder.
Woodard, J. L. (1997). Controlled and automatic Biological Psychiatry, 45(7), 905-916.
fonns of memory and attention: Process purity Savage, C. R., Deckersbach, T., Wilhelm, S.,
and the uniqueness of age-related influences. Rauch, S. L., Baer, L., Reid, T., et al. (2000).
Journals of Gerontology. Series B: Psychological Strategic processing and episodic memory im-
Sciences and Social Sciences, 52B(5), P216-P228. pairment in obsessive compulsive disorder.
Salthouse, T. A., Toth, J., Daniels, K., Parks, C., Neuropsychology, 14(1), 141-151.
Pak, R., Wolbrette, M., et al. (2000). Effects of Savage, C. R., Deckersbach, T., Heckers, S.,
aging on efficiency of task switching in a variant Wagner, A. D., Schacter, D. L., Alpert, N. M.,
of the Trail Making Test. Neuropsychology, et al. (2001). Prefrontal regions supporting
14(1), 102-111. spontaneous and directed application of verbal
594 REFERENCES
learning strategies. Evidence from PET. Brain, Regional cerebral blood flow in Down syndrome
124(1), 219-231. adults during the Wisconsin Card Sorting Test:
Savage, R. M., & Gouvier, W. D. (1992). Rey Exploring cognitive activation in the context of
Auditory-Verbal Learning Test: The effects of poor performance. Biological Psychiatry, 45(9),
age and gender, and norms for delayed recall 1190-1196.
and story recognition trials. Archives of Clinical Schatz, P. (2001). Commentary on "Psychometric
Neuropsychology, 7, 407-414. concerns in neuropsychological testing'' [Special
Savard, R. J., Rey, A. C., & Post, R. M .. (1980). Issue: Controversies in neuropsychology]. Neuro-
Halstead-Reitan Category Test in bipolar and Rehabilitation, 16(4), 303.
unipolar affective disorders. Relationshi~ to age Schear, J. M. (1984). Neuropsychological assess-
and phase of illness. Journal of Nervaus and ment of the elderly in clinical practice. In P. E.
Mental Disease, 168(5), 297-304. Logue & J. M. Schear (Eds.), Clinical neuro-
Sawrie, S. M., Martin, R. C., Gilliam, F. G., psychology: A multidisciplinary approach.
Faught, R. E., Maton, B., Hugg, J. W., et al. Springfield, IL: Thomas.
(2000). Visual confrontation naming and hippo- Schear, J. M. (1986). Utility of half-credit scoring of
campal function: A neural network study using Russell's revision of the Wechsler Memory
quantitative 1H magnetic resonance spectros- Scale. Journal of Clinical Psychology, 42(5),
copy. Brain, 123(4), 770-780. 783-787.
Saxton, J., Ratcliff, G., Munro, C. A., Coffey, E. C., Schear, J. M., & Sato, S.D. (1989). Effects of visual
Becker, J. T., Fried, L., et al. (2000a). Normative acuity and visual motor speed and dexterity on
data on the Boston Naming Test and two cognitive test performance. Archives of Clinical
equivalent 30-item short forms. Clinical Neuro- Neuropsychology, 4(1), 25-32.
psychologist, 14(4), 526-534. Schloesser, R., Hutchinson, M., Joseffer, S.,
Saxton, J., Ratcliff, G., Newman, A., Belle, S., Rusinek, H., Saarimaki, A., Stevenson, J., et al.
Fried, L., Yee, J., et al. (2000b). Cognitive test (1998). Functional magnetic resonance imaging
performance and presence of subclinical cardio- of human brain activity in a verbal fluency task.
vascular disease in the cardiovascular health Journal of Neurology, Neurosurgery, and Psy-
study. Neuroepidemiology, 19(6), 312-319. chiatry, 64(4), 492-498.
Saxton, J. A., Becker, J. T., & Wisniewski, S. (2003). Schmidt, I. W., Brouwer, W. H., Vanier, M., &
The ROCF and dementia. In J. A. Knight (Ed.), Kemp, F. (1996). Flexible adaptation to chang-
The handbook of Rey-Osterrieth Complex Figure ing task demands in severe closed head injury
usage: Clinical and research application~. Lutz, patients: A driving simulator study. Applied
FL: Psychological Assessment Resources. Neuropsychology, 3(3-4), 155-165.
Scarone, S., Abbruzzese, M., & Gambini, 0. (1993). Schinidt, M. (1996). Rey Auditory and Verbal
The Wisconsin Card Sorting Test discriminates Learning Test: A handbook. Los Angeles: West-
schizophrenic patients and their siblings. em Psychological Services.
Schizophrenia Research, 10(2), 103-107. Schmidt, S. L., Oliveira, R. M., Rocha, F. R., &
Schaie, K. W. (1983). The Seattle Longitudinal Abreu-Villaca, Y. (2000). Influences of handed-
Study: A 21-year exploration of psychometric ness and gender on the grooved pegboard test.
intelligence in adulthood. InK. W. Schaie (Ed.), Brain and Cognition, 44(3), 445-454.
Longitudinal studies of adult psychological de- Schlnitt,F.A., Bigley,J. W., McKinnis,R., Logue,P. E.,
velopment (pp. 64-135). New York: Guilford. Evans, R. W., & Drucker, J. L. (1988). Neu-
Schaie, K. W., & Parham, I. A. (1977). Cohort- ropsychological outcome of zi.dovudine (AZT)
sequential analyses of adult intellectual de- treatment of patients with AIDS and AIDS-related
velopment. Developmental Psychology, 13(6), complex. New England Journal of Medicine,
649-653. 319(24), 1573-1578.
Schaie, K. W., & Strother, C. R. (1968a). Cognitive Schmitter-Edgecombe, M., Vesneski, M., &
and personality variables in college graduates of Jones, D. W. R. (2000). Aging and word-finding:
advanced age. In G. A. Talland (Ed.), Human A comparison of spontaneous and constrained
aging and behavior. New York: Acadeinie Press. naming tests. Archives of Clinical Neuropsy-
Schaie, K. W., & Strother, C. R. (1968b). A cross- chology, 15(6), 479-493.
sectional study of age changes in cognitive be- Schneider, W. (1989). Enhancing a standard ex-
havior. Psychological Bulletin, 70, 671-680. perimental delivery system (MEL) for advanced
Schapiro, M. B., Berman, K. F., Alexander, G. E., psychological experimentation. Behavioral Re-
Weinberger, D. R., & Rapoport, S. I. (1999). search Methods, 1nstroments, and Computers.
REFERENCES 595
Schonfield, A. D., Davidson, H., & Jones, H. (1983). Searight, H. R., Dunn, E. J., Grisso, T., Margolis,
An example of age-associated interference in R. B., & Gibbons, J. L. (1989). The relation of
memorizing. Journal ofGerontology, 38,204-210. the Halstead-Reitan Neuropsychological Battery
Schreiber, D. J., Goldman, H., Kleinman, K. M., to ratings of everyday functioning in a geriatric
Goldfader, P. R., & Snow, M. Y. (1976). The sample. Neuropsychology, 3, 135-145.
relationship between independent neuropsy- Segalowitz, S. J., Unsal, A., & Dywan, J. (1992).
chological and neurological detection and local- CNV evidence for the distinctiveness of frontal
ization of cerebral impairment. Journal of and posterior neural processes in a traumatic
Neroous and Mental Disease, 162(5), 360-365. brain-injured population. Journal of Clinical and
Schreiber, H., Rothmeier, J., Becker, W., Jurgens, R., Experimental Neuropsychology, 14, 545-565.
Born, J., Stolz-Bom, G., et al. (1995). Compara- Seidel, W. T. (1994). Applicability of the Hooper
tive assessment of saccadic eye movements, psy- Visual Organization Test to pediatric population:
chomotor and cognitive performance in Preliminary findings. Clinical Neuropsychologist,
schizophrenics, their first-degree relatives and 8, 59-68.
control subjects. Acta Psychiatrica Scandinavica, Seidenberg, M., Gamache, M. P., Smith, M.,
91, 195-201. Sackellares, J. C., Beck, N.C., Giordani, B., et al.
Schreiber, H. E., Javorsky, D. J., Robinson, J. E., & (1984). Subject variables and performance on
Stem, R. A. (1999). Rey-Osterrieth Complex the Halstead Neuropsychological Test Battery: A
Figure performance in adults with attention multivariate analysis. Journal of Consulting and
deficit hyperactivity disorder: A validation study Clinical Psychology, 52(4), 658-Q62.
of the Boston Qualitative Scoring System. Clin- Seidenberg, M., Hermann, B., Noe, A., & Wyler,
ical Neuropsychologist, 13(4), 509-520. A. R. (1995). Depression in temporal lobe epi-
Schretlen, D. J., Munro, C. A., Anthony, J. C., & lepsy: Interaction between laterality of lesion
Pearlson, G. D. (2003). Examining the range of and Wisconsin Card Sort performance. Neuro-
normal intraindividual variability in neu- psychiatry, Neuropsychology, and Behavioral
ropsychological test performance. Journal of the Neurology, 8(2), 81-87.
International Neuropsychological Society, 9(6), Seidenberg, M., Hermann, B., Wyler, A. R., Da-
864-870. vies, K., Dohan, F. C., Jr., & Leveroni, C. (1998).
Schuepbach, D., Goenner, F., Staikov, I., Mattie, H. P., Neuropsychological outcome following anterior
Hell, D., & Brenner, H.-D. (2002a). Temporal temporal lobectomy in patients with and without
modulation of cerebral hemodynamics under the syndrome of mesial temporal lobe epilepsy.
prefrontal challenge in schizophrenia: A tran- Neuropsychology, 12(2), 303-316.
scranial Doppler sonography study. Psychiatry Seidman, L. J., Yurgelun-Todd, D., Kreman, W. S..
Research: NeuroiTTUJging, 115(3), 155-170. Woods, B. T., Goldstein, J. M., Faraone, S. V.,
Schuepbach, D., Merlo, M. C. G., Goenner, F., et a!. (1994). Relationship between prefrontal
Staikov, 1., Mattie, H. P., Dierks, T., et al. and temporal lobe MRI measures to neu-
(2002b). Cerebral hemodynamic response in- ropsychological performance in chronic schizo-
duced by the Tower of Hanoi puzzle and the phrenia. Biological Psychiatry, 35, 235-246.
Wisconsin Card Sorting Test. Neuropsychologia, Seidman, L. J., Benedict, K. B., Biederman, J.,
40(1), 39-53. Bernstein, J. H., et al. (1995). Performance of
Schultheis, M. T., Caplan, B., Ricker, J. H., & children with ADHD on the Rey-Osterrieth
Woessner, R. (2000). Fractioning the Hooper: A Complex Figure: A pilot neuropsychological
multiple-choice response format. Clinical Neu- study. Journal of Child Psychology and Psychi-
ropsychologist, 14(2), 196-201. atry and Allied Disciplines, 36(8), 1459-1473.
Schwartz, M. S., & Ivnik, R. J. (1980, September). Seidman, L. J., Biederman, J., Faraone, S. V., We-
Wechsler Menwry Scale 1: Toward a nwre objec- ber, W., et al. (1997). Toward defining a neuro-
tive and systematic scoring system for the Logical psychology of attention deficit-hyperactivity
Menwn1 and Visual Reproduction subtests. Paper disorder: Performance of children and adoles-
presented at the meeting of the America! Psy- cents from a large clinically referred sample.
chological Association, Montreal, Canada. Journal of Consulting and Clinical Psychology,
Schwartz, R. H., Gruenewald, P. J., Klitzner, M., & 65(1), 150-160.
Fedio, P., (1989). Short-term memory impair- Sellers, A. H., & Nadler, J. D. (1992). A survey of
ment in cannabis-dependent adolescents. Amer- current neuropsychological assessment proce-
ican Journal of Diseases of Children, 143(10), dures used for different age groups. Psycho-
1214-1219. therapy in Private Practice, 11(3), 47-57.
596 REFERENCES
Seines, 0. A., Jacobson, L., Machado, A. M., and Experimental Neuropsychology, 7(3), 231-
Becker,J. T., Wesch,J., Miller, E. N.,etai: (1991). 238.
Nonnative data for a brief neuropsychological Shichita, K., Hatano, S., Ohashi, Y., Shibata, H., &
screening battery. Perceptual and Motot Skills, Matuzaki, T. (1986). Memory changes in the
73, 539--550. Benton Visual Retention Test between ages 70
Shapiro, A.M., Benedict, R. H. B., Schretlen, D., & and 75. Journal of Gerontology, 41 (3), 385-386.
Brandt, J. (1999). Construct and concurrent Shimamura, A. P., Salmon, D.P., Squire, L. R., &
validity of the Hopkins Verbal Learning Test- Butters, N. (1987). Memory dysfunction and
Revised. Clinical Neuropsychologist, 13(3), word priming in dementia and amnesia. Behav-
348--358. ioral Neuroscience, 101, 347-351.
Shapiro, D. M., & Harrison, D. W. (1990). Alter- Shipley, J. E., Kupfer, D. J., Spiker, D. G., Shaw,
nate forms of the AVLT: A procedure and test of D. H., Coble, P. A., Neil, J. F., et al. (1981).
form equivalency. Archives of Clinical ·Neuro- Neuropsychological assessment and EEG sleep
psychology, 5, 405-410. in affective disorders. Biological Psychiatry,
Shawaryn, M.A., Schiaffino, K. M., LaRocca, N. G., 16(10), 907-918.
& Johnston, M. V. (2002). Determinants of Shoqeirat, M. A., Mayes, A., MacDonald, C.,
health-related quality of life in multiple sclerosis: Meudell, P., et al. (1990). Performance on tests
The role of illness intrusiveness. Multiple Scle- sensitive to frontal lobe lesions by patients with
rosis, 8(4), 310-318. organic amnesia: Leng & Parkin revisited. British
Shea, B., Dube, C., & Moher, D. (2001). A¥essing Journal of Clinical Psychology, 29(4), 401-408.
the quality of reports of systematic revie~: The Shore, C., Shore, H., & Pihl, R. 0. (1971). Corre-
QUOROM statement compared to other tools. lations between performance on the category
In M. Egger, G. D. Smith, & D. G. :Altman test and the Wechsler Adult Intelligence Scale.
(Ed.), Systenwtic reviews in health care: Meta- Perceptual and Motor Skills, 32, 70.
analysis in context. London: BMJ. Shores, E. A., & Carstairs, J. R. (2000). The Mac-
Shean, G., Burnett, T., & Eckman, F. S. (2002). quarie University Neuropsychological Nonnative
Symptoms of schizophrenia and neurocognitive Study (MUNNS): Australian norms for the
test performance. Journal of Clinical Psychology, WAIS-R and WMS-R. Australian Psychologist,
58(7), 723-731. 35(1), 41--59.
Shear, P. K., Wells, C. T., & Brock, A.M. (2000). The Shorr, J., Delis, D., & Massman, P. (1992). Memory
effect of semantic cuing on CVLT performance in for the Rey-Osterrieth Figure: Perceptual clus-
healthy participants. Journal of Clinical and Ex- tering, encoding, and storage. Neuropsychology,
perimental Neuropsychology, 22(5), 649--e55. 6, 43--50.
Sherer, M., Nick, T. G., Millis, S. R., & lYovack, Shum, D. H., McFarland, K. A., & Bain, J. D.
T. A. (2003). Use of the WCST and the WCST- (1990). Construct validity of eight tests of at-
64 in the assessment of traumatic brain injury. tention: Comparison of normal and closed head
Journal of Clinical and Experimental N~ropsy injured samples. Clinical Neuropsychologist,
chology, 25(4), 512-520. 4(2), 151-162.
Shennan, A. M., & Massman, P. J. (1999). Preva- Shum, D., Murray, R., & Eadie, K. (1997). Effect
lence and correlates of category versus letter of speed of presentation on administration of the
fluency discrepancies in Alzheimer's disease. Logical Memory subtest of the Wechsler Mem-
Archives of Clinical Neuropsychology, 14(5), ory Scale-Revised. Clinical Neuropsychologist,
411-418. 11(2), 188-191.
Shennan, D. S., Boone, K., Lu, P., & Razani, J. Shure, G. H., & Halstead, W. C. (1958). Cerebral
(2002). Re-examination of a Rey Auditory Verbal localization of intellectual processes. Psychological
Learning Test/Rey Complex Figure discriminant Monographs: General and Applied, 72(12), 1-40.
function to detect suspect effort. Clinictd Neu- Shute, G. E., & Huertas, V. (1990). Developmental
ropsychologist, 16(3), 242-250. variability in frontal lobe function. Develop-
Shennan, E. M. S., Strauss, E., & Spellacy, F. mental Neuropsychology, 6(1), 1-11.
(1997). Validity of the Paced Auditol)'\ Serial Shuttleworth, E. C., & Huber, S. J. (1988). The
Addition Test (PASAT) in adults referred for naming disorder of dementia of Alzheimer type.
neuropsychological assessment after head injury. Brain and Language, 34, 222-234.
Clinical Neuropsychologist, 11(1), 34-45. Siegert, R. J., & Cavana, C. M. (1997). Norms for
Sherrill, R. E. (1985). Comparison of three short older New Zealanders on the Trail-Making Test.
forms of the Category Test. Journal of Clinical New Zealand Journal of Psychology, 26(2), 25--31.
REFERENCES 597
Silver, H., Shlomo, N., Turner, T., & Gur, R. C. norms. Journal of the International Neu-
(2002). Perception of happy and sad facial ex- ropsychological Society, 3(4), 317-326.
pressions in chronic schizophrenia: Evidence for Sliwinski, M., Lipton, R., Buschke, H., & Wasyly-
two evaluative systems. Schizophrenia Research, shyn, C. (2003). Optimizing cognitive test norms
55(1-2), 171-177. for detection. In R. Petersen (Ed.), Mild cogni-
Silverstein, A. B. (1962). Perceptual, motor, and tive impairment: Aging to Alzheimer's disease.
memory functions in the Visual Retention Test. New York: Oxford University Press.
American Journal of Mental Deficiency, 66, Small, B. J., Graves, A. B., McEvoy, C. L., Craw-
613--fil7. ford, F. C., Mullan, M., & Mortimer, J. A.
Silverstein, A. B. (1963). Qualitative analysis of (2000). Is APOE-epsilon4 a risk factor for cog-
performance on the Visual Retention Test. nitive impairment in normal aging? Neurology,
American Journal of Mental Deficiency, 68, 54(11), 2082-2088.
109--113. Small, G. W., La Rue, A., Komo, S., Kaplan, A., &
Simard, M., van Reekum, R., & Myran, D. (2003). Mandelkern, M. A. (1995). Predictors of cogni-
Visuospatial impairment in dementia with Lewy tive change in middle-aged and older adults with
bodies and Alzheimer's disease: A process anal- memory loss. American Journal of Psychiatry,
ysis approach. International Journal of Geriatric 152(12), 1757-1764.
Psychiatry, 18, 387-391. Smalley, S. L., Wolkenstein, B. H., LaRue, A.,
Simkins-Bullock,J., Brown, G. G., Greiffenstein, M., Woodward, J. A., Jarvik, L. F., & Matsuyama, S. S.
Malik, G. M., & McGillicuddy (1994). Neu- (1992). Commingling analysis of memory perfor-
ropsychological correlates of short-term memory mance in offspring of Alzheimer patients. Genetic
distractor tasks among patients with surgical re- Epidemiology, 9(5), 333--345.
pair of anterior communicating artery aneu- Smith, G. E., & Ivnik, R. J. (2003). Normative
rysms. Neuropsychology, 8(2), 246--254. neuropsychology. In: R. C. Petersen (Ed.), Mild
Sjogren, P., Olsen, A. K., Thomsen, A. B., & Dal- cognitive impairment: Aging to Alzheimer's dis-
berg, J. (2000). Neuropsychological performance ease. New York: Oxford University Press.
in cancer patients: The role of oral opioids, pain Smith, G. E., Ivnik, R. J., Malec, J. F., Petersen, R. C.,
and performance status. Pain, 86(3), 237-245. Kokmen, E., Tangalos, E. G., et al. (1992).
Ska, B., Poissant, A., & Joanette, Y. (1990). Line Mayo's Older Americans Normative Studies
orientation judgment in normal elderly and (MOANS): Factor structure of a core battery.
subjects with dementia of Alzheimer's type. Psychological Assessment, 4(3), 382-390.
Journal of Clinical and Experimental Neuropsy- Smith, R. L., Goode, K. T., La Marche, J. A., &
chology, 12(5), 695-702. Boll, T. J. (1995). Selective Reminding Test short
Skelton-Robinson, M., & Jones, S. (1984). Nominal form administration: A comparison of two
dysphasia and the severity of senile dementia. through twelve trials. Psychological Assessment,
British Journal of Psychiatry, 145, 168-171. 7(2), 177-182.
Skenazy, J. A., & Bigler, E. D. (1984). Neu- Smith, S., Murdoch, B., & Chenery, H. (1989).
ropsychological findings in diabetes mellitus. Semantic abilities in dementia of the Alzheimer's
Journal of Clinical Psychology. 40(1), 246--258. type. Brain and Language, 36, 314-324.
Slay, D. K. (1984). A portable Halstead-Reitan Smith, Y., Giordani, B., Lajiness-O'Neill, & Zu-
category test. Journal of Clinical Psychology, bieta, J. (2001). Long-term estrogen replacement
40(4), 1023-1027. is associated with improved nonverbal memory
Slick, D. J., Iverson, G. L., & Green, P. (2000). and attentional measures in postmenopausal
California Verbal Learning Test indicators of women. Fertility and Sterility, 76(6), 1101-1107.
suboptimal performance in a sample of head- Snodgrass, J. G. (1984). Concepts and their surface
injury litigants. Journal of Clinical and Experi- representations. Journal of Verbal Learning and
mental Neuropsychology, 22(5), 569--579. Verbal Behavior, 23, 3--22.
Slick, D. J., Hinkin, C. H., van Gorp, W. G., & Satz, Snodgrass, J. G., & Vanderwart, M. (1980). A
P. (2001). Base rate of WMS-R Malingering standardized set of 260 pictures: Norms for
Index in a sample of non-compensation-seeking name agreement, image agreement, familiarity
men infected with HIV-1. Applied Neuropsy- and visual complexity. Journal of Experimental
chology, 8(3), 185-189. Psychology: Human Learning and Merrwry, 6,
Sliwinski, M., Buschke, H., Stewart, W. F., Masur, D., 174-215.
et al. (1997). The effect of dementia risk factors on Snow, J. H. (1998). Clinical use of the Benton Vi-
comparative and diagnostic selective reminding sual Retention Test for children and adolescents
598 REFERENCES
with learning disabilities. Archives aJ Clinical Spencer, W. D., & Raz, N. (1994). Memory for
Neuropsychology, 13(1), 629--636. facts, source, and context: Can frontal lobe dys-
Snow, W. G. (1987). Standardization of test function explain age-related differences? Psy-
administration and scoring criteria: s<lme short- chology and Aging, 9(1), 149-159.
comings of current pmctice with thei Halstead- Spreen, 0., & Benton, A. L. (1969). Neurosensory
Reitan Test Battery. Clinical Neuropsychologist, Center comprehensive examination for aphasia:
1(3), 250-262. Manual of directions. Victoria: Neuropsychology
Snyder, K. A., & Harrison, D. W. (1997). The Af- Laboratory, University of Victoria.
fective Auditory Verbal Learning Test. Archives Spreen, 0., & Strauss, E. (1991). A compendium of
of Clinical Neuropsychology, 12(5), 477-482. neuropsychological tests. New York: Oxford
Snyder, K. A., Harrison, D. W., & Sh~al, B. V. University Press.
(1998). The Affective Auditory Verbal Learning Spreen, 0., & Strauss, E. (1998). A Compendium of
Test: Peripheral arousal correlates. Archives of neuropsychological tests (2nd ed.). New York:
Clinical Neuropsychology, 13(3), 25H258. Oxford University Press.
Snyder, P. J., & Cappelleri, J. C. (2001)1 Informa- Squire, L. R., & Shimamura, A. P. (1986). Char-
tion processing speed deficits may ~be better acterizing amnesic patients for neurobehavioral
correlated with the extent of white rriatter scle- study. Behavioral Neuroscience, 100, 866--877.
rotic lesions in multiple sclerosis than previously Stanczak, D. E., & Triplett, G. (2003). Psycho-
suspected [Special Issue: TENNET X~: Theoret- metric properties of the Mid-Range Expanded
ical and Experimental NeuropsychofJgy, June Trail Making Test: An examination of learning-
15-17, 2000]. Brain and Cognitio11, 46(1-2), disabled and non-learning-disabled children.
279-284. Archives of Clinical Neuropsychology, 18(2),
Snyder, P. J., Cappelleri, J. C., Archibald., C. J., & 107-120.
Fisk, J. D. (2001). Improved detectiOn of dif- Stanczak, D. E., Lynch, M. D., McNeil, C. K., &
ferential information-processing spe¥1 deficits Brown, B. (1998). The Expanded Trail Making
between two disease-course types of multiple Test: Rationale, development, and psychometric
sclerosis. Neuropsychology, 15(4), 611,.625. properties. Archives of Clinical Neuropsychol-
Soininen, H. S., Partanen, K., Pitkanen, ~., Vainio, ogy, 13(5), 473-487.
P., Hanninen, T., Hallikainen, M., et hl. (1994). Stanczak, E. M., Stanczak, D. E., & Templer, D. I.
Volumetric MRI analysis of the amygdala and (2000). Subject-selection procedures in neu-
the hippocampus in subjects with age~sociated ropsychological research: A meta-analysis and
memory impairment: Correlation to ~sual and prospective study. Archives of Clinical Neuro-
verbal memory. Neurology, 44(9), 16®--1668. psychology, 15(7), 587--601.
Solari, A., & Filippini, G., (1995). ~out self- Stanczak, D. E., Stanczak, E. M., & Awadalla, A. W.
assessment of neurological disability ip multiple (2001). Development and initial validation of an
sclerosis. International Journal of Epil/emiology, Arabic version of the Expanded Trail Making
24,466. I Test: Implications for cross-cultural assessment.
Somerville, J., Tremont, G., & Stem, R.;A. (2000). Archives of Clinical Neuropsychology, 16(2),
The Boston Qualitative Scoring Syjtem as a 141-149.
measure of executive functioning in Rey-Os- Stanton, B. A., Jenkins, C. D., Savageau, J. A., &
terrieth Complex Figure performance.)oumal of Zyzanski, S. J. (1984). Age and educational dif-
Clinical and Experimental Neur~ychology, ferences on the Trail Making Test and Wechsler
22(5), 613-621. Memory Scales. Perceptual and Motor Skills, 58,
Sonobe, N., Kanno, M., Ito, M., Uchiy~a, M., Ta- 311-318.
kahashi, Y., Yashima, Y., et al. (199p. Lateral Steenhuis, R. E., & Ostbye, T. (1995). Neu-
asymmetry of eye movements in temporal lobe ropsychological test performance of specific di-
epileptic patients with unilateral foci.lnfemational agnostic groups in the Canadian Study of Health
Journal of Psychophysiology, 11(3), 253--256. and Aging (CHSA). Journal of Clinical and Ex-
Soukup, V. M., Ingram, F., Grady, J. J., & Schiess, perimental Neuropsychology, 17(5), 773--785.
M. C. (1998). Trail Making Test: Issues in nor- Stefanova, E. D., Kostic, V., Ziropadja, L., Marko-
mative data selection. Applied Neuropsychology, vic, M., & Ocic, G. (2002). Serial position learn-
5(2), 65-73. ing effects in patients with aneurysms of the
Sovcikova, E., & Bronis, M. (1985). Eviluation of anterior communicating artery. Journal of Clin-
mental workload by Stroop Colour-~ord Test. ical and Experimental Neuropsychology, 24(5),
Studia Psychologica, 27, 245--248. 687-694.
REFERENCES 599
Steffens, D. C., Wagner, H. R., Levy, R. M., Hom, version of the California Verbal Learning Test.
K. A., & Krishnan, K. R. R. (2001). Perfonnance Clinical Neuropsychologist, 15(1), 95-108.
feedback deficit in geriabic depression. Biologi- Stewart, R., Richards, M., Brayne, C., & Mann, A.
cal Psychiatry, 50(5), 358--363. (2001). Cognitive function in UK community-
Stein, M. B., Hanna, C., Vaerum, V., & Koverola, dwelling African Caribbean elders: Normative
C. (1999). Memory functioning in adult women data for a test battery. International Journal of
traumatized by childhood sexual abuse. Journal Geriatric Psychiatry, I6(5), 518-527.
ofTraumatic Stress, 12(3), 527-534. Stillhard, G., Landis, T., Schiess, R., Regard, M., &
Stein, M. B., Kennedy, C. M., & Twamley, E. W. Sigler, G. (1990). Bitemporal hypoperfusion in
(2002). Neuropsychological function in female transient global amnesia: 99m-Tc-HM-PAO
victims of intimate partner violence with and SPECT and neuropsychological findings during
without posttraumatic stress disorder. Biological and after an attack. Journal of Neurology, Neu-
Psychiatry, 52(11), 1079-1088. rosurgery, and Psychiatry, 53(4), 339-342.
Steindl, S. R., & Boyle, G. J. (1995). Use of the Stolar, N., Berenbaum, H., Banich, M. T., & Barch,
Booklet Category Test to assess abstract concept D. (1994). Neuropsychological correlates of
fonnation in schizophrenic disorders. Archives of alogia and affective flattening in schizophrenia.
Clinical Neuropsychology, 10(3), 205-210. Biological Psychiatry, 35(3), 164-172.
Stem, R. A., Javorsky, D. J., Singer, E. A., Singer, Stone, B. J., Gray, J. W., Dean, R. S., & Wheeler, T.
N. G., Duke, L. M., Somerville, J. A., et al. E. (1988). An examination of the Wechsler Adult
(1999). The Boston Qualitative Scoring System Intelligence Scale (WAIS) subtests from a neu-
for the Rey-Osterrieth Complex Figure (version ropsychological perspective. International Jour-
3.1). Odessa, FL: Psychological Assessment Re- nal of Neuroscience, 40(1-2), 31-39.
sources. Storandt, M., & Hill, R. D. (1989). Very mild senile
Stem, R. A., Singer, E. A., Duke, L. M., Singer, N. dementia of the Alzheimer type: II. Psychomet-
G., Morey, C. E., & Daughtrey, E. W. (1994). ric test perfonnance. Archives of Neurology, 46,
The Boston Qualitative Scoring System for the 383--386.
Rey-Osterrieth Complex Figure: Description Stratta, P., Rossi, A., Mancini, F., Cupillari, M.,
and interrater reliability. Clinical Neuropsy- Matteri, & Casacchia (1993). Wisconsin Card
chologist, 8, 309-322. Sorting Test perfonnance and educational level
Stem, Y., Andrews, H., Pittman, J., Sano, M., Ta- in schizophrenic and control samples. Neuro-
temichi, T., Lantigua, R., et al. (1992). Diagnosis psychiatry, Neuropsychology, and Behavioral
of dementia in a heterogeneous population. Ar- Neurology, 6(3), 149-153.
chives of Neurology, 49,453-460. Strauss, E., & Spreen, 0. (1990). A comparison of
Stem, Y., Tang, M. X., Jacobs, D. M., Sano, M., the Rey and Taylor figures. Archives of Clinical
Marder, K., et al. (1998). Prospective compara- Neuropsychology, 5, 417-420.
tive study of the evolution of probable Alzhei- Strauss, E., & Wada, J. (1988). Hand preference
mer's disease and Parkinson's disease dementia. and proficiency and cerebral speech dominance
Journal of the International Neuropsychological determined by the carotid amytal test. Journal of
Society, 4(3), 279-284. Clinical and Erperimental Neuropsychology, 10,
Stem, Y., Albert, S., Tang, M.-X., & Tsai, W-Y. 169-174.
(1999). Rate of memory decline in AD is related Strauss, E., Spellacy, F., Hunter, M., & Berry, T.
to education and occupation: Cognitive reserve? (1994). Assessing believable deficits on measures of
Neurology, 53(9), 1942-1947. attention and information processing capacity. Ar-
Sterne, J., Egger, M., & Smith, G. D. (2001). In- chives ofClinical Neuropsychology, 9{6), 483-490.
vestigating and dealing with publication and Strenge, H., Niederberger, U., & Seelhorst, U.
other biases in meta-analysis. British Medical (2002). Correlation between tests of attention
Journal, 323, 101-105. and perfonnance on Grooved and Purdue Peg-
Stevens, M. C. (2000). Pictorial presentation of boards in nonnal subjects. Perceptual and Motor
verbal stimuli: A semantic memory study using Skills, 95(2), 507-514.
an adaptation of the California Verbal Learning Sbicker, J. L., Brown, G. G., Wixted, J., Baldo, J. V., &
Test. Dissertation Abstracts International. Sec- Delis, D. C. (2002). New semantic and serial
tion B: The Sciences and Engineering, 60(8-B), clustering indices for the California Verbal Learn-
4254. ing Test-second edition: Background, rationale,
Stevens, M. C., Fein, D. A., & Markus, E. (2001). and fonnulae. Journal of the International Neu-
Connecticut Pictorial Learning Test: A pictorial ropsychological Society, 8(3), 425-435.
600 REFERENCES
Strickland, T., D'Elia, L., James, R., & Stein, R. before and after seizure surgery. Journal of
(1997). Stroop Color-Word performance of Af- Clinical and Experimental Neuropsychology,
rican Americans. Clinical Neuropsychologist, 11, 25(2), 190-200.
87-90. Suhr, J. A. & Boyer, D. (1999). Use of the Wis-
Stroop, J. (1935). Studies of interference in serial consin Card Sorting Test in the detection of
verbal reactions. Journal of Experimetttal Psy- malingering in student simulator and patient
chology, 18, 643--662. samples. Journal of Clinical and Experimental
Stuss, D. T. (1987). Contribution of fro,tal lobe Neuropsychology, 21, 701-708.
injury to cognitive impairment after cl0$ed head Sullivan, E. V., Mathalon, D. H., Zipursky, R. B.,
injury: Methods of assessment and re~nt find- Kerteen-Tucker, Z., Knight, R. T., & Pfeffer-
ings. In H. S. Levin, J. Grafman & It M. Ei- baum, A. (1993). Factors of the Wisconsin Card
senberg (Eds.), Neurobehavioral recovtry from Sorting Test as measures of frontal-lobe function
head injury. New York: Oxford Univers~ Press. in schizophrenia and in chronic alcoholism. Psy-
Stuss, D. T., Benson, D. F., Kaplan, E. F., eir, W. chiatry Research, 46, 175-199.
S., Naiser, M. A., Lieberman, 1., & F rrill, D. Sullivan, K., & Bowden, S. C. (1997). Which tests
(1983). The involvement of orbitofrontal cere- do neuropsychologists use? Journal of Clinical
brum in cognitive tasks. Neuropsycho{iJgia, 21, Psychology, 53(1), 657-661.
235-248. Sullivan, K., Keane, B., & Deffenti, C. (2001).
Stuss, D. T., Ely, P., Hugenholtz, H., Richard, M. Malingering on the RAVLT: Part I. Deterrence
T., LaRochelle, S., Poirier, C. A., et ai. (1985). strategies. Archives of Clinical Neuropsychology,
Subtle neuropsychological deficits in '.patients 16(1), 627-641.
with good recovery after closed hea4 injury. Sullivan, K., Deffenti, C., & Keane, B. ("2002).
Neurosurgery, 17(1), 41-47. Malingering on the RAVLT: Part II. Detection
Stuss, D. T., Stethem, L. L., & Poirier, C. A. (1987). strategies. Archives of Clinical Neuropsychology,
Comparison of three tests of attention apd rapid 17(3), 223-233.
information processing across six age:. groups. Sutton, A., Abrams, K. R., Jones, D. R., Sheldon,
Clinical Neuropsychologist, 1, 139-152.· T. A., & Song, F. (2000). Methods for meta-
Stuss, D. T., Stethem, L. L., & Pelchat, a (1988). analysis in medical research. Chichester: Wiley.
Three tests of attention and rapid inf~rmation Swan, G. E., Morrison, E., & Eslinger, P. (1990).
processing: An extension. Clinical ~ropsy Interrater agreement on the Benton Visual Re-
chologist, 2, 246-250. tention Test. Clinical Neuropsychologist, 4(1),
Stuss, D. T., Stethem, L. L., Hugenhol~, H. & 37-44.
Richard, M. T. (1989). Traumatic braill injury: Swan, G. E., Dame, A., & Carmelli, D. (1991).
A comparison of three clinical tests and analysis Involuntary retirement, type A behavior, and
of recovery. Clinical Neuropsychologist, 3, current functioning in elderly men: 27-year fol-
145-156. low-up of the Western Collaborative Group
Stuss, D. T., Alexander, M. P., Hamer,) L., Pa- Study. Psychology and Aging, 6(3), 384-391.
lumbo, C., Dempster, R., Binns, ML et al. Swan, G. E., DeCarli, C., Miller, B. L., Reed, T.,
(1998). The effects of focal anterior and poste- Wolf, P. A., Jack, L. M., et al. (1998). Association
rior brain lesions on verbal fluency. Joornal of of midlife blood pressure to late-life cognitive
the International Neuropsychological Society, decline and brain morphology. Neurology, 51(4),
4(3), 265-278. : 986-993.
Stuss, D. T., Levine, B., Alexander, M.P., Hong, J., Sweeney, J. A., Haas, G. L., Keilp, J. G., & Long, M.
Palumbo, C., Hamer, L., et al. (2000). Wisconsin (1991). Evaluation of the stability of neu-
Card Sorting Test performance in patients ' with ropsychological functioning after acute episodes of
focal frontal and posterior brain damage; Effects schizophrenia: One-year followup study. Psychia-
of lesion location and test structure on Sfparable try Research, 38(1), 63--76.
cognitive processes. Neuropsychologid,. 38(4), Sweet, J. J., & King, J. H. (2002). Category test
388-402. validity indicators: Overview and practice rec-
Stuss, D. T., Bisschop, S. M., Alexander, M. P., ommendations. Journal of Forensic Neuropsy-
Levine, B., Katz, D., & lzukawa, D. (20~)1). The chology. Special Issue: Detection of Response
Trail Making Test: A study in focal le$ion pa- Bias in Forensic Neuropsychology: Part II,
tients. Psychological Assessment, 13(2), ~0-239. 3(1-2), 241-274.
Suchy, Y., Sands, K., & Chelune, G. J.i (2003). Sweet, J. J., Moberg, P. J., & Westergaard, C. K.
Verbal and nonverbal fluency pe;a,rmance (1996). Five-year follow-up survey of practices
REFERENCES 601
and beliefs of clinical neuropsychologists. Clini- Tarter, R. E., & Parsons, 0. A. (1971). Conceptual
cal Neuropsychologist, 10(2), 202-221. shifting in chronic alcoholics. Journal of Abnor-
Sweet, J. J., Moberg, P. J., & Suchy, Y. (2000a). mal Psychology, 77, 71-75.
Ten-year follow-up survey of clinical neuropsy- Taylor, A. E., Saint-Cyr, J. A., & Lang, A. E. (1986).
chologists: Part I. Practices and beliefs. Clinical Frontal lobe dysfunction in Parkinson's disease.
Neuropsychologist, 14(1), 18--37. Brain, 109, 845-883.
Sweet, J. J., Wolfe, P., Sattlberger, E., Numan, B., Taylor, D. J., Hunt, C., & Glaser, B. (1990). A
Rosenfeld, J. P., Clingerman, S., et al. (2000b). crossvalidation of the revised Category Test.
Further investigation of traumatic brain injury Psychological Assessment: A Journal of Consult-
versus insufficient effort with the California ing and Clinical Psychology, 2(4), 486-488.
Verbal Learning Test. Archives of Clinical Neu- Taylor, E. M. (1959). Psychological appraisal of
ropsychology, 15(2), 105-113. children with cerebral defects. Cambridge, MA:
Sweetland, J., Kertesz, A., Prato, F. S., & Nantau, K. Harvard University Press.
(1987). The effect of magnetic resonance imag- Taylor, H. C., & Russell, J. T. (1939). The rela-
ing on human cognition. Magnetic Resonance tionship of validity coefficients to the practical
Imaging, 5(2), 129-135. effectiveness of tests in selection: Discussion and
Swerdlow, N. R., Filion, D., Geyer, M. A., & tables. Journal of Applied Psychology, 23.
Braff, D. L. (1995). "Normal" personality cor- 565-578.
relates of sensorimotor, cognitive, and visuospa- Taylor, J. M., Goldman, H., Leavitt, J., & Klein-
tial gating. Biological Psychiatry, 37, 286-299. man, K. N. (1984). Limitations of the briefform
Swets, J. A. (1996). Signal detection theory and of the Halstead Category Test. Journal of Clini-
ROC analysis in psychology and diagnostics: cal Neuropsychology, 6(3), 341-344.
Collected papers. In Scientific psychology series. Taylor, L. B. (1969). Localization of cerebral lesions
Hillsdale, NJ: Laurence Erlbaum. by psychological testing. Clinical Neurosurgery,
Swiercinsky, D. P. (1978). Manual for the adult 16, 269-287.
neuropsychological evaluation. Springfield, IL: Taylor, L. B. (1979). Psychological assessment of
Thomas. neurosurgical patients. Functional Neurosur-
Swiercinsky, D. P. (1979). Factorial pattern de- gery, 165-180.
scription and comparison of functional abilities Taylor, R. (1998a). Order effects within the Trail
in neuropsychological assessment. Perceptual Making and Stroop tests in patients with neu-
and Motor Skills, 48(1), 231-241. rologic disorders. Journal of Clinical and Ex-
Takashima, Y., Yao, H., Koga, H., Endo, K., perimental Neuropsychology, 20(5), 750-754.
Matsumoto, T., Uchino, A., et al. (2003). Frontal Taylor, R. (1998b). Continuous norming: Improved
lobe dysfunction caused by multiple lacunar in- equations for the WAIS-R. British Journal of
farction in community-dwelling elderly subjects. Clinical Psychology, 37(4), 451-456.
Neurological Sciences, 214, 37-41. Taylor, S. F., Kornblum, S., Lauber, E. J.,
Tallent, K. A., & Gooding, D. C. (1999). Working Minoshima, S., & Koeppe, R. A. (1997). Isolation
memory and Wisconsin Card Sorting Test per- of specific interference processing in the Stroop
formance in schizotypic individuals: A replica- task: PET activation studies. Neuroimage, 6,
tion and extension. Psychiatry Research, 89(3), 81-92.
161-170. Teknos, K. S., Bernstein, J. H., & Seidman, L. J.
Tamkin, A. S., & Hyer, L. A. (1984). Testing for (2003). ROCF performance of attention-deficit!
cognitive dysfunction in the aging psychiatric hyperactivity disordered children. In J. A. Knight
patient. Military Medicine, 149(7), 397-399. (Ed.), The handbook of Rey-Osterrieth Complex
Tamkin, A. S., & Jacobsen, R. (1984). Age-related Figure usage: Clinical and research applications.
norms for the Hooper Visual Organization Test. Lutz, FL: Psychological Assessment Resources.
Journal of Clinical Psychology, 40(6), 1459-1463. Tenhula, W. N., & Sweet, J. J. (1996). Double
Tang, C., & Liu, Y. (1993). Impairment of visual cross-validation of the Booklet Category Test in
form discrimination in Parkinson's disease. Acta detecting malingered traumatic brain injury.
Psychologica Sinica, 25, 258-263. Clinical Neuropsychologist, 10(1), 104-116.
Tang, H. W., Liang, Y. X., Hu, X. H., & Yang, H. G. Teunissen, C. E., De Vente, J., von Bergmann, K.,
(1995). Alterations of monoamine metabolites Bosma, H., vanBoxtel, M.P. J., De Bruijn, C.,etal.
and neurobehavioral function in lead-exposed (2003). Serum cholesterol, precursors and metab-
workers. Biomedical and Environmental Sci- olites and cognitive performance in an aging pop-
ences, 8(1), 23-29. ulation. Neurobiology of Aging, 24(1), 147-155.
602 REFERENCES
Thompson, L. L., & Heaton, R. K. (1989). Com- complex figures. Journal of Clinical and Experi-
parison of different versions of the Boston mental Neuropsychology, 13, 587-599.
Naming Test. Clinical NeuropsycholofPt, 3(2), Tombaugh, T. N., & Hubley, A. M. (1997). The 60-
184-192. item Boston Naming Test: Norms for cognitively
Thompson, L. L., & Heaton, R. K. (1991}. Pattern intact adults aged 25 to 88 years. Journal of
of performance on the Tactual Perlbrmance Clinical and Experimental Neuropsychology,
Test. Clinical Neuropsychologist, 5(4), 322-328. 19(6), 922-932.
Thompson, L. L., & Parsons, 0. A. (19$). Con- Tombaugh, T. N., Hubley, A.M., Faulkner, P., &
tribution of the TPT to adult neuropsychological Schmidt, J. P. (1990). The Rey-Osterrieth and
assessment: A review. Journal of ClirrJcal and Taylor complex figures: Comparative studies,
Experimental Neuropsychology, 7(4), ~- modified figures and normative data for the
Thompson, L. L., Heaton, K. R., Matthe~. C. G., Taylor figure. Paper presented at the 18th an-
& Grant, I. (1987). Comparison of pref$Ted and nual meeting of the International Neuropsy-
nonpreferred hand performance on fottr neuro- chology Society, Orlando, FL.
psychological motor tasks. Clinical Weuropsy- Tombaugh, T. N., Faulkner, P., & Hubley, A. M.
chologist, 1(4), 324-334. (1992a). Effects of age on the Rey-Osterrieth
Thompson, M. D., Scott, J. G., DicksoQ, S. W., and Taylor complex figures: Test-retest data
Schoenfeld, J. D., Ruwe, W. D., & A~s. R. L. using an intentional learning paradigm. Journal
(1999). Clinical utility of the Trail Making Test of Clinical and Experimental Neuropsychology,
practice time. Clinical NeuropsychologiJt, 13(4), 14, 647-661.
450-455. Tombaugh, T. N., Schmidt, J. P., & Faulkner, P.
11lorndike, E. L., & Lorge, I. (1944). The Teacher's (1992b). A new procedure for administering the
word book of30,000 words. New York: 'teacher's Taylor Complex Figure: Normative data over a
College, Columbia University. , 60-year age span. Clinical Neuropsychologist,
11lurstone, L. L. (1944). A factorial studt of per- 6(1), 63-79.
ception. Chicago: University of Chicago Press. Tombaugh, T. N., Kozak, J., & Rees, L. (1999).
Thurstone, L. L., & 11lurstone, T. G. (1949). Exam- Normative data stratified by age and education
iner manual for the SRA Primary Mentai, Abilities for two measures of verbal fluency: F AS and
Test. Chicago: Science Research Associaies. animal naming. Archives af Clinical Neuropsy-
Thurstone, L. L., & Thurstone, T. G. (1002). Pri- chology, 14(2), 167-177.
mary mental abilities (Rev.). Chicago: Science Tomer, R., & Levin, B. E. (1993). Differential ef-
Research Associates. fects of aging on two verbal fluency tasks. Per-
Tierney, M. C., Nores, A., Snow, W. G., Fisher, ceptual and Motor Skills, 76, 465-466.
R. H., Zorzitto, M. L., & Reid, D. W. (1994). Use Tomer, R., Fisher, T., Giladi, N ., & Aharon-Peretz, J.
of the Rey Auditory Verbal Learnin~ Test in (2002). Dissociation between spontaneous and
differentiating normal aging from ~eimer's reactive flexibility in early Parkinson's disease.
and Parkinson's dementia. Psychological Assess- Neuropsychiatry, Neuropsychology, and Behav-
ment, 6, 129--134. ioral Neurology, 15(2), 106--112.
Tiersky, L.A., Cicerone, K. D., Natelson, B. H., & Torralva, T., Dorrego, F., Sabe, L., Chemerinski, E., &
DeLuca, J. (1998). Neuropsychological func- Starkstein, S. E. (2000). Impairments of social
tioning in chronic fatigue syndrome and mild cognition and decision making in Alzheimer's
traumatic brain injury: A comparison. Clinical disease. International Psychogeriatrics, 12(3),
Neuropsychologist, 12(4), 503-512. 359--368.
Toglia, M. P., & Battig, W. F. (1978). Handbook of Torres, I. J., Flashman, L. A., O'Leary, D. S., &
word norms. Hillsdale, NJ: Lawrence Erlbaum. Andreasen, N. C. (2001). Effects of retroactive
Tombaugh, T. N. (1999). Administrative manual and proactive interference on word list recall in
for the Adjusting-Paced Serial Addi~n Test schizophrenia. Journal of the International
(Adjusting-PSAT). Ottawa: Ottawa Cancussion Neuropsychological Society, 7(4), 481-490.
Clinic, Carleton University. Torres, I. J., Mundt, A. J., Sweeney, P. J., Uanes-
Tombaugh, T. N. (2004). Trail Making Test A and Macy, S., Dunaway, L., Castillo, M., et al. (2003).
B: Normative data stratified by age and educa- A longitudinal neuropsychological study of par-
tion. Archives af Clinical Neuropsychology, 19, tial brain radiation in adults with brain tumors.
203-214. Neurology, 60(7), 1113-1118.
Tombaugh, T. N., & Hubley, A. M. (1991). Four Toshima, T., Toma, C., Demic, J., & Wapner, S.
studies comparing the Rey-Osterrieth and Taylor (1992). Age and cross-cultural differences in
REFERENCES 603
processes underlying sequential cognitive activ- S., et al. (1995). Cognitive function and biological
ity. In B. Wilpert, H. Motoaki, & J. Misumi correlates of cognitive performance in schizoty-
(Eds.), General psychology and environmental pal personality disorder. Psychiatry Research,
psychology: Proceedings of the 22nd Interna- 59(1-2), 127-136.
tional Congress of Applied Psychology (p. 189). Trichard, C., Martinot, J. L., Alagille, M., Masure,
Hillsdale, NJ: Lawrence Erlbaum. M. C., Hardy, P., Ginestet, D., et al. (1995).
Toshima, T., Demick, J., Miyatani, M., Ishii, S., & Time course of prefrontal lobe dysfunction in
Wapner, S. (1996). Cross-cultural differences severely depressed inpatients: A longitudinal
in processes underlying sequential cognitive ac- neuropsychological study. Psychological Medi-
tivity. Japanese Psychological Research, 38(2), cine, 25, 79-85.
90-96. Triggs, W. J., Calvanio, R., Levine, M., Heaton,
Touradji, P., Manly, J. J., Jacobs, D. M., & Stern, Y. R. K., & Heilman, K. M. (2000). Predicting hand
(2001). Neuropsychological test performance: A preference with performance on motor tasks.
study of non-Hispanic white elderly. Journal of Cortex, 36(5), 679-689.
Clinical and Experimental Neuropsychology, Troester, A. 1., Fields, J. A., Testa, J. A., Paul, R. H.,
23(5), 643-649. Blanco, C. R., Hames, K. A., et al. (1998). Cor-
Trahan, D. E. (1998). Judgment of line orientation tical and subcortical influences on clustering and
in patients with unilateral cerebrovascular le- switching in the performance of verbal fluency
sions. Assessment, 5(3), 227-235. tasks. Neuropsychologia, 36(4), 295--304.
Trahan, D. E., & Larrabee, G. J. (1993). Clinical Troyer, A. K. (2000). Normative data for clustering
and methodological issues in measuring rate of and switching on verbal fluency tasks. Journal of
forgetting with the verbal selective reminding Cltnical and Experimental Neuropsychology,
test. Psychological Assessment, 5(1), 67-71. 22(3), 370-378.
Trahan, D. E., Patterson, J., Quintana, J., & Biron, Troyer, A. K., & Wishart, H. A. (1997). A com-
R. (1987). The Finger Tapping Test: A reex- parison of qualitative scoring systems for the
amination of traditional hypotheses regarding Rey-Osterrieth Complex Figure Test. Clinical
normal adult performance. Paper presented at Neuropsychologist, 11(4), 381-390.
the 15th annual meeting of the International Troyer, A. K., Moscovitch, M., & Winocur, G.
Neuropsychological Society, Washington, DC. (1997). Clustering and switching as two compo-
Trahan, D., Quintana, J., Willingham, A., & nents of verbal fluency: Evidence from younger
Goethe, K. (1988). The Visual Reproduction and older healthy adults. Neuropsychology,
subtest: Standardization and clinical validation of 11(1), 138-146.
a delayed recall procedure. Neuropsychology, Troyer, A. K., Moscovitch, M., Winocur, G., Alex-
2(1), 29-39. ander, M. P., & Stuss, D. (1998a). Clustering and
Trautt, G. M., Chavez, E. L., Brandon, A. D., & switching on verbal fluency: The effects of focal
Steyaert, J. (1983). Effects of test anxiety and sex frontal- and temporal-lobe lesions. Neuropsy-
of subject on neuropsychological test perfor- chologia, 36(6), 499-504.
mance: Finger Tapping, Trail Making, Digit Troyer, A. K., Moscovitch, M., Winocur, G.,
Span, and Digit Symbol tests. Perceptual and Leach, L., & Freedman, M. (1998b). Clustering
Motor Skills, 56, 923-929. and switching on verbal fluency tests in Alzhei-
Tremont, G., Hoffman, R. G., Scott, J. G., & Ad- mer's and Parkinson's disease. Journal of the
ams, R. L. (1998). Effect of intellectual level on International Neuropsychological Society, 4(2),
neuropsychological test performance: A re- 137-143.
sponse to Dodrill (1997). Clinical Neuropsy- Tsai, C. H., Lu, C. S., Hua, M. S., Lo, W. L. & Lo,
chologist, 12(4), 560-567. S. K. (1994). Cognitive dysfunction in early onset
Treneny, M. R., Crosson, B., DeBoe,J., & Leber, W. R. parkinsonism. Acta Neurologica Scandinavica,
(1989). Stroop Neuropsychological Screening 89,9-14.
Test, manual. Odessa, FL: Psychological Assess- Tsunoda, M., Kurachi, M., Yuasa, S., Kadono, Y.,
ment Resources. Matsui, M., & Shimizu, A. (1992). Scanning eye
Treneny, M. R., Crosson, B., DeBoe, J., & Leber, movements in schizophrenic patients: Relation-
W. R. (1990). Visual Search and Attention Test. ship to clinical symptoms and regional cerebral
Odessa, FL: Psychological Assessment Re- blood flow using 123I-IMP SPECT. Schizo-
sources. phrenia Research, 7(2), 159-168.
Trestman, R. L., Keefe, R. S. E., Mitropoulou, V., Tucha, 0., Smely, C., & Lange, K. W. (1999).
Harvey, P. D., deVegvar, M. L., Lees-Roitman, Verbal and figural fluency in patients with mass
604 REFERENCES
lesions of the left or right frontal lob~s. Journal Vakil, E., & Blachstein, H. (1993). Rey Auditory-
of Clinical and Experimental Neuropsychology, Verbal Learning Test: Structure analysis. Journal
21 (2), 229-236. of Clinical Psychology, 49, 883-890.
Tuokko, H., & Woodward, T. S. (1996): Develop- Vakil, E., & Blachstein, H. (1994). A supplementary
ment and validation of a demographiclcorrection measure in the Rey AVLT for assessing inci-
system for neuropsychological measures used in dental learning of temporal order. Journal of
the Canadian Study of Health and Aging. Jour- Clinical Psychology, 50(2), 240-245.
nal of Clinical and Experimental Neuropsychol- Vakil, E., & Blachstein, H. (1997). Rey AVLT:
ogy, 18(4), 479-616. Developmental norms for adults and the sensi-
Tupler, L. A., Welsh, K. A., Asare-Aboagye, Y., & tivity of different memory measures to age.
Dawson, D. V. (1995). Reliability of the Rey- Clinical Neurqpsychologist, 11(4), 356-369.
Osterrieth Complex Figure in use with memory- Valdois, S., Poissant, A., & Joanette, Y. (1989).
impaired patients. Journal of Clinictfz and Ex- Visual form discrimination in normal aging
perimental Neuropsychology, 17(4), $6--579. and dementia of the Alzheimer's type. Journal
Tupper, D. E. (1999). Introduction: Neuropsycho- of Clinical and Experimental Neuropsychology,
logical assessment apres Luria. [Special Issue: 11, 91.
Part II: International extentions of L.,ria's neu- van Boxtel, M. P. J., ten Tusscher, M. P. M.,
ropsychological investigation]. NeunYpsychology Metsemakers, J. F. M., Willems, B., & Jolles, J.
Review, 9(2), 57-61. (2001). Visual determinants of reduced perfor-
Turner, M. A. (1999). Generating novel ideas: mance on the Stroop Color-Word Test in normal
Fluency performance in high-functi(ming and aging individuals. Journal of Clinical and Ex-
learning diabled individuals with autiSJll. Journal perimental Neuropsychology, 23(5), 620-627.
of Child Psychology and Psychiatry bnd Allied Van den Broek, M. D., Bradshaw, C. M., &
Disciplines, 40(2), 189-201. Szabadi, E. (1993). Utility of the Modified Wis-
Uchiyama, C. L., D'Elia, L. F., Dellinger, A. M., consin Card Sorting Test in neuropsychological
Becker, J. T., et al. (1995). Alternate f~rms of the assessment. British Journal of Clinical Psychol-
Auditory-Verbal Learning Test: Isst.les of test ogy, 32(3), 333--343.
comparability, longitudinal reliability,i and mod- Vanderploeg, R. D., LaLone, L. V., Greblo, P., &
erating demographic variables. A(-chives of Schinka, J. A. (1997). Odd-even short forms of
Clinical Neurqpsychology, 10(2), 133+145. the Judgment of Line Orientation Test. Applied
Unverzagt, F. W., Hall, K. S., TorkE, A. M., Neuropsychology. 4, 244-246.
Rediger, J. D., et al. (1996). Effects of age, Vanderploeg, R. D., Schinka, J. A., Jones, T., Small,
education and gender on CERAD neu- B. J., Graves, A. B., & Mortimer, J. A. (2000).
ropsychological test performance in an African Elderly norms for the Hopkins Verbal Learning
American sample. Clinical Neuropsychologist, Test-Revised. Clinical Neuropsychologist, 14(3),
10(2), 180-190. 318-324.
Unverzagt, F. W., Morgan, 0. S., Thesiger, C. H., Vanderploeg, R. D., Crowell, T. A., & Curtiss, G.
Eldemire, D. A., Luseko, J., Pokuri~ S., et al. (2001). Verbal learning and memory deficits in
(1999). Clinical utility of CERAD ne~ropsycho traumatic brain injury: Encoding, consolidation,
logical battery in elderly Jamaicans. Journal of and retrieval. Journal of Clinical and Experi-
the International Neuropsychologic4l Society, mental Neurqpsychology, 23(2), 185-195.
5(3), 255-259. Van Gorp, W. G., & McMullen, W. J. (1997).
Uttl, B., & Graf, P. (1997). Color-Word Stroop Test Possible sources of bias in forensic neuropsy-
performance across the adult life sp~. Journal chological evaluations. Clinical Neurqpsycholo-
of Clinical and Experimental Neurop,sychology, gist, 11(2), 180-187.
19(3), 405-420. . Van Gorp, W. G., Satz, P., Kiersch, M. E., &
Uzzell, B. P., & Oler, J. (1986). Chron~ low-level Henry, R. (1986). Normative data on the Boston
mercury exposure and neuropsycholOgical func- Naming Test for a group of normal older adults.
tioning. Journal of Clinical and E~erimental Journal of Clinical and Experimental Neurqpsy-
Neurqpsychology, 8(5), 581-593. chology, 8(6), 702-705.
Vaisman, N., Voet, H., Akivis, A., & Vakil, E. Van Gorp, W. G., Satz, P., Miller, E., & Visscher, E.
(1996). Effect of breakfast timing on:the cogni- (1989). Neuropsychological performance in
tive functions of elementary school students. HIV-1 immunocompromised patients: A pre-
Archives of Pediatric and Adolescent; Medicine, liminary report. Journal of Clinical and Experi-
150(10), 1089-1092. ; mental Neuropsychology, 11(5), 763-773.
REFERENCES 605
Van Gorp, W. G., Satz, P., & Mitrushina, M. Visser, R. S. H. (1973). Manual of the Complex
(1990). Neuropsychological processes associated Figpre Test (CFT). Amsterdam: Swets & Zei-
with normal aging. Developmental Neuropsy- tlinger.
chology, 6(4), 279-290. Vlahou, C. H., & Kosmidis, M. H. (2002). The
van Spaendonck, K. P. M., Berger, H. J. C., Hor- Greek Trail Making Test: Preliminary nonnative
stink, M. W. I. M., Bonn, G. F., & Cools, A. R. data for clinical and research use. Psychology:
(1995). Card sorting performance in Parkinson's Journal of the HeUenic Psychological Society,
disease: A comparison between acquisition 9(3), 336--352.
and shifting perfonnance. Journal of Clini- Vollant, M., Lafitte, M. L., & Rapin, J. R. (1986).
cal and Experimental Neuropsychology, 17(6), Some specific errors in VRT of Benton in de-
918-925. tection of senile dementia of Alzheimer type. In:
Varney, N. R. (1981). Letter recognition and visual A. Bes (Ed.), Senile demmtia: Early detection
form discrimination in aphasic alexia. Neu- (pp. 631~). John Libbey Eurotext.
ropsychologia, 19, 795-800. Volz, H. P., Gaser, C., Haeger, F., Rzanny, R.,
Varney, N. R., Roberts, R. J., Struchen, M. A., Mentzel, H. J., Kreitschmann-Andermahr, 1.,
Hanson, T. V., Franzen, M. D., & Connell et al. (1997). Brain activation during cognitive
(1996). Design fluency among normals and pa- stimulation with the Wisconsin Card Sorting
tients with closed head injmy. Archives of Clin- Test-A functional MRI study on healthy vol-
ical Neuropsychology, 11(4), 345--353. unteers and schizophrenics. Psychiatric Re-
Vega, A., & Parsons, 0. A. (1967). Cross-validation search: Neuroimaging, 75, 145-157.
of the Halstead-Reitan Tests for brain damage. Waber, D. P., & Holmes, J. M. (1985). Assessing
Journal of Consulting Psychology, 31, 619-625. children's copy productions of the Rey-Osterrieth
Verma, S. K., Pershad, D., & Khanna, R. (1993). Complex Figure. Journal of Clinical and Experi-
Hooper's Visual Organization Test: Item analysis mental Neuropsychology, 7(3), 264-280.
on Indian subjects. Indian Journal of Clinical Waber, D. P., & Holmes, J. M. (1986). Assessing
Psychology, 20(1), 5-10. children's memory productions of the Rey-
Vernon, P. A. (1993). Der Zahlen-Verbindungs- Osterrieth Complex Figure. Journal of Clinical
Test and other trail-making correlates of general and Experimental Neuropsychology, 8, 563--580.
intelligence. Personality and Individual Differ- Waber, D., Bernstein, J., & Merola, J. (1989).
ences, 14(1), 35-40. Remembering the Rey-Osterrieth Complex
Veroff, A. E. (1980). The neuropsychology of aging: Figure: A dual-code, cognitive neuropsychologi-
Qualitative analysis of visual reproductions. cal model. Developmental Neuropsychology,
Psychological Research, 41, 259-268. 5, 1-15.
Vickers, D., & Lee, M. D. (1998). Never cross the Waber, D. P., Isquith, P. K., Kahn, C. M., & Ro-
path of a traveling salesman: The neural network mero, I. (1994). Metacognitive factors in the vi-
generation of Halstead-Reitan trail making tests. suospatial skills of long-term survivors of acute
Behavior Research Methods, Instruments, and lymphoblastic leukemia: An experimental ap-
Computers, 30(3), 423-431. proach to the Rey-Osterrieth Complex Figure
Vickers, D., Vincent, N., & Medvedev, A. (1996). Test. Developmental Neuropsychology, 10(4),
The geometric structure, construction, and in- 349-367.
terpretation of path-following (trail-making) tests. Waber, D. P. (2003). Parsing children's productions
Journal of Clinical Psychology, 52(6), 651-661. of the ROCF: What develops? In J. A. Knight
Vilkki, J. (1989). Differential perseverations in (Ed.), The handbook of Rey-Osterrieth Complex
verbal retrieval related to anterior and posterior Figure usage: Clinical and research applications.
left hemisphere lesions. Brain and Langpage, Lutz, FL: Psychological Assessment Resources.
36(4), 543-554. Wahlin, T.-B. R., Baeckman, L., Wahlin, A., &
Villardita, C., Cultrera, S., Cupone, V., & Mejia, R. Winblad, B. (1996). Trail Making Test perfor-
(1985). Neuropsychological test performances mance in a community-based sample of healthy
and normal aging. International Workshop: very old adults: Effects of age on completion
Psychiatry in aging and dementia. Archives of time, but not on accuracy. Archives of Geron-
Gerontology and Geriatrics, 4(4), 311-319. tology and Geriatrics, 22(1), 87-102.
Vingerhoets, G., Lannoo, E., & Wolters, M. (1998). Waldmann, B. W., Dickson, A. L., Monahan, M. C.,
Comparing the Rey-Osterrieth and Taylor & Kazelskis, R. (1992). The relationship between
Complex Figures: Empirical data and meta- intellectual ability and adult performance on
analysis. Psychologica Belgica, 38(2), 109-119. the Trail Making Test and the Symbol Digit
606 REFERENCES
Modalities Test. Journal of Clinical Psychology, and encoding strategy in nonnal aging. Journal
48(3), 360-363. of Clinical and Experimental Neuropsychology,
Walsh, P. F., Lichtenberg, P. A., & R(lwe, R. J. 22(4), 455-464.
(1997). Hooper Visual Organization Test per- Weible, J. A., Nuest, B. D., Welty, J., Pate, W. E., &
fonnance in geriatric rehabilitation patients. Turner, M. L. (2002). Demonstrating the effects
Clinical Gerontologist, 17(4), 3-11. of presentation rate on aging memory using the
Wang, P. J. (1977). Visual organization ability in California Verbal Learning Test (CVLT). Aging,
brain damaged adults. Perceptual firid Motor Neuropsychology, and Cognition, 9(1), 38-47.
Skills, 45, 723-728. Weigl, E. (1941). On the psychology of so-called
Ward, T. (1997). A note of caution for clinicians processes of abstraction. Journal of Abnormal
using the Paced Auditory Serial Addition task. and Social Psychology, 36, 3-33.
British Journal of Clinical Psycholdgy, 36(2), Weigner, S., & Donders, J. (1999). Perfonnance on
303--307. the Wisconsin Card Sorting Test after traumatic
Warkentin, S., c!L Passant, U. (1997). functional brain injury. Assessment, 6, 179-187.
imaging of the frontal lobes in organic: dementia. Weinberger, D. R., Bennan, K. F., & Zec, R. F.
Dementia and Geriatrtc Cognitive l)isorders, (1986). Physiologic dysfunction of dorsolateral
8(2), 105-109. prefrontal cortex in schizophrenia. I. Regional
Warner, M. H., Ernst, J., Townes, B. D., Peel, J. H., cerebral blood How evidence. Archives of Gen-
& Preston, M. (1987). Relationships between IQ eral Psychiatry, 43, 114-124.
and neuropsychological measures in ·neuropsy- Weinberger, D. R., Bennan, K. F., Iadarola, M.,
chiatric populations: Within-laboratory and Driesen, N., & Zec, R. F. (1988). Prefrontal
cross-cultural replications using WAIS and cortical blood How and cognitive function in
WAIS-R. Journal of Clinical and Experimental Huntington's disease. Journal of NeurolorJ.J.
Neuropsychology, 9(5), 545-562. : Neurosurgery, and Psychiatry, 51, 94-104.
Warrington, E. (1984). Recognition Merr»:Jry Test- Weinberger, D. R., Bennan, K. F., & Illowsky,
Faces. Windsor, UK: NFER-Nelson. B. P., (1989). Physiologic dysfunction of dorso-
Warrington, E., & Rabin, P. (1970). Perceptual lateral prefrontal cortex in schizophrenia. III. A
matching in patients with cerebral leSions. Neu- new cohort and evidence for a monoaminergic
ropsychologia, 8(4), 475-487. mechanism. Archives of General Psychiatry, 45,
Wechsler, D. (1945). A standardized memory scale 609--615.
for clinical use. Journal of Psychology, 19,87-95. Weinberger, D. R., Aloia, M. S., Goldberg, T. E., &
Wechsler, D. (1955). WAlS Technical Monual. San Bennan, K. F. (1994). The frontal lobes and
Antonio, TX: Psychological Corporation. schizophrenia. Journal of Neuropsychiatry and
Wechsler, D. (1987). Wechsler Memory Scale- Clinical Neuroscience, 6, 419-427.
Revised. San Antonio, TX: Psychological Corpo- Weingartner, H., Burns, S., Diebel, R., &
ration/Harcourt Brace Jovanovich. LeWitt, P. A. (1984). Cognitive impainnents in
Wechsler, D. (1997). WMS-III. Administration and Parkinson's disease: Distinguishing between
scoring manual. San Antonio, TX: Ps;chological effort-demanding and automatic cognitive pro-
Corporation/Harcourt Brace JovanoviCh. cesses. Psychiatry Research, 11, 223-235.
Wechsler, D. (2002a). WAlS-lll, WMS..Jll techni- Weinstein, C., Kaplan, E., Casey, M., & Hurwitz, I.
cal manual~pdated. San Antonio, TX: Psy- (1990). Delineation of female perfonnance on
chological Corporation. the Rey-Osterrieth Complex Figure. Neuropsy-
Wechsler, D. (2002b). WMS-III abbreviated- chology, 4, 117-127.
Manual. San Antonio, TX: Psychological Corpo- Weiss, K. M. (1996). A simple clinical assessment of
ration. attention in schizophrenia. Psychiatry Research,
Wecker, N. S., Kramer, J. H., Wisniewski, A., Delis, 60(2-3), 147-154.
D. C., & Kaplan, E. (2000). Age ·effects on Welch, L. W., Doineau, D., Johnson, S., & King, D.
executive ability. Neuropsycholol!lJ, 14(3), (1996). Educational and gender normative data
4{K}-414. . for the Boston Naming Test in a group of older
Wedding, D., & Faust, D. (1989). Clinical judg- adults. Broin and Language, 53, 260-266.
ment and decision making in neuropsychology. Welsh, K. A., Butters, N., Hughes, J., Mobs, R., &
Archives of Clinical NeuropsycholJgy, 4(3), Heyman, A. (1991). Detection of abnonnal
233-265. memory decline in mild cases of Alzheimer's
Wegesin, D. J., Jacobs, D. M., Zubin, N. R., Ven- disease using CERAD neuropsychological mea-
tura, P. R., & Stem, Y. (2000). Souree memory sures. Archives of Neurology, 48(3), 278-281.
REFERENCES 607
Welsh, K. A., Butters, N., Mobs, R. C., Beeldy, D., Wiens, A. N., McMinn, M. R., & Crossen, J. R.
Edland, S., Fillenhaum, G., et al. (1994). The (1988). Rey Auditory-Verbal Learning Test:
Consortium to Establish a Registry for Alzhei- Development of norms for healthy young adults.
mer's Disease (CERAD): V. A normative study Clinical Neuropsychologist, 2(1), 67-87.
of the neuropsychological battery. Neurology, Wiens, A. N., Tindall, A. G., & Crossen, J. R.
44(4), 609-614. (1994). California Verbal Learning Test: A nor-
Welsh, K. A., Fillenbaum, G., Wilkinson, W., mative data study. Clinical Neuropsychologist,
Heyman, A., et al. (1995). Neuropsychological 8(1), 75-90.
test performance in African-American and White Wiens, A. N., Fuller, K. H., & Crossen, J. R. (1997).
patients with Alzheimer's disease. Neurology, Paced Auditory Serial Addition Test: Adult
45(12), 2207-2211. norms and moderator variables. Journal of Clin-
Wentworth-Rohr, I., Mackintosh, R. M., & Fialk- ical and Experimental Neuropsychology, 19(4),
off, B. S. (1974). The relationship of Hooper 473-483.
VOT score to sex, education, intelligence and Wiggs, C. L., Weisberg, J., & Martin, A. (1999).
age. Journal of Clinical Psychology, 30(1), 73-75. Neural correlates of semantic and episodic mem-
Wetzel, L., & Boll, T. (1987). Short Category Test, ory retrieval. Neuropsychologta, 37(1), 103-118.
Booklet Format. Los Angeles, CA: Western Wilde, M. C., Boeke, C., & Sherer, M. (2000).
Psychological Services. Wechsler Adult Intelligence Scale-Revised block
Wetzel, L., & Murphy, G. S. (1991). Validity of the design broken configuration errors in non-
use of a discontinue rule and evaluation of dis- penetrating traumatic brain injury. Applied
criminability of the Hooper VISual Organization Neuropsychology, 7, 208-214.
Test. Neuropsychology, 5(2), 119--122. Wildgruber, D., Kischka, U., Fassbender, K., &
Wheeler, L., & Reitan, R. M. (1963). Discriminant Ettlin, T. M. (2000). The Frontal Lobe Score.
functions applied to the problem of predicting Part II: Evaluation of its clinical valdity. Clinical
cerebral damage from behavioral tests: A cross- Rehabilitation, 14(3), 272-278.
validation study. Perceptual and Motor Skills, 16, Williams, A. D. (2001). Psychometric concerns in
681-701. neuropsychological testing. NeuroRehabtlitation,
Whelihan, W. M., & Lesher, E. L. (1985). Neu- 16(4), 221-224.
ropsychological changes in frontal functions with Williams, B. W., Mack, W., & Henderson, V. W.
aging. Developmental Neuropsychology, 1, (1989). Boston Naming Test in Alzheimer's dis-
371-380. ease. Neuropsychologta, 27(8), 1073-1079.
White, A. J. (1984). Cognitive impairment of acute Williamson, J. B., & Harrison, D. W. (2003). Func-
mountain sickness and acetazolamide. Avia- tional cerebral asymmetry in hostility: A dual task
tion, Space and Environmental Medicine, 55, approach with fluency and cardiovascular regu-
589-603. lation. Brain and Cognition, 52(2), 167-174.
Whitfl.eld, K. E., Fillenbaum, G. G., Pieper, C., Willis, L., Yeo, R., Thomas, P .• & Garry, P. G.
Albert, M. S., Berkman, L. F., Blazer, D. G., (1988). Differential declines in cognitive function
et al. (2000). The effect of race and health- with aging: The possible role of health status.
related factors on naming and memory. The Developmental Neuropsychology, 4(1), 23-28.
MacArthur Studies of Successful Aging. Journal Wdson, B. A., & Watson, P. (2003). Performance of
of Aging and Health, 12(1), 69--89. people with nonprogressive brain injury and or-
Wiederholt, W. C., Cahn, D., Butters, N. M., Sal- ganic memory impairment on the ROCF. In J. A.
mon, D.P., et al. (1993). Effects of age, gender Knight (Ed.), The handbook of Rey-Osterrieth
and education on selected neuropsychological Complex Figure usage: Clinical and research ap-
tests in an elderly community cohort. Journal of plications. Lutz, FL: Psychological Assessment
the American Geriatrics Society, 41(6), 639--647. Resources.
Wiegner, S., & Donders, J. (1999). Performance on Wilson, B. A., Cockburn, J., & Halligan, P. (1987).
the California Verbal Learning Test after trau- Behaoioml Inattention Test. Titch6.eld, UK:
matic brain injury. Journal of Clinical and Ex- Thames Valley Test CoJGaylord, MI: National
perimental Neuropsychology, 21(2), 159--170. Rehabilitation Services.
Wiens, A.M., & Matarazzo, J.D. (1977). WAIS and Winegarden, B. J., Yates, B. L., Moses, J. A., Jr.,
MMPI correlates of the Halstead-Reitan Neu- Benton, A. L., & Faustman, W. 0. (1998). De-
ropsychology Battery in normal male subjects. ,velopment of an optimally reliable short form for
Journal of Neroous and Mental Disease, 164(2), Judgment of Line Orientation. Clinical Neuro-
112-121. psychologist, 12(3), 311-314.
608 REFERENCES
Wingenfeld, S. A., Holdwick, D. J., Jr., Davis, J. L., Woodard, J. L., Dunlosky, J., & Salthouse, T. A.
& Hunter, B. B. (1999). Normative data on (1999a). Task decomposition analysis of intertrial
computerized paced auditory serial addition task free recall performance on the Rey Auditory
performance. Clinical Neuropsychologist, 13(3), Verbal Learning Test in normal aging and Alz-
268-273. heimer's disease. Journal of Clinical and Exper-
Witjes-Ane, M.-N. W .• Vegter-van der Vlis, M., van imental Neuropsychology, 21(5), 666--676.
Vugt. J. P. P., Lanser, J. B. K., Hermans, J., Woodard, J. L., Goldstein, F. C., Roberts, V. J., &
Zwinderman, A., et al. (2003). Cognitive and McGuire, C. (1999b). Convergent and discrimi-
motor functioning in gene carriers for Hunting- nant validity of the CVLT (dementia version).
ton's disease: A baseline study. Journal of Journal of Clinical and Experimental Neuropsy-
Neuropsychiatnj and Clinical Neurosciences, chology, 21(4), 553-558.
15(1), 7-16. Woodruff-Pak, D. S., & Finkbiner, R. G. (1995).
Wolf, F. M. (1986). Meta-analysis: Quantitative Larger nondeclarative than declarative deficits in
methods for research synthesis. Newbury Park, learning and memory in human aging. Psychol-
CA: Sage. ogy and Aging, 10(3), 416-426.
Wolf, L. E., Comblatt, B. A., Roberts, S. A., Sha- Woodward, A. C. (1982). The Hooper Visual
piro, B. M., & Erlenmeyer-Kimling, L. (2002). Organization Test: A case against its use in
Wisconsin Card Sorting deficits in the offspring neuropsychological assessment. Journal of
of schizophrenics in the New York liigh-Risk Consulting and Clinical Psychology, 50(2), 286-
Project. Schizophrenia Research, 57(2-3), 288.
173-182. Woodward, J. L., Benedict, R. H. B., Roberts, V. J.,
Wong, T. M. (2000). Neuropsychological assess- Goldstein, F. C., Kinner, K. M., Capruso, D. X.,
ment and intervention with Asian Americans. In et al. (1996). Short-form alternatives to the
E. Fletcher-Janzen, T. L. Strickland, et al. Judgment of Line Orientation Test. Journal of
(Eds.), Handbook of cross-cultural r&europsy- Clinical and Experimental Neuropsychology, 18,
chology. Critical issues in neuropsychology. 898-904.
Amsterdam: Kluwer. Worrall, L. E., Yiu, E.M-L., Hickson, L. M. H., &
Wood, F. B., Ebert, V., & Kinsbourne, M. (1982). Barnett, H. M. (1995). Normative data for the
The episodic-semantic memory distinction in Boston Naming Test for Australian elderly.
memory and amnesia: Clinical and experimental Aphasiology, 9(6), 541-551.
observations. In L. Cermak (Ed.), Memory and Xavier, F., Ferraz, M., Trentini, C., Freitas, N., &
amnesia. Hillsdale, NJ: Lawrence Erlbaum. Moriguchi, E. (2002). Bereavement-related cog-
Wood, W. D .• & Strider, M.A. (1980). Comparison nitive impairment in an oldest-old community-
of two methods of administering the Halstead dwelling Brazilian sample. Journal of Clinical and
Category Test. Journal of Clinical Psrchology, Experimental Neuropsychology, 24(3), 294--301.
36(2), 476-479. Yaffe, K., Cauley, J., Sands, L., & Browner, W.
Woodard, J. L. (1994). Personal program- (1997). Apolipoprotein E phenotype and cog-
Computerized version of the Wisconain Card nitve decline in a prospective study of elderly
Sorting Test. community women. Archives of Neurology,
Woodard, J. L., Axelrod, B. N., & Henry, R. R. 54(9), 1110-1114.
(1992). Interrater reliability of scoring :parame- Yaffe, K., Grady, D., Pressman, A., & Cummings, S.
ters for the Design Fluency Test. Neuropsy- (1998). Serum estrogen levels, cognitive perfor-
chology, 6(2), 173-178. mance, and risk of cognitive decline in older
Woodard, J. L., Benedict, R. H. B., Roberts, V. J., community women. Journal of the American
Goldstein, F. C., Kinner, K. M., Caprulio, D. X., Geriatrics Society, 46(7), 816--821.
et al. (1996). Short-form alternative~ to the Yaffe, K., Blackwell, T., Gore, R., Sands, L., Reus,
Judgment of Line Orientation Test. Journal of V., & Browner, W. S. (1999a). Depressive
Clinical and Experimental Neuropsychology, symptoms and cognitive decline in nondemented
18(6), 898-904. elderly women: A prospective study. Archives of
Woodard, J. L., Benedict, R. H. B., Salthouse, T. A., General Psychiatry, 56(5), 425-430.
Toth, J. P., Zgaljardic, D. J., & Hancodc, H. E. Yaffe, K., Browner, W., Cauley, J., Launer, L., &
(1998). Normative data for equivalent. parallel Harris, T. (1999b). Association between hone
forms of the Judgment of Line Orientation Test. mineral density and cognitive decline in older
Journal of Clinical and Experimental N~ropsy women. Journal of the American Geriatrics So-
chology, 20(4), 457-462. ciety, 47(10), 1176-1182.
REFERENCES 609
Yaffe, K., Lui, L.-Y., Zmuda, J., & Cauley, J. (2002). Youngjohn, J. R., Larrabee, G. J., & Crook, T. H.
Sex honnones and cognitive function in older (1992). Discriminating age-associated memory
men. Journal of the American Geriatrics Society, impainnent from Alzheimer's disease. Psycho-
50(4), 707-712. logical Assessment, 4(1), 54-59.
Yamazaki, A. (1985). Interference in the Stroop Youngjohn, J. R., Larrabee, G. J., & Crook, T. H.
color-naming task. Japanese Journal of Psychol- (1993). New adult age- and education-correction
ogy, 56, 185-191. nonns for the Benton Visual Retention Test.
Yeates, K. 0., Patterson, C. M., Waber, D. P., & Clinical Neuropsychologist, 7(2), 155-160.
Bernstein, J. H. (2003). Constructional and figural Yurgelun-Todd, D. A., & Kinney, D. K. (1993).
memory skills following pediatric closed-head Patterns of neuropsychological deficits that
injury: Evaluation using the ROCF. In J. A. Knight discriminate schizophrenic individuals from
(Ed.), The handbook of Rsy-Osterrieth Complex siblings and control subjects. Journal of Neu-
Figure usage: Clinical and research applications. ropsychiatry and Clinical Neurosciences, 5(3),
Lutz, FL: Psychological Assessment Resources. 294-300.
Yehuda, R., Keefe, R. S. E., Harvey, P. D., Le- Zable, M., & Harlow, H. F. (1946). The perfor-
vengood, R. A., Gerber, D. K., Geni, J., et al. mance of rhesus monkeys on a series of object
(1995). Learning and memory in combat veter- quality and positional discriminations and dis-
ans with posttraumatic stress disorder. American crimination reversals. Journal of Comparotive
Journal of Psychiatry, 152(1), 137-139. Psychology, 39, 1.
Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, 0., Zachary, R. A., & Gorsuch, R. L. (1985). Contin-
Huang, V., Adey, M., et al. (1983). Development uous nonning: Implications for the WAIS-R.
and validity of a Geriatric Depression Scale: A Journal of Clinical Psychology, 41(1), 86-94.
preliminary report. Journal of Psychiatric Re- Zakzanis, K. K. (1998). The reliability of meta-
search, 17, 37-49. analytic review. Psychological Reports, 83(1),
Yeudall, L. T., Fromm-Auch, D., & Davies, P. 215-222.
(1982). Neuropsychological impainnent of per- Zalewski, C., Thompson, W., & Gottesman, I.
sistent delinquency. Journal of Nervous and (1994). Comparison of neuropsychological test
Mental Disease, 170(5), 257-265. perfonnance in PTSD, generalized anxiety dis-
Yeudall, L. T., Reddon, J. R., Gill, D. M., & Ste- order, and control Vietnam veterans. Assess-
fanyk, W. 0. (1987). Nonnative data for the ment, 1(2), 133-142.
Halstead-Reitan Neuropsychological Tests strati- Zappala, G., Measso, G., Cavarzeran, F., Grigo-
fied by age and sex. Journal of Clinical Psy- letto, F., Lebowitz, B., Pirozzolo, F., et al.
chology, 43(3), 346--367. (1995). Aging and memory: Corrections for age,
Yeudall, L. T., Fromm, D., Reddon, J. R., & Stefanyk, sex and education for three widely used memory
W. 0. (1986). Normative data stratified by age and tests. Italian Journal of Neurological Sciences,
sex for 12 neuropsychological tests. Journal of 16(3), 177-184.
Clinical Psychology, 43(3), 918-946. Zec, R. F., Landreth, E. S., Fritz, S., Grames, E.,
Ylikoski, R., Ylikoski, A., Erkinjuntti, T., Sulkava, Hasara, A., Fraizer, W., et al. (1999). A com-
R., Raininko, R., & Tilvis, R. (1993). White parison of phonemic, semantic, and alternating
matter changes in healthy elderly persons cor- word fluency in Parkinson's disease. Archives of
relate with attention and speed of mental pro- Clinical Neuropsychology, 14(3), 255-264.
cessing. Archives of Neurology, 50, 818-824. Zeitlhofer, J., Asenbaum, S., Spiss, C., Wimmer, A.,
York, C. D., & Cennak, S. A. (1995). VISual percep- Mayr, N., Wolner, E., et al. (1993). Central
tion and praxis in adults after stroke. American nervous system function after cardiopul-
JournalofOccupatianalTherapy,49(6),543-550. monary bypass. European Heart Journal, 14(1),
York, M. K., Levin, H. S., Grossman, R. G., Lai, 885-890.
E. C., & Krauss, J. K. (2003). Clustering and Zhou, W., Liang, Y., & Christiani, D. C. (2002).
switching in phonemic fluency following pallid- Utility of the WHO neurobehavioral core test
otomy for the treatment of Parkinson's disease. battery in Chinese workers-A meta-analysis.
Journal of Clinical and Experimental Neuropsy- Environmental Research, 88(2), 94-102.
chology, 25(1), 110-121. Zondennan, A. B., Giambra, L. M., Arenberg, D.,
Young, M. H., & Justice, J. (1998). Neuropsy- Resnick, S., & Costa, P. (1995). Changes in im-
chological functioning of inmates referred for mediate visual memory predict cognitive im-
psychiatric treatment. Archives of Clinical Neu- painnent. Archives of Clinical Neuropsychology,
ropsychology, 13(3), 303-318. 10(2), 111-123.
Appendix 1: Where to Buy the Tests
Several of the tests mentioned in this book are indicates that the company listed at left is also
available from more than one distributor. An the primary publisher of the test.
asterisk placed before the name of the test
TEST PUBLISHER/DISTRIBUTOR TEST NAME

Editorial Medica Panamericana 0 Boston Naming Test, Spanish version
Alberto Alcocer, 24-6a
28036 Madrid
Spain
(3491)-457-0203 [Phone]
www.medicapanamericana.com
Lafayette Instrument 0 Lafayette Hand Dynamometer
3700 Sagamore Parkway North 0 Grooved Pegboard Test
P.O. Box 5729
Lafayette, IN 47903
1-800-428-7545 [Phone orders]
1-765-423-4111 [Fax orders]
www.Lafayetteinstrument.com
Normative Data. com 0 WHO-UCLA Auditory-Verbal Learning Test
35 S. Raymond Ave., #304 0 Stroop (Comalli/Kaplan versions in Spanish
Pasadena, CA 91105-1993 and/or English)
1-626-304-9995 [Fax orders only]
www.NormativeData.com
611
612 APPENDIX 1
Psychological Assessment Resources •color Trails Test (Adult & Children's

16204 N. Florida Ave. versions)
Lutz, FL 33549 Grooved Pegboard Test
1-800-331-TEST [Phone orders] Finger Tapping Test
1-866-727-2884 [24-hour order line] Lafayette Hand Dynamometer
1-800-727-9329 [Fax orders] Boston Naming Test-Revised
www.parinc.com •Rey Complex Figure & Recognition
•Intermediate Booklet Category Test
•Portable Tactual Performance Test
•stroop Test (Trenerry version)
Stroop Color & Word Test (Golden version)
Ruff Figural Fluency Test
•wisconsin Card Sorting Test (64 SP,
64 V2, and V4 versions)
Boston Qualitative Scoring System for the
Rey-Osterrieth Complex Figure
Hopkins Verbal Learning Test-Revised
Rey Auditory-Verbal Learning Test-A
Handbook
Benton Judgment of Line Orientation
Benton Visual Form Discrimination Test
Ruf£2&7
Digit Vigilence Test
The Psychological Corporation Boston Naming Test-Revised
555 Academic Court Rey Complex Figure & Recognition
San Antonio, TX 78204-2498 •Wechsler Memory Scale-Revised
1-800-211-8378 [Phone orders] •Wechsler Memory Scale-III, IliA
1-800-232-1223 [Fax orders] California Verbal Learning Test, 2nd ed.
www.PsychCorp.com Wisconsin Card Sorting Test (64 V2 and
V3 versions)
Paced Auditory Serial Addition Test
Reitan Neuropsychological Laboratory •category Test
P.O. Box 66080 •Finger Tapping Test
Tucson, AZ 85728 •Tactual Performance Test
1-520-577-2970 [Phone orders] •Trail Making Test (1945 version)
1-520-577-2940 [Fax orders] Hand Dynamometer
www.ReitanLabs.com
Riverside Publishing Boston Naming Test-Revised
425 Spring Lake Drive Stroop Color & Word Test (Golden version)
Itasca, IL 60143-2079 Lafayette Hand Dynamometer
1-800-323-9540 [Phone orders] Grooved Pegboard Test
1-630-467-7192 [Fax orders]
www.riversidepublishing.com
Western Psychological Services •short Category Test, Booklet Format
12031 Wilshire Blvd. •Stroop Color & Word Test (Golden version)
Los Angeles, CA 90025-1251 •Hooper Visual Organization Test
1-800-648-8857 [Phone orders] Rey Auditory-Verbal Learning Test
1-310-478-7838 [Fax orders] Wisconsin Card Sorting Test
www.wpspublish.com
Appendix 2a: Subject Instructions for
ACT According to Boone et al. (1990)
and Boone (1999)
The examiner instructs the patient: ''I'm going letters. In other words, you have to do two
to say three letters, and I want you to say them things at once-hold the letters in mind while
back to me. Ready? Q, L, X." you are counting backward by 3s. It is a dif-
Administer first five trials. Write down pa- ficult task, and it is normal to make mistakes.
tient's responses in the response column. In Let's try one. X, C, P, 194. Start counting."
the next column, indicate the number of cor- 'nle numbers in the "delay" column indi-
rect responses (i.e., 0, 1, 2, or 3). If more than cate how many seconds the person should
one or two errors are present, the test should count backward by 3s. Begin timing when the
be discontinued. If the patient cannot repeat patient actually says a number out loud. You
the letters reliably, this would point to a lin- may allow the patient to attempt to consoli-
guistic or hearing problem, in which case date the information for approximately 1-
frontal lobe functioning cannot be assessed 2 seconds prior to the actual commencement
with this test. of counting; do not allow more time than this.
"Now I'm going to say three letters, and A perseveration is scored if the person says
then I'm going to say a number. After I say the a letter that is incorrect and one which was
number, I want you to count backward out said in the trial directly preceding the current
loud by 3s from the number. For example, if one; a total of 57 perseverations are possible.
the number was 100, you would say, '100, 97, A sequence error occurs when the patient says
94,' etc. After a few seconds of counting, I the letters out of sequence; a total of 20 se-
will stop you and will want you to tell me the quence errors are possible.
613
Append ix 2b: Auditory Consonant Trigrams
(Boone et al., 1990; Boone, 1999)
Starting Delay
Stimulus umbe r (Seconds) Response #Correct Perseveration Sequence
QLX - 0
SZB - 0
HJT - 0
CPW - 0
DLH - 0
X p 194 l
DJ 75 9
FXB 28 3
J 1 0 9
BCQ 167 1
KM 20 3
RA'T 18 l
KF 82 9
MBW 47 3
TDH 141 9
LRP 51 3
zw 117 18
PHQ 9 9
XGD 15 l
ZQ 91 3
umber orrect:
0" Delay /15
3" Delay / 15
9" D elay / 15
1 " Delay /15
TOTAL / 60
614
Appendix 2c: Subject Instructions for ACT
According to Stuss et al. (1987, 1988)
The examiner starts by saying: "I am going to make sure that the patient is counting dur-
say three letters and when I am through, I am ing the delayed period. Some patients tend
going to knock like this. When I do I want you to repeat the letters after the examiner in-
to say the letters back to me." The examiner stead of starting counting immediately. In
says letters out loud at the rate of 1 per second this instance, the patient should be told not
and records the patient's answers. to repeat the letters but to start counting as
After the five trials with "0" delay, the examiner soon as he or she hears the number. The
continues the test. "This time, I am going to say examiner may have to encourage him or
three letters followed immediately by a number. her to count out loud by counting with him
As soon as you get the number, I want you to or her at the beginning or if the patient
start counting backward out loud until I knock as hesitates.
before." (Examiner demonstrates by knocking on Some patients have great difficulty counting
the desk) "When I knock, I want you to recall the by 3s backward. In such cases, the patient is
three letters. Do you have any questions?" asked to count backward by ls instead. How-
If the instructions are clearly understood, ever, this procedure is nonstandard and should
the examiner starts with the delayed trials. If be noted.
not understood, repeat with examples. For On this test, only one presentation for each
this part of the test there are three delayed- trial is allowed.
recall conditions, which are "3," "9," and "18"
randomly alternating. All trials are presented
independently of the patient's performance. SCORING
The examiner says the letters and the num-
bers and immediately starts the stopwatch For each trial, the letters given by the patient
until the corresponding delayed-recall period are recorded verbatim (i.e., in the order re-
elapses. He then knocks on the desk and reported by the patient) in the first column of
cords the letters reported by the patient. the score sheet. The number of correct letters
identified is noted in the second column. The
number of correct letters for each delayed
condition over 15 is registered at the bottom
SUPPLEMENTARY INSTRUCTIONS
of the score sheet. The summation of these
On this test, it is important to maintain in- subscores constitutes the total score of the test
terference conditions. The examiner must over 60.
615
Appendix 2d: Auditory Consonant Trigrams
(Stuss et al., 1987, 1988)
Starting Delay Responses Number

Stimulus Number (Seconds) Correct
QLX - 0
SZB - 0
HJT - 0
GPW - 0
DLH - 0
XCP 194 18
NDJ 75 9
FXB 128 36
JCN 180 9
BGQ 167 18
KMC 120 36
RXT 188 18
KFN 82 9
MBW 147 36
TDH 141 9
616
APPENDIX 20 617
Starting Delay Responses Number

Stimulus Number (Seconds) Correct
LRP 151 36
zws 117 18
PHQ 89 9
XGD 158 18
CZQ 191 36
Last
Number Correct Two Two
0" Delay
9" Delay
18" Delay
36" Delay
TOTAL
Appendix 3: WHO-UCLA Auditory Verbal
Learning Test: Instructions and Test Forms
GENERAL INSTRUCTIONS list. At the end of each trial write down the
FOR EXAMINERS total numbers of correct responses, repeti-
tions, and intrusions.
Instructions to Subjects: The ins~ons to
subjects are printed on the test fonn. Be sure If the subject makes an intrusion, wait until he
to indicate before Trial II that the: subject indicates that he is finished, then prompt the
should try to remember as many wo~ as he subject by saying, "Yau said __ , __ was
can, ". . . tncludtng the words yau remem- not on the list."
bered on the first trial." After the last trial
If the subject makes an intrusion that may
(Trial V), you should remember to ·tell the
reflect poor hearing on the part of the subject
subject that "I want yau to try to ~eras
or poor pronunciation on the part of the ex-
many of those words as possible bectuse I'm
aminer (such as saying "Pie" instead of
going to ask yau abaut them again a little
"Eye"), count the item as correct the first
later." These instructions are printed on the
time, and correct the subject by saying, "Yau
test fonn for "Recall Following Interference."
said Pie, the correct word is Eye." If the
Instructions to Examiners: Use a clipboard subject continues to produce the same intru-
to hold the test fonns out of the subject's view. sion on subsequent trials, correct the subject
Read the words at the rate of approxi~ately 1 in the same manner but score the response as
word per second. On the last word, 4rop the an intrusion. Make allowances for translation
pitch of your voice to indicate that you are and pronunciation difficulties on the part of
finished. If necessary, prompt the s~ject to nonnative speakers of the language being used
begin recall. In general, you should not look at for test administration.
the subject while reading the list or while re-
If a subject asks if a particular response is
cording his responses since this kind of eye
correct, answer him truthfully. If a subject asks
contact makes many people nervous.
if he has already said a particular word, again,
Code all responses by placing a check mark in answer truthfully, but count the word as a
the relevant box. Place additional chedc marks repetition if he has already said it once. In
when items are repeated. If the subject gives a general, feel free to answer any of the subject's
word that is not on the list, record ~ intru- questions about the task, including the number
sion in the spaces provided below the ;IS-item of trials and the number of words on the list.
618
APPENDIX 3 619
If the subject is not producing at least 10 re- etc.), and record responses under Trial VI.
sponses by the third trial, encourage the Unlike earlier trials, you should not correct
subject to try a little longer. any intrusions that the subject makes.
Instructions for Trial VII: After the subject
has recalled as much as possible from the in-
ACQUISITION TRIALS terference list, say, "Now I'd like to see how
Trial 1: "The next task may seem a bit difficult many words you can recall from the first list-
in the beginning, but usually it gets easier as the one we went through five times. Tell me as
we go along. I am going to read for you a long many words as you can remember from the
list of words. Once I'm done, I'd like to see first list." Record responses under Trial VII.
how many of the words you can recall. You
can repeat the words in any order that you
prefer; you don't have to use the same order 30-MINUTE DELAYEO RECALL
that I use. Then, I am going to read the same AND RECOGNITION
list for you a few more times, to see how many
of the words you can eventually learn. Instructions for Trial VIII: Without reading
Ready?" the list again, say, "Remember the long list of
words I read to you five times? I'd like you
Trial II: "That was a good beginning. Now now to tell me as many of the words from that
I'm going to read the same list again, and list as you can remember." Do not correct the
again I would like to see how many of the subject if he/she makes any intrusions.
words you can recall, including the words
you remembered on the first trial. Again, listen Instructions for Trial IX: Immediately fol-
very carefully. Ready?" lowing the delayed recall say, "Next I would
like to see how many of the words you can
Trials 111-V: "Very good. I'm going to read recognize. Say Yes if you hear a word you
the list again. Again, listen carefully and try to think was part of the original list we went
remember as many words as you can. Ready?" through five times. If you think that the word
is not from that list, say No. Make sure you
only say Yes to those words you are sure you
INTERFERENCE LIST, RECALL remember as being a part of that list we went
FOLLOWING INTERFERENCE through five times." Read the words in order
from left to right. Circle the word if the sub-
Instructions for Trial VI: After Trial V of the
ject says 'Yes.' If the subject hesitates and
primary word list, say, "Very good. I want you
fails to answer within a few seconds, say, "If
to try to remember as many of those words as
you are not sure, just make your best guess."
possible because I'm going to ask you about
Words from the original list are capitalized.
them again a little later." Then say, "Now I am
going to read for you a different list of words. Scoring: "Correct Recognitions" is the total
Once again, when I'm done, I'd like to see how number of circled words that are capitalized.
many of the words you can recall. Ready?" "False Identifications" is the total number of
Read the interference list (boot, monkey, circled words that are not capitalized.
620 APPENDIX 3
Trial I Trial II Trial II Trial IV Trial V
Ann Arm
Cat Cat
AJ.e AJ.e
Bed Bed
Plane Plane
Ear Ear
Dog Dog
Hammer Hammer
Chair Chair
Car Car
Eye Eye
Horse Horse
Knife Knife
Clock Clock
Bike Bike
Correct:
Repeats:
Intrusions:
Copyright © 1990 by Paul Satz, Ph.D., Alexander Chervinsky, Ph.D., and Louis F. D'Elia, Ph.D. All rights reserved.
Fonn design by E.N. Miller.
APPENDIX 3 621
Trial VI Trial VII
Boot Arm
Monkey Cat
Bowl Axe
Cow Bed
Finger Plane
Dress Ear
Spider Dog
Cup Hammer
Bee Chair
Foot Car
Hat Eye
Butterfly Horse
Kettle Knife
Mouse Clock
Hand Bike
Correct:
Repeats:
Intrusions:
Copyright © 1990 by Paul Satz, Ph.D., Alexander CheJVinsky, Ph.D., and Louis F. D'Elia, Ph.D. All rights reserved.
Form design by E.N. Miller.
622 APPENDIX 3
Trial VIII Trial IX-Oral Recognition
Ann
Cat
. mirror
HAMMER
KNIFE
HORSE
leg
DOG
truck
EYE
fish
candle table EAR
Axe motorcycle CAT BIKE
Bed AXE lips snake
CLOCK tree stool
Plane CHAIR ARM bus
PLANE nose BED
Ear turtle sun CAR
Dog Correct Recognitions: _ __

False Identifications: _ __
Hammer
Chair
Car
Eye
Horse
Knife
Clock
Bike
Correct:
Repeats:
Copyright © 1990 by Paul Satz, Ph.D., Alexander Chervinsky, Ph.D., and Louis F. D'Eiia, Ph.D. All rights reserved.
Form design by E.N. Miller. ·
Appendix 4: Locator and Data Tables
for the Trailmaking- Test (TMT)
Study numbers and page numbers provided in Locator table also provides a reference for
these tables refer to study numbers and de- each study to a corresponding data table in
scriptions of studies in the text of Chapter 4. this appendix.
Table A4.1. Locator Table for the Trail-Making Test (TMT)

Study Age• n Sample Composition IQ/Education• Location
TMT.l Davies, 1968 540 British adults: England

page 72 20-39 180 50 M,40 F
Table A4.2 40-49 90 in each decade of age.
50-59 90 Mean times
60-69 90 corresponding to several
70-79 90 percentile ranges are
presented.
TMT.2 Goul & 106 Ss were Canadian workers' Educ. Canada
Brown, 1970 20-29 26 compensation non-brain- 6-13
page 73 30-39 25 injured, hospitalized patients.
Table A4.3 40-49 24 Data are stratified by five age
50-59 16 groups.
1»-72 15
TMT.3 Wiens & 48 All males, neurologically normal. Educ. Portland, OR
Matarazzo, 1977 23.6 Divided into two equal groups. 13.7
page 73 24.8 Random sample of 29 were 14.0
Table A4.4 retested 14-24 weeks later. FSIQ
117.5
118.3
TMT.4 Eson et a!., 63 Older participants. Data are USA
personal communication 63.2 15 provided for 4 age groups.
page 74 67.0 16
Table A4.5 72.0 16
78.3 16
(continued)
623
624 APPENDIX 4
Table A4.1. (Contd.)

TMT.5 Harley et a!., 55-79 193 V.A.-hospitalized patients Educ. Wisconsin

1980 (with chronic brain 8.8
page 74 syndrome were included).
Table A4.6 55-79 160 Participants in alcohol-
equated sample. Both
samples are divided into
5 age groups.
TMT.6 Anthony et a!., 38.88 100 Healthy control group. Educ. Colorado
1980 (15.80) Data are provided 13.33 (2.56)
page 74 for Trails B only. FSIQ
Table A4.7 113.5 (10.8)
TMT. 7 Bak & Greene, 50-62 15 6 M, 9F Educ. Texas
1980 55.6 13.7 (1.91)
page 75 (4.4)
Table A4.8 67-86 15 5 M, 10 F 14.9 (2.99)
74.9 healthy participants
(6.0)
TMT.8 Kennedy, 1981 150 Ss were employees of a mental Educ. (est. IQ) Canada
page 75 2~29 30 health center. Five age groups 13.73 (123.43)
Table A4.9 ~9 30 are represented; equal M/F ratio. 13.53 (127.10)
~9 30 13.11 (127.40)
~9 30 11.59 (123.30)
~9 30 12.50 (128.54)
TMT.9 Fromm-Auch & 15-64 193 111 M, 82 F volunteers Educ. Canada
Yeudall, 1983 25.4 described as nonpsychiatric ~26
page 76 (8.2) & nonneurological; 83% 14.8 (3.0)
Table A4.10 15-17 32 are R-handed. Sample FSIQ
18-23 75 partitioned into 5 age groups. 119.1 (8.8)
24-32 57
33-40 18
41-64 10
TMT.10 Bornstein, 1985 1~9 365 178M, 187 F paid volunteers. Educ. Canada
page 76 43.3 Neurologically healthy. 12.3 (2.7)
Table A4.11 (17.1) Data are presented in age x (5-20 years)
2~9 gender x education ceUs. < high school
~9 2: high school
~9
TMT.ll Heaton eta!., 15-81 553 356M, 197 F. Data for Trails B Educ. Colorado,
1986 39.3 are reported. Sample was 13.3 (3.4) Callfornia,
page 77 (17.5) divided into 3 age groups & < 12 (132) Wisconsin
Table A4.12 <40 3 education categories; % of Ss 12-15 (249)
~9 classified as normal is provided. 2:16 (172)
2:60
TMT.12 Alekoumbides 19-82 118 Ss were medical and psychiatric Educ. Southern
eta!., 1987 46.85 V.A. inpatients without cerebral 1-20 California
page 77 (17.17) lesions or histories of alcoholism 11.43 (3.20)
Table A4.13 or cerebral contusions. All Ss FSIQ
except for one were male. 105.9 (13.5)
TMT.13 Bornstein 17--52 23 9 M, 14 F volunteers. Data on VIQ Canada/
eta!., 1987a 32.3 3-week retest are provided. 105.8 (10.8) Ohio
page 78 (10.3) PIQ
Table A4.14 105.0 (10.5)
APPENDIX 4 625

TMT.l4 Dodrill, 1987 27.77 120 60 M, 60 F volunteers. Educ. Washington

page 78 (11.04) Data for various intelligence 12.28 (2.18)
Table A4.15 levels are presented. FSIQ
100 (14.3)
TMT.l5 Ernst, 1987 65-75 no 51 M, 59 F volunteers. Time Educ. Brisbane,
page 79 to completion and number of 10.3 Australia
Table A4.16 errors are provided.
TMT.l6 Stuss et al., 1987 60 Canadian English- or French- Educ. Canada
page 80 16-19 10 speaking Ss; 55% male, 18% 14.3 (2.62)
Tables A4.17, A4.18 20-29 10 L-handed; 6 age groups $12
30-39 10 represented. Data for test/ >12
40-49 10 retest ( 1 week) are provided.
50-59 10
60-69 10
TMT.l1 Yeudall et al., 15-20 225 Normal adults; 127 M, Educ. Canada
1987 21-25 98 F, data are stratified by 10-17
page 80 26-30 4 age groups x gender. 14.55 (2.78)
Table A4.19 31-40 FSIQ
ll2.25 (10.25)
TMT.18 Bomstein & Suga, 55-70 134 Healthy elderly paid volunteers; Educ. Canada
1988 62.7 49 M, 85 F. No history of 11.7
page 81 (4.3) neurological or psychiatric (2.9)
TableA4.20 disorders. Data are divided
into 3 education groups: Range, mean:
17M, 29 F 5-10,8.5
16M, 28 F 11-12, 11.7
16M, 28 F >12, 15.0
TMT.l9 Stuss et al., 90 Data are divided into 3 Canada
1988 16-29 30 age groups for original test Educ.frange
page 81 30-49 30 and 1-week retest. 14.1 (1.34)
Table A4.21 50-69 30 Expansion of the TMT.16 11-18
study 14.9 (3.95)
5-20
13.2 (2.38)
8-18
TMT.20 Van Gorp et al., 57-SS 156 Elderly Ss with no history Educ. California
1990 57-65 28 of neurological or 14.4 (2.86)
page 82 66-65 45 psychiatric disorders; FSIQ
Table A4.22 71-75 57 61% F; 4 age groups 117.21
76-85 26 presented. (21.59)
TMT.lU Heaton et al., 42.1 486 Volunteers; 65% of the Educ. California,
1991,2004 (16.8) sample were males. Data 13.6 (3.5) Washington,
page 82 20-34 are presented in T-score FSIQ Texas,
Data are not 35-39 equivalents for M and F 113.8 Oklahoma,
reproduced 40-44 separately in 10 age groupings (12.3) Wisconsin,
in this book 45-49 by 6 education groupings. Illinois,
50-54 In the 2004 edition, age Michigan.
55-59 range is expanded to New York,
60-64 include 85 years, and the Virginia,
65-69 data are presented for Massachusetts,
70-74 African-American and Canada
75-80 Caucasian participants
separately.
(continued)
626 APPENDIX 4

TMT.U Seines et al., 733 Ss from MACS study. Sero- <College MACS centers
1991 25--34 negative homosexual and College at Baltimore,
page 83 35--44 bisexual males. Data are >CoDege Chicago,
Table A4.23 45-54 stratified by 3 age Los Angeles.
groups and 3 education &: Pittsburgh
levels.
TMT.i3 Elias et al., 15--24 427 Healthy volunteers; 187 M, 240 F. Educ. Maine
1993 25--34 Data are stratified by 6 age 12-19
page 84 35--44 groups x gender.
Table A4.24 45-54
55-&4
~65
TMT.!4 Cahn et al., 78.4 238 Cognitively intact elderly 13.8 California
1995 (6.8) participants in Rancho Bernardo (2.6)
page 84 Study. Data for the entire sample
Table A4.25 and optimal cutoffs are provided.
TMT.215 Ivnik et al., 1996 359 167 M, 192 F; 332 R-handed; MAYOFSIQ Minnesota
page 85 56-59 54 normal elderly volunteers. 106.2 (14.0)
TableA4.26 60-64 81 The article provides tables
65-69 65 for age correction and a
70-74 57 regression equation for
75--79 53 education correction.
80-84 27 Tables are not reproduced.
85--89 17
90-94 5
TMT.I6 Richardson&: 81.5 101 All autonomously living elderly Educ. New Haven,
Marottoli, 1996 (3.3) Ss, current drivers; 53 M, 48 F. 11.0 (3.7) CT
page 86 76-80 Data are provided for Trails B <12
Tables A4.27, A4.28 81-91 for younger-old and older-old ~12
by two educational levels.
TMT.I7 Hoff et al., 32.1 54 Paid male volunteers, control 15.4 New York
1996 (9.7) group; strict selection criteria. (2.4)
page 86
Table A4.29
TMT.28 Salthouse 18-39 40 Healthy adults, 47% M. 15.5 Atlanta,
et al., 1997 Data are stratified into 3 (1.7) GA
page 87 40-59 38 age groups. 15.2
Table A4.30 (2.5)
60-78 37 15.3
(2.6)
TMT.I9 Rasmusson 60-69 203 Nondemented elderly sample, 16.0 Baltimore, MD
et al., 1998 70-79 262 participants in Bl.SA study, (2.9)
page 87 80-89 179 majority are males; sample
Table A4.31 90-96 23 is partitioned into 4 age groups.
TMT.30 Miner &: 21.7 110 Undergrad. students, 22 M, 88 F. Undergrad. North Dakota
Ferraro, 1998 (5.24) Performance was compared for students
page 88 A-B and B-A order of presentation.
Table A4.32
TMT.31 Crowe, 23.4 98 Undergrad. students, 49 M, 49 F 14.0 Australia
1998b (3.1) (2.3)
page 88
Table A4.33
APPENDIX 4 627

Study Age" n Sample Composition IQ/Education• Location
TMT.32 Tremont 16-74 157 Patients referred for evaluation with 13.12 Oklahoma
et al., 1998 negative findings; 71 M, 86 F. (3.26)
page88 Data are stratified by 3 levels of
TableA4.34 intelligence.
TMT.33 Basso 32.50 50 Data for healthy men on 2 testing 14.98 Tulsa, OK/
et al., 1999 (9.27) probes over a 12-month interval. (1.93) Ohio
page 89
TableA4.35
TMT.34 Crews 20.20 30 Control sample of 30 women 14.40 Blacksburg,
et al., 1999 (3.47) VA
page 89
Table A4.36
TMT.315 Dikmen 34.2 384 Normal and neurologically stable 12.1 Washington,
et al., 1999 (16.7) adults; some had neurological (2.6) Colorado,
page 90 conditions; 66% M. Data on California
Table A4.37 test-retest reliability and practice
effect are provided.
TMT.36 Binder 82.3 125 Normal elderly sample, aged 70 or above, 13.5 St. louis, MO
et al., 1999 (4.4) 25% M, 87% Caucasian. Data on the (3.0)
page 91 number of lines drawn within 180 sec.
Table A4.38 for both parts A and B and time to
completion are reported.
TMT.37 Ruffolo 29.1 49 Control sample consisting of students 14.3 Rhode Island
et al., 2000 (12.1) and employees of social services (1.9)
page 91 agency. Time to completion and
TableA4.39 number of errors are reported.
TMT.38 Saxton 73.63 357 Elderly volunteers who participated in 13.23 Washington
et al., 2000 (4.45) the multicenter Memory and Aging (2.85) County, MD
page 91 Study; 44.9% male. Pittsburg, PA
Table A4.40
TMT.39 Chen et al., 74.9 483 Control elderly volunteers who 31.9% Pennsylvania
2000 (4.4) participated in the MoVIES < high school
page 92 study, 37.5% male. Data are
Table A4.41 reported for the entire sample.
Results of ROC analysis are
reported.
TMT.40 Small et al., 72.90 413 Normal elderly volunteers, approximately 13.76-14.58 South Florida
2000 60-86 equal number of M and F.
page 92 Data are stratified by 2 age groups
Table A4.42 and 2 APOE genotype groups.
TMT. 41 Stuss et al., 53.4 19 Control group; 8 M, 11 F. Time to 13.7 Canada
2001 (13.6) completion, B-A difference and (2.5)
page 93 (8-A)/A proportion are reported.
TableA4.43
TMT.42 BeD et al., 34.4 29 Sample included friends, relatives, 13.0 Wisconsin
2001 (12.5) and spouses of TLE patients; (1.7)
page93 28% male.
Table A4.44
TMT.43 Stein et al., 29.4 22 Control group of women. Tune to 13.9 California
2002 (10.7) completion and B-A difference (1.5)
page 94 are reported.
Table A4.45
(continued)
628 APPENDIX 4

TMT.44 Drane et al., 18-20 18 286 healthy adults; 205 M, 80 F. 12.98 USA
2002 00-29 39 'rime to completion, 8-A difference, (2.65)
page 94 30-39 53 Jnd B:A ratio scores are reported.
Table A4.46 40-49 46
50-59 38
60-69 36
70-79 36
80-90 19
TMT.45Grady 66.3 517 D~ for women with established 12.7 Multicenter,
et al., 2002 (6.4) toronary
disease; 2 groups: (2.7) USA
page 95 i:RT treatment and placebo.
Table A4.47 67.3 546 12.7
(6.3) (2.7)
TMT.46 Miller, 38.0 949 Se"negative homosexual 16.3 MACS
2003 (an update (7.5) ~d bisexual males from (2.4) centers
on Seines et al., tfle MACS study, native
1991) page 95 25--34 inglish speakers. Data <16
Table A4.48 35-44 li-e partitioned by age x 16
45-59 ctiucation. >16
TMT.47 Tombaugh, 18-24 911 Heftthy Canadian volunteers 0-12 Canada
2004 25--34 418 M, 503 F. Data are presented 12+
page 96 35-44 ~ age x education groups.
Table A4.49 45-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
• Age column and IQ/Education column contain Information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, whichever information is provided by authors.
Table A4.2. [TMT.l] Davies, 1968: Mean Time in Seconds Corresponding to lOth, 25th, 50th, 75th, and
90th Percentile Ranks for Trails A and 8 per Age Group for a Sample of British Adults
Trails A Percentile Trails B Percentile
n Age 10 25 50 75 90 10 25 50 75 90
180 20s+30s 50 42 31 26 21 129 94 69 55 45

90
90
90
90
40s
50s
60s
70s
59
67
104
168
45
49
67
105
.
3l
38
so
28
29
35
54
22
25
29
38
151
177
282
450
100
135
172
292
78
98
119
196
57
75
89
132
49
55
64
79
APPENDIX 4 629
Table A4.3. [TMT.2] Goul and Brown, 1970: Data for Trails A, B, and Total Time (A+ B) for a Sample of
Non-Brain-Injured, Hospitalized Canadian Patients
Trails•
n Age FSIQ VIQ PIQ A B A+B
26 20-29 103.8 102.6 104.7 36.1 85.7 121.8

(12.1) (12.6) (13.0) (10.0) (38.7) (42.2)
Range 19-62 47-245 72-290
Median 35.0 76.0 111.5
25 30-39 110.1 106.4 113.7 35.5 79.6 114.0
(8.9) (11.3) (8.9) (9.4) (20,4) (25.3)
Range 19--58 32-115 51-173
Median 34.0 80.0 115.0
24 40-49 105.3 105.8 104.5 40.0 105.2 145.0
(7.9) (9.1) (9.2) (13.3) (42.2) (48.5)
Range 16--70 51-225 72-277
Median 38.0 99.5 142.0
16 50-59 112.7 111.2 111.6 45.3 103.2 148.4
(8.6) (11.5) (9.0) (13.6) (43.3) (53.0)
Range 22-68 5~190 80-253
Median 46.0 98.5 140.0
15 60-72 104.2 103.2 105.4 68.9 158.8 227.7
(12.2) (13.2) (11.6) (21.2) (49.5) (63.2)
Range 35-120 ~272 123--347
Median 68.0 147.0 219.0
106 Total 107.0 105.6 107.9 42.9 101.7 144.3
(10.6) (11.8) (11.2) (17.3) (46.2) (58.6)
•Mean time in seconds, standard deviations, ranges, and medians.
Table A4.4. [TMT.3] Wiens and Matarazzo, 1977: Data for Two Equal Groups of Male Applicants to a
Patrolman Program and Retest Data 14-24 Weeks Later for a Random Subsample
Trails•
Initial Test Retest
n Age Education FSIQ VIQ PIQ A B A B
24 23.6 13.7 117.5 117.4 115.4 23.83 56.42

(21-27) (12-16) (8.3) (8.4) (10.5) (6.61) (12.79)
(1~139)
24 24.8 14.0 118.3 116.4 118.2 20.54 51.04

(21-28) (12-16) (6.8) (6.9) (8.6) (4.43) (11.46)
(1~131)
29 24 14 118 116 118 21.76 54.17 21.72 51.28

(21-28) (12-16) (5.65) (12.54) (5.86) (12.29)
•Mean time in seconds and standard deviations.

630 APPENDIX 4
Table A4.5. [TMT.4] Eson et al., Personal Communi-

cation: Data for a Sample of Older Participants
Trails•
n Age A B
15 63.2 38.9 115.3

(15.6) (57.8)
16 67.0 42.1 134.9
(14.2) (57.7)
16 72.0 46.2 145.2
(24.0) (134.9)
16 78.3 80.1 234.8
(64.8) (117.0)
Table A4.6. [TMT.S] Harley et al., 1980: Data for the Whole Sample and for an Alcohol-Equated Sample of
Veterans Administration Inpatients with Various Psychiatric Diagnoses and Alcoholism, Stratified into Five
Age Groups, Patients with chronic brain syndrome included
WAIS TraiJs•
Age n FSIQ VIQ PIQ Education A 8
TotaltGmple
55-59 56 98.57 99.39 97.00 10.1 67.04 175.5
(11.43) (12.92) (10.65) (29.75) (99.98)
80-129 77-131 72-129 ~160 60-676
60-64 45 98.58 101.27 95.00 9.8 63.67 158.67
(9.93) (11.42) (9.82) (24.30) (55.11)
80-121 78-123 78-116 ~114 70-275
~9 35 97.51 100.37 93.66 8.7 87.89 219.43
(11.18) (12.51) (10.20) (75.60) (120.60)
80-130 80-135 68-120 27-470 ~8
70-74 37 100.41 102.95 97.24 8.8 95.24 237.16

(9.92) (11.81) (10.08) (6U7) (126.90)
82-125 80-133 75-114 35-353 83-606
75-79 20 101.75 101.40 102.15 6.5 85.80 225.15
(10.18) (11.40) (9.95) (43.64) (81.16)
81-119 77-117 83-119 28-180 100-410
Akohol-equmed _...,.
55-59 47 99.00 100.00 98.00 10.1 65.89 178.43
(11.73) (13.02) (11.13) (28.21) (107.29)
80-129 77-131 72-129 ~140 60-176
60-64 33 96.00 99.00 93.00 9.3 71.82 174.97
(9.43) (11.33) (9.30) (23.08) (52.68)
80-117 78-123 78-112 ~114 80-275
65-69 23 99.00 102.00 95.00 8.8 88.78 210.26
(12.06) (13.06) (11.52) (88.34) (135.24)
80-130 80-135 68-120 27-470 80-678
APPENDIX 4 631

WAIS Trails•
Age n FSIQ VIQ PIQ Education A 8
70-74 37 100.00 103.00 97.00 8.8 95.24 237.16

(9.92) (11.91) (10.08) (64.17) (126.90)
82-125 80-133 75-114 35-353 83-606
75-79 20 102.00 101.00 102.00 6.5 85.80 255.15
(10.18) (11.40) (9.95) (43.64) (81.16)
31-119 77-117 ~119 28-180 100-410
"Mean time in seconds, standard deviations, and ranges.
Table A4.7. [TMT.6] Anthony et al., 1980: Data for Trails B for a
Control Group
n Age Education FSIQ VIQ PIQ Trails 8"
100 38.88 13.33 113.54 113.24 112.26 68.58

(15.80) (2.56) (10.83) (11.59) (10.88) (32.72)
"Mean time in seconds and standard deviations.
Table A4.8. [TMT.7] Bak and Greene, 1980: Data for Healthy Right-Handed Older Adults
Trails" WAISt
Age n Education M/F ratio A 8 Info Arithmetic Block Design Digit Symbol
55.6 15 13.7 6/9 32.53 81.67 20.13 11.87 34.33 54.67

(4.44) (1.91) (12.58) (30.76) (3.38) (1.92) (6.79) (12.19)
50-62
74.9 15 14.9 5/10 41.60 109.00 21.07 13.60 28.07 39.47
(6.04) (2.99) (10.33) (38.84) (3.84) (2.97) (5.36) (12.11)
67--86

twechsler Adult Intelligence Scale data are presented in raw scores.
632 APPENDIX 4
Table A4.9. [TMT.8] Kennedy, 1981: ~a for Trails A, B, and Total Time (A+ B) for a
Sample of Healthy Canadian Employees q£ a Mental Health Center: Males and Females are
Equally Represented within each Age G$>uping
I
Trails*
n Age• Education IQ Estimatet A B A+B
30 20-29 13.73 123.43 25.03 59.58 84.62

25.77 (8.94) (28.78) (33.11)
30 30-39 13.53 ! 127.10 28.88 70.28 99.13
34.34 (9.70) (27.79) (34.57)
'
30 40-49 13.11 : 127.40 29.68 78.80 108.48
45.79 (7.67) (26.81) (30.32)
30 50-59 11.59 ; 123.30 37.73 96.01 133.74
53.74 (19.01) (39.25) (55.98)
30 60-69 12.50 '128.54 35.22 95.02 130.23
64.24 (12.36) (34.62) (40.67)
•Age range and mean age are provided for eaci. group.
trntelligence quotient estimate is based on pe+rmance on the Ammons Quick Test.
*Mean time in seconds and standard deviatioiiSI
lI
I
I
Table A4.10. [TMT.9] Fromm-Auch ani Yeudall, 1983: Data for a
Sample of Healthy Canadian Adults• '
n Age J\ B
32 15-17 23.. 47.7

(5.9) (10.4)
15.~9.0 25.4-81.0
76 1~23 26.!7 51.3
(14.6)
~
12. . .1 23.3-101.0
57 24.• 53.2
(7.f) (15.6)
11.~.0 29.1-98.0
18 27.; 62.1
(8.. ) (17.5)
16.0-$2.7 39.0-111.0
'
10 41-64 29.7 73.6
(8.-4) (19.4)
16.5--f.O 41.9-102.0
"Mean education 14.8 years, mean full-scale int~gence quotient 119.1.

tMean time in seconds, standard deviations, and ;ranges.
APPENDIX 4 633
Table A4.11. [TMT.10] Bomstein, 1985: Data for a Sample of Healthy Canadian Adults for Trails A and B"
Stratified by Three Age Groupings and Two Educational Levels (< High School, ~High School) for Males
and Females Separately
Male Female
<HS ?:HS <HS ?:HS
Age A B A B A B A B
20-39 28.3 70.0 23.8 53.9 23.2 56.4 22.0 53.5

(8.4) (28.7) (6.8) (18.3) (5.5) (21.3) (6.0) (20.5)
n=21 n=86 n=13 n=50
40-59 38.9 107.8 28.6 74.1 30.5 76.7 27.3 65.6
(12.5) (52.2) (9.6) (35.8) (9.2) (25.7) (9.2) (28.5)
n=13 n=17 n=22 n=43
60-69 37.6 119.4 35.3 78.3 40.7 96.4 34.5 87.4
(8.5) (42.3) (10.2) (26.1) (12.9) (27.3) (8.9) (27.1)
n=l6 n=23 n=22 n=34
Table A4.12. [TMT.ll] Heaton et al., 1986: Data for a Normal Control
Sample on Trails B and Percent Classified as Normal using Russell
et al:s {1970) Criteria, Stratified by Age and by Education
Trails Percent WAIS SS

n Age Education B• Normal Meant
319 <40 58.5 91.5 11.9

134 40-59 78.3 74.6 11.2
100 ?:60 116.8 33.0 9.7
132 <12 102.2 54.6 9.5
249 12-15 69.7 79.9 11.2
172 ?:16 57.9 89.5 12.9
"Mean time in seconds.

tMean scaled scores for Wechsler Adult Intelligence Scale subtests.
Table A4.13. [TMT.l2] Alekoumbides et al., 1987: Data for a Sample of Medical and
Psychiatric Veterans Administration Patients without a History of Neurological Disorder:
Mostly Males and Mostly Inpatients
Trails•
n Age Education FSIQ VIQ PIQ A B
118 46.85 11.43 105.89 107.03 103.31 48.60 120.49

(17.17) (3.20) (13.47) (14.38) (13.02) (23.79) (78.90)

634 APPENDIX 4
Table A4.14. [TMT.13] Bornstein et al., 1987a: Data for a Sample of Healthy Volunteers for Trails A and
B• over Two Testing Probes 3 Weeks Apart
Raw Score Median Raw Mean IIJ

Test Retest Changet Score Change of Change
n Age VIQ PIQ A B A B A B A B A B

23 32.3 105.8 105.0 25.6 52.1 21.5 47.4 3.1 6.9 3 6.5 13 9
(10.3) (10.8) (10.5) (6.8) (15.1) (5.6) (16.5) (4.9) (11.4)
17-41 34-97 15-35 25-73
•Mean time in seconds, standard deviations, and ranges.

t Change from test to retest.
Table A4.15. [TMT.14] Dodrill, 1987: Data for the Whole Sample of Healthy Participants and
for the Different Intelligence Levels
Trails•
n Age Education M/F Ratio FSIQ VIQ PIQ A B
120 27.73 12.28 60/60 100.00 100.92 98.25 25.37 66.02

(11.04) (2.18) (14.35) (14.73) (13.39) (9.17) (34.17)
Trails Trails
n FSIQ A B FSIQ A B
7 130 20 50 >89 <30 <76

18 125 21 47 80-89 30-39 76-103
34 120 20 48 70-79 40-49 104-180
64 115 22 49 <70 >49 >180
93 110 23 53
101 105 24 56
75 100 25 60
60 95 26 62
48 90 26 68
33 85 28 82
19 80 30 99
10 75 33 135
70 39 159

APPENDIX 4 635
Table A4.16. [TMT.l5] Ernst, 1987: Data for Neurologically

Healthy Older Australian Adults•
Trailst
n Gender A B
51 Male 40.5 98.2

(21.1) (52.9)
Errors 0.3 0.8
(0.6) (1.1)
59 Female 42.3 108.7
(14.4) (79.2)
Errors 0.3 0.7
(0.5) (0.9)
•Mean education is 10.3 years.

tMean time in seconds with standard deviations and number of errors
with standard deviations.
Table A4.17. [TMT.l6a] Stuss et al.,l987: Data for the Initial Test, Retest 1 Week Later, and Both Testing
Probes Combined for a Sample of Healthy Canadian Adults Partitioned into Six Age Groups
Trails•
Gender Handedness A B
n Age M F R L Education Combined Test 1 Test 2 Combined Test 1 Test 2
10 17.3 5 5 8 2 12.3 21.8 21.8 22.2 45.5 49.0 41.8

(0.95) (0.95) (7.0) (5.3) (8.7) (17.2) (21.2) (13.1)
16-19 11-13
10 23.0 6 4 6 4 16.2 17.3 18.5 16.2 37.8 41.6 34.0
(2.67) (1.39) (5.1) (5.1) (5.0) (12.1) (11.4) (12.7)
20-29 14-18
10 33.9 5 5 7 3 16.7 20.1 21.9 18.4 46.5 46.3 46.8
(2.88) (3.86) (4.6) (6.3) (2.9) (12.6) (13.7) (11.4)
30-39 10-20
10 44.2 6 4 9 1 15.5 27.9 29.2 26.5 64.3 64.1 64.4
(3.12) (2.88) (8.2) (9.0) (7.5) (19.7) (16.3) (23.0)
40-49 10-20
10 55.3 6 4 9 1 11.7 35.6 38.5 32.7 77.3 83.1 71.4
(2.98) (2.41) (20.9) (18.2) (23.8) (42.8) (44.3) (41.3)
50-59 8-16
10 63.7 5 5 10 0 14.3 33.6 37.3 29.9 70.3 73.3 67.3
(3.13) (2.00) (11.8) (14.7) (8.9) (21.7) (20.3) (23.2)
60-69 12-17

636 APPENDIX 4
Table A4.18. [TMT.16b] Stuss et al., 1987: Data for the

Whole Sample Stratified by Gender and iEducation
Trails•
n A 8
Males 33 25.80, 55.40

(14.85. (30.97)
Females 27 26.32. 55.10
(10.08) (21.27)
::;High school 27 28.56: 63.58
(15.06) (32.33)
>High school 33 23.98 51.63
(10.37~ (20.44)
•Mean time in seconds and standard deviatiolll.
Table A4.19. [TMT.17] Yeudall et al., Ujs7: Data for Healthy Canadian Adults Stratified by Age for the
Entire Sample and for Males and Femal~ Separately
Trails
Age %Right- WAIS-R
n Group AW! Education Handed FSIQ A 8
Entire....,. (n=JJS)
62 15-20 17.16 12.16 79.03 111.75 24.75 49.17
(I. lis) (1.75) (10.16) (8.19) (15.21)
73 21-25 22.'lb 14.82 86.30 109.79 24.53 50.36
(1.-1)) (1.88) (9.97) (7.93) (12.96)
48 26-30 28.06 15.50 89.58 113.95 24.49 51.94
(I.$) (2.65) (10.61) (7.22) (15.75)
42 31-40 34.~ 16.50 90.48 116.09 25.74 59.35
(2 ...,) (3.11) (9.51) (7.53) (17.1J)
225 15-40 24.~ 14.55 85.78 112.25 24.81 52.05
(6.~) (2.78) (10.25) (7.75) (15.36)
Femala(n=98)
30 15-20 17.7r3 12.10 73.33 110.32 25.74 47.69
(UN) (1.52) (10.64) (9.10) (15.11)
36 21-25 22.83 14.53 83.33 107.28 25.71 51.76
(l.S.) (1.99) (9.14) (9.16) (12.39)
16 26-30 28.69 14.94 93.75 113.10 25.71 51.47
(1.~) (2.32) (11.37) (7.15) (11.52)
16 31-40 33.88 16.19 87.50 114.27 25.49 57.29
(2.~) (2.29) (11.32) (6.00) (12.38)
98 15-40 24.0J 14.12 82.65 119.19 25.69 51.37
(5.~) (2.43) (10.46) (8.28) (13.34)
APPENDIX 4 637

Trails
Age %Right- WAIS-R
n Group Age Education Handed FSIQ A B
Males (n =121)
32 15-20 17.78 12.22 84.38 113.00 23.83 50.56
(2.09) (1.96) (9.72) (7.26) (15.41)
37 21-25 22.57 15.11 89.19 112.30 23.34 48.99
(1.96) (1.74) (10.27) (6.39) (13.52)
32 26-30 27.75 15.78 87.50 114.38 23.88 52.18
(1.57) (2.79) (lo.43) (7.28) (17.65)
26 31-40 34.69 16.69 92.31 117.31 25.89 60.63
(2.41) (3.55) (8.21) (8.45) (19.60)
127 15-40 25.15 14.87 88.19 113.87 24.12 52.57
(6.29) (2.99) (9.83) (7.27) (16.79)
Table A4.20. [TMT.l8] Bornstein and Suga, 1988: Data for a

Sample of Healthy Older Canadian Volunteers Stratified by Educa-
tion
Gender Trails*
n Education• Aget M F A B
46 5-10 62.3 17 29 38.9 102.0

8.5 (11.5) {39.5)
44 11-12 62.9 16 28 33.6 82.5
11.7 (10.3) (34.5)
44 >12 63.0 16 28 34.0 80.9
15.0 (10.7) {30.9)
"Range and mean number of years for education.

range for the sample is 55-70 years.
t Age
*Mean time in seconds and standard deviations.
638 APPENDIX 4
Table A4.21. [TMT.l9] Stuss et al., 1988: Data for a Sample of Healthy Canadian Adults Stratified into
Three Age Groups, for Two Testing Sessions One Week Apart
Trails•
Gender Handedness A B
n Age M F R L Education Test Retest Test Retest
30 22.43 16 14 22 8 14.1 21.48 19.68 48.77 42.18

(2.67) (1.34) (6.44) (7.32) (18.66) (15.54)
1~29 11-18
30 40.63 14 16 26 4 14.9 27.58 22.95 61.30 61.52
(2.97) (3.95) (9.43) (6.23) (17.88) (22.79)
30-49 5-20
30 61.77 14 16 28 2 13.2 36.73 29.30 76.97 67.10
(3.0) (2.38) (13.68) (14.73) (30.52) (28.37)
~9 ~18
Table A4.22. [TMT.20] Van Gorp et al., 1990: Data for Four Age
Groups and for the Whole Sample of Healthy Older Adults•
Trailst
n Age VIQ PIQ A B
28 57--ffi 117.20 109.20 41.50 84.40

(11.33) (11.56) (7.38) (24.60)
45 ~70 114.80 111.47 43.20 105.20
(17.03) (16.83) (14.98) (43.43)
57 71-75 122.88 115.08 50.08 97.79
(11.38) (11.94) (12.88) (30.40)
26 7~ 110.55 101.00 59.73 153.09
(11.25) (8.78) (15.95) (62.60)
156 57-&5 117.65 110.62 48.70 107.55
(13.53) (13.49) (14.47) (45.63)
•Mean education for the sample is 14.14 (2.86) years, 61% females.
tMean time in seconds and standard deviations.
APPENDIX 4 639
Table A4.23. [TMT.22] Seines et al., 1991: Data for a Sample of Seronegative Homosexual/Bisexual Males
Participating in the Multi-Center AIDS Cohort Study, Stratified by Age and Education
Trails•
A B
Percentiles Percentiles
Age n Mean Age Education Mean (SD) 5th lOth Mean (SD) 5th lOth
By age
25--34 309 31.0 16.1 19.0 27 24 49.5 80 74
(2.6) (2.2) (5.9) (17.1)
35-44 290 39.3 16.4 20.8 32 29 52.5 83 78
(2.9) (2.3) (5.5) (18.6)
45-54 97 48.5 16.7 23.1 37 35 53.9 87 79
(2.6) (2.6) (7.3) (20.3)
By education
<College 229 36.1 13.7 22.8 31 30 51.8 87 79
(7.4) (1.2) (7.1) (20.7)
College 202 35.6 16.0 19.2 32 25 51.4 83 75
(7.2) (0.0) (5.8) (17.1)
>College 302 38.4 18.6 20.1 30 28 50.2 79 73
(7.8) (1.3) (5.5) (15.8)
Table A4.24. [TMT.23] Elias et al., 1993: Data for a Sample of Healthy Adults Stratified by Age and
Gender
Men Women
Trails• Trails•
Age Group n Education A B n Education A B
15-24 37 14.81 21.46 52.73 24 14.50 20.17 43.00

(1.54) (1.25) (4.60) (1.53) (.85) (2.34)
25--34 40 14.85 29.92 58.85 56 14.62 23.34 52.75
(1.70) (2.47) (3.50) (1.68) (1.07) (2.72)
35-44 36 14.56 25.89 62.06 56 14.64 29.57 56.96
(1.70) (1.46) (4.21) (2.08) (1.93) (2.36)
45-54 25 14.72 26.24 58.04 46 14.89 27.72 56.50
(2.26) (1.65) (3.21) (2.38) (1.30) (1.87)
55-64 25 14.72 37.48 75.36 35 14.40 34.31 78.54
(1.57) (2.10) (3.70) (1.90) (1.64) (4.46)
~65 24 14.54 39.58 86.46 23 14.65 40.70 86.57
(2.15) (3.78) (10.22) (2.52) (3.24) (6.71)

640 APPENDIX 4
Table A4.27. [TMT.26a] Richardson and Marottoli,

Table A4.25. [TMT.24] Cahn et al., 1995: Data for
1996: Demographic Characteristics of the Healthy
a Control Sample of Cognitively Intact Elderly
Elderly Sample
Trails•
Total
n Age Education M/F ratio A B Sample Younger-Old Older-Old
238 78.4 13.8 97/141 47.9 123.5 n 101 50 51

(6.8) (2.6) (1.4) (3.4) Age 81.47 78.80 84.08
(3.30) (1.07) (2.56)
Education 11.02 10.44 11.59
(3.68) (3.86) (3.45)
Mini-Mental 26.97 26.56 27.37
State Exam (2.55) (3.03) (1.92)
Table A4.26. [TMT.25] Ivnik et al.,
1996: Demographic Description of %Female 47.5% 46.0% 49.0%
the Healthy Sample Partitioned into %White 90.1% 82.0% 98.0%
Groups Used in TMT Testing %Black 9.9% 18.0% 2.0%
Characteristic n
Age group
56-59 54 Table A4.28. [TMT.26b] Richardson and Marot-
60--64 81 toli, 1996: Data for Trails B• Stratified into Two
65-69 65
Age Groups By Two Education Groups (in Years of
70--74 57
Schooling)
75-79 53
80--84 27 Age
85-89 17
90--94 5 7&-80 81-91
95+ 0
Education
Education
~7 2 <12 ~12 <12 ~12
8-11 33
n 26 24 18 33
12 135
13--15 87 Trails B 197.17 119.17 195.47 137.30
1&-17 67 (71.03) (33.47) (69.70) (55.93)
~18 35
Gender
Male 167
Female 192
Race
Caucasian 358 Table A4.29. [TMT.27] Hoff et al., 1996: Data for
Black 1 the Control Sample
Handednea
Right 332
Left 17 n Age Education A B
Mixed 10
54 32.1 15.4 22.7 55.7
Total 359 (9.7) (2.4) (8.7) (19.3)

APPENDIX 4 641
Table A4.30. [TMT.28] Salthouse et al., 1997:

Data for a Sample of Healthy Participants Stratified
into Three Age Groups
Trails•
Age Mean
Group Age n Education %Male A B
18-39 29.0 40 15.5 42.5 21.0 53.6

(4.8) (1.7) (4.6) (20.3)
40-59 49.1 38 15.2 50.0 26.2 66.7
(5.1) (2.5) (7.9) (21.3)
69.2 37 15.3 48.6 32.9 87.2
60-78 (5.1) (2.6) (10.5) (36.6)
Table A4.31. [TMT.29] Rasmusson et al., 1998: Data for a Sample of Nondemented Highly
Educated Elderly Participants• Partitioned into Four Age Groups
Trailst
Age Group Mean Age n Education A B %A Errors* % B Errors*
60-69 64.8 203 15.7 32.1 81.2 9.8 34.5

(2.8) (11.0) (35.2)
8-20
70-79 74.2 262 16.0 40.7 103.3 13.4 39.0
(3.0) (15.5) (51.1)
8-20
80-89 83.4 179 16.4 48.6 132.1 8.8 48.2
(2.9) (17.5) (55.9)
7-20
90-96 91.5 23 16.6 52.1 153.0 5.3 57.9
(2.9) (19.6) (68.2)
12-20
"Majority of the sample are males.

*Percent of participants who made errors on parts A and B.
Table A4.32. [TMT.30] Miner and Ferraro, 1998: Data for Two
Administration Conditions (A-B, B-A) for a Sample of Undergraduate
Students
Trails•
M/F
TMTOrder n Age Ratio A B
Total 110 21.7 22/88
sample (5.24)
A-B 55 22.93 45.68
(6.25) (10.74)
B-A 55 21.04 49.17
(6.99) (13.04)
•Mean time in seconds and standard devilitions.

642 APPENDIX 4
Table A4.33. [TMT.31] Crowe, 1998: Data for a Sample of Under-

graduate Students
Trails•
n Age Education M/F Ratio A B
98 23.4 14.0 49/49 24.7 50.3

(3.1) (2.3) (5.9) (11.8)
Table A4.34. [TMT.32] Tremont et al., 1998: Data for Patients

Referred for Evaluation which Yielded Negative Findings, Stratified
by Three Levels of Intelligence
WAIS-R FSIQ
Below Above
Average Average Average
n 35 84 38
Age 34.03 40.55 41.71
(13.8) (16.73) (14.65)
Education 11.53 12.62 15.63
(2.76) (2.76) (3.37)
FSIQ 84.89 99.15 119.92
(4.84) (8.05) (7.55)
VIQ 85.74 98.32 118.79
(7.29) (9.13) (8.86)
PIQ 86.06 101.12 117.42
(7.92) (9.22) (10.24)
Trails A" 40.43 34.84 28.76
(18.07) (14.41) (10.61)
Trails B" 124.51 93.20 68.21
(76.49) (49.11) (32.72)
Table A4.35. [TMT.33] Basso et al., 1999: Data for a Sample of Healthy Men on Two Testing Probes over a
12-Month Interval
Trails"
Test Retest
n Age Education WAIS-R FSIQ A B A B
50 32.50 14.98 109.30 21.52 48.70 21.32 47.72

(9.27) (1.93) (12.29) (7.54) (17.76) (7.36) (19.33)

'
APPENDIX 4 643
Table A4.36. [TMT.34] Crews et al., 1999: Data for a Control

Sample of Women
Trailst
WAIS-R
n Age Education Vocabulaty" A B
30 20.20 14.40 13.50 21.13 47.43

(3.47) (1.33) (2.08) (5.78) (13.33)
•Wechsler Adult Intelligence Scale-Revised Vocabulary scaled score.

Table A4.37. [TMT.35] Dikmen et al., 1999: Test-Retest Data for Normal and Neurologically Stable
Adults•
Trails A1 Trails B
WAIS Test-retest
n Age Education M/F Ratio FSIQt interval Time 1 Tlme2 Time 1 Time2
384 34.2 12.1 66/34 108.8 9.1 26.52 25.56 72.05 68.19
(16.7) (2.6) (12.3) (3.0) (11.66) (11.66) (45.22) (46.13)
•A number of participants had preexisting conditions that might affect test performance, the most significant being alcohol
abuse and a significant traumatic brain injwy.
twechsler Adult Intelligence Scale full-scale intelligence quotient (Wechsler, 1955).
1Mean time in seconds and standard deviations.
Table A4.38. [TMT.36] Binder et al., 1999: Data for a Normal Elderly Sample
Trails
A B
% % cos· Blessedt
n Age Education Male Caucasian score score Time1 Lines' Time Lines
125 82.3 13.5 25 87 1.8 2.1 53.5 24.0 124.1 22.6

(4.4) (3.0) (1.8) (2.1) (25.3) (0.3) (39.9) (3.6)
"Geriatric Depression Scale.

tshort Blessed Orientation-Memory-Concentration Test.
'Mean time in seconds and standard deviations.
'Number of lines correctly drawn within the time limit of 180 seconds for both parts A and B.
644 APPENDIX 4
Table A4.39. [TMT.37] Ruffolo et al., 2000: Data for the Control
Sample
Trails
A 8
n Age Education Errors• Timet Errors T'une
49 29.1 14.3 0.14 26.6 0.47 57.2

(12.1) (1.9) (0.41) (7.9) (0.77) (17.2)
•Mean number of performance errors and sta$lard deviations.

tMean time in seconds and standard deviationS.
Table A4.40. [TMT.38] Saxton et al., ~: Data for a Sample of

Elderly Free of Cardiovascular Disease ·
Trails•
n Age Education %~e A 8
357 73.63 13.23 44~ 43.61 114.53

(4.45) (2.85) (1.00) (3.10)
Table A4.41. [TMT.39] Chen et al., 2000: Pata for the Control Sample of
Nondemented Elderly•
n Age %Male A 8
483 74.9 37.5 48.02 130.49

(4.4) (17.09) (62.21)
•Participants with lower than high school educatton, 31.9%.

tMean time in seconds and standard deviations. '
APPENDIX 4 645
Table A4.42. [TMT.40] Small et al., 2000: Data for a Sample of Normal Elderly Stratified by
Two Age Groups and the Presence of the APOE-£4 Allele
Trails•
Group n Age %Male Education A B
APOE-84-nega&e
Young-old 156 67.56 50.6 14.58 38.34 89.47
(3.63) (2.84) (13.24) (34.27)
Old-old 166 78.27 51.8 13.92 46.33 128.34
(3.01) (2.78) (18.38) (66.18)
APOE-84-poaitive
Young-old 46 66.91 37.0 13.78 38.04 109.58
(3.53) (2.53) (13.71) (51.25)
Old-old 45 77.76 57.8 13.76 43.91 132.55
(2.96) (2.60) (15.69) (77.56)
Table A4.43. [TMT.41] Stuss et al., 2001: Data for the Control Sample
Trails"
M/F NART Difference Proportion
n Age Education Ratio Score A 8 8-A (8-A)/A
19 53.4 13.7 8/11 113.2 30.8 64.2 33.4 1.3

(13.6) (2.5) (5.6) (17.0) (26.2) (17.6) (.8)

NART, National Adult Reading Test.
Table A4.44. [TMT.42] Bell et al., 2001: Data for Table A4.45. [TMT.43] Stein et al., 2001: Data for
the Control Sample the Control Sample of Women
Trailst Trails•
Difference
n Age Education 'If> Male FSIQ" A 8 n Age Education A 8 8 -A
29 34.4 13.0 28 97.7 24.3 57.9 22 29.4 13.9 23.9 55.0 31.1
(12.5) (1.7) (6.4) (6.3) (18.7) (10.7) (1.5) (8.6) (18.3) (14.9)
•Wechsler Adult Intelligence Scale-ill full-scale intelli- "Mean time in seconds and standard deviations.
gence quotient based on seven-subtest short form.
646 APPENDIX 4
Table A4.46. [TMT.44] Drane et al., 2002: Data Table A4.48. [TMT.46] Miller, 2003 (Update
for a Sample• of Healthy Adults Partitioned into on Seines et al., 1991): Data for a Sample of Sero-
Eight Age Groups negative Homosexual/Bisexual Males Participating
in the MACS Study, Stratified by Age x Education
Trailst
Age Difference Ratio Trails•
Group n A B B-A B:A
Age Education A B
18-20 18 23.22 52.94 29.72 2.31
(6.56) (20.10) (16.21) (0.58) 25-34 <16 Mean 28.41 58.76
20-29 39 26.12 60.92 35.31 2.36 (SD) (12.86) (18.39)
(9.78) (33.17) (27.72) (0.78) n 46 123
16 Mean 23.50 50.80
30-39 53 28.02 72.30 44.13 2.72
(SD) (7.37) (18.70)
(8.78) (28.55) (26.72) (1.21)
n 36 96
40-49 46 31.00 81.26 50.04 2.80 >16 Mean 23.03 47.08
(11.21) (23.69) (20.28) (0.93) (SD) (7.38) (17.01)
50-59 38 36.29 103.42 67.24 2.94 n 35 110
(16.41) (50.26) (39.35) (0.88) Total Mean 25.29 52.53
(SD) (10.14) (18.67)
60-69 36 39.60 105.23 65.60 2.70
(12.14) (41.15) (33.84) (0.77) n 117 329
70-79 45.58 152.59 109.14 35-44 <16 Mean 28.52 65.52

36 3.49
(88.42) (1.76) (SD) (7.86) (26.87)
(18.91) (73.87)
n 63 124
80-90 19 56.37 170.21 113.84 3.05 16 Mean 27.80 59.64
(20.20) (84.68) (70.73) (1.05) (SD) (8.65) (22.31)
n 59 121
•Mean education for the sample= 12.98 (2.65) years. >16 Mean 24.60 53.53
tMean time in seconds and standard deviations. (SD) (6.68) (16.65)
n 80 177
Total Mean 26.76 58.80
(SD) (7.83) (22.22)
n 202 422
Table A4.47. [TMT.45] Grady et al., 2002: Data 45-59 <16 Mean 31.25 70.65
for Trails B for a Sample• of Women with (SD) (10.51) (23.60)
Established Coronary Disease, Stratified into n 40 60
16 Mean 33.11 64.65
Estrogen/Progestin Replacement Treatment and
(SD) (15.04) (23.53)
Placebo Groups n 27 40
% Trails >16 Mean 29.84 60.05
st (SD) (9.15) (17.90)
Group n Age Education White
n 62 96
Treatment 517 66.3 12.7 90.9 156.2 Total Mean 30.96 64.23
(6.4) (2.7) (77.5) (SD) (11.00) (21.37)
Placebo 546 67.3 12.7 90.5 151.5
n 129 196
(6.3) (2.7) (77.5) Total <16 Mean 29.22 63.82
(SD) (10.33) (23.52)
• All participants were younger than 80 years of age. n 149 307
tMean time in seconds and standard deviations. 16 Mean 27.70 57.12
(SD) (10.58) (21.77)
n 122 257
>16 Mean 26.12 53.31
(SD) . (8.21) (17.68)
n 177 383
Total Mean 27.58 57.75
(SD) (9.69) (21.29)
n 448 947

>
"'C
"'C
m
z
a
X
•
Table A4.49. [TMT.47] Tombaugh, 2004: Data for a Sample of Healthy Canadian Adults Stratified by Age and Education•
Age Group
18-24 25-34 35-44 45-54 55-59 60-64 61)..$ 70-74 75-79 80-84 85-89
Education Group
0-12 12+ 0-12 12+ 0-12 12+ 0-12 12+ 0-12 12+ 0-12 12+ 0-12 12+
n 155 33 39 41 58 37 55 31 65 32 76 30 74 34 84 34 16 13
Age 20.17 29.42 39.74 48.54 56.90 57.05 62.33 61.94 67.04 67.22 71.99 72.07 77.32 77.21 81.94 81.56 86.38 86.31
(1.48) (2.87) (2.94) (2.96) (1.31) (1.45) (1.28) (1.50) (1.63) (1.43) (1.4) (1.60) (1.35) (1.49) (1.41) (1.52) (1.50) (1.65)
Education 12.92 14.18 13.59 13.68 11.05 15.32 10.84 15.45 10.87 15.91 10.50 15.43 10.80 15.29 10.48 15.50 9.88 16.23
(1.01) (1.61) (2.06) (2.80) (1.05) (1.93) (1.27) (1.31) (1.71) (1.87) (1.72) (2.21) (1.50) (1.80) (1.54) (2.54) (1.96) (2.45)
Trails At 22.93 24.40 28.54 31.78 35.10 31.72 33.22 31.32 39.14 33.84 42.47 40.13 50.81 41.74 58.19 55.32 57.56 63.46
(6.87) (8.71) (10.09) (9.93) (10.94) (10.14) (9.10) (6.96) (11.84) (6.69) (15.15) (14.48) (17.44) (15.32) (23.31) (21.28) (21.54) (29.22)
Trails st 48.97 50.68 58.46 63.76 78.84 68.74 74.55 64.58 91.32 67.12 109.95 86.27 130.61 100.68 152.74 132.15 167.69 140.54
(12.69) (12.36) (16.41) (14.42) (19.09) (21.02) (19.55) (18.59) (28.89) (9.31) (35.15) (24.07) (45.74) (44.16) (65.68) (42.95) (78.50) (75.38)
•Male/female ratio for the sample is 408/503.

~
Appendix 4m: Meta-Analysis
Tables for the Trailmaking Test (TMT)
Table A4m.1. Results of the Meta-Analysis and Predicted Scores for the
TMT, Trails A
(Relevant values are weighted on the standard error for the test mean)
Description of the aggregate sample
Number of studies included in the analysis 28

Years of publication 1980-2004
Number of data points used in the analysis 89
(a data point denotes a study or a cell
in education/gender-stratified data)
Total number of participants 6,317
Variable n• xt sot Range

sample me
Mean 89 49.33 59.29 10-483
Age
Mean 89 53.94 21.92 1&-91.5
SD 89 3.52 3.02 0.5-16.7
Education
Mean 69 13.87 1.49 8.5-16.7
SD 64 2.38 0.65 0.3-3.9
IQ
Mean 21 116.69 7.48 97.7-128.5
SD 16 10.64 3.25 5.6-17.0
Percent male 70 46.16 29.98 0-100
Tellt score meGne
Combined mean 89 35.79 11.53 19.0-60.5
Combined SD 89 12.21 5.56 0.9-25.3
•Number of data points differs for different analyses due to missing data.
tweighted means and standard deviations.
648
APPENDIX 4M 649
Table A4m.1. (Contd.)
Predicted number of seconds to completion and SDs per age group• (TMT-A)
95%CI 95%CI
Age Predicted Lower Upper Predicted Lower Upper

&nge Score Band Band SD Band Band
16-19 23.97 22.33 25.62 7.63 6.59 8.67

20-24 24.05 23.07 25.02 7.63 7.09 8.18
25-29 24.46 23.74 25.18 7.78 7.15 8.39
30-34 25.23 24.24 26.21 8.05 7.03 9.07
35-39 25.34 25.04 27.65 8.48 7.10 9.86
40-44 27.81 26.28 29.33 9.04 7.40 10.68
45-49 29.62 27.99 31.24 9.75 7.95 11.54
50-54 31.78 30.17 33.40 10.59 8.74 12.44
55-59 34.30 32.74 35.86 11.58 9.76 13.40
~ 37.16 35.59 38.74 12.71 10.98 14.45
65-69 40.38 38.56 42.19 13.98 12.34 15.63
70-74 43.94 41.57 46.31 15.40 13.74 17.05
75-79 47.85 44.63 51.07 16.95 15.08 18.82
80-84 52.11 47.78 56.44 18.65 16.30 20.99
85-89 56.73 51.07 62.38 20.49 17.43 23.54
•Based on the equations:
Predicted teat score =26.50094 - 0.2665049 • age+ 0.0069935 • age2

Predicted SD = 8.760348- 0.1138093 • age+ 0.0028324 • age2
Correction for education
Years of Correction
Education Factor
12 +2.62
13 + 1.31
14 0
15 -1.31
16 -2.62
17 -3.93
With every year of education above or below 14, we

suggest correcting the obtained score by adding or
subtracting 1.31 to or from the predicted score given in
the table for the relevant age group. SD for the person's
actual age group should be used with the education-
corrected scores. Extrapolation of this correction outside
the boundaries of 12-17 years of education should be
made with caution as empirical data are not available
beyond these educational ranges.
(continued)
650 APPENDIX 4M
Significance tests for regression with .the test scores
Ordinary least-squares regression or tail means on age

(quadratic)
Number of observations 89
Number of clusters 28
R2 0.905
F<dO· p F<2.27> =146.47, p < 0.000.
Term Coefficient SE p 95%CI
Age -0.2665049 0.166 -1.60" 0.121" - 0.608 to 0.075

Age2 0.0069935 0.002 3.85 0.001 0.003 to 0.011
Constant 26.50094 3.179 8.34 0.000 19.98 to 33.02
"Significance test for age centered (sample means - aggregate mean): t = 12.80, p = 0.000.
Prediction
Predicted age range 16-90 yeara
Mean predicted score 35.05 (10.~)
SEe 1.05
95%CI 32.99-37.11
eo
50
40
30
i ,i
20 t1
age
Figure A4m.1. A scatterplot illustrating the dispersion of the data points around the regression line for
TMT-A. Tbe size of the bubbles reHects the weight of the data point, with larger bubbles indicating larger
standard error and smaller weight.
Tests for assumptioos and model 8t
Tests for heterogeneity in the 8aal dataset

Pooled estimates for fixed effect 37.898
Pooled estimates for random effect 32.166
QcciO•P Q(88) =721.02, p < 0.000
Moment-based estimate of
between-study variance 130.721
Tests for model &t-wlition of a quadratic term
Model BIC BIC'
Linear 0.838 0.836 134.376 -157.335

Quadratic 0.905 0.903 91.403 -200.307
BIC' difference of 42.972 provides very strong support for the quadratic model.
Tests for parameter speeJ8ealioas
Normality of the residuals
Shapiro-Wille W test W = 0.990, p = 0.706
Homoscedasticity
White's general test 22.246, p < 0.000
Significance tests for regression with SDs
Ordinary least-squares regression of SD1 on age (quadratic)

R2 0.602
F<dO· p F<2.27l = 61.58, p < 0.000
Age -0.1138093 0.129 -0.89• 0.384• - 0.378 to 0.150

Age2 0.0028324 0.001 2.18 0.038 0.000 to 0.006
Constant 8.760348 2.385 3.67 0.001 3.87 to 13.65
•Significance test for age centered (sample means- aggregate mean): t = 9.75, p < 0.000.
Prediction
Mean predicted SD 11.91 (4.28)
SEe 0.82
95%CI 10.31-13.51
Effects of demographic variables
Edueation
Est. tau2 without education 83.81
Est. tau2 with education 79.48
Regression of test means on education and age
R2 0.917
(continued)
652 APPENDIX 4M
Education -1.308 0.435 -3.00 0.006 -2.21 to -0.41
IQ
Regression of test means on IQ and age
R2 0.946
IQ 0.0678098 0.045 1.51 0.181 -0.042 to 0.177
Gender
t-test by gender
n X male X female M-F difference p
17M, 15F 26.449 27.153 -0.704 -0.343 0.367
TMT, Trails B

(a data point denotes a study or a cell in
education/gender-stratified data)
Sampk size
Mean 89 50.69 65.44 10-483
Age
Mean 89 56.69 22.03 16-91.5
so 89 3.56 3.06 0.5-16.7
Education
Mean 69 13.82 1.56 8.5-16.7
so 64 2.38 0.65 0.~.9
IQ
Mean 21 116.88 7.80 97.7-128.5
so 16 11.25 3.26 5.6-17.0
APPENDIX 4M 653
Variable n• x·t sot Range
Percentmak 70 48.08 29.47 0-100
Tat acore means

Combined mean 89 95.08 38.93 43.0-170.2
Combined SD 89 38.97 23.27 1.9-88.4
"Number of data points differs for different analyses due to missing data.
Predicted number of seconds to completion and SDs per age group• (TMT-B)
95%CI 95%CI
Age Predicted Predicted
Btmge Score Lower Band Upper Band SD Lower Band Upper Band
1~19 53.92 49.21 58.63 20.12 14.76 25.48

.20-24 53.77 50.54 56.99 19.19 14.93 23.45
.25-.29 54.72 51.69 57.74 18.87 15.09 22.65
30-34 56.84 52.89 60.80 19•.29 15.30 23.29
35-39 60.15 55.05 65.25 20.46 15.88 25.03
40-44 64.63 58.50 70.77 .2.2.37 17.11 27.62
45--49 70.29 63.29 77.30 25.02 19.10 30.94
50-IU 77.13 69.40 84.86 28.42 21.86 34.97
SS-S9 85.15 76.76 93.54 32.55 25.37 39.74
60-64 94.34 85.24 103.45 37.44 29.55 45.33
65-69 104.71 94.71 114.72 43.07 34.33 51.80
10-14 116.26 105.03 127.50 49.44 39.63 59.24
15-19 128.99 116.11 141.88 56.55 45.38 67.72
80-84 142.90 127.86 157.94 64.41 51.54 77.28
85-89 157.98 140.26 175.71 73.01 58.07 87.95
"Based on the equations:

Predicted te1t •core =64.07469- 0.9881013•age+0.023558l•age2
Predicted SD = 29.8444-0.8080508 • age+ 0.0148732 • age2
Years of Correction
Education Factor
12 +12.90
13 +6.45
14 0
15 -6.45
16 -12.90
17 -19.35
With every year ofeducation above or below 14, we suggest correcting the obtained score by adding or
subtracting 6.45 to or from the predicted score given in the table for the relevant age group. Standard
deviation for the person's actual age group should be used with the education-corrected scores.
Extrapolation of this correction outside the boundaries of 12-17 years of education should be made
with caution as empirical data are not available beyond these educational ranges.
(continued)
654 APPENDIX 4M
Significance tests for regression with the test scores
Orclioary least squares regression o£ test means on age (quadratic)

R2 0.876
F<dO• p F(2.28l ~74.01, p < 0.000
Tenn Coefficient SE p 95%CI
Age -0.9881013 0.428 -2.31• 0.029• -1.865 to - 0.111

Age2 0.0235581 0.004 5.35 0.000 0.014 to 0.032
II
Constant 64.07469 8.368 7.66 0.000 46.93 to 81.22
•significance test for age centered (sample means - aggregate mean): t = 11.92, p = 0.000.
Predietion
Predicted age range 16-90 yeal's
Mean predicted score 88.12 (34.13)
SEe 4.26 •
95%CI 79.77-96.f7
200
150
100
50
20 30 40 50 70 80
age
Figure A4m.2. A scatterplot illustrating the ~rsion of the data points around the regression line for
TMT-8. The size of the bubbles reftects the weight of the data point, with larger bubbles indicating larger
Tests for assumptioas and model &t
Tests for heterogeneity in the 8aal datf4 set

Pooled estimates for 6xed effect 87.131
Q(dO• p Q(88) = 114431.30, p < 0.000
--~--
APPENDIX 4M 655
Tests for model &t--'lition of a quadratic term
Model BIC BIC'
Linear 0.810 0.808 365.177 -143.195

Quadratic 0.876 0.873 331.763 -176.610
BIC' difference of 33.414 provides veey strong support for the quadratic model.
Tests for parameter speei&eatioas
Sbapiro-Wilk W test W =0.979, p =0.166
Homoscedasticity
White's general test 23.806, p <0.000
Signiftcance tests for regression with the SDs
Ordinary least squares regression of SDs on age (quaclratie)

R2 0.676
F<dO• p F<2.28) = 28.44, p < 0.000
Term Coefficient SE t p 95%CI
Age -0.8080508 0.332 -2.43" 0.022" -1.488 to - 0.128

Age2 0.0148732 0.003 4.43 0.000 0.008 to 0.022
Constant 29.8444 7.159 4.17 0.000 15.18 to 44.51
"Significance test for age centered (sample means -aggregate mean): t = 7.53, p < 0.000.
Predietion
SEe 3.82
95%CI 27.86--42.83
Education
# of observations 68
R2 0.907
Education -6.446 2.515 -2.56 0.017 -11.63 to-1.27
IQ
# of observations 21
~ Q~
(continued)
656 APPENDIX 4M
IQ 0.4133416 0.286 1.44 0.192 -0.264 to 1.091
Gender
t-test by gender:

17M,15F 59.809 59.431 0.379 0.084 0.467
Table A4m.3. Summary Table of Predicted Values for the TMT

Part A PartB
Age
.Range Time SD Time SD
16-19 23.97 7.63 53.92 20.12

20-24 24.05 7.63 53.77 19.19
25-!9 24.46 7.78 54.72 18.87
30-34 25.23 8.05 56.84 19.29
3S-;J9 25.34 8.48 60.15 20.46
40-44 27.81 9.04 64.63 22.37
45-49 29.62 9.75 70.29 25.02
50-54 31.78 10.59 77.13 28.42
55-59 34.30 11.58 85.15 32.55
60-64 37.16 12.71 94.34 37.44
65-69 40.38 13.98 104.71 43.07
10-14 43.94 15.40 116.26 49.44
15-19 47.85 16.95 128.99 56.55
80-84 52.11 18.65 142.90 64.41
85-89 56.73 20.49 157.98 73.01
Correction for education•
Yean o£ Education Part A Part 8
12 +2.62 +12.90
13 +1.31 +6.45
14 0 0
15 -1.31 -6.45
16 -2.62 -12.90
17 -3.93 -19.35
•To be added to or subtracted from the predicted score

for the respective age group.
for the Color Trails Test
Table A5.1. Locator Table for the Color Trails Test (CIT)
Study Age• n Sample Composition Education• Location
CIT.l D'Elia et al., 18-29 Medically & psychiatrically Data are stratified USA
1994 30-44 healthy adults residing in at each age category
page 103 45-59 a variety of settings by 6 education
Data are not reproduced 60-74 categories: < 8 years,
in this book 75--89 9-11 years, 12 years,
1,528 Sample is 88% male 13-15 years, 16 years,
~17 years
CIT.2 Ponton et al., 16-29 42 180 female, 120 male Data are stratified Southern
1996 30-39 66 medically & psychiatrically at each age category California
page 104 40-49 27 healthy Hispanics by< 10 and ~10
Table A5.2, A5.3 50-75 45 years of education
CIT.3 Hsieh & Riley, 30-39 43 Urban-dwelling, Mandarin- Data collected on Mainland
1997 40-49 39 speaking adults; 93 males individuals with 1-17 China
page 105 50-59 33 and 84 females years of education;
Table A5.4-A5.6 60-69 32 however, data
70-83 30 reported by age
categories only
CIT.4 LaRue et al., 64-74 240 Community-dwelling, Data are stratified New Mexico
1999 75-97 106 non-Hispanic and Hispanic by age and educational
page 105 men and women who are level. Education
Table A5.7-A5.10 bilingual Spanish/ English. categories: 0-6 years,
Data reflect number of circles 7-9 years, 10-12 years,
with digits that were correctly and > 12 years for
completed in 60 seconds. Hispanics; 0-12 and
> 12 for non-Hispanics
•Age column and education column contain information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, whichever is provided by the authors.
657
658 APPENDIX 5
Table A5.2. [CIT.2a] Ponton et al., 1996: Data for a Sample of Spanish-Speaking Adult Females Stratified
by Two Age Groups and Two Education Groups (Means and Standard Deviations)
Age Group
16-29 30-39 40-49 50-75
Years of Education
<10 2:10 <10 2:10 <10 2:10 <10 ~10

(n= 12) (n=30) (n=22) (n=44) (n=16) (n=11) (n=25) (n=20)
Color Trails 1 49.33 34.43 49.62 34.96 56.38 34.09 70.28 44.20
(30.75) (12.69) (13.76) (11.91) (14.16) (6.44) (44.10) (15.73)
Color Trails 2 128.58 80.03 114.05 84.48 134.06 79.46 146.72 99.20
(45.19) (23.43) (49.80) (27.40) (54.15) (28.42) (62.21) (25.75)
Raven's Total 38.33 42.73 34.77 42.48 29.19 41.36 23.36 42.30
(13.24) (10.28) (13.82) (9.70) (12.49) (9.52) (7.63) (10.38)
Table A5.3. [CIT.2b] Ponton et al., 1996: Data for a Sample of Spanish-Speaking Adult Males Stratified by
Two Age Groups and Two Education Groups (Means and Standard Deviations)
Age Group
16-29 30-39 40-49 50-75
Years of Education
<10 2:10 <10 2:10 <10 2:10 <10 ~10

(n=11) (n=ll) (n= 13) (n= 18) (n=l2) (n=l7) (n=l8) (n=6)
Color Trails 1 49.27 37.28 49.23 39.28 59.08 36.88 53.06 55.17
(15.70) (13.20) (10,92) (10.44) (15.22) (8.12) (20.00) (21.86)
Color Trails 2 100.64 88.36 116.31 84.33 129.17 91.24 136.22 113.83
(20.38) (26.39) (39.35) (17.11) (39.64) (16.12) (34.71) (39.78)
Raven's Total 35.09 42.68 38.08 46.28 28.58 45.94 31.78 42.50
(10.00) (11.86) (10.44) (7.76) (13.84) (6.62) (8.63) (10.17)
Table A5.4. [CIT.3a] Hsieh and Riley, 1997: Data for a Sample of
Mandarin-Speaking Adults in China: Sample Size for Each Age/
Education Category
Years of Education
Age 1-6 7-10 11-17
30-39 6 14 22
40-49 6 19 14
50-59 6 15 12
60-69 22 6 4
70-83 28 2 0
Total 68 57 52
APPENDIX 5 659
Table A5.5. [CIT.3b] Hsieh and Riley, 1997: Data for a Sample of Mandarin-Speaking
Adults in China: Effect of Age on Test Performance
Age Group
30--39 40-49 50-59 60-69 70-83

(n =43) (n=39) (n=33) (n=32) (n =30)
Color Trails 1 42.05 50.97 56.76 129.66 162.97

(15.69) (20.66) (27.63) (80.99) (98.55)
Color Trails 2 89.95 104.74 138.58 225.31 306.47
(37.47) (35.65) (67.75) (103.06) (188.38)
Interference index 1.24 1.28 1.59 0.98 1.01
(0.80) (0.97) (l.li) (0.73) (0.75)
Table A5.6. [CIT.3c] Hsieh and Riley, 1997: Data for a Sample
of Mandarin-Speaking Adults in China: Effect of Education on
Test Performance
Years of Education
0--6 7-9 10--17

(n=96) (n=59) (n=22)
Color Trails 1 132.9 55.6 48.2

(92.45) (30.42) (25.5)
Color Trails 2 243.10 128.0 97.9
(154.30) (77.17) (39.90)
Interference index 1.12 1.28 1.56
(0.73) (1.08) (1.11)
Table A5.7. [CTT.4a] LaRue et al., 1999: Data for a Sample of Bilingual Hispanics in
New Mexico, aged 65-74 years: Effects of Age and Education on Test Performance•
Years of Education
0--6 7-9 10--12 >12
Color Trails 1 18.44 20.73 21.67 22.77

(5.26) (4.42) (4.14) (4.50)
Sample size n=39 n=56 n=92 n=53
Color Trails 2 10.19 11.93 12.50 14.09
(3.84) (3.37) (4.00) (4.72)
Sample size n=37 n=55 n=92 n=53
•Performance was measured as the number of digits correctly traced in 60 seconds.

660 APPENDIX S
Table A5.8. [CIT.4b] LaRue et al., 1999: Data for a Sample of Bilingual
Hispanics in New Mexico, aged 75-97 years: Effects of Age and Education
on Test Perfonnance•
Years of Education
0-6 7-9 10-12 >12
Color 14.41 17.15 19.08 20.73

Trails 1 (5.77) (5.50) (4.48) (4.05)
Sample size n=37 n=34 n=24 n=ll
Color 7.84 9.33 9.71 11.73
Trails 2 (3.25) (3.25) (4.12) (4.82)
Sample size n=32 n=33 n=24 n=ll
•perfonnance was measured as the number of digits correctly traced in 60 seconds.
Table A5.9. [CIT.4c] LaRue et al., 1999: Data for a Sample of Non-
Hispanic Caucasians in New Mexico, aged 65-74 years: Effects of Age
and Education on Test Perfonnance•
Years of Education
0-12 >12
Color Trails 1 23.46 23.49

(3.14) (2.91)
Sample size n=84 n=181
(4.38) (4.81)
Sample size n=84 n=l81
"Perfonnance was measured as the number of digits correctly traced in 60 seconds.
Table A5.10. [CIT.4d] LaRue et al., 1999: Data for a Sample of Non-
Hispanic Caucasians in New Mexico, aged 75-97 years: Effects of Age
and Education on Test Perfonnance•
Years of Education
0-12 >12

(4.90) (5.24)
(3.62) (4.10)
•perfonnance was measured as the number of digits correctly traced in 60 seconds.

Appendix 6: locator and Data
Tables for the Stroop Test
Table A6.1. Locator Table for the Stroop Test

Study/Version Age• n Sample Composition t IQ/Education• Locationt
STROOP.l/Golden 15--45 USA

Golden, 1978 46-64
page 116 65-80
Data are not reproduced
in this book
STROOP..2/Trenerry 18-49 106 43 males, 14.68 (2.44) USA
Trenerry et al., 1989 x=30.34 (8.57) 63 females;
page 117 50-79 50 26 males, 14.70 (3.24)
Data are not reproduced x=62.68 (7.93) 24 males;
in this book Nonneurological,
nonpsychiatric
STROOP.3/Comalli 17-19 18 Age 17-44 years, MA
Comalli et al., 1962 25-34 14 college students
page 117 35-44 16
Table A6.2 65-80 15 Age 65-80 from
community
old-age club
STROOP.4/Comalli 63.2 15 NY
Eson, personal 67.0 16
communication 72.0 16
page 118 78.3 16
Table A6.3
(continued)
661
662 APPENDIX 6

Study/Version Age• n Sample Compositiont IQ/Education• Locationt
STROOP.5/Comalli 29.2 (12.0) 20 13 males, 12.5 (2.0) Canada

Stuss et al., 1985 7 females, 106.6 (13.4)
page US English/French
Table A6.4
STROOP.6/Comalli 50-59 25 25 males, 14.34 (2.63) S.CA
Boone et al., 1990 60-69 21 36 females; 113.79 (13.51)
page US 70-79 15 51 white,
Table A6.5 4 black,
3 Asian,
3 Hispanic;
fluent English;
Nonneurological,
nonpsychiatric,
no substance abuse
STROOP.7/ComalJi 35.8 (13.7) 16 7 males, 15.2 (2.8) S.CA
Boone et al., 1991 9 females; 2LD
page 119 19% left-handed; 109.1 (10.9)
TableA6.6 14 white,
2 Asian;
fluent English;
no substance abuse,
nonneurological,
nonpsychiatric
STROOP.8/Coma1Ji 231 Community- Boston,
Demick & Harkins, 1997 20-39 24M, 32 F dwelling High school MA
page 120 40-59 21M, 31 F individuals plus some
Tables A6.7-A6.10 60-74 23M, 31 F in good college
75+ 38M, 28 F health
STROOP.9/Comalli <65 53 males, 14.57 (2.55) S.CA
Boone, 1999 Average IQ 33 102 females;
page 120 High average 1q• 23 nonneurological,
Table A6.11 Superior IQ : 35 nonpsychiatric,
?:65 I no substance abuse;
I
Average IQ 20 ftuent English
High average uj 16
SuperioriQ 24
STROOP.IO/Comalli 34.32 22 Data for 13.36 S.CA
Boone et al., 2001 (14.81) male controls; (2.15)
page 121 fluent English
Table A6.12
STROOP.ll/Kaplan 40-49 118 White male pilots, 16.1 (1.9) USA
D'Elia et al., unpublished 50-59 79 passed med exam 15.6 (2.0)
data
page 121
Table A6.13
STROOP.l2/Kaplan 18-20 50 28 males, 13.36 (0.63) S.CA
Schiltz, personal 22 females; range 13-15
communication college students;
page 122 no head injury with
Table A6.14 loss of consciousness;
native English-speaking
old-age club
APPENDIX 6 663

Study/Version Age• n Sample Compositiont IQ/Education• I.ncationt
STROOP.l3/Kaplan 19--41 42 15 males, 14.76 (2.2) S.CA

Strickland et a!., 1997 30.17 27 females;
page 122 (6.34) African American;
Table A6.15 nonneurological,
nonpsychiatric,
no substance abuse
STROOP.l4/Kaplan 40.57 692 Seronegative homosexual 16.3 MACS
Miller, 2003 (7.5) and bisexual males from (2.3) centers
page 122 MACS, native English <16
Table A6.16 speakers; data partitioned 16
by age x education >16
STROOP.l5/Golden 24-36 46 College students and College Israel
Ingraham et a!., 1988 28.4 (3.2) college-educated adults;
page 123 28 males,
Table A6.17 18 females;
Golden version with new
randomization, bold typeface,
and Hebrew lettering
STROOP.l6/Golden 18-32 40 17 males, College wv
Connor et a!., 1988 23 females students
page 123
TableA6.18
STROOP.l7/Golden 72.9 (8.3) 36 13 males, 14.6 (2.7) S.CA
Fisher et a!., 1990 23 females;
page 124 no ocular disease
Table A6.19
STROOP.l8/Golden 2()....35 70 38 males, 12.36 (2.09) Canada
Daigneault et a!., 1992 27.71 (4.05) 32 females;
page124 45-65 58 30 males, 12.11 (3.63)
Table A6.20 56.62 (5.29) 28 females;
French language;
no substance abuse,
nonneurological,
nonpsychiatric;
unskilled
blue-collar to
professional
STROOP.l9/Golden 72 Normal controls;
Swerdlow et a!., 1995 35 males,
page 125 38 females;
TableA6.21 sample divided by
MMPI criteria and
gender; interference
ratio reported in
addition to means
STROOP.20/Golden 56-59 54 165 males, :s;7=2 MN
Ivnik et a!., 1996 60-64 81 191 females; 8-11 (n=34)
page125 65-69 65 355 white, 12 (n= 133)
Table A6.22 70-74 57 1 black; 13-15 (n =86)
75-79 52 329 right-handed, 16-17 (n=66)
80-84 27 17 left-handed, ~18 (n=35)
85-89 16 10 mixed-handed;
90-94 4 nonneurological,
nonpsychiatric
(CXJntinued)
664 APPENDIX 6

Study/Version Age• n Sample Compositiont IQ/Education• Locationt
STROOP.21/Golden 34.4 (13.1) 30 Vietnamese 14.3 (3.5) S.CA

Doan &: Swerdlow, speakers,
1999 31.2 (11.9) 30 English 15.4 (1.6)
page 126 speakers,
Table A6.23 Vietnamese
translation used
STROOP.U/Golden 33.2 32 Control sample; 14.8 (2.5)
Rapport et al., 2001 (13.2) 19 males, FSIQ
page 127 13 females; 108.0
Table A6.24 undergraduate (7.7)
students
STROOP.23/Golden 31.98 (13.14) 71 English-Spanish 14.92 (2.35) S.FL
Rosselli et al., 2002 35.90 (13.08) 40 bilinguals, 15.35 (2.45)
page 127
Table A6.25
40.91 (15.17)
.
~1 English
monolinguals,
14.25 (3.49)
Spanish
monolinguals
STROOP.24/Golden 28.23 us All-male sample; 6.66 Los Angeles,
Lopez-Carlos et al., 2003 (8.74) monolingual Spanish- (2.54) Mexico
page 128 lS-29 speaking Latino manual 0-6
Tables A6.26-A6.29 30-49 laborers; data 7-10
partitioned by age,
education, age x education,
and country;
Spanish version
administered
STROOP.25/Golden 30.5 2o Male control 11.8 Massachusetts
Cohen et al., 2003 (10.7) group (3.3)
page 129 FSIQ
Table A6.30 100.7
(11.0)
STROOP.26/Golden 60-84 2a6 African-American <12 Tampa.
Moering et al., 2004 older adults; 12 FL
page 129 data stratified >12
Table A6.31 by 2 age x gender x
education
STROOP.27/Dodrill 27.34 (8.41) so 30 males, 11.96 (2.01) WA
Dodrill, 1978a 20 females;
page 130 49 white,
Table A6.32 1 nonwhite;
9 students,
26 unemployed,
15 employed;
nonneurological
STROOP.28/Dodrill 22.4 (5) 12 Male, normal vision 13.7 (2.3) Australia
Sacks et al., 1991 range 1S-32 109.1(9.5)
page 130
Table A6.33
STROOP.29/Victoria
Regard. 1981, cited in
20-35
x=26.7
.,
I
Right-handed
young adults
Average IQ Canada
Spreen &: Strauss, 1991

page 131
Table A6.34
APPENDIX 6 665
Study/Version Age" n Sample Compositiont IQ/Education•
STROOP.30/Victroria 50-59 19 Healthy volunteers 13.2 (3.1) Canada

Spreen & Strauss, 1991 60-69 28
page 131 70-79 24
Table A6.35 80-94 15
STROOP.31/Trenerry 79.04 369 Data collected on 11.25 Australia
Anstey et al., 2000 (6.59) old and very old adults (2.79)
page 131 living in retirement
Table A6.36 villages and hostels; data 0-9
presented in raw scores 10-12
for the entire sample and 13+
in percentile distribution
for the sample stratified
by age x education;
14% male
"Age column and IQ/education column contain information regarding range and/or mean and standard deviation for the
ts.CA, Southern California; S.FL, Southern Florida.
Table A6.2. [Stroop.3] Comalli et al., 1962 (Comalli Version): Data for a
Nonclinical Sample
Age Groupings
17-19 25-34 35--44 65-80
n 18 14 16 15
Word Reading 40.5 39.4 42.6 45.1
Color Naming 56.1 60.9 57.9 68.9
Color Interference 103.0 106.2 109.9 165.1
Table A6.3. [Stroop.4] Eson, Personal Communication (Comalli Version):

Data for a Nonclinical Sample of Older Participants
Mean Age
63.2 67.0 72.0 78.3
n 15 16 16 16
Stroop A 46.9 51.8 53.4 67.9
(11.4) (18.6) (20.3) (23.3)
StroopB 64.9 71.2 71.6 83.9
(16.2) (18.4) (20.4) (20.3)
Stroop C 148.9 165.2 177.4 231.1
(45.2) (59.2) (63.8) (72.2)
666 APPENDIX 6
Table A6.4. [Stroop.S] Stuss et al., Table A6.6. [Stroop.7] Boone et al.,
1985 (Comalli Version): Data for the 1991 (Comalli Version): Data for the
Control Sample Control Sample
n 20 n 16
Age 29.2 Age 35.8
(12.0) (13.7)
Education 12.5 Education 15.2
(2.0) (2.2)
WAIS IQ 106.6 WAIS-R FSIQ 109.1
(13.4) (10.9)
Color Interference 64.0 Color Interference 112.9
(12.9) (22.5)
Table A6.5. [Stroop.6] Boone et al., 1990 (Comalli Table A6.7. [Stroop.8a] Demick and Harkins, 1997
Version): Data for the Control Sample• : (Comalli Version): Data for a Sample of Healthy
Adults•
Age Groupinrf
20-39 Age Grouping
50-59 60-69 70-79
Number of errors
n 25 21 15 CardA M 0.5
Word .Reading (SD) (0.8)
Time 40.25 46.05 44.07 Range 0.0-3.0
(4.95) (9.02) ' (6.39) Card 8 M 1.4
Errors 0.42 0.55 0.40 (SD) (1.5)
Range 0.0-6.0
(0.72) (1.15) ' (0.63)
M 3.5
CardC
Color Naming (SD) (2.9)
Time 57.75 63.30 74.27 Range 0.0-14.0
(9.74) (8.25) (17.16) Total Time(s)
Errors 2.00 1.55 2.00 CardA M 45.0
(2.11) (1.99) (2.27) (SD) (6.7)
Range 28.0-65.0
Color I~ Card B M 62.1
Time 120.79 135.50 148.67 (SD) (12.0)
(38.07) (29.81) (41.09) Range 41.0-91.0
Errors 1.75 2.20 1.60 CardC M 109.4
(1.42) (2.12) (1.50) (SD) (29.5)
Range 70.0-233.0
•The sample included 25 men and 36 womep; mean
education 14.34 (2.63)years; mean WAIS-R FSIQ 113.79 Color difficulty factor M 1.4
(total time on 8/total (SD) (0.3)
(13.51); 51 participants were white, 4 African Apterican,
time on A) Range 0.8-2.3
3 Asian, 3 Hispanic.
Interference factor M 47.3
(total time on C - total (SD) (24.6)
time on B) Range 14.0-151.0
•Mean age 24.2 (6.2) years, n=56; 24M, 32 F.

APPENDIX 6 667
Table A6.8. [Stroop.Bb] Demick and Harkins, 1997 (Comalli Version):

Data for a Sample of Healthy Adults•
40-59 Age Grouping
Number of errors
CardA M 0.3
(SO) (0.9)
Range 0.0--5.0
Card B M 0.9
(SO) (1.3)
Range 0.0--7.0
CardC M 1.3
(SO) (1.8)
Range 0.0--8.0
Total Time(s)
CardA M 43.7
(SO) (8.8)
Range 32.4-85.0
Card B M 59.2
(SO) (11.5)
Range 36.0--93.0
CardC M 104.5
(SO) (20.6)
Range 63.0--155.0
Color difficulty factor (total M 1.4
time on B/total time on A) (SO) (0.2)
Range 1.0--2.0
Interference factor (total M 45.3
time on C - total time on B) (SO) (15.4)
Range 17.0--82.0
•Mean age 48.9 (5.5) years; n =55; 21 M, 34 F.
Table A6.9. [Stroop.&] Demick and Harkins, 1997 (Comalli Version):

Data for a Sample of Healthy Adults•
60--74 Age Grouping
Number of errors
CardA M 0.5
(SO) (0.9)
Range 0.0--4.0
Card B M 1.7
(SO) (2.0)
Range 0.0--11.0
CardC M 2.8
(SO) (3.5)
Range 0.0--13.0
Total Time(s)
CardA M 48.5
(SO) (11.2)
Range 25.0--86.0
(continued)
668 APPENDIX 6

60-74 Age Grouping
Card B M 69.4
(SD) (14.1)
Range 46.0-123.0
CardC M 142.4
(SD) (26.2)
Range 88.0-204.0
time on 8/total time on A) (SD) (0.3)
Range 0.8-2.2
time on C -total time on B) (SD) (23.1)
Range 32.0-142.0
•Mean age (68.9) (3.8) years; n =54; 23M, 31 F.
Table A6.10. [Stroop.Bd] Demick and Harkins, Table A6.11. [Stroop.9] Boone, 1999 (Comalli
1997 (Comalli Version): Data for a Sample of Version): Data for a Nonclinical Sample• of Adults
Healthy Adults• Ranging in Age from 45 to 84, Partitioned by
Age x IQ
75+ Age Grouping
High
Number of errors Average IQ Average IQ Superior IQ
CardA M 0.6
(SD) (2.1)
Age
Range 0.0-16.0
Card 8 M 2.1 <65 132.64 128.65 110.29
(SD) (2.2) (34.51) (26.87) (22.37)
Range 0.0-8.0 (n=33) (n=23) (n=35)
CardC M 3.2 ~65 164.65 153.75 137.08
(SD) (4.7) (51.90) (56.99) (33.14)
Range 0.0-229.0 (n=20) (n=16) (n=24)
Total Time(s)
•Mean education for the sample is 14.57 (2.55); mean
CardA M 50.2
WAIS-R FSIQ 115.41 (14.11); 53 males, 102 females.
(SD) (9.1)
Range 33.0-76.0
Card 8 M 77.6
(SD) (19.0)
Range 44.0-133.0
CardC M 156.6
(SD) (49.8)
Range 66.0-344.0
time on Bltotal time on A) (SD) (0.3)
Range 1.0-2.7
time on C- total time on B) (SD) (40.4)
Range 2.0-216.0
•Mean age 80.7 (3.6) years; n=66; 38M, 28 F.

APPENDIX 6 669
Table A6.12. [Sbuop.lO] Boone et al.,

2001 (Comalli Version): Data on 22 Male
Controls
Age 34.32
(14.81)
Education 13.36
(2.15)
WAIS-R FSIQ 107.14
(15.89)
Word Reading 43.64
(8.16)
Color Naming 58.91
(9.96)
Color Interference 112.36
(21.48)
Table A6.13. [Sbuop.ll] D'Elia et al., Unpublished Data (Kaplan

Version): Data for a Sample of Male Pilots Stratified into Two Age
Groups
Age Groupings
40-49 50-59
n 118 79
Education 16.1 15.6
(1.9) (2.0)
Color NIJflling
T'une 60.0 63.9
(10.2) (15.7)
Near-misses 0.63 0.33
(1.0) (0.63)
Errors 0.39 0.39
(0.67) (0.67)
Word Beading
Time 45.8 47.2
(8.8) (9.1)
(0.41) (0.58)
Errors 0.26 0.20
(0.52) (0.49)
Color Irtterferenee
Time 105.7 112.4
(21.4) (22.6)
(1.4) (1.7)
Errors 0.70 0.95
(1.1) (1.8)
670 APPENDIX 6
Table A6.14. [Stroop.12] Schiltz, Persoqru Communication (Kaplan

Version): Data for a Sample of 50 HealtJty Undergraduate Students
First Half Total
Color Naming 23.02

I
I
51.14
(3.28) (6.79)
(range 16-30) I (range 37-67)
I
Word Reading 17.62 38.46
(2.78) ·~ (5.10)
(range 13-25) I (range 30-54)
Interference 43.22 89.40
(9.23) (17.76)
(range 27-74) (range 58-132)
Table A6.15. [Stroop.13] Strickland et ., 1997 (Kaplan Version):

Data for a Sample of 42 Healthy African- erican Adults• (15 Male,
27 Female) Between 19 and 41 Years of
Tune Near-Misses
Color Naming 59.26 0.43 1.28

(17.57) 0.74) (1.29)
Word Reading 43.62 0.24 0.50
(7.17) 0.53) (0.67)
Color Interference 109.98 1.05 2.74
(23.42) .82) (2.07)
•Mean age for the sample is 30.17 (6.34) years, d mean education is 14.76
(2.24) years.
APPENDIX 6 671
Table A6.16. [Stroop.l4) Miller, 2003 (Kaplan Version): Data for a Sample of Sero-
negative Homosexual/Bisexual Males Participating in MACS, • Stratified by Age x
Education
Color Word
Age Education (years) Naming Reading Interference
25-34 <16 Mean 59.56 45.31 115.41

(SD) (10.31) (9.06) (27.23)
n 32 32 46
16 Mean 55.03 43.25 99.84
(SD) (9.57) (8.31) (22.02)
n 32 32 55
>16 Mean 54.41 40.33 99.78
(SD) (8.82) (5.60) (23.79)
n 27 27 45
Total Mean 56.44 43.11 104.73
(SD) (9.80) (8.07) (25.20)
n 91 91 146
35--44 <16 Mean 59.08 45.97 115.16
(SD) (9.48) (9.30) (29.22)
n 73 73 99
16 Mean 55.97 42.68 103.21
(SD) (9.72) (6.56) (22.98)
n 79 79 97
>16 Mean 53.44 40.19 102.90
(SD) (9.69) (6.35) (25.16)
n 87 86 125
Total Mean 55.69 42.79 106.78
(SD) (9.78) (7.78) (26.39)
n 239 238 321
45-59 <16 Mean 59.85 46.11 120.34
(SD) (11.30) (8.13) (30.88)
n 54 54 65
16 Mean 61.92 47.75 113.52
(SD) (12.93) (8.19) (23.16)
n 36 36 46
>16 Mean 57.81 42.56 108.27
(SD) (12.19) (7.86) (24.60)
n 102 102 114
Total Mean 59.16 44.53 112.83
(SD) (12.13) (8.25) (26.69)
n 192 192 225
Total <16 Mean 58.98 45.89 116.82
(SD) (10.26) (8.82) (29.28)
n 159 159 210
16 Mean 57.22 44.05 104.67
(SD) (10.83) (7.63) (23.21)
n 147 147 198
>16 Mean 55.62 41.33 104.56
(SD) (11.01) (7.10) (24.85)
n 216 215 284
Total Mean 57.10 43.49 108.31
(SD) (10.81) (8.03) (26.41)
n 522 521 692
•Multi-Center AIDS Cohort Study.

672 APPENDIX 6
Table A6.17. [Stroop.15] Ingraham et al., Table A6.20. [Stroop.18] Diagneault et al., 1992
1988 (Golden Version): Data for 46 College- (Golden Version): Data for Two Age Groups of
Educated Israelis on a Hebrew Version French-Speaking Canadian Adults
(18 Female, 28 Male)
Age Group
Age 28.4
(3.2) 20-35 45-65
24-36
n 70 58
Reading 99.6
Mean age 27.71 56.62
(11.0)
(4.05) (5.29)
Naming colors 77.0
Mean education 12.36 12.11
(10.4)
(2.09) (3.63)
Color Interference 47.1
Male/female ratio 38132 30128
(10.1)
Color Interference 48.80 37.87
Golden Interference score 3.9
(8.63) (7.67)
(8.3)
Table A6.18. [Stroop.16] Connor et al., 1988

(Golden Version): Data for a Sample of 40 College Table A6.21. [Stroop. 19] Swerdlow et al., 1995
Students (17 Male, 23 Female)• (Golden Version): Data on Normals Divided into
"Psychosis-Prone" vs. "Non-Psychosis-Prone" Groups
Pretest Post-Test FoHow-Upt and Male vs. Female
Word 113.52 123.22 130.87 Normal Abnormal
(14.72) (19.28) (17.00) MMPI MMPI Male Female
Color 81.22 93.80 99.18
(9.38) (16.85) (14.67) n 46 26 34 38
Color-word 49.75 70.62 75.07 Word 112.36 106.39 106.24 113.76

(7.53) (15.74) (16.30) (2.59) (3.16) (3.13) (2.53)
Color 80.74 78.15 76.53 84.72
• Age range for the sample is 18-25, with the exception of (1.56) (3.02) (2.43) (1.64)
one 32-year-old participant.
Interference 49.70 44.00 46.12 49.00
'The test was administered at the baseline (pretest), after (1.16) (1.75) (1.54) (1.34)
6ve practice sessions (post-test), and at 1-week follow-up. Interference 1.65 1.80 1.69 1.71
ratio (0.03) (0.06) (0.05) (0.04)
TableA6.19. [Stroop.17] Fisheretal.,1990 (Golden

Version): Data for the Control Sample
n 36
Age 72.9
(8.3)
Education 14.6
(2.7)
Male/female ratio 13123
Word 96.6
(15.8)
Color 64.9
(13.9)
Color-word 33.4
(10.8)
Interference scores -5.2
(8.6)
APPENDIX 6 673
Table A6.22. [Stroop.20] Ivnik et al., Table A6.23. [Stroop.21] Doan and Swerdlow,
1996 (Golden Version): Demographic 1999 (Golden Version): Data for 30 Vietnamese-
Description of the Sample Used in the speaking Participants (on a Vietnamese Version)
Mayo Older Americans Normative and 30 English-Speaking Participants
Studies
English Vietnamese
n
Age 31.2 34.4
Age groups (11.9) (13.1)
56-59 54 Education 15.4 14.3
60-64 81 (1.6) (3.5)
65-69 65 Gender 12M, 18F 13M, 17F
70-74 57
Word 108.5 103.85
75-79 52
(12.22) (18.68)
80-84 27
85--89 16 Color 76.25 72.10
90-94 4 (10.79) (18.06)
Interference 44.50 43.20
Education
(9.93) (14.27)
<7 2
Color-interference ratio 1.77 1.72
8--11 34
(0.35) (0.42)
12 133
13-15 86 "Cost" 1.25 0.94
16-17 66 (8.45) (11.19)
>18 35
Gender
Male 165
Female 191
.Race
Table A6.24. [Stroop.22] Rapport et al.,
Caucasian 355 2001 (Golden Version): Data for a Sam-
Black 1
ple of 32 Controls (19M, 13 F)
Handetlna.
Age 33.2
Right 329 (13.2)
Left 17
Mixed 10 Education 14.8
(2.5)
FSIQ 108.0
(7.7)
Word 100.9
(13.4)
Color 80.3
(10.4)
674 APPENDIX 6
Table A6.25. [Stroop.23] Rosselli et al., 2002 (Golden stimuli but scores
are time to complete all items): Data for Spanish Monolinguals, Spanish-
English Bilinguals, and English Monolinguals
Monolingual Monolingual
Spanish Bilingual English
n 11 71 40
Age 40.91 31.98 35.90
(15.17) (13.14) (13.80)
Education 14.25 14.92 15.35
(3.49) (2.35) (2.45)
Gender 3M, SF 32M, 39F 13M, 27F
Spani.h Stroop
Word Reading
Time 45.73 46.89
(5.39) (10.01)
Errors 00.00 0.09
(0.00) (0.42)
Color Naming
Time 63.56 68.76
(12.18) (16.14)
Errors 0.09 0.28
(0.26) (0.72)
Interference
Time 97.91 112.85
(27.44) (30.18)
Errors 0.64 0.76
(0.93) (1.20)
Engliah Stroop
Word Reading
Time 47.20 43.68
(14.34) (8.59)
Errors 0.42 0.15
(0.20) (0.70)
Color Naming
Time 72.07 61.98
(17.94) (12.53)
Errors 0.23 0.20
(0.54) (0.56)
Interference
Time 114.24 108.40
(32.22) (30.17)
Errors 0.59 0.68
(1.69) (1.07)
APPENDIX 6 675
Table A6.26. [Stroop.24a] Lopez-Carlos et al., 2003: Data for Monolingual

Spanish Speakers with :510 Years of Education Stratified by Age Group
Age Block Color-

Group n Design• Word Color Word
18-29 71 29.94 102.00 67.60 39.97

(10.35) (20.87) (17.29) (13.65)
30-49 44 28.41 104.67 64.81 36.42
(12.06) (25.27) (18.31) (12.94)
•wechsler Adult Intelligence Scale-III Block Design raw scores (Mexican version).
Table A6.27. [Stroop.24b] Lopez-Carlos et al., 2003: Data for Monolingual

Spanish Speakers with :510 Years of Education Stratified by Education
Group
Education Block Color-

Group n Design• Word Color Word
0-6 56 25.75 99.96 61.16 33.44

(10.74) (26.10) (17.65) (11.53)
7-10 59 32.78 105.91 71.64 43.53
(10.22) (18.39) (16.22) (13.36)
•wAIS-111 Block Design raw scores (Mexican version).
Table A6.28. [Stroop.24c] Lopez-Carlos et al., 2003: Data for Monolingual Spanish Speakers
with :510 Years of Education Stratified by Age and Education Group
Age x Education
Group n Block Design" Word Color Color-Word
18-29
0-6 30 26.50 97.40 62.53 33.47
(9.46) (25.86) (16.63) (10.98)
7-10 41 32.46 105.45 71.40 44.85
(10.20) (15.65) (16.99) (13.54)
30-49
0-6 26 24.88 103.04 59.52 33.40
(11.97) (26.59) (19.02) (12.40)
7-10 18 33.50 106.94 72.17 40.61
(10.52) (23.88) (14.80) (12.85)
•wAIS-111 Block Design raw scores (Mexican version).

676 APPENDIX 6
Table A6.29. [Stroop.24d] Lopez-Carlo~ et al., 2003: Data for Monolingual Spanish
Speakers with ~10 Years of Education Statified by Country Group
Country Block Color-
Group n Desigr( Word Color Word
Los Angeles, USA 65 27.69~ 103.25 64.92 36.48

(10.26) (24.64) (17.76) (13.32)
Jalisco, Mexico 50 31.52 102.72 68.58 41.32
(11.66! (19.90) (17.50) (13.24)
•wAIS-111 Block Design raw scores (Mexican frsion).
Table A6.30. [Stroop.25] Cohen et al.,

2003 (Golden Version): Data for a
Male Control Sample
Age 30.5
(10.7)
Education 11.8
(3.3)
FSIQ 100.7
(11.0)
Stroop Interference 56.2
(6.3)
Table A6.31. [Stroop.26] Moering et al.~ 2004 (Golden Version): Data for a
Sample of Elderly African Americans I
Years of Education
<12 : 12 >12 Total
Age rcmge 60-71

Male
n 37 i1 6 54
Education 9.73 l2·00 16.33 10.93
(0.81) :- (1.51) (2.28)
Age 65.11 t.82 65.83 65.13
(2.93) .82) (1.60) (2.77)
Word 70.49 'f3.18 81.83 72.30
(13.64) (:(4.55) (18.80) (14.57)
Color 31.46 ~.82 35.67 32.20
(13.98) (9.77) (11.11) (12.81)
Color-word 20.57 ~.73 25.17 21.11
(6.38) {4.73) (7.28) (6.24)
Female
n 30 ]j; 11 57
Education 9.37
(1.22)
too 15.00
(1.79)
11.19
(2.48)
Age 65.53 ~.63 64.18 64.46
(3.61) <. .93) (3.16) (3.33)
Word 74.93 ~.87 87.45 79.02
(10.93) ( .93) (8.24) (11.94)
Color 34.67 ~.50 53.91 38.33
(12.34) .42) (14.52) (13.55)
Color-word 21.87 2$.44 27.82 24.02
(5.85) (~.49) (4.09) (5.66)
APPENDIX 6 677

Years of Education
<12 12 >12 Total
Age nmge 7!-84

Male
n 48 4 2 54
Education 7.29 12.00 17.00 8.00
(2.19) (1.41) (3.00)
Age 77.31 77.75 74.50 77.24
(2.13) (2.63) (3.54) (2.22)
Word 54.75 65.75 80.50 56.52
(17.27) (7.37) (7.78) (17.31)
Color 27.50 29.00 45.50 28.28
(10.49) (5.42) (21.92) (10.96)
Color-word 19.96 20.75 27.00 20.28
(3.89) (3.40) (15.56) (4.52)
Female
n 56 11 4 71
Education 8.27 12.00 15.50 9.25
(1.88) (1.92) (2.67)
Age 77.61 76.64 79.75 77.56
(2.73) (3.04) (3.10) (2.83)
Word 66.36 75.27 89.50 69.04
(14.59) (10.39) (10.41) (14.94)
Color 28.48 32.64 56.25 30.69
(7.95) (5.87) (24.07) (11.01)
Color-word 20.57 23.91 31.50 21.70
(4.38) (4.21) (10.75) (5.46)
Allagu
n 171 42 23 236
Education 8.50 12.00 15.61 9.82
(1.90) (1.75) (2.90)
Age 72.75 68.30 68.21 71.53
(6.56) (7.01) (6.66) (6.91)
Word 65.50 75.95 85.74 69.33
(16.28) (12.84) (11.81) (16.69)
Color 29.94 33.05 48.83 32.33
(11.18) (6.86) (17.20) (12.54)
Color-word 20.63 23.36 27.70 21.80
(5.02) (4.72) (7.13) (5.63)
678 APPENDIX 6
Table A6.32. [Stroop.27] Dodrill, 1978a

(Dodrill Version): Data for a Control Sant>le
of 50 Participants•
Part I 88.62
(17..23)
Part II 225$0
(59~)
Part I+ part II 314.~

(71.04)
Part II- part I 137.18
(50.74)
•Mean age 27.34 (8.41) years, mean educat+m

11.96 (2.01) years.
Table A6.33. [Stroop.28] Sacks et al., l~~n (Dodrill Version): Data for a
Sample of 12 Male Australian University ~dents•
Alternate Fonns
DodrillFonn 1 2 3 4 5
72.6 63.7 68.8 68.6 67.4 66.0

(18.2) (13.4) (21.0) (13.3) (18.1) (18.0)
•Mean age 22.4 (5.0) years, mean education 13.t (2.3) years, mean FSIQ 109.1 (9.5).
Table A6.34. [Stroop.29] Regard, 1981, ~ted in

Spreen and Strauss, 1998 (Victoria Versioq'): Data
for a Sample of 40 Right-Handed Young\Adults,
Age 20-35
Nalldng color of cloe. (D)
Time 10.10
(2.01)
Errors 0.03
(0.16)
Nalldng color print of noncolor tDOrda (W)

Time 12.00
(2.49)
Errors 0.03
(0.16)
NIJfldng color print of color tDOrda (C) I

Time i 19.25
I (5.18)
Errors : 0.23
: (0.53)
APPENDIX 6 679
Table A6.3S. [Stroop.30] Spreen and Strauss, 1991 (Victoria Version): Data for a Sample of
Healthy Older Participants•
Age Groupings
50-59 60-69 70-79 80-94
n 19 28 24 15
Nflllling color of doe. (D)
Time 13.74 12.71 15.00 18.87
(2.58) (1.90) (5.07) (4.67)
Errors 0.08 0.20
(0.28) (0.56)
Naming color print of noncolor worcla
Time 16.58 16.32 19.04 24.13
(3.34) (3.33) (5.10) (5.13)
Errors 0.04 0.13
(0.19) (0.35)
Naming color print of color worcla
Tune 28.90 31.82 38.38 61.13
(7.62) (9.86) (13.29) (30.94)
Errors 0.42 0.36 0.71 2.73
(0.77) (0.68) (1.16) (2.46)
•Mean education for the sample is 13.2 (3.1) years.
Table A6.36. [Stroop.31] Anstey et al., 2000 (Treneny Version): Data for 259 Retired Australian Elderly
Partitioned by Age Group and Education
Age Group
62-69 70-79 80-89 90-95
Education Group
Percentile 0-9 10-12 13+ 0-9 10-12 13+ 0-9 10-12 13+ 0-9 10-12 13+
Color Naming
5 47 42 52 49 47 42 48 50 44 56 56 49
10 47 42 52 49 48 49 51 52 45 56 56 49
25 60 42 54 56 50 57 57 58 50 60 59 49
50 64 49 66 63 56 56 63 64 56 61 69 59
75 74 52 71 73 64 58 76 75 63 85 81
90 86 73 65 88 86 71
95 120 86 75 103 103 108
n 7 7 6 39 56 26 31 44 25 7 4 3
Ini#Jrforence
5 42 83 47 39 31 51 5 8 5 6
10 42 83 47 45 47 54 17 13 32 6
25 59 83 72 54 63 63 41 44 58 5 3 22
50 89 99 93 78 77 76 69 59 73 8 37 53
75 99 112 100 92 93 96 90 80 85 55 78 83
90 103 105 112 100 90 99
95 112 112 112 110 95 103
n 7 7 6 39 58 29 28 45 24 7 4 5
Appendix 6m: Meta-Analysis Tables
for the Stro<)p Test (Golden Version,
Interference! Condition)
I
I
Table A6m.1. Results of the Meta-Anal~is and Predicted Scores for the Stroop Test,
Interference Condition ;
(Relevant values are weighted on the standlfd error for the test mean)
I
Description of the aggregate sample ,
Number of studies included in the analyiU 6

Years of publieation 1988-2004
Number of data points used in the analyiU 10
Total number ol participants 490
Variable n• xt snt Range
SGmp~.e.-
Mean 10 44.40 17.14 .20-71
Age
Mean 10 ~ 48.95 22.13 22.0-77.6
SD 10 6.44 3.93 2.0-11.9
Edueadon
Mean 9 12.53 2.30 8.0-15.4
SD 9 2.58 .68 Uh'3.6
Percent male 4 44.48 8.66 36.1-54.3
Tat acore met1t11
Combined mean 10 38.71 12.71 20.3-56.2
Combined SD 10 7.90 3.00 4.5-10.8
•Number of data points differs for different ~s due to missing data.

tweighted means and standard deviations. !
680
APPENDIX 6M 681
Predicted scores and SDs per age group• (Stroop, Golden

version, Interference condition)
95%CI
Age Predicted
Btmge Score Lower Band Upper Band
25-29 49.66 44.43 54.88

30-34 47.10 42.36 51.85
3S-39 44.55 40.03 49.07
40-44 41.99 37.39 46.60 Standard deviation for all age
4S-49 39.45 34.48 44.42 groups is 7.90.
50-54 36.89 31.33 42.46
55-59 34.34 28.01 40.67
60-64 31.79 24.59 38.99
~ 29.24 21.09 37.39
10-14 26.68 17.53 35.84
•Based on the equation:

Predict«l teat score= 63.69403-0.5104828 •age
Ordinary least-squares regression of test means on age

(linear)
R2 0.791
F<df)• p F(I.S) = 18.07, p = 0.008
Age -0.5104828 0.120 -4.25 0.008 - 0.819 to- 0.202

Constant 63.69403 5.184 12.29 0.000 50.369 to 77.019
Prediction
Predicted age range 25-74 years
SEe 3.08
95%CI 32.12-44.22
(continued)
682 APPENDIX 6M
0
40
30
0 0
20 0
20 30 40 50 60 70 80
age
Figure A6m.1. A scatterplot illustrating the dispersion of the data points around the regression line for the
Stroop Test. The size of the bubbles reflects the weight of the data point, with larger bubbles indicating larger
Tests for assumptions and model &t
Tests for heterogeneity in the 8nal data set

Q(dO·P Q(9) = 1659.82, p <0.000
Tests for model 8t--addition of a quadratic term
Model Adjusted R2 BIC BIC'
linear 0.791 0.764 44.113 -13.330

Quadratic 0.791 0.731 46.415 -11.028
BIC' difference of 2.302 provides positive support for the linear model.
Tests for parameter speclfleations

Normality of residuals
Shapiro-WJ.Ik W test W = 0.941, p = 0.562
Homoscedasticity
White's general test 2.171, p=0.338
APPENDIX 6M 683
A regression of SDs on age yielded an R2 of 0.015 (Fo.sl = 0.06, p = 0.822). Therefore,

the SD for the aggregate sample is suggested for use with all age groups.
Education
R2 0.792
Education 0.772 1.308 0.59 0.586 - 2.858 to 4.403
Table A6m.2. Comparison of means and standard deviations for the aggregate
sample for three conditions•
Word Color Color
Reading Naming Interference
Mean score 92.09 58.07 34.24

(14.32) (11.83) (8.12)
Aggregate n 342
Number of studies 4
Number of data points 7
Mean age 54.19
(23.68)
Mean education 12.79
(2.34)
•Four studies containing data for all three conditions are compared.
Appendix 7; Locator and Data Tables
for Auditory Consonant Trigrams
I
Study numbers and page numbers pro~ded in Locator table also provides a reference for
these tables refer to study numbers fi1d de- each study to a corresponding data table in
I
Table A7.1. Locator Table for Auditory q,nsonant Trigrams (ACf)
Sample IQ/ Version

Study Age• n Composition Education• Used Location
Acr.I Stuss et al., 60

.
Data presented by age, 14.5 9,18,36 Canada
1987 1~19 10 ! gender, and education (2.63)
page 137 20-29 10 : groups; 33 male, 27
Tables A7.2, A7.3 30-39 10 i
female; 49
40-49 10 right-handed;
50-59 10 I
nonpsychiatric,
60-69 10 1nonneurological,
1 no substance abuse
Acr.! Stuss et al., 90 Data presented by age 9,18,36 Canada

1988 1~29 30 igroups; 44 male, 46 14.1
page 137 30-49 30 'female; 76 (1.34)
Tables A7.4, A7.5 5()...81 30 1right-handed; 14.9
:nonpsychiatric, (3.95)
jnonneurological 13.2
(2.38)
ACf.3 Boone et al., 61 r;ta presented by age 14.34 3,9, 18 California
1990 50-59 25 ~ups; 25 men, 36 (2.63)
page 138 60-69 21 :WOmen; all
TableA7.6 70-79 15 but 10 white;
nonpsychiatric IQ 113.8
~onneurological, (13.5)
~o substance abuse,
~ major medical
ndition
I
684
APPENDIX 7 685

Sample IQ/ Version
Study Age" n Composition Education• Used Location
Acr.4 Boone 35.8 16 Data presented 15.2 3, 9,18 California

et al., 1991 (13.7) for healthy (2.8)
page 139 controls; 7 men,
Data are not 9women;
reproduced nonpsychiatric,
in this book nonneurological, IQ 109.1
no substance abuse, (10.9)
no major medical
condition; total score
for the sample provided
Acr.s Boone, 155 Data presented by 14.6 3, 9,18 California
1999 <65, IQ and age groups; (2.6)
page 139 Average IQ 32 53 male, 102 female;
Table A7.7 High average IQ 23 no substance abuse, IQ 115.4
Superior IQ 37 nonpsychiatric, (14.1)
~65. nonneurological
Average IQ 20
High average IQ 16
SuperioriQ 23
Acr.&Anil 236 Volunteers were 8-10 3,9, 18 Turkey
et al., 2003 16-25 assessed in Turkey 11-14
page 140 26-45 using a translated and >14
Table A7.8 46-65 modified version
of the ACT; data
partitioned by
3 age x 3 education
groups
•Age column and IQ/education column contain information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, whichever information is provided by the authors.
Data for the 9-, 18-, and 36-Second Delay Version
Table A7.2. [ACf.la] Stuss et al., 1987: Data for 60 Healthy Canadian
English- or French-Speaking Volunteers, Partitioned by Age Group•
Age Group
Delay
Interval 16-19 20-29 30-39 40-49 50-59 60-09
Combined cialcl
9 seconds 12.4 12.6 12.7 11.1 10.9 12.3
(1.6) (2.3) (2.6) (2.6) (2.8) (1.6)
18 seconds 12.0 12.2 12.6 11.0 10.5 11.2
(2.1) (2.7) (2.8) (2.5) (3.2) (1.8)
36 seconds 10.0 9.8 11.3 9.9 8.2 9.9
(3.0) (2.8) (2.8) (2.8) (3.6) (2.1)
(continued)
686 APPENDIX 7

Age Group
Delay
Interval 1~19 20-29 30--39 40-49 50-59 60-69
Fint ciait
9 seconds 11.8 12.2 12.8 11.0 10.9 12.2
(1.9) (2.5) (1.8) (2.8) (2.5) (1.6)
18 seconds 11.8 11.7 12.5 10.0 9.8 10.5
(2.2) (2.9) (3.0) (2.8) (3.4) (1.7)
36 seconds 9.9 8.7 11.0 8.7 8.2 10.0
(2.3) (3.3) (2.5) (3.4) (3.1) (1.8)
Second ciait (l tHeA: later)

9 seconds 13.0 13.0 12.6 11.2 11.0 12.3
(1.4) (2.0) (3.5) (2.5) (3.0) (1.7)
18 seconds 12.2 12.6 12.7 12.0 11.1 11.8
(2.4) (2.6) (2.5) (2.2) (2.9) (2.0)
36 seconds 10.1 10.9 11.5 11.0 8.2 9.8
(3.6) (2.3) (3.0) (2.1) (4.1) (2.4)
"Mean age for the sample 39.6 (2.62) years, and mean education is 14.5 (2.63) years;
33 males, 27 females.
Table A7.3. [ACf.1b] Stuss et al., 1987: Data Collapsed Across Age
Groups, Partitioned by Gender and Educational Level
Delay ~High >High
Interval Males Females School School
9 seconds 11.71 12.35 11.46 12.44

(2.65) (1.94) (2.60) (2.06)
18 seconds 11.17 12.04 11.13 11.91
(2.74) (2.31) (2.55) (2.56)
36 seconds 9.20 10.61 9.32 10.26
(3.13) (2.57) (3.25) (2.67)
Table A7.4. [ACf.2a] Stuss et al., 1988: Demographic Characteristics for the Sample of
Healthy Canadian Volunteers Stratified into 11uee Age Groups
Gender Hand Preference Age Education (Years)
Group n M F R L Mean (SD) Mean (SD) Range
1 30 16 14 22 8 22.43 14.10 11-18

(2.67) (1.34)
2 30 14 16 26 4 40.63 14.90 5-20
(2.97) (3.95)
3 30 14 16 28 2 61.77 13.20 8-18
(3.0) (2.38)
APPENDIX 7 687
Table A7.5. [ACf.2b] Stuss et al., 1988: Performance Across Three Age Groups for the Initial
Test and Retest 1 Week Later
Age Group
16-29 30-49 50-69
Delay Interval Test Retest Test Retest Test Retest
9 seconds 12.0 12.6 12.0 12.1 11.5 11.7

(2.2) (2.0) (2.5) (2.9) (2.3) (2.3)
18 seconds 11.4 12.3 10.5 12.0 10.2 10.7
(2.8) (2.4) (3.1) (2.6) (2.5) (2.9)
36 seconds 9.4 10.9 9.9 11.1 8.7 8.6
(2.7) (2.9) (3.0) (2.4) (2.9) (3.5)
Data for the 3-, 9-, and 18-Second Delay Version
Table A7.6. [ACf.3] Boone et al., 1990: Data for the Control Sample
Stratified into Three Age Groups•
Age Group
50-59 60-09 70-79
n 25 21 15
Total 44.76 48.15 42.50
(7.36) (8.02) (7.70)
Perseverative errors 6.36 4.20 5.71
(3.81) (2.78) (2.87)
Altered sequence 2.00 1.85 2.71
(1.50) (1.73) (2.55)
3 seconds 12.56 12.95 11.21
(2.02) (2.42) (3.17)
9 seconds 9.44 10.75 8.93
(3.79) (3.34) (2.70)
18 seconds 8.20 9.65 7.50
(3.56) (3.59) (3.32)
"Mean education for the sample 14.34 (2.63), mean WAIS-R FSIQ 113.79 (13.51); 25
male, 36 female; 51 Caucasian, 4 African American, 3 Asian, 3 Hispanic.
Table A7.7. [ACf.S] Boone, 1999: Data for a

Nonclinical Sample• of Adults Ranging in Age from
45 to 84, Partitioned by Age x IQ
High
Age Average IQ Average IQ Superior IQ
<65 45.81 45.91 50.38

(6.05) (6.45) (8.01)
(n=32) (n=23) (n=37)
39.95 43.31 49.22
(9.99) (9.23) (6.02)
(n=20) (n=l6) (n=23)
"Mean education for the sample 14.57 (2.55); mean

WAIS-R FSIQ 115.41 (14.11); 53 males, 102 females.
688 APPENDIX 7
Table A7.8. [ACT.6] Anil eta!., 2003: Data Collected in Turkey on 236 Volunteers who Were Tested with a
Translated Version of the ACT
Age Group
16-25 26-45 46-65
Education Group•
M H u M H u M H u
Delay interval:
0 seconds 14.9 14.9 15.0 14.6 14.9 15.0 14.6 14.7 15.0
(0.3) (0.3) (0.0) (0.9) (0.3) (0.0) (0.8) (0.6) (0.0)
3 seconds 12.8 13.4 14.5 9.7 11.9 14.0 11.3 12.4 13.6
(3.0) (1.9) (1.0) (2.8) (2.3) (1.3) (2.8) (1.7) (1.7)
9 seconds 8.9 11.0 12.5 8.2 10.1 11.5 8.6 8.3 9.9
(2.7) (3.0) (2.1) (3.0) (2.4) (3.0) (3.1) (2.5) (2.5)
18 seconds 8.8 11.2 12.4 6.6 9.0 11.4 5.9 7.3 9.6
(4.0) (3.3) (2.8) (3.6) (3.3) (3.1) (2.8) (2.5) (3.2)
"M, middle school; H, high school; U, university.

Appendix 8: Locator and Data Tables for
the Paced Auditory Serial Addition Test
Table A8.1. Locator Table for the Paced Auditory Serial Addition Test (PASAT)
Sample IQ/
Study/Version Age• n Composition Education• Location
PASAT.l/ 14-.55 60 Control group: New Zealand

Gronwall 10 individuals who
Gronwall, 1977 had experienced accidents
page 146 but no head injuries, 10 naval
TableA8.2 "ratings" and 40 1st-year
university students;
no exclusion criteria
PASAT.!/ 1~19 60 M 33, F 27; community 12.30 Ottawa, Canada
Gronwall ~29 volunteers; non-neurological (0.95)
Stuss et al., 1987 30-39 or psychiatric conditions; 16.20
page 146 40-49 native language either (1.39)
TableA8.3 50-59 French or English 16.70
60-09 (3.86)
15.50
(2.88)
11.70
(2.41)
14.30
(2.00)
(continued)
689
690 APPENDIX 8

Sample IQ/
StudyNersion Age• n Composition Education• Ulcation
PASAT.3/ 1~29 90 M 46, F 44; see above 14.10 Ottawa,

Gronwall 30-49 study (PASAT.2) for (1.34) Canada
Stuss et al., 1988 50-69 description 14.90
page 147 (3.95)
Table A8.4 13.20
(2.38)
PASAT.4/ 42.8 40 M 10, F 30; no neurological, 14.0 Wisoonsin
Gronwall (8.1) physical, or psychiatric (2.3)
Rao et al., 1989 conditions
page 148 and neurological exams;
Table A8.5 normal brain imaging WAIS-R
Verbal IQ:
108.1
(6.3)
PASAT.5/ 29.7 26 M 20, F 6; no 13.2 Ottawa,
Gronwall (12.4) neurological or (3.0) Canada
Stuss et al., 1989 17-57 psychiatric conditions 7-20
page 148
TableA8.6
PASAT.6/ 46.0 100 M 25, F 75; paid volunteers; 13.3 Wisconsin
Gronwall (11.6) no neurological or psychiatric (2.0)
Rao et al., conditions; 99 out of 100
1991a were Caucasian
page 149
Table A8.7
PASAT.7/ 23.70 10 M 4, F 6; undergraduate 15.21 Victoria, Canada
Gronwall (2.53) students from University (0.79)
Strauss et al., 20-35 of Victoria
1994
page 149
Table A8.8
PASAT.8/ (38.0) 241 Nonpsychiatric and 13.6
Gronwall non neurological
Zalewski et al., veterans;
1994 189 Caucasians, 35
page 149 African Americans,
Table A8.9 11 Hispanics, and 6 "other"
PASAT.9/ 25.00 152 M 77, F 75; no neurological 12.97 United Kingdom
Gronwall (3.27) or psychiatric conditions; (2.86)
Crawford et al., 1~29 no systemic illness
1998b WAIS-R
page 150 38.10 FSIQ:
Table A8.10 (5.67) 105.0
30-49
60.70
(7.41)
50-74
APPENDIX 8 691

Sample IQ/
Study/Version Age" n Composition Education• Location
PASAT.IO/ 44.4 45 Neurologically normal; 12.8 Various locations, USA

Gronwall (11.4) recruited from (1.9)
Prevey et al., 1998 nonmedical hospital staff;
page 150 no serious medical, neurological,
Table A8.11 or psychiatric disorders;
no substance abuse
PASAT.ll/ 19.8 20 Undergraduate students;
Gronwall (computerized) (3.85) native English speakers;
Holdwick& no hearing problems; no
Wmgenfeld, 1999 history of repeating grades;
page 151 no neurological problems,
Table A8.12 head trauma, substance abuse
attention problems, learning
disability, or current medication
use
PASAT.l2/ 32.5 76 HIV-negative males, 14.6
Gronwall (6.3) 13.2% and 13.4% had (2.4)
Honn et al., 1999 history of marijuana abuse
page 151 or dependence; no
TableA8.13 intravenous drug use,
head injuries, or neurological or
psychiatric conditions
PASAT.l3/ 21.0 168 M 80, F 88; college College Arkansas
Gronwall (computerized) (5.1) students; native English students
Wingenfeld 17-48 speakers; no neurological or
et al., 1999 psychiatric conditions;
page 152 no emotional problems, learning
Table A8.14 disability, attentional
problems; 88% Caucasian,
4% African American, 4% Asian
American, and 4% "other"
PASAT.l4/ 30.2 35 M 20, F 15; no psychiatric 12.6 Australia
Gronwall (10.3) or neurological conditions; (2.0)
Bate et al., no intellectual disability,
2001 substance abuse, or hemiplegia Verbal IQ
page 152 of the dominant hand; (NART-R)
Table A8.15 native English speakers 101.1
(9.1)
PASAT.IS/ 45.8 140 M 62, F 78; community <9 years Amsterdam
Gronwall 22-73 volunteers; no neurological or 9-10 years
Boringa et al., 2001 psychiatric conditions; no >10years
page 153 substance abuse, learning
Table A8.16 disability, or serious
head injury
PASAT.l6/ Young M: 60 M 30, F 30; college students; 113.0 London, UK
Gronwall 21.1 nonpsychiatric; no medication; (1.5)
Fluck et al., 2001 (0.4) four groupings based on
page 153 age and gender
Table A8.17 YoungF: 112.4
20.9 (1.7)
(0.2)
(continued)
692 APPENDIX 8

Sample IQ/
StudyNersion Age•
• Composition Education• Location
Middle-aged M: 117.7
57.5 (1.8)
{1.3)
Middle-aged F: 113.3
60.3 (2.2)
(0.7)
PASAT.l7/ 37.97 35 M 9, F 26; 14.1 Nova Scotia,
Gronwall (12.94) staff, volunteer workers, and (2.3) Canada
Snyder et al., 2001 students recruited from the
page153 Queen Elizabeth II Health WAIS-R
TableA8.18 Science Centre, Dalhousi Vocabulary
University, and Multiple Sclerosis score:
Society; no psychiatric 54.5
or neurological (7.0)
conditions
PASAT.l8/ 21.0 526 M 233, F 293; 13.0
Levin
Brittain et al., 1991 (2.1) healthy volunteers; (1.3)
page 154 <25 no psychiatric or Shipley IQ:
Tables A8.19, A8.20 neurological conditions; 105.0
391 Caucasian and 135 (9.1)
non-Caucasian
31.4 14.0
(4.1) (2.2)
25-39 Shipley IQ:
103.0
(10.4)
46.2 13.0
(4.3) (3.1)
40-54 Shipley IQ:
101.0
(12.6)
67.0 12.0
(7.1) (2.5)
60-75 Shipley IQ:
106.0
{15.1)
PASAT.l9/ 19.0 143: M 66, F 77; 12.0 Texas
Levin {1.6) undergraduate students (0.77)
Roman et al., 1991 18-27 and employees of FSIQ:
page 155 Baylor University, 110
Table A8.21 business organizations, (12.3)
and senior centers;
40.0 no psychiatric or 15.0
(5.1) neurological conditions (2.6)
33-50 IQ:
110
(12.3)
69.0 15.0
(4.1) (3.2)
60-75 IQ:
107
(11.0)
APPENDIX 8 693

Sample IQ/
StudyNersion Age• n Composition Education• Location
PASAT.20/ 33.3 40 Normal controls: 14.9 New Jersey

Levin (12.4) no neurological or (2.2)
Cicerone, 1997 18-59 psychiatric conditions
page 155
Table A8.22
PASAT.21/ 20-49 821 M 672, F 149; 14.6 Pacific
Levin no neurological or (1.5) Northwest, USA
Wiens et al., 1997 M: 29.2 psychiatric conditions; FSIQ:
page 156 (6.1) 699 Caucasian, 46 Mrican 106.6
Tables A8.23, A8.24 F: 29.2 American, 31 Hispanic, (11.0)
(5.6) and 13 Native American H.6)
FSIQ:
105.4
(11.1)
PASAT.22/ 37.1 20 All female; from 15.0 New Jersey
Levin (2.4) local community; (0.55)
Tiersky et al., 1998 no neurological or
page 156 psychiatric conditions
Table A8.25
PASAT.23/ 29.4 22 All female; fluent in 13.9 California
Levin (10.7) English; no psychiatric (1.5);
Stein et al., 2002 or neurological conditions
page 156
Table A8.26
PASAT.24/ 40.9 22 Recruited from 15.4 New Jersey
Levin (8.9) Kessler Institute and its local (2.2)
Diamond et al., 31--56 community; no psychiatric or
1997 neurological conditions;
page 157 no substance abuse
Table A8.27
PASAT.25/ 39.7 566 M 61%, F 39%; 14.2 California
PASAT-50 (12.1) no psychiatric or (2.6)
Diehr et al., 1998 20--68 neurological conditions; 9-20
page 157 55% Mrican American,
Table A8.28 45% Caucasian
PASAT.26/ 39.7 560 M 342, F 218; 14.2 California
PASAT-50, (12.1) No psychiatric or (2.6)
-100, -200 20--68 neurological conditions 9-20
Diehr et al., 2003
page 158
Table A8.29
• Age column and IQ/education column contain information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, whichever information is provided by authors.
694 APPENDIX 8
Gronwall's Administration Version
Table A8.2. [PASAT.l] Gronwall, 1971a (Gronwall's Administration Version):

Data for Two Testing Probes Round&\ to the Nearest Whole Number for a
Sample of 60 New Zealander Controls•
PASATTrials
Testing Occasion 2.4 sec 2.0 sec 1.6 sec 1.2 sec
Test 46 40 32 22
(6) (7) (8) (5)
Retest 50 45 39 31
(5) (5) (6) (4)
"Age range for the sample is 14-55. The majOrity of participants are 17-25 years of age.
Table A8.3. [PASAT.2] Stuss et al., 19$7 (Gronwall's Administration Version): Data for a Sample of
Canadian Volunteers Stratified into Six Aie Groups
PASATTrials
Age Mean MIF
Group Age n ratio Education 2.4 sec 2.0 sec 1.6sec 1.2 sec
lat clait
16-19 17.3 10 515 12.3 45.7 42.8 35.3 24.0
(0.915) (0.95) (12.3) (15.2) (14.4) (9.6)
20-29 23.0 10 614 16.2 51.3 45.7 41.3 33.6
(2.67) (1.39) (6.2) (8.8) (12.0) (9.2)
30-39 33.9 10 515 16.7 44.8 42.7 35.6 27.6
(2.88) (3.86) (8.1) (10.4) (11.7) (8.1)
40-49 44.2 10 614 15.5 49.1 45.0 30.6 23.0
(3.12) (2.88) (7.3) (6.5) (13.8) (15.0)
50-59 55.3 10 614 11.7 35.3 25.5 17.6 8.2
(2.98) (2.41) (15.6) (16.4) (15.1) (11.5)
60-69 63.7 10 515 14.3 47.7 41.7 40.3 30.7
(3.13) (2.00) (8.8) (9.6) (10.7) (8.7)
2nd clait
16-19 52.0 47.4 40.6 28.7
(10.2) (11.6) (13.2) (11.7)
20-29 56.8 55.3 48.8 35.0
(4.6) (5.0) (6.0) (7.9)
30-39 51.9 48.7 43.2 32.0
(8.1) (10.3) (13.7) (11.8)
40-49 55.1 50.6 43.0 30.9
(5.2) (7.9) (9.2) (9.7)
50-59 41.1 32.6 24.4 14.5
(15.8) (19.8) (16.6) (12.4)
60-69 54.0 51.9 45.7 37.2
(5.1) (6.4) (8.3) (7.2)
APPENDIX 8 695
Table A8.4. [PASAT.3] Stuss et al., 1988 (Gronwall's Administration Version): Data for a Sample of
Canadian Volunteers Stratified into Three Age Groups
PASAT Trials
Age Mean MIF
Group Age n ratio Education 2.4 sec 2.0 sec 1.6 sec 1.2 sec
let Nit
1~29 22.43 30 16114 14.1 47.40 42.00 35.97 27.40
(2.67) (1.34) (10.12) (12.50) (12.97) (9.86)
30-49 40.63 30 14/16 14.9 43.43 41.87 33.10 24.63
(2.97) (3.95) (10.16) (10.16) (12.20) (10.55)
50-69 61.77 30 14/16 13.2 43.47 35.60 30.83 21.20
(3.00) (2.38) (13.60) (14.58) (15.85) (14.44)
.hdmait
1~29 53.73 50.23 43.37 31.20
(7.30) (9.17) (11.02) (10.24)
30-49 52.57 49.23 41.93 31.60
(7.89) (8.67) (10.56) (10.12)
50-69 49.20 45.00 37.10 27.27
(11.40) (15.30) (15.18) (13.50)
Table A8.5. [PASAT.4] Rao et al., 1989 (Gronwall's Administration Version):

Data for a Normal Control Sample
PASATTrials
WAIS-R
n MIF Ratio Age Education Verbal IQ MMSE• 3 sec 2sec
40 10130 42.8 14.0 106.5 29.8 48.5 37.1

(8.1) (2.3) (5.8) (0.5) (9.6) (10.1)
•Mini-Mental State Exam (Folstein et al., 1975).
Table A8.6. [PASAT.S] Stuss et al., 1989 (Gronwall's Administration Version): Data for
Two Testing Probes for Canadian Normal Controls
MIF Mean Mean PASAT Visit 2•
n Ratio Age Education Trials VISit 1 (Retest)
26 2016 29.7 13.2 2.4 sec 47.1 51.3

(12.4) (3.0) (10.9) (11.0)
2.0 sec 41.8 47.2
(11.7) (11.9)
1.6 sec 36.1 40.2
(11.8) (12.3)
1.2 sec 26.3 29.3
(10.2) (12.0)
•Test-retest interval was approximately 1 week.

696 APPENDIX 8
Table A8.7. [PASAT.6] Rao et al., 1991a (Gronwall's Administration Version):

Data for a Normal Control Sample
PASAT Trials
n MIF Ratio Age Education Premorbid IQ• 3 sec 2 sec
100 'lJ5n5 46.0 13.3 107.2 48.5 37.1

(11.6) (2.0) (11.2) (9.6) (10.1)
• Barona et a!. (1984).
Table A8.8. [PASAT.7] Strauss et al., 1994 (Gronwall's Administration

Version): Data for a Control Sample of Canadian Undergraduate Students
PASAT Trials
n MIF Ratio Age Education 2.0 sec 1.6sec
10 4/6 23.7 15.21 43.6 35.1

(2.58) (0.79) (9.1) (9.1)
Table A8.9. [PASAT.8] Zalewski et al., 1994 (Gronwall's Administration

Version): Total Score for Trials 2.4 and 1.2 for a Control Sample of
Veterans•
n Age Education Total PASAT Score
241 38.0 13.6 111.16

(52.4)
•Ethnic composition: White (n = 189), Black (n = 35), Hispanic (n = 11), Other

(n=6).
Table A8.10. [PASAT.9] Crawford et al., 1998b (Gronwall's Administration Version): Data for a Sample of
Healthy Volunteers from United Kingdom Stratified into Three Age Groups
Age MIF Total PASAT WAIS-R NART-
Group n Ratio Age Education Score for 4 Trials FSIQ Errors
Total sample 152 11n5 40.21 12.97 151.6 105.0 18.0

(13.89) (2.86) (40.32) (14.08) (9.01)
16-29 38 'lJ5.00 169.2
(3.27) (30.12)
30-49 78 38.10 149.8
(5.67) (40.29)
50-74 36 60.70 136.9
(7.41) (43.79)
•National Adult Reading Test (Nelson, 1982).

APPENDIX 8 697
Table A8.11. [PASAT.10] Prevey et al., 1998 (Gronwall's Administration

Version): Data for Two Trials for the Control Sample
PASATTrials
n Age Education 2.4 sec 2.0 sec
45 44.4 12.8 31.0 30.0

(11.4) (1.9) (11.4) (10.8)
Table A8.12. [PASAT.ll] Holdwick and Wingenfeld, 1999 (Gronwall's Admin-

istration Version, Computerized): Data for a Sample of College Students•
PASAT Trials
n % Femalet 2.4 sec 2.0 sec 1.6 sec 1.2 sec
20 65 19.8 45.60 44.60 40.20 33.25

(3.85) (7.68) (8.56) (7.32) (7.74)
"Ethnic distribution: 82.5% Caucasian, 10% African American, 3.8% Hispanic, 3.8% other.
tvalue represents information for the entire sample, not just for the control group.
Table A8.13. [PASAT.12] Honn et al., 1999 (Gronwall's Administration Version): Data
for Three Trials for an HIV-Negative Male Control Sample Subdivided into Exerciser
and Non-Exerciser Groups
PASAT trials
Group n Age Education WAIS-R IQ 2.4 sec 2.0 sec 1.6 sec
Exercisers 38 31.1 15.0 107.7 42.3 40.3 32.5

(8.7) (2.2) (11.1) (12.5) (10.4) (9.4)
Non-exercisers 38 32.9 14.3 105.9 39.8 33.6 30.2
(9.9) (2.3) (14.2) (11.7) (10.9) (8.5)
Table A8.14. [PASAT.13] Wingenfeld et al., 1999 (Gronwall's Administration

Version, Computerized): Data for a Sample of College Students Stratified by Gender
and Two Age Groups
Gender Age Group
PASAT
Trials Male (n=80) Female (n = 88) 17-29 (n = 156) 30-48 (n = 12)
2.4 sec 48.37 45.46 47.30 41.00

(8.78) (10.10) (9.17) (12.88)
2.0 sec 44.33 41.01 42.83 39.42
(9.30) (10.53) (9.86) (12.50)
1.6 sec 38.25 36.09 37.42 33.25
(9.58) (10.19) (9.47) (14.72)
1.2 sec 30.47 27.66 29.29 25.25
(9.73) (10.02) (9.69) (12.86)
698 APPENDIX 8
Table A8.15. [PASAT.14] Bate et al., 2001 (Gronwall's Administration Version): Data
for an Australian Control Sample
PASAT Trials
MIF Verbal IQ
n Ratio Age Education (NART-R)• 2.4 sec 2.0 sec 1.6 sec 1.2 sec
35 20/15 30.2 12.6 101.1 43.1 40.8 36.7 29.5

(10.3) (2.0) (9.1) (12.0) (10.6) (10.1) (6.0)
•National Adult Reading Test-Revised (Crawford, 1992).
Table A8.16. [PASAT.15] Boringa et al., 2001 (Gronwall's Administration

Version): Data for Two Trials Collected in Amsterdam on a Sample of
Healthy Volunteers
PASAT Trials
n MIF Ratio Age 3.0 sec 2.0 sec
140 62178 45.8 48.7 38.2

(range 22-73) (10.7) (11.0)
Table A8.17. [PASAT.16] Fluck et al., 2001 (Gronwall's Administration

Version): Data for Healthy Volunteers Collected in London
Young Middle-Aged
Male Female Male Female
n 15 15 15 15
Age 21.1 20.9 57.5 60.3
(0.4) (0.2) (1.3) (0.7)
WAIS-R FSIQ 113.0 112.4 117.7 113.3
(1.5) (1.7) (1.8) (2.2)
PASAT Trials
2.4 sec 55.4 53.6 48.4 41.6
(1.3) (1.8) (2.5) (3.4)
2.0 sec 45.6 46.9 38.9 32.5
(2.3) (2.1) (2.3) (3 ..2)
1.6 sec 38.7 38.2 30.1 .24.4
(2.4) (2.8) (2.7) (.2.4)
1.2 sec .28.6 31.9 2.2.5 17.5
(2.7) (2.7) (2.1) (1.9)
APPENDIX 8 699
Table A8.18. [PASAT.l7] Snyder et al., 2001 (Gronwall's Administration

Version): Data for the Canadian Control Group
M/F Averaget Dyadt
n Ratio Age• Education• PASAT PASAT
35 9/26 37.97 14.1 35.53 11.66

(12.94) (2.3) (9.85) (6.17)
"Data obtained from initial study (Fisk & Archibald, 2001).

tTraditional scoring method in which the total correct responses for each trial are
summed and divided by 4.
:Based on the dyad scoring method in which pairs of correct responses were counted;
these scores were then summed and divided by 4.
Levin's Administration Version
Table A8.19. [PASAT.l8a] Brittain et al., 1991 {Levin's Administration version):

Demographic Characteristics of the Sample of Healthy Participants
Age Mean M/F Race Shipley
Group Age n Ratio White/Other Education IQ
<25 21.0 145 55190 79/66 13.0 105.0

(2.1) (1.3) (9.1)
25-39 31.4 164 67197 114150 14.0 103.0
(4.1) (2.2) (10.4)
40-54 46.2 95 50/45 79/16 13.0 101.0
(4.3) (3.1) (12.6)
>55 67.0 122 82182° 11913 12.0 106.0
(7.1) (2.5) (15.1)
•The authors report the gender ratio as 82/82, but this is likely a misprint given that n = 122.
700 APPENDIX 8
I
Table A8.20. [PASAT.18b] Brittain et; al., 1991 (Levin's Administration
Version): Data• ror a Healthy Sample Stfcttifl.ed by Four Age and Three IQ
Levels 1
I
Shipley IQ
PASAT '
Age Group Trial <94 90-109 >109
(n='h (n=89) (n=49)

<25 2.4 sec 15."* 10.79 16.63
(9.4f) (10.79) (7.24)
2.0 sec 18.7 14.84 11.06
(9.1 ) (8.66) (8.81)
1.6 sec 'JJJ.7 19.10 12.59
(8.7 ) (8.43) (8.28)
. 1.2 sec 29. 24.22 18.59
(7.2) (7.77) (7.52)
Total trials 83. 68.88 49.08
(29. ) (29.68) (28.23)
(n=1) (n=95) (n=S4)
25-39 2.4 sec 18. 9.78 6.37
(11. (8.16) (6.08)
2.0 sec 'JJJ.4 14.98 10.74
(10. (8.28) (6.81)
1.6 sec 24.8 19.26 13.31
(8. (8.47) (7.66)
1.2 sec 28.6 25.64 'JJJ.61
(5.4 (7.14) (7.18)
Total trials 92.~ 69.65 51.26
(31.9 (28.45) (24.13)
(n = 1-;t (n=47) (n=31)
2.4 sec 26.351 10.79 5.65
(9.8~ (9.35) (4.53)
2.0 sec 28.06: 15.15 9.97
(8.57) (8.48) (7.18)
1.6 sec 29.82 18.36 14.84
(6.64) (9.21) (7.48)
1.2 sec 32.76j 25.34 21.61
(6.90) (6.68) (7.13)
Total trials 117.00: 69.43 51.90
(28.39) (30.08) (23.26)
(n=18) (n=S4) (n=SO)
>54 2.4 sec 28.83: 21.78 14.24
(6.84) (10.45) (10.36)
2.0 sec 28.39: 23.61 16.54
(6.70~ (7.86) (9.32)
1.6 sec 28.50: 25.04 'JJJ.90
(4.87i (6.08) (8.50)
1.2 sec 32.441 29.26 25.12
(4.76~ (6.42) (7.22)
Total trials 118.06' 99.76 76.82
('JJJ.37)~ (27.01) (31.59)
•Mean error scores.

APPENDIX 8 701
Table A8.21. [PASAT.19] Roman et al., 1991 (Levin's Administration

Version): Raw Scores and Percent Correct for Three Adult Age Groups
Young Adult Middle-Aged Older Adult
(18-27 years) (33-50 yean) (60-75 years)
n 62 40 41
%female 58 50 51
Age 19 40 69
(1.60) (5.1) (4.1)
Education 12 15 15
(0.77) (2.6) (3.2)
Estimated IQ• 110 110 107
(12.30) (12.3) (11.0)
PASATnawacora
2.4 sec 45 44 37
(4.3) (4.9) (9.1)
2.0 sec 39 38 31
(7.8) (7.6) (9.2)
1.6sec 36 33 27
(7.7) (8.9) (8.5)
1.2sec 28 28 20
(6.7) (9.3) (6.1)
Total (4 trials) 148 144 115
(23.5) (27.0) (29.9)
PASATpen:entcorrect
2.4 sec 91.68 90.70 75.95
(8.9) (10.3) (18.6)
2.0 sec 79.53 77.50 63.04
(16.2) (15.3) (18.9)
1.6sec 72.79 68.22 54.95
(15.9) (18.2) (17.4)
1.2 sec 57.92 57.22 41.63
(13.8) (18.9) (12.2)
Total (4 trials) 75.00 73.00 59.00
(12.0) (14.0) (15.3)
•Prorated IQ using the Vocabulary and Block Design subtests of the WAIS-R.
Table A8.22. [PASAT.20] Cicerone, 1997 (Levin's

Administration Version): Data for a Control Sample
Total PASAT
n Age Education Score for 4 trials
40 33.3 14.9 144.0
(12.4) (2.2) (27.0)
702 APPENDIX 8
Table A8.23. [PASAT.21a] Wiens et al., 1997

(Levin's Administration Version): Data for a Sample Age Group
WAIS-R PASAT
of Healthy Participants Stratified by Gertder
FSIQ Trials 20-29 30--39 40-49
Male Female
(n= 192) (n=94) (n =11)
n 672 149 100-109 2.4 sec 44.2 41.8 38.4
(4.4) (6.5) (8.8)
Age 29.2 29.2 2.0 sec 40.4 37.8 36.4
(6.1) (5.6) (6.3) (7.1) (10.6)
Education 14.6 14.5 1.6 sec 35.6 34.1 33.7
(1.5) (1.6) (7.1) (8.0) (10.8)
WAIS-R FSIQ 106.6 105.4 1.2 sec 29.9 28.5 26.4
(11.0) (11.1) (6.8) (7.8) (10.9)
Total correct 150.1 142.2 135.1
PASAT trials (20.5) (25.0) (38.7)
2.4 sec 43.2 43.5 (n=95) (n =72) (n = 11)
(5.6) (5.8) 110-119 2.4 sec 44.7 43.9 41.9
2.0 sec 39.3 39.1 (4.4) (4.8) (7.2)
(7.3) (6.8) 2.0 sec 40.8 39.4 39.6
(6.8) (7.5) (7.0)
1.6 sec 35.1 34.9
(8.1) (8.1) 1.6 sec 37.6 36.4 35.0
(7.2) (7.0) (7.5)
1.2 sec 29.2 27.8 1.2 sec 31.3 30.0 27.9
(7.9) (7.3) (7.3) (7.4) (8.2)
Total trials 146.8 145.2 Total correct 154.5 149.7 144.4
(25.2) (24.5) (22.8) (23.5) (27.2)
(n =33) (n=44) (n = 12)
120-129 2.4 sec 46.9 44.8 47.2
(2.5) (5.4) (2.3)
2.0 sec 43.7 40.7 43.7
Table A8.24. [PASAT.2lb] Wiens et al., 1997 (4.4) (6.7) (3.6)
(Levin's Administration Version): Data fot a Sample 1.6 sec 40.3 38.5 39.7
of Healthy Participants Stratified by Age and IQ (6.4) (7.0) (7.6)
1.2 sec 35.1 31.7 32.1
Age Group (7.2) (7.4) (7.6)
WAIS-R PASAT Total correct 166.0 155.6 162.7
FSIQ Trials 20-29 30-39 40-49 (18.4) (22.4) (17.0)
(n=27) (n=10) (n= 12) (n=7) (n=4)
80-89 2.4 sec 42.1 40.9 ~130 2.4 sec 48.0 46.0 45.0
(5.6) (4.8) (1.4) (3.9) (3.7)
2.0 sec 35.7 35.1 2.0 sec 46.5 39.7 43.8
(6.1) (6.8) (2.2) (6.8) (6.6)
1.6 sec 28.2 31.1 1.6 sec 42.2 38.1 35.0
(9.1) (8.5) (6.3) (6.8) (5.2)
1.2 sec 24.3 24.9 1.2 sec 35.2 30.7 28.8
(7.4) (7.0) (5.3) (6.8) (4.1)
Total correct 130.3 132.0 Total correct 171.8 154.6 152.5
(22.4) (24.0) (13.4) (22.8) (16.1)
(n= 116) (n=72) (n=7)
90-99 2.4 sec 42.0 40.8 33.7
(5.5) (6.0) (16.3)
2.0 sec 37.6 36.1 32.7
(6.9) (8.2) (11.8)
1.6 sec 32.5 31.6 27.1
(7.5) (9.4) (10.5)
1.2 sec 26.1 24.6' 19.0
(7.0) (7.9) (6.5)
Total correct 138.2 133.1 112.6
(22.9) (27.9); (39.9)
APPENDIX 8 703
Table A8.25. [PASAT.22] Tiersky et al., 1998 (Levin's Administration

Version): Data for an All-Female Control Sample
n Age Education Total PASAT Score
20 37.1 15.00 143.40

(2.4) (0.55) (5.08)
Table A8.26. [PASAT.23] Stein et al., 2002 (Levin's Administration

Version): Data for an All-Female Control Sample
n Age Education Total PASA,.. Score
22 29.4 13.9 124.4

(10.7) (1.5) (29.8)
•Mean and SD based on n=20.
Table A8.27. [PASAT.24] Diamond et al., 1997 (Levin's Administration Version): Data for a Control
Sample
PASAT
Verbal IQ
n Age Education (NAART)• 2.4 sec 2.0 sec 1.6sec 1.2 sec Total
22 40.9 15.4 113.6 42.0 38.0 33.0 27.0 140.0

(8.9) (2.2) (13.0) (7.3) (8.5) (8.0) (7.9) (29.0)
•North American Adult Reading Test (Blair ~ Spreen, 1989).

704 APPENDIX 8
PASAT-50, PASAT-100 and PAS.t\II'-200 Administration Versions
Table A8.28. [PASAT.25a] Diehr et al.,l998 (PASAT-200 Administration Version):

Data for a Control Sample Stratified by Ethnicity, Education, and Age Groupings
Age Group
n Educatiqn 20-34 35-49
Etludclty-
African American 172 13.2 111.5 106.4 92.8
(28.3) (29.7) (26.0)
(n=77) (n=60) (n=35)
Caucasian 132 13.3 135.1 123.5 109.9
(26.0) (27.7) (30.1)
(n=51) (n=47) (n=34)
Education level
9-11years 70 99.3 97.9 84.3
(27.2) (33.6) (31.7)
(n=24) (n=26) (n=20)
12-15 years 304 120.9 113.9 101.2
(29.7) (29.9) (29.2)
(n= 128) (n=107) (n=69)
16-20 years 192 129.4 128.9 120.8
(27.8) (35.2) (31.2)
(n=64) (n=SO) (n=48)
• A smaller portion of the sample who had 12-45 years of education was reported.
Table A8.29. [PASAT.26] Diehr et al., 2003 (PASAT-50, PASAT-100, and PASAT-200 Administration
Versions): Data for the Control Sample
%Caucasian/ PASAT-50 PASAT-100 PASAT-200

n Age Education African Anierican (3 sec) (3 and 2.4 sec) (3, 2.4, 2.0, and 1.6 sec)
560 39.7 14.2 45/sS 37.4 68.4 115.9

(12.1) (2.6) (9.6) (18.0) (32.6)
for the Cancellation Tests
Study numbers and page numbers provided in Chapter 9. Locator table also provides a re-
these tables refer to study numbers and ference for each study to a corresponding data
descriptions of the studies in the text of table in this appendix.
Table A9.1. Locator Table for the Cancellation Tests

Study Age• n Sample Composition IQ!Education• Location
RUFF.2&7.1 Ruff 16-24 360 180M, 180 F; no <12 California,

& Allen, 1996 25-39 exclusion 13-15 Michigan
page 164 40-45 criteria provided 16
Data are not 55-70
reproduced
in this book
RUFF.2&7.2 Ruff 16-24 68 107 M,152 F; ~12 California,
et al., 1986a 25-39 83 no exclusion 13-15 Michigan
page 165 40-45 60 criteria provided 2::16
Table A9.2 55-70 48
RUFF.2&7.3 Ruff 31.2 60 Normal controls; 12.9 California, Michigan,
et al., 1992 (4.1) no neurological (1.5) New York
page165 or psychiatric
TahleA9.3 condition
RUFF.2&7.4 Bate 30.2 35 20M, 15 F; participants 12.6 Australia
et al., 2001 (10.3) were native English (2.0)
page 166 speakers; no neurological
Table A9.4 or psychiatric conditions
(continued)
705
706 APPENDIX 9

Study Age• n ~ Sample Composition IQ/Education• Location
DVf.l Heaton 20-34 200 bvr data in 7-8 California,

et al., 1991 & 35-39 1~1 manual; 9-11 Washington DC,
2004 40-44 !ovrdatain 12 Colorado, Texas,
page 166 45--49 manual; 13-15 Oklahoma,
Data are not 50-54 ly half were 16-17 Wisconsin, Illinois,
reproduced 55-59 ~can-American 18-20 Michigan, New York,
in this book 60-64 ~halfwere Vuginia.
65-69 ucasian in 2004 Manitoba (Canada)
70-74 sakple; no neurological or
75-79 $hiatric conditions;
80-85
;by. . . .
a
are presented
scores
cation, and
for African-
~erican and
DVf.2 Prigatano
et al., 1983
page 167
59.6
(9.0)
25 H £-w.,.-e
ely
adult volunteers
screened
~illnesses that
10.5
(3.3)
Wmnipeg.
Canada;
Oklahoma
TableA9.5 t interfere
~their neuropsychological
g. COPD, medications
eart or lung disease,
DVf.3 Grant
et al., 1987
63.1
(10.3)
99 75
p
i
an11 diabetes
24 F; no neurological or
hiatric conditions
10.2
(3.6)
USA
page 168
TableA9.6
DVf.4 Kelland & 20.0 20 10 ~. 10 F; college students; 13.1
Lewis, 1994 (2.8) nti neurological or (1.3)
page 168 psf<:hiatric conditions
0
Table A9.7
DVf.5 Barncord &
Wanlass, 1999
page 169
Table A9.8
19.80
(3.95)
10
""£' """"" """"""'

n exclusion criteria
PI "ded
I
12.80
(0.63)
DVf.6Smith 65.0 16 All-f-tmale sample; 15.0 Michigan

et al., 2001 (4.0) ~tmenopausal; (2.0)
page 169 67.0 13 nor neurological or 16.0
Table A9.9 (6.0) p hiatric (3.0)
ditions; data
rted for
enopausal
en on HRT and
not on HRT
DVf.7 Stein 29.4 22 13.9 California
et al., 2002 (10.7) (1.5)
page 169
TableA9.10
•Age column and IQ/education column con · information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, which r information is provided by authors.
APPENDIX 9 707
Table A9.2. [RUFF 2&7.2] Ruff et al., 1986a: Data for a Sample of Healthy Volunteers Stratified into
Three Age and Three Education Groups
Education (years)
:512 13-15 :2:16
Age Group x MIF MIF MIF

Condition n Ratio Score n Ratio Score n Ratio Score
Automatic Defection•
16-24 32 16/16 148.9 26 13113 164.1 10 515 165.1
(29.8) (28.9) (29.4)
25-39 26 10/16 142.8 30 11/19 156.3 27 12/15 153.5
(27.1) (23.5) (40.2)
40-45 18 5/13 131.7 17 7/10 132.3 25 11/14 152.2
(26.9) (29.6) (31.9)
55-70 14 3111 124.7 14 3111 131.3 20 11/9 137.1
(21.8) (27.5) (24.2)
ControllBd Search•
16-24 32 16/16 134.2 26 13113 142.0 10 515 143.7
(24.6) (24.9) (20.6)
25-39 26 10/16 126.4 30 11/19 140.3 27 12/15 138.1
(24.9) (20.5) (22.1)
40-45 18 5/13 119.2 17 7110 122.4 25 11/14 131.9
(19.5) (22.1) (21.3)
55-70 14 3/11 113.7 14 3111 122.0 20 11/9 123.0
(18.4) (21.7) (19.2)
•capital letters are used as distractors in the Automatic Detection condition and digits other than 2 and 7 are used as
distractors in the Controlled Search condition.
Table A9.3. [RUFF 2&7.3] Ruff et al., 1992: Speed and Accuracy Data for
a Sample of Healthy Volunteers
n Age Education Speed• Accuracl
60 31.2 12.9 284.4 94.4

(4.1) (1.5) (47.2) (4.7)
"Speed: sum of hits for both conditions (i.e., digit-digit and digit-letter conditions).
t Accuracy: (total hits for both conditions- total errors for both conditions)/(total hits
for both conditions).
Table A9.4. [RUFF 2&7.4] Bate et al., 2001: Data for a Sample of Australian Volunteers
MIF Verbal IQ Digits Correctly
n Ratio Age Education (NART-R)• Cancelled
35 20115 30.2 12.6 101.1 262.5

(10.3) (2.0) (9.1) (44.7)
•National Adult Reading Test-Revised (Crawford, 1992).

708 APPENDIX 9
Table A9.5. [DVI'.3] Prigatano et al., ~983: Data for a Sample of

Healthy Volunteers
n Age Education! DVI"' Total TIDle
25 59.6 10.5 390.9

(9.0) (3.3) (100.0)
•ovr data reported for 21 participants.
Table A9.6. [DVI'.4] Grant et al., 1987: pata for a Sample of Nonpatient
Control Volunteen;
MIF DVfTotal
n Ratio Age Education Time
99 75124 63.1 10.2 424.6

(10.3) (3.6) (100.6)
Table A9.7. [DVI'.S] Kelland and Lewis,i1994: Data for a Sample of College Students
Standard Form Alternate Form
n MIF Ratio Age Week 1 Week2 Week 1 Week2
20 10/10 20.0 ~.1 314.62 289.00 309.79 284.42

(2.8) {1.3) (57.20) (49.10) (52.30) (43.30)
Table A9.8. [DVI'.6] Barncord and WanJa5s, 1999:

Data for a Sample of College Students Table A9.10. [DVI'.8] Stein et al., 2002: Data for
an All-Female Sample
ovr
MIF Total DVfTotal
n Ratio Age Education Items n Age Education Time
10 515 19.80 12.80 t 360.20 22 29.4 13.9 350,4

(3.95) (0.63) : (23.22) (10.7) (1.5) (80.2)
Table A9.9. [DVI'.7] Smith et al., 2001: Data for

Postmenopausal Females on HRT and ·not on
HRT•
DVI'Total
Groups n Age Education Errors
HRT 16 65.0 15.0 I 4.63
(4.0) (2.0) I (4.1)
NoHRT 13 fj{,O 16.0 12.58
(6.0) (3.0) (13.1)
•HRT, hormone replacement therapy.

for the Boston Naming Test (BNT)
Study numbers and page numbers provided in ter 10. Locator table also provides a reference
these tables refer to study numbers and de- for each study to a corresponding data table in
scriptions of the studies in the text of Chap- this appendix.
Table A10.1. Locator Table for the Boston Naming Test (BNT)
IQ/
Study Age" n Sample Composition Education• Location
BNT.l Van Gorp 59-95 78 Normal, FSIQ: 122 Los Angeles, CA

et al., 1986 independently
page 183 59-64 12 living elderly (29M, 49F)
Table A10.2 ~ 20
70-74 24
75-79 13
;:::so 9
BNT.i Farmer, 1990 20-69 125 Neurologically intact 8-22 California
page 183 43.9 males, data are presented 14.6
Table A10.3 (14.3) for 5 age decades (2.2)
20-29 25
30-39 25
40-49 25
50-59 25
60-69 25
BNT.3 Boone 63.06 110 Control group (52M, 58F), High IQ Los Angeles, CA
et al., 1995 (9.19) rigorous exclusion criteria 14.84
page 184 (2.61)
Table A10.4
(continued)
709
710 APPENDIX 10
Sample IQ/
Study Age• n Composition Education• Location
BNT.4 Mitrushina & 57-$ 122 Normal, independently 14.1 Los Angeles, CA
Satz, 1995 living elderly (49M, 73F) (2.7)
page 184 57-65 19 Test-retest data over 3 FSIQ: 118.2
Tables A10.5, A10.6 ~70 40 annual probes are provided (13.0)
71-75 47
76-$ 16
BNT.5 Neils 65-97 323 Neurologically intact volunteers 6-9 Northern
et al., 1995 (244 F, 79 M); 167 10-12 Kentucky,
page 185 65-74 were living independently, >12 Cincinnati, OH
Table A10.7 75-84 156 were institutionalized in
85-97 extended-care facilities; data is
presented in age-by-education-
by-living environment matrix
BNT.6 Ross et al., 1995 123 Geriatric medical inpatients from 11.3 Michigan
page 185 70-74 40 an urban rehab hospital with a (3.1)
Table A10.8 75-79 40 variety of physical diagnoses, some
;::so 43 of whom were 2-3 weeks
post-orthopedic surgery
BNT.7 Worrall 70,43 136 Independently living, Australia
et al., 1995 (7.8) neurologically intact elderly
page 186 (74.3% F); interrater reliability
Table A10.9 55-59 7 data are provided; error
60-64 29 analysis was performed
65-69 35
70-74 31
75-79 14
80-84 13
;::85 7
BNT.8 LaHeche & 76.2 20 Control group (9M, llF), 14.7 Massachusetts
Albert, 1995 nondemented elderly
page 186
Table A10.10
BNT.9 Ivnik et al., 1996 663 Normal elderly volunteers, Mayo Minnesota
page 187 56-59 tables for age correction FSIQ:
Table AlO.ll 60-64 and a regression equation 106.2
65-69 for education correction; (14.0)
70-74 tables are not reproduced
75-79
80-84
85-89
90-94
;::95
BNT.IO Welch 74.0 176 Neurologically intact 12.28 Middle
et al., 1996 volunteers (74M, 102F) (3-18) Tennessee
page 188 60-64 representative of the regional
Tables Al0.12-Al0.14 65-69 population across most
70-74 demographic parameters;
75-79 data are presented for 5 age
80-93 groups, age x education,
and age x gender cells;
suggested cutoffs
are presented
APPENDIX 10 711
Sample IQ/
Study Age" n Composition Education• Location
BNT.ll Hoff 32.1 54 Paid male volunteers, 15.4 New York

et al., 1996 (9.7) strict selection criteria (2.4)
page 188
Table A10.15
BNT.ll Ponton 38.4 300 Spanish-speaking 10.7 Los Angeles, CA
et al., 1996 (13.5) healthy volunteers; (5.1)
page 189 MIF ratio 40/60%;
Table A10.16 16-29 data are partitioned by <10
30-39 gender (2) x age (4) x >10
40-49 education (2);
50-75 30-item Spanish version
was administered
BNT.l3 Tombaugh &: 25-88 219 Community-dwelling volunteers 12.9 Canada
Hubley, 1997 25-34 22 (46% male) with no known (2.3)
page 189 35-44 28 history of neurological or
Tables A10.17, A10.18 45-54 33 psychiatric illness, head injury,
55-59 24 or stroke
60-M 19
65-69 22
70-74 18
75-79 24
80-88 29
BNT.l4 Henderson 17-87 100 50 African-American and 10-17 South Carolina
et al., 1998 50 Caucasian participants,
page 190 with 25M, 25F in each group;
Table A10.19 rigorous exclusion criteria
BNT.l5 Stuss 54.4 37 Control group (19M, 18F) 13.9 Canada
et al., 1998 (14.4) without neurological or (2.3)
page 190 psychiatric disorder
Table Al0.20
BNT.l6 Fastenau 72.2 108 Nonneurological, community- 14.1 Indiana
et al., 1998 (7.0) dwelling adults (47% female); (2.4)
page 191 rigorous exclusion criteria
Table Al0.21 57--*3 35
69-76 38
77-85 35
BNT.l7 Ross &: 76.1 233 Neurologically intact medical 11.1 South Carolina
Uchtenberg, 1998 (7.1) patients from an urban hospital (3.2)
page 191 (73% female), data are presented
Table A10.22 65-69 in age-by-education cells <12
70-74 ~12
75-79
80-84
85-95
BNT.l8 Kahnert 20.82 100 Bilingual young adults of 14.4 San Diego, CA
et al., 1998 (2.6) Mexican-American descent (1.7)
page 192 recruited from Univ. CA San Diego,
Table A10.23 Univ. CA Santa Barbara,
and San Diego community, with
Spanish as the primary language;
test was administered in both
Spanish and English
(continued)
712 APPENDIX 10

Sample IQ/
Study Age• n Composition Education• Ux:ation
BNT.l9 Randolph 73.6 719 Neurologically nonnal elderly 13.4 Minnesota

et al., 1999 (10.3) (60% female), almost (2.9)
page 193 exclusively white; data are
Table A10.24 stratified by gender and <12
broken down by age group
based on overlapping 12
midpoints technique;
data are also presented by >12
education
BNT.20 Killgore & 45.7 62 Patients referred for 13.1 Massachusetts,
Adams,l999 (15.1) neuropsychological (2.7) Oklahoma
page 193 evaluation, which yielded FSIQ
Table Al0.25 no evidence of neurological 95.1
impainnent (28M, 34F) (12.0)
BNT.il Heaton 68.71 96 Healthy community residents 13.5 San Diego, CA
et al., 1999 (5.47) (51.8% male), rigorous exclusion (2.45)
page 194 60-80 criteria
Table Al0.26
BNT.U Rosselli 61.3 18 Monolingual Spanish-speaking 13.3 Florida
et al., 2000 (8.1) (4M,14F), (4.8)
page 194 63.4 45 Monolingual English-speaking 16.6
Table Al0.27 (10.1) (15M, 30F), (2.4)
60.6 19 Bilingual, healthy participants 14.5
(9.7) (9M, IOF); data are reported (3.6)
for three language groups
BNT.23 Schmitter- 18-22 26 Healthy students and community- 13.38
Edgecombe et al., 2000 dwelling individuals (0.75)
page 194 58-74 26 (7M, 19F in each age group) 17.58
Table Al0.28 (2.23)
75-93 26 16.65
(3.03)
BNT.24 Saxton 73.63 357 Elderly volunteers who 13.23 Washington
et al., 2000 (4.45) participated in the (2.85) County, MD
page 195 multicenter Memory and Pittsburgh, PA
Table Al0.29 Aging Study (44.9% male)
BNT.25 Bell 34.4 29 Sample included friends, 13.0 WJSClODSin

et al., 2001 (12.5) relatives, and spouses of (1.7)
page 195 temporal lobe epilepsy
Table A10.30 patients (28% male)
BNT.26 Roberts 37.1 42 Monolingual English speakers 16.29 Florida, Canada

et al., 2002 (8.99) (3.05)
page 196 39.6 32 Spanish/English bilinguals 17.91
Table A10.31 (6.59) (3.38)
34.9 49 French/English bilinguals; 15.0
(8.14) data reported for (2.93)
3 language groups
APPENDIX 10 713

Sample IQ/
Study Age" Composition Education• Location
"
BNT.27 Coffey 74.85 320 Neurologically healthy elderly 12.98 Pittsburgh, PA
et al., 2001 (4.95) from Cardiovascular Health (2.87) Hagerstown, MD
page 196 Study (38% male); data are
Table A10.32 reported for males and females
separately
BNT.IS Giovannetti 23.2 31 21M, lOF healthy adults Education: New York,
et al., 2003 6.07 recruited from a medical 15.0 Pennsylvania
page 197 center community (1.48)
Table A10.33 FSIQ:
109.3
(11.51)
Table A10.2. [BNT.1] Van Gorp et al., 1986: Data for a Sample of Healthy Elderly
Stratified into Five Age Groups
Age Groups
59-64 65-69 70-74 75-79 80+
n 12 20 24 13 9
Education 13.58 14.40 15.25 14.23 15.23
(2.37) (2.38) (3.30) (3.58) (4.59)
WAIS-R Verbal IQ 122 123 130 115 118
BNT score 56.75 55.60 54.46 51.69 51.56
(3.05) (4.29) (5.17) (6.20) (7.00)
Cutoff 51 47.12 44.12 39.29 37.56
Table A10.3. [BNT.2] Farmer, 1990: Data for Five Age Groups and the
Entire Sample of Healthy Adult Males
Age Group n Age Education BNT Scores
20-29 25 24.08 14.88 56.04

(2.53) (1.67) (3.60)
30-39 25 33.92 15.04 57.04
(3.35) (2.32) (2.25)
40-49 23 44.60 15.12 57.76
(2.50) (1.94) (2.19)
50-59 25 53.56 14.40 58.24
(2.77) (2.26) (1.88)
60-69 25 63.68 13.64 58.28
(2.93) (2.50) (3.19)
All 125 43.97 14.62 57.47
(14.31) (2.19) (2.79)
714 APPENDIX 10
Table A10.4. [BNT.3] Boone et al., 1995: Data for the Control Group
n Age Education Men Women VIQ BNT Scores
110 63.06 14.84 52 58 115.78 54.97

(9.19) (2.61) (14.18) (6.30)
Table A10.5. [BNT.4a] Mitrushina and Satz, 1995: Demographic Characteristics for the
Total Sample of Healthy Elderly and for Four Age Groups•
Age Groups
All 57-65 66-70 71-75 7&-85
Education 14.1 14.4 13.7 14.5 14.0

(2.7) (2.0) (1.8) (3.1) (3.6)
Age 70.4 62.6 68.2 72.9 78.3
(5.0) (2.5) (1.2) (1.4) (2.5)
WAIS-R FSIQ 118.2 115.0 119.4 119.9 114.4
(13.0) (12.1) (15.2) (11.3) (12.3)
n 122 19 40 47 16
"The sample included 49 males and 73 females.
Table A10.6. [BNT.4b] Mitrushina and Satz, 1995: Boston Naming

Test Scores for Four Age Groups and for the Total Sample of Healthy
Elderly Over Three Testing Probes
Age Group Time 1 Time2 Time3
57-65 56.0 56.2 56.0

(3.3) (2.8) (2.4)
66-70 56.1 56.0 56.1
(3.1) (2.9) (2.9)
71-75 53.7 54.6 54.2
(7.3) (5.3) (6.9)
7&-85 51.2 51.1 51.4
(7.3) (8.6) (7.9)
Total 54.5 54.7 54.8
(5.9) (6.2) (5.7)
APPENDIX 10 715
Table A10.7. [BNT.S] Neils et al., 1995: Data for the Elderly Sample Stratified
Into Three Age Groups, Three Educational Levels, and Two Living Environment
Settings: Noninstitutionalized/Institutionalized
Educational Level
Age Group 6-9 10--12 12+ All
Noninttitutiontllized
(n= 167)
65-74 47.58 53.00 53.10 51.83
(6.14) (6.63) (6.55) (6.77)
n=12 n=22 n=20 n=54
75-84 42.79 50.73 48.55 47.54
(10.99) (5.72) (7.96) (8.89)
n=19 n=22 n=20 n=61
85-97 36.00 45.53 49.88 43.75
(12.46) (10.70) (7.19) (11.72)
n=17 n=19 n=16 n=52
Total 41.58 49.95 50.55
(11.36) (8.29) (7.40)
n=48 n=63 n=56
I~
(n=l56)
65-74 35.14 46.95 49.54 44.09
(6.77) (8.78) (6.42) (9.59)
n=14 n=l9 n=13 n=46
75-84 36.90 39.95 48.30 41.82
(11.84) (10.05) (6.62) (10.71)
n=19 n=19 n=20 n=58
85-97 34.53 38.11 40.20 37.40
(9.78) (7.48) (7.62) (8.60)
n=19 n=18 n=15 n=52
Total 35.56 41.73 46.10
(9.80) (9.51) (7.87)
n=52 n=56 n=48
Table A10.8. [BNT.6] Ross et al., 1995: Data for a Sample of Geriatric
Medical Inpatients From an Urban Rehabilitation Hospital with a Variety of
Physical Diagnoses, Some of Whom Were 2--3 Weeks Post-Orthopedic
Surgery
Age n Education DRS• Scores BNT Scores
70-74 40 11.3 133.2 43.1

(3.1) (4.5) (11.7)
75-79 40 10.6 133.4 40.1
(3.3) (4.8) (10.9)
~80 43 10.2 131.4 35.8
(3.2) (4.8) (11.3)
•DRS, Depression Rating Scale.

716 APPENDIX 10
Table A10.9. [BNT.7] Worrall et al., 1995: Data Table A10.11. [BNT.9] Ivnik et al.,
for a Sample of Healthy Older Australians 1996: Demographic Description of the
Sample Partitioned into Groups Used
Age BNT BNT Recommended
in Boston Naming Testing
Groups n Score Range Cutoff
Agegroupa n
55-59 7 52.57 47-57 46.37
{3.10) 56-59 55
60-64 87
60-64 29 53.65 36-00 42.45
65-69 79
{5.60)
70-74 85
65-69 35 54.17 39-59 45.23 75-79 125
{4.47) 80-84 132
70-74 31 52.29 34-00 39.53 85-89 71
{6.38) 90-94 24
95+ 5
75-79 14 49.43 32-60 33.41
{8.01) Etluct.mon
80-84 13 47.46 33-58 32.38 ~7 4
{7.54) 8-11 103
85+ 7 47.14 39-57 34.90 12 218
{6.12) 13-15 163
16-17 115
2:;18 60
Gender
Males 263
Females 400
Table A10.1 0. [BNT.S] Lafleche and Albert, 1995: Bace

Data for the Control Sample Caucasians 662
Blacks 1
n Age Education MIF Ratio BNT Scores
Handetlneu
20 76.2 14.7 9/11 50.75 Right 607
{7.40) Left 27
Mixed 29
Total 663
Table A10.12. [BNT.10a] Welch et al., 1996:

Boston Naming Test Scores for Five Age Groups•
of Healthy Elderly
Age Group n BNT Score Cutoff
60-64 20 51.6 45
{5.4)
65-69 37 53.4 45
{4.7)
70-74 40 50.1 42
{8.5)
75-79 28 42.9 35
{12.4)
80-93 51 44.7 35
{9.6)
•The sample included 74 males and 102 females with a

mean education of 12.28 years (range 3-18).
APPENDIX 10 717
Table A10.13. [BNT.10b] Welch et al., 1996: Boston Naming Test

Scores for Five Age Groups by Two Educational Levels
<12 years ~12 years
Age BNT BNT

Group n Score n Score
60-64 6 49.8 14 52.4

(5.4) (5.5)
65--69 5 49.2 32 54.0
(3.6) (4.5)
70-74 7 38.4 33 52.6
(12.3) (4.7)
75-79 17 36.6 11 53.4
(10.9) (4.9)
80-93 20 40.7 31 47.2
(11.3) (7.5)
Table A10.14. [BNT.lOc] Welch et al., 1996: Boston Naming Test

Scores for Five Age Groups for Males and Females Separately
Males Females
Age Group n M n M
60-64 10 54 10 49.2
(4.3) (5.6)
65--69 22 54.5 15 51.7
(3.7) (5.4)
70-74 11 52.9 29 49.0
(8.8) (8.2)
75-79 11 46.7 17 40.4
(12.0) (12.4)
80-93 31 45.7 20 43.9
(9.7) (9.7)
Table A10.15. [BNT.ll] Hoff et al., 1996: Data

for the Control Sample
n Age Education BNT Scores
54 32.1 15.4 56.3

(9.7) (2.4) (3.5)
718 APPENDIX 10
Table A10.16. [BNT.12] Ponton et al., 1996: Boston Naming Test Scores• for a Sample of 300 Spanish-
Speaking, Healthy Participants Stratified by Gender, Age, and Education
Age Group
16-29 30-39 40-49 50-75
Education (Years)
<10 >10 <UO >10 <10 >10 <10 >10
Males
n 11 25 13 18 12 17 18 6
x 22.09 22.52 22!39 26.56 22.92 27.82 24.11 27.17
(SD) (3.42) (3.28) (2l29) (2.68) (3.68) (1.94) (2.14) (2.56)
Femala
n 12 30 22 44 16 11 25 20
x 20.17 21.77 21.59 24.57 21.81 25.64 20.52 24.15
(SD) (2.29) (3.95) (2~) (2.82) (2.29) (1.57) (4.12) (2.85)
"Ponton-Satz 30-item Spanish version was ~tered.

I
---rable A10.17. [BNT.13a] Tombaugh ancl Hubley, 1997: Boston Naming

-!fest Scores for a Sample of Healthy AduJts Stratified by Age, Education,
and Gender '
n SR SR+SC SR+SC+PC•
Age
25-34 22 55.9 56.0 58.4
(2.8) (2.9) (2.5)
35-44 28 55.5 56.1 58.2
(3.9) (3.6) (2.2)
45-54 33 54.8 55.4 58.5
(4.1) (3.6) (1.8)
55-59 24 55.2 56.0 58.8
(3.6) (3.1) (1.3)
60-M 19 55.6 56.6 58.7
(3.5) (2.9) (1.7)
65-69 22 54.9 55.8 58.4
(3.9) (3.5) (1.7)
70-74 18 52.5 54.3 57.2
(4.6) (4.4) (2.7)
75-79 24 51.7 53.4 57.8
(5.5) (4.6) (1.9)
80-88 29 53.1 54.3 58.1
(4.0) (3.8) (1.6)
EtluctJtion
9-12 123 53.4 54.5 57.8
(4.4) (3.9) (2.1)
13-21 96 55.6 56.3 58.8
(3.7) (3.2) (1.7)
APPENDIX 10 719
n SR SR+SC SR+SC+PC"
Gender
Male 100 54.9 55.9 58.5
(4.3) (3.7) (1.8)
Female 119 53.9 54.8 58.0
(4.1) (3.7) (2.0)
Total 219 54.3 55.3 58.3

(4.2) (3.7) (2.0)
SR =Spontaneous Response; SR + SC =Sum of Spontaneous Response (SR)+ Stimulus

Cue (SC); SR + SC+PC =Sum of Spontaneous Response (SR) +Stimulus Cue (SC) +
Phonemic Cue (PC). Because scores on the Boston Naming Test are not normally
distributed, standard deviations should not be used to compute normative data.
© Swets & Zeitlinger, 1997.
Table A10.18. [BNT.l3b] Tombaugh and Hubley, 1997: Boston Naming Test Norms Expressed as
Percentiles for Age and Education
Age Group
25-69 70-88
Education (Years)
9-12 13-21 9-12 13-21 Total

Percentiles (n=78) (n=70) (n=45) (n=26) (n=219)
Sll
90 59 60 58 58 59
75 58 59 56 56 58
50 55 58 53 54 56
25 53 55 48 52 53
10 49 53 45 48 49
Sll+SC
90 59 60 59 59 60
75 58 60 58 58 58
50 56 58 55 56 57
25 54 56 52 53 54
10 51 53 47 49 51
Sll+SC+PC
90 60 60 60 60 60
75 60 60 39 60 60
50 59 60 58 59 59
25 57 59 57 58 58
10 56 58 54 57 56
DemognJplaica
Age 52.1 (12.2) 47.5 (12.4) 78.0 (4.8) 78.0 (4.7) 59.0 (16.9)
Education 11.3 (0.9) 15.1 (1.9) 11.2 (1.0) 14.9 (1.4) 12.9 (2.3)
Vocabulary• 51.0 (9.0) 57.2 (7.4) 53.3 (9.1) 58.6 (6.9) 54.3 (8.8)
%Male 41% 41% 56% 54% 46%
"Raw scores on the Wechsler Adult Intelligence Scale-Revised Vocabulary subtest. SR, spontaneous response; SC,
stimulus cue; PC, phonemic cue.
© Swets & Zeitlinger, 1997.
720 APPENDIX 10
Table A10.19. [BNT.14] Henderson et al., 1998: Table A1 0.21. [BNT.16] Fastenau et al., 1998: Data
Demographic Information for the Race by Gender for Three Age Groups and the Whole Sample• of
Subgroups of a Healthy Adult Sample and the Healthy Elderly
Mean Boston Naming Test Scores for the Race,
Age Mean % BNTt
Gender, and Education Groups
Group n Age Education Female Scores
n Age Education BNT
57-68 35 64.3 14.8 49 56.8
(3.2) (2.5) (2.8)
Caucaaiaru
69-76 38 72.2 14.1 47 54.0
Males 25 41.36 15.48
(2.3) (2.5) (4.4)
(17.95) (2.21)
Females 25 49.76 14.58 77-85 35 80.3 13.5 46 53.8
(19.74) (1.88) (2.4) (2.3) (5.4)
Total 108 72.2 14.1 47 54.8
African Americaru
sample (7.0) (2.4) (4.5)
Males 25 32.84 14.52
(13.88) (1.87) "The sample was predominantly Caucasian (95%).
Females 25 38.84 15.16 tBoston Naming Test administration procedure was not
(12.49) (1.95) standard: the test items were reorganized.
Race
African American 50 56.54
(5.40)
Caucasian 50 55.56
(4.34) Table A10.22. [BNT.17] Ross and Lichtenberg,
Gender 1998: Boston Naming Test Scores Stratified by Age
Female 50 56.12 and Education for Neurologically Intact Medical
(4.89) Patients from an Urban Hospital
Male 50 55.98
(4.95) Education
Education <12 ~12
Noncollege 23 51.91
Age n BNT Score n BNT Score
(6.37)
College 77 57.28 65--69 16 41.1 18 48.7
(3.58) (12.9) (7.8)
70-74 27 38.1 39 45.8
(11.3) (9.8)
75-79 26 37.0 36 44.9
Table A10.20. [BNT.15] Stuss et al., 1998: Data (10.5) (8.9)
for the Control Sample 80-84 22 36.4 15 43.4
(10.3) (11.9)
M/F NART" BNT
n Age Education Ratio IQ Score 85-95 17 29.6 17 41.7
(10.1) (11.7)
37 54.4 13.9 19/18 113.8 55.5
(14.4) (2.3) (6.1) (4.1)
"NART, National Adult Reading Test.

APPENDIX 10 721
Table A10.23. [BNT.l8] Kohnert et al., 1998: Data for the Spanish and English
Versions of the Test•
BNT Scorest
n Age Education MIF Ratio Spanish English
100 20.82 14.4 41.159 32.00 46.66

(2.6) (1.7) (8.83) (6.64)
15.23-48.78 34.06-59.26
• All participants were b11ingual, with Spanish as the primary home language, and a mean age
of English acquisition of 4.6 (3.0) years.
tBoston Naming Test scores with SDs and 95'11 confidence intervals.
Table A10.24. [BNT.l9] Randolph et al., 1999: Data for

the Sample of 719 Participants Stratified by Gender and
Broken Down into Age Groups Using Overlapping Midpoints
Technique
Males Females
n BNT Score n BNT Score
Age nange
50-60 53 56.4 45 55.0
(3.2) (4.3)
~ 67 55.9 80 55.3
(3.4) (3.7)
56-66 85 55.5 98 55.2
(3.4) (3.7)
59-69 93 55.5 109 54.5
(3.2) (3.8)
62-72 90 55.1 111 54.3
(3.7) (3.8)
65-75 99 54.9 112 53.8
(4.1) (4.3)
68-78 99 54.4 127 52.6
(4.2) (5.2)
71-S1 95 53.5 156 51.6
(5.3) (5.9)
74-S4 100 52.3 173 50.6
(5.9) (6.4)
77-S7 87 51.3 177 49.5
(6.4) (6.9)
80-90 65 50.9 157 48.8
(7.1) (7.4)
83+ 41 50.5 115 47.3
(6.7) (8.2)
Education (,_..) n BNTScore
<12 102 49.1
(7.0)
12 235 52.2
(6.1)
>12 382 53.4
(6.0)
722 APPENDIX 10
Table A10.25. [BNT.20] Killgore and Adams, 1999: Data for a Sample of
Patients Referred for a Neuropsychological Evaluation but Cleared of Neuro-
logical Impairment
WAIS-R BNT
n Age Education WF Ratio FSIQ Score
62 45.7 13.1 95.1 52.8
(15.1) (2.7) (12.0) (6.2)
Table A10.26. [BNT.21] Heaton et al., 1999: Data for a Healthy Elderly
Sample
% % BNT
n Age Education Male Caucasian Score
96 68.71 13.50 111.8 82.4 54.01
(5.47) (2.45) (4.43)
Table A10.27. [BNT.22] Rosselli et al., 2000: Data for Monolingual Spanish-Speaking.
Monolingual English-Speaking, and BilingUal Participants
BNT Scores
Language Group n Age £ducation MIF Ratio Spanish English

Spanish 18 61.3 13.3 4/14 51.1
(8.1) (4.8) (4.1)
English 45 63.4 16.6 15130 54.9
(10.1) (2.4) (4.8)
Bilingual 19 60.6 14.5 9/10 52.9 52.4
(9.7) (3.6) (6.1) (7.1)
Table A10.28. [BNT.23] Schmitter-Edgecombe et al., 2000: Data for a Sample of Healthy
Adults Stratified into Three Age Groups
Age Group n Age Educalion MIF Ratio FSIQ" BNT Scores
18-22 26 18.93 13.38 7/19 113.15 53.54

(0.7$) (7.88) (3.39)
58-74 26 66.29 17.58 7/19 114.58 57.50
(2.28) (9.60) (2.03)
75-93 26 79.19 16.65 7/19 115.04 55.35
(3.0$) (9.88) (3.53)
"Estimated Wechsler Adult Intelligence Scale-Revised full-scale IQ based on four subtests.

APPENDIX 10 723
Table A10.29. [BNT.24] Saxton et al., 2000: Data for a Sample of Free of
Cardiovascular Disease Elderly
n Age Education %Male BNT Score
357 73.63 13.23 44.9 52.95

(4.45) (2.85) (0.33) 0
"It is unclear whether this value represents SD.
Table A10.30. [BNT.25] Bell et al., 2001: Data for the Control Group
n Age Education %Male FSIQ" BNT Score
29 34.4 13.0 28 97.7 53.6

(12.5) (1.7) (6.4) (3.2)
"Wechsler Adult Intelligence Scale-III full-scale IQ based on seven-subtest short fonn.
Table A10.31. [BNT.26] Roberts et al., 2002: Data for Monolingual English
Speakers, Spanish/English Bilinguals, and French/English Bilinguals
Language Group n Age Education BNT Scores
English 42 37.1 16.29 50.9

(8.99) (3.05) (3.45)
Spanish! 32 39.6 17.91 42.6
English (6.59) (3.38) (8.04)
French! 49 34.9 15.0 39.5
English (8.14) (2.93) (7.43)
Table A10.32. [BNT.27) Coffey et al., 2001: Data for a Sample of Neurologically Healthy Elderly
MMSE" Vocabularyt BNT
n Age Education scores scores scores
Whole 320 74.85 12.98 28.29 47.52 51.08

sample (4.95) (2.87) (1.50) (13.26) (7.50)
Males 122 75.20 13.30 28.16 47.83 51.45
(5.36) (3.09) (1.53) (13.73) (8.69)
Females 198 74.63 12.78 28.38 47.33 50.86
(4.68) (2.91) (1.48) (13.99) (6.69)
"MMSE, Mini-Mental State Exam.

twAIS-R Vocabulary.
Table A10.33. [BNT.28] Giovannetti et al., 2003: Data for the Control Sample
n Age Education MIF Ratio WAIS-R FSIQ BNT Score
31 25.2 15.0 21110 109.3 54.7

(6.07) (1.48) (11.51) (3.7)
Appendix 1Om: Meta-Analysis Tables
for the Boston Naming Test (BNT)
Table A1Om.1. Results of the Meta-Analysis and Predicted Scores for the Boston
Naming Test
or
Number studies included in the analysis 14
Years or
publieation 1~2003
or
Number data points used in the
analysis 42
Total number of partieipants 1,684
Sample aize
Mean 42 29.51 27.31 7-173
Age
Mean 42 67.91 15.26 24.1-87.0
SD 41 2.53 2.50 1.(~14.4
Education
Mean 35 13.79 1.50 11.0-16.6
SD 34 2.64 1.04 0.5-4.0
IQ
Mean 6 116.10 2.60 113.8-119.9
SD 6 11.77 2.65 6.1-15.2
Percmt lllllle 20 79.93 34.34 0-100
Te81 acore mearu

Combined mean 42 52.25 3.26 45.7-58.3
Combined SD 42 5.92 2.03 1.9-9.7
724
APPENDIX 1OM 725
Predicted Seores and SDs per age group• (BNT)
95%CI 95%CI

Bange Score Band Band SD Band Band
.25-.29 55.71 54.93 57.00 3.08 2.47 3.69
30-34 56.37 55.60 57.14 3.09 2.71 3.47
35-39 56.76 55.93 57.59 3.18 2.78 3.58
40-44 56.89 55.97 57.81 3.35 2.82 3.88
45-49 56.75 55.77 57.74 3.61 2.98 4.24
50-S4 56.35 55.34 57.37 3.95 3.29 4.62
SIJ.-.59 55.68 54.69 56.68 4.38 3.74 5.02
60-64 54.75 53.82 55.69 4.89 4.32 5.46
~ 53.56 52.74 54.38 5.48 5.00 5.97
70-74 suo 51.42 52.77 6.16 5.65 6.67
75-79 50.37 49.85 50.89 6.92 6.18 7.66
80-84 48.39 47.91 48.86 7.77 6.64 8.91
Predkted ,_, ecore = 47.36842 + 0.4489501 • age- 0.0052924 • age2

Predkted SD = 4.542304- 0.0992503 • age+ 0.0016771 • ag~
Ordioary least square regressioa or test - on age (quadratic)

R2 0.850
F<dl)• p F<2.13l = 149.10, p<O.OOO
0.4489501 0.070 6.45• o.ooo• 0.299 to 0.599

-0.0052924 0.001 -8.46 0.000 -0.007 to -0.004
47.36842 1.567 30.23 0.000 43.98 to 50.75
•Significance test for age centered (sample means- aggregate mean): t= -14.74, p =0.000.
Prediction
Predicted age range 2&-84years
SEe 0.41
95%CI 53.66-55.28
(continued)
726 APPENDIX 1OM
Table A10rn.1. (Contd.)
60
0 0
0
55
50
0
~ ~---.-----.----~------~-----.-----.-----
30 40 50
••eo. 70 60
Figure A10m.1. A scatterplot illustrating the ~rsion of the data points around the regression line for the
Boston Naming Test The size of the bubbles ~re8ects the weight of the data point, with larger bubbles
indicating larger standard error and smaller we~t.
Tests for assumptions and model 8t
I
Tests for heteropneity In the 8aal data jlet
Pooled estimates mr fixed effect I
55.083
Pooled estimates fOr random effect I 53.961
p
Q(dO· P<41) = 465.64, p < 0.000
Moment-based estimate of 1
between-study variance ' 4.654
Tests for model &t-edclition of a ~ term
Model Adjusted~ BIC BIC'
Linear 0,611 0.602 28.180 -35.968

Quadratic 0.850 0.842 ' -8.091 -72.238
BIC' difference of 36.270 provides very strong s;pport for the quadratic model.
Tests for parameter speclflcatioas
Shapiro-Wilk W test W=0.914, p =0.442
Homoscedasticity
White's general test 2.397, f = 0.663
APPENDIX 1OM 727
OnliDary least-squares regression of SDs on age (quadratic)

R2 0.583
Fcdfl• p Fc2.I3l = 45.34, p<O.OOO
Term Coefficient SE p 95% CI
Age -0.0992503 0.089 -I.n• 0.287• -0.292 to 0.094

Age2 0.0016771 0.001 1.97 0.071 -0.000 to 0.004
Constant 4.542304 2.067 2.20 0.047 0.08 to 9.01
•Significance test for age centered (sample means -aggregate mean): t = 4.61, p< 0.000.
Predietion
SE., 0.31
95% CI 4.05-5.26
Education
R2 0.854
Education -0.050 0.136 -0.36 0.722 -0.35 to 0.25
IQ
Regression of test means and IQ on age
R2 0.961
Term Coefficient SE t p 95% CI
IQ 0.1061451 0.140 0.76 0.527 -0.495 to 0.707
Gender
t-test by gender
6M,6F 51.617 48.167 3.450 1.246 0.121

Appendix 11 : Locator and Data Tables
for the Verbal Fluency Test
Study numbers and page numbers provided in Locator tab] al o provid s a re~ rene for
these tables refer to study numbers and de- each stud to a corre pond.ing data table in thi
scriptions ofthe studies in the text of Chapter 11. appendix.
728
Table A11.1. Locator Table for the Verbal Fluency Test
>
-o
Study Age• n Sample Composition IQ!Education • Version Location -o
m
VF.l Cauthen, 1978b 20--59 51 Subjects in younger FSIQ: Letters Alberta, z
0
page 209 ~60 64 group gathered from 115.6 S, G, U,N, Canada
X
Table A11.2 different sources ( 12 M, (8.7) F, T,J, P
39 F), older subjects lived 111.5
primarily in institutional (13.1)
settings (28 M, 36 F)
VF.2 Yeudall et al., 1986 15-40 225 127 M, 98 F volunteers; Education: FAS Alberta,
page 210 data presented in 4 14.6 Canada
Tables All.3-A11.5 15--20 62 age groups for M and F (2.8)
21-25 73 separately and combined FSIQ:
26-30 48 118.6
31-40 42 (8.8)
VF.3 Gordon & 26.5 250 90 M, 160 F university RJ14.0 Letter Fluency Pittsburgh
Lee, 1986 students (1.0)
page 211
Table All.6
VF.4 Bolla et al., 39-89 199 80 M, 119 F volunteer 8-22 FAS Maryland
1990 64.3 participants in a study 14.7
page 211 (13.5) on normal aging; data (3)
Table A11.7 divided by 3 levels
of raw WAIS-R
Vocabulary scores
VF.5 Seines et al., 733 Subjects from MACS; <College (229) FAS, Animals Baltimore,
1991 seronegative homosexual Chicago,
page 212 25--34 309 males; data presented College (202) Los Angeles,
Table A11.8 35--44 290 for 3 age groups and 3 Pittsburgh
45--54 97 educational levels; mean >College (302)
education= 16 years
VF.6 Axelrod & 50-89 80 Healthy, independently Education: FAS Michigan
Henry, 1992 living volunteers
page 212 50--59 10M 15.4
Table A11.9 55.3 lOF (2.5)
(2.5)
60-69 10M 14.4
65.2 10 F (2.5) ....
N
(2.6) ID
(continued)
Table A11.1. (Contd.) ....
Study Age• n Sample Composition IQ!Education • Version Location
=
IIW
70-79 10M 14.5

74.3 10 F (4.2)
(2.9)
80--89 8M 14.5
83.4 12 F (4.1)
(3.0)
VF.7 Monsch 71.2 53 17 M, 36 F healthy 13.6 FAS San Diego,
et al., 1992 (7.9) elderly volunteers (2.7) 3 categories CA
page 212 (including
Table A1l.l0 Animals, Fruits
and Vegetables),
first names,
supermarket
VF.8 Simkins- 52.6 19 10M, 9 F healthy Education: FAS Michigan
Bullock, 1994 (15.6) volunteers 13.26 categories
page 213 (2.5) (includes
Table A1l.ll FSIQ: Animals and
102.0 Fruits or
(12.85) Vegetables)
VF.9 Parkin & 71.9 22 4 M, 18 F healthy Education: FAS UK
Lawrence, 1994 (4.8) elderly volunteers 9.4
page 213 (1.3)
Table A1l.l2 NART FSIQ:
106.1
(12.6)
VF.10 Friedman 35.8 24 Healthy volunteers FAS Ohio
et al., 1995 (11.0) recruited primarily
page 214 from hospital staff
Table A11.13
Education: FAS,
>
VF.ll Kozora & 50-89 174 Volunteers screened USA '"0
'"0
Cullum, 1995 50-59 41 for major medical and 14.3 (2.3) Animals, m
page 214 60-69 43 psychiatric disorders 14.2 (2.3) Supermarket, z
Tables A11.14, A11.15 · 70-79 47 14.3 (3.1) First names, 0
14.9 (3.3) U.S. States X
80-89 43
VF.12 Norris et al., 129 3 samples were used: FAS Texas >
-c
1995 60--86 54 I. Community elderly 16.7 -c
m
page 215 73.1 living independently (2.3) z
Table A11.16 (6.1) 0
62-89 35 2. Institutionalized 12.4 X
75.3 elderly with MMSE (3.7)
(7.5) scores ~20
18-28 40 3. Undergraduate students 13.6
19.4 (1.1)
(1.8) Interrater reliability
data are provided.
VF.13 Cahn et al., 78.4 238 Cognitively intact elderly 13.8 FAS California
1995 (6.8) participants of Rancho Bernardo (2.6)
page 216 Study; data for the entire
Table A11.17 sample and optimal cutoffs
are provided
VF.14 lvnik et al., 56 to 95+ 743 Normal elderly volunteers; Mayo COWA Minnesota
1996 tables for age correction FSIQ:
page 216 and a regression equation 106.2
Table A11.18 for education correction; (14.0)
data tables not reproduced
VF.l5 Ruff et al., 16-70 360 Native English speakers: Education COWA California,
1996 180 M, 180 F; data are 7-22 Michigan,
page 217 reported for 3 education groups: eastern
Tables A11.19, A11.20 16-24 90 groups, M & F separately; :::>12 seaboard
25-39 90 tables for data conversion 13--15
40-54 90 to T scores and percentile ~16
55-70 90 ranks are provided; test-
retest and internal
consistency data reported
VF.16 Hoff et al., 32.1 54 Paid male volunteers, 15.4 COWA New York
1996 (9.7) strict selection criteria (2.4)
page 219
Table A11.21
(continued)
.......
IIW
Table A11.1. (Contd.) .....
~
N
Study Age• n Sample Composition IQ!Education• Version Location
VF.l7 Ponton 38.4 300 Spanish-speaking 10.7 FAS Los Angeles, CA

et al., 1996 (13.5) healthy volunteers; (5.1)
page 219 MIF ratio 40%160%;
Table A11.22 16-29 data partitioned by <10
30--39 gender (2) x age (4) x >10
40-49 education (2)
50-75
VF.l8 Crossley 628 Community-dwelling seniors, 4 education FAS, Canada
et al., 1997 65-74 (635) cognitively normal groups: Animal
page 220 75--84 0-6 Naming
Table A11.23 85+ 7-9
10--12
13+
VF.l9 Beatty 73.7 38 18M, 20 F elderly Education: FAS, Oklahoma
et al., 1997 (8.7) volunteers screened 13.4 Animal
page 220 for health problems (3.4) Naming
Table A11.24
VF.20 Nyberg 77.3 39 Healthy elderly 13.6 FAS Canada.
et al., 1997 Sweden
page 220
Table A11.25
VF..21 Salthouse 18-39 40 Healthy adults, 47% M; 15.5 CFL, Atlanta
et al., 1997 data stratified into (1.7) Animals,
page 221 40--59 38 3 age groups 15.2 Furniture,
Table Al1.26 (2.5) Vegetables
60--78 37 15.3
(2.6)
VF..22 Kempler 73.0 317 Healthy elderly volunteers 10.3 Animal California
et al., 1998 (7.6) from 5 ethnic groups (5.0) Naming
page 221 (Chinese, Hispanic >
"tt
Table Al1.27 54-74 Vietnamese, white, 0--8 "tt
m
75-99 and African-American) ;:::9 z
were assessed in their 0
native language; X
data grouped by age,
education, gender, ethnicity
VF.I3 Stuss 54.4 37 Control group (19M, 18 F) 13.9 FAS, Canada
et al., 1998 (14.4) without neurological or (2.3) Animal >
"tl
page222 psychiatric disorder Naming "tl
m
Table A11.28 z
VF.24 Johnson- Data for 3 ethnic groups of FAS, USA
0
X
Selfridge et al., male veterans: Animal
1998 37.9 200 white, 13.5 Naming
page222 (2.6) (2.3)
Table All.29 37.8 200 black, 12.9
(2.7) (2.1)
37.9 200 hispanic; 13.3
(2.6) sample includes (2.4)
medically and psychiatrically
ill participants
VF..25 Dikmen 28.5 81 Normal or neurologically 12.2 FAS Washington
et al., 1999 (12.2) stable adults; 20% had (1.9)
page 223 neurological conditions;
Table All.30 60% M; data on test-retest
reliability and practice effect
are provided
VF.26 Manly 2:;65 187 Illiterate and literate nondemented 0-3 Animals, New York
et al., 1999 elders; 74% F; data are presented for Food,
page 224 education-matched and uneducated Clothing
Table A11.31 samples for English-and Spanish-
speaking participants separately
VF.27 Boone, 45-84 155 53 M, 102 F healthy Education: FAS California
1999 elderly volunteers; 14.57
page224 63.07 data stratified by (2.55)
Table A11.32 (9.29) FSIQ level (average, high FSIQ:
average, superior) 115.41
(14.11)
VF..28 Demakis, 22.5 21 Undergraduate 13.6 COWA Illinois
1999 (7.99) students, 67% F (1.46)
page 225
Table A11.33
VF.29 Epker 70.6 65 22 M, 43 F healthy 14.3 FAS, Texas
et al., 1999 (4.7) elderly volunteers (2.9) Animal
page 225 Naming ......
Table A11.34 ~
~
(continued)
.....
~
....
Study Age• n Sample Composition IQ/Education• Version Location
VF.30 Tombaugh, 1&-95 FAS 895 Large samples 0-21 FAS, Canada
et al., 1999 of healthy subjects Animal
page 226 1&-19 Animal735 obtained from 2 0-8 Naming
Tables A11.35-A11.37 20-29 studies; data are 9-12
30-39 stratified by 3 13-16
40-49 age x 3 educational 17-21
50-59 levels, as well as by
60-69 9 age groups, 4
70-79 education groups, and
80-89 2 genders
90-95
VF.31 Basso 32.50 50 Data for healthy men 14.98 FAS Tulsa, OK/
et al., 1999 (9.27) on 2 testing probes (1.93) Ohio
page 226 over a 12-month interval
Table A11.38
VF.32 Gladsjo 20-34 768 Healthy adults, 52% M; <11 FAS, San Diego,
et al., 1999 35-49 data stratified by age (3) x 12-15 Animal CA
page 227 50+ education (3); data are also 1&-20 Naming
Tables All.39, A11.40 presented for African Americans
and Caucasians separately
VF.33 Binder 82.3 125 Normal elderly sample, aged 13.5 Animal St. Louis,
et al., 1999 (4.4) 70 or above, 25% M, (3.0) Naming MO
page 227 87% Caucasian
Table A11.41
VF.34 Fama 66.7 51 Healthy elderly, 16.4 FAS, California
et al., 2000 (7.4) rigorous exclusion (2.3) Animals,
page 228 criteria Inanimate
Table Al1.42 Objects
VF.35 Troyer, 18-91 FAS 257 Healthy adults, 5-21 FAS/CFL, Canada
2000 59.8 30%M 13.9 Animals,
>
"'tt
"'tt
page228 (20.7) CFL 154 (2.9) Supermarket m
Table All.43 z
Animal407 0
X
Supermarket 156
VF.36 Acevedo 69.1 424 English 14.4 Animals, Florida >
"tl
et al., 2000 (6.9) speakers, (2.5) Vegetables, "tl
m
page 229 26% Male; Fruits
Tables A11.44-A11.47 64.9 278 Spanish 13.4
z
0
(7.7) speakers, (3.2) ><
30.8% M;
50-59 data stratified 8-12
60-69 by age, education, 13-16
70-79 and gender 2:17
VF.37Chen 74.9 483 Control elderly volunteers 31.9% Total for Pennsylvania
et al., 2000 (4.4) who participated in the <HS letters P and S
page 230 MoVIES project, 37.5% M Total for
Table A11.48 Animals and
Fruits
VF.38 Anstey 79.04 280 Normative data collected on 11.25 FAS Australia
et al., 2000 (6.59) old and very old adults (2.79)
page 230 living in retirement villages
Tables A11.49, All.SO and hostels; data presented
in raw scores for the entire sample
and in percentile distribution for
the sample stratified by age x
education; 14% M
VF.39 Brady 66.41 235 Healthy elderly, 14.03 Animals Boston
et al., 2001 (6.73) all-male sample (2.62)
page 231
Table A11.51
VF.40 Rosselli 50-84 82 28M, 54 F; 2-23 FAS, Florida
et al., 2002a 61.76 English monolingual, 14.8 Animals
page 231 (9.3) Spanish monolingual, (3.6)
Table All.52 bilingual
VF.41 Grady 66.3 517 Data for women with 12.7 Animals Multicenter,
et al., 2002 (6.4) established coronary (2.7) USA
page 232 disease, 2 groups:
Table All.53 67.3 546 ERT and placebo 12.7
(6.3) (2.7)
(continued)
.....
"'-!
c.n
;:j
cr-

Study Age• n Sample Composition IQ!Education• Version Location
VF.42 Giovannetti 25.2 31 21 M, 10 F healthy Education: Animals New York,

et al., 2003 6.07 adults recruited from 15.0 Pennsylvania
page 233 a medical center (1.48)
Table A11.54 community FSIQ:
109.3
(11.51)
VF.43 Lopez- 28.23 115 All-male sample; 6.66 PMR, Los Angeles, CA:
Carlos et al., (8.74) monolingual Spanish- (2.54) Animals Jalisco,
2003 speaking Latino manual Mexico
page 233 18-29 laborers; data 0-6
Tables All.55-A11.58 30-49 partitioned by age, 7-10
education, country, and
age (2) x education (2)
VF.44 Miller, 37.5 728 Seronegative homosexual 16.3 FAS, MACS
2003 (an update (6.9) and bisexual males from (2.3) Foods Centers
on Seines et al., MACS, native
1991) 25-34 English speakers; data <16
page 234 35-44 partitioned by age x 16
Table A11.59 45-59 education >16
VF.4S Ravdin 6()...$ 34 Healthy elderly, 32M, 117 F; 15.57 CFL, New York
et al., 2003 70-79 80 data partitioned by (2.67) Animals,
page234 80-92 35 3 age groups est.VIQ Fruits,
Table A11.60 120.44 Vegetables
(5.74)
>
"tl
•Age column and IQ/education column contain information regarding range and/or mean and standard deviation for the whole sample and/or separate groups, whichever information is "tl
rn
provided by the authors. z
0
><
APPENDIX 11 737
Table A11.2. [VF.l] Cauthen, 1978b: Number ofWords for Eight Letters• Generated by Two Age Groups
Age 20--59 (n =51)
s G u N F T J p
IQ range 100-140 M 16.2 11.7 6.1 9.3 12.9 13.3 8.1 13.4
X FSIQ = 115.6 (8.7) so (4.6) (3.7) (2.0) (2.7) (3.8) (3.7) (3.1) (3.4)
Age ~60 (n = 64)
s G u N F T p
IQ range
80--106 (n=21) M 8.8 6.9 3.1 5.5 8.0 7.6 3.7 7.4
so (4.4) (3.1) (1.8) (2.8) (2.9) (3.5) (2.3) (4.1)
107-118 (n = 21) M 10.9 8.9 3.9 6.0 10.1 9.8 4.7 10.0
so (3.9) (3.3) (1.7) (3.0) (2.7) (2.7) (1.8) (3.3)
119--140 (n=22) M 13.9 10.4 5.5 8.7 12.9 13.0 7.1 13.6
so (4.8) (3.7) (2.0) (3.1) (4.3) (3.5) (2.9) (4.7)
•Data for the younger group (12M, 39 F) are presented for the whole sample. Data for the older group (28 M, 36 F,
mostly institutionalized elderly) are stratified by the full-scale IQ (FSIQ) level.
Table A11.3. [VF.2a] Yeudall et al., 1986: Data• for the Whole Sample of Healthy
Adults and for Four Age Groups
Age Group
15-20 21-25 26--30 31-40 15-40
n 62 73 48 42 225
Age 17.76 22.70 28.06 34.38 24.66
(1.96) (1.40) (1.52) (2.46) (6.16)
Education 12.16 14.82 15.50 16.50 14.55
(1.75) (1.88) (2.65) (3.11) (2.78)
WAIS FSIQ 118.14 116.45 120.03 121.87 118.56
(8.73) (8.57) (9.12) (8.17) (8.81)
F 13.82 14.99 15.65 16.83 15.15
(4.36) (4.37) (4.42) (4.04) (4.41)
A 12.48 13.33 13.08 14.50 13.26
(3.87) (4.89) (3.41) (3.66) (4.13)
s 15.87 16.63 16.54 18.10 16.67
(4.52) (4.97) (4.70) (4.89) (4.80)
Average of 3 trials 14.06 14.98 15.09 16.48 15.03
(3.82) (4.29) (3.34) (3.61) (3.90)
•Total number of words generated for each letter and average for three letters.
738 APPENDIX 11
Table A11.4. [VF.2b] Yeudall et al., 1986: Data for Males

Age Group
15-20 21-25 26-30 31-40 15-40
n 32 37 32 26 127
Age 17.78 22.57 27.75 34.69 25.15
(2.09) (1.26) (1.57) (2.41) (6.29)
Education 12.22 15.11 15.78 16.69 14.87
(1.96) (1.74) (2.79) (3.55) (2.99)
WAIS FSIQ 119.21 118.61 120.30 122.92 119.96
(8.36) (8.83) (8.97) (7.06) (8.45)
F 14.03 14.83 15.84 16.58 15.25
(4.48) (4.84) (4.07) (4.31) (4.50)
A 13.00 13.22 13.03 14.50 13.38
(3.91) (5.52) (3.31) (4.13) (4.34)
s 15.81 16.94 17.44 17.88 16.98
(4.79) (5.05) (5.19) (4.41) (4.90)
(3.96) (4.73) (3.37) (3.77) (4.04)
Table A11.5. [VF.2c] Yeudall et al., 1986: Data for Females

Age Group
15-20 21-25 26-30 31-40 15-40
n 30 36 16 16 98
Age 17.73 22.83 28.69 33.88 24.03
(1.84) (1.54) (1.25) (2.53) (5.95)
Education 12.10 14.53 14.94 16.19 14.12
(1.52) (1.99) (2.32) (2.29) (2.43)
WAIS FSIQ 116.91 114.29 119.29 120.30 116.79
(9.15) (7.86) (9.78) (9.73) (8.99)
F 13.60 15.14 15.25 17.25 15.03
(4.29) (3.90) (5.16) (3.64) (4.31)
A 11.93 13.44 13.19 14.50 13.11
(3.82) (4.24) (3.71) (2.85) (3.88)
s 15.93 16.31 14.75 18.44 16.29
(4.31) (4.94) (2.91) (5.72) (4.68)
(3.71) (3.86) (3.27) (3.43) (3.73)
APPENDIX 11 739
Table A11.6. [VF.3] Gordon and Lee,

1986: Scores on a Letter Fluency Test
for a Sample of University Students
Aged 18--35 Years
n Score•
Males 90 40.76
(11.46)
Females 160 43.12
(10.84)
"Total number of words for three letters

(with SD).
Table A11.7. [VF.4] Bolla et al., 1990: FAS Data for a Healthy Elderly Sample Divided into
Three Groups Based on Vocabulmy Scores, Stratified by Gender
Males on Vocabulary Females on Vocabulary
~53 54-60 2:61 ~53 54-60 2:61
n 32 25 23 33 39 47
Age 61 63 65 61 65 69
(12) (15) (17) (11) (15) (17)
Education 13 14 17 13 15 16
(03) (03) (03) (03) (03) (03)
Vocabulary• 47 57 65 45 52 65
(OS) (02) (02) (06) (02) (02)
FASt 38 43 47 42 46 49
(12) (12) (09) (09) (12) (12)
•Raw Wechsler Adult Intel1igence Scale-Revised Vocabulary scores.

tTotal number of words for three letters (with SD).
740 APPENDIX 11
Table A11.8. [VF.5] Seines et al., 1991: Data for a Sample of Seronegative Homosexual/Bisexual Males
Participating in the Multi-Center AIDS Cohort Study, Stratified by Age and Education
FAS" Animals
Percentiles Percentiles
Age n Mean Age Education Mean (SD) 5th lOth Mean (SD) 5th lOth
By age
25--34 309 31.0 16.1 45.7 26 30.5 23.4 15 17
(2.6) (2.2) (12.7) (5.8)
35-44 290 39.3 16.4 46.1 26 29 23.4 14 17
(2.9) (2.3) (12.6) (5.4)
45-54 97 48.5 16.7 45.9 25 29 23.3 15 17
(2.6) (2.6) (12.3) (4.7)
By education
<College 229 36.1 13.7 41.7 23 26 22.0 13 15
(7.4) (1.2) (11.6) (5.3)
College 202 35.6 16.0 46.2 28 31 23.1 16 17
(7.2) (0.0) (12.3) (4.8)
>College 302 38.4 18.6 49.0 29 32 24.6 16 18
(7.8) (1.3) (12.4) (5.7)
"Total number of words for three letters.

APPENDIX 11 741
Table A11.9. [VF.6] Axelrod and Heruy, 1992: FAS data for Four Age Groups of
Healthy Adults
Age Groups
Variables 50s 60s 70s 80s
Age
M 55.3 65.2 74.3 83.4
SD (2.5) (2.6) (2.9) (3.0)
EtluccJtion
M 15.4 14.4 14.5 14.5
SD (2.5) (3.0) (4.2) (4.1)
Malelfornale ratio 10/10 10/10 10/10 8/12
Etlmicity 48, 16C 48, 16C 48, 16C 28, 18C

(Black/Caucasian)
WAIS-B Vocabulcary•
M 10.3 10.3 9.8 10.0
SD (2.1) (2.3) (2.8) (2.5)
Health n~tingt
M 4.3 4.0 4.0 3.9
SD (0.7) (0.9) (0.9) (0.8)
Number of playskian appointmenta
M 2.4 2.7 3.6 3.1
SD (3.0) (2.6) (3.2) (2.5)
FtDOf'd.
M 14.6 14.2 11.8 13.1
SD (3.8) (4.7) (3.2) (4.1)
A tDOf'd8
M 11.5 11.2 10.8 11.5
SD (4.3) (3.7) (4.3) (5.1)
s tDOf'd8
M 14.0 14.2 13.4 13.2
SD (3.9) (3.8) (3.9) (5.4)
Total FAS tDOf'd8
M 41.1 39.6 36.0 37.8
SD (9.9) (10.7) (9.3) (14.0)
•Wechsler Adult Intelligence Scale-Revised scaled scores.

tself-rating of health on a 5-point scale, where higher values indicate better health.
Table A11.10. [VF.7] Monsch et al., 1992: Data• for the Control Group on FAS, Category,
First Names, and Supennarket Fluency
n Age Education MIF Ratio FAS Categoryt First Names Supennarket
53 71.2 13.6 17/36 41.2 48.4 22.6 22.8

(7.9) (2.7) (12.5) (9.8) (5.8) (4.7)
•Total number of words for all trials per task.

teategory fluency task included animals, fruits, and vegetables trials.
742 APPENDIX 11
Table A11.11. [VF.8] Simkins-Bullock et al., 1994: Data for the Control Sample
n Age Education fSIQ M1F Ratio
19 52.6 13.26 JP2.0 1019 43.58 37.95

(15.6) (2.5) (12.85) (9.63) (6.54)
•Total number of words for three letters.

teategory fluency condition included animals Fd fruits or vegetables trials.
i
Table A11.12. [VF.9] Parkin and Lawreece, 1994:
Data for the Control Sample
Est Mffl
n Age Education FSIQ" Ratie FASt
22 71.9 9.4 106.1 4/lli 36.9

(4.8) (1.3) (12.6) (10.7)
i
"Full-scale IQ was estimated using the NatiOnal Adult
Reading Test. :
tTotal number of words for three letters.
Table A11.13. [VF.10] Friedman et al., '

1995: Data for the Control Sample
n Age
24 35.8 44.29
(11.0) (12.5)
Table A11.14. [VF.lla] Kozora and Cull9m, 1995: Demographic Characteristics for a
Sample of Healthy Adults Partitioned int~ Four Age Groups
Age Group
50-59 60-69 70-79 80-89
n 41 43 47 43
Mean age (SD) 54.5 64.6 74.6 83.8
(3.0) (2.8) (2.5) (3.1)
Mean education (SD) 14.3 14.2 14.3 14.9
(2.3) (2.3) (3.1) (3.3)
I
Male/female ratio 21120 16127 15132 16127
Mean Vocabulary score• (SD) 57.15 58.8 60.02 58.79
(6.37)' (5.63) (8.72) (8.87)
•Wechsler Adult Intelligence Scale-Revised VoCabulary raw scores.

APPENDIX 11 743
Table A11.15. [VF.llb] Kozora and Cullum,1995: Data for the Letter
and Category Fluency Conditions for the Four Age Groups
Age Group
50-09 60-69 70-79 80-89
Letter Total 41.23 45.76 46.49 40.74

(12.10) (14.26) (10.46) (11.19)
F 14.05 15.69 15.98 14.21
(4.55) (5.25) (3.91) (3.69)
A 12.98 14.17 14.40 12.95
(4.08) (4.92) (4.26) (4.44)
s 14.13 15.91 16.11 13.49
(5.19) (5.36) (4.51) (4.40)
Category Total 108.55 105.13 92.53 82.63
(17.08) (18.40) (16.23) (17.36)
Animals 20.95 21.07 18.96 15.81
(4.16) (5.08) (4.67) (4.51)
Supermarket 26.85 25.58 22.60 19.93
(6.75) (5.11) (5.27) (5.41)
First names 29.21 26.76 23.73 20.47
(5.67) (6.74) (5.90) (5.27)
U.S. states 30.77 31.60 26.67 27.81
(7.24) (7.40) (5.78) (8.05)
Table A11.16. [VF.12] Norris et al., 1995: Data for a Sample of

Independently Living Elderly (60-86 Years Old), Institutionalized El-
derly (62--89 Years Old), and Undergraduate Students (18-28 Years
Old)
Group
Old Old
Variable (Community) (Institution) Young
n 54 35 40
Age 73.1 75.3 19.4
(6.1) (7.5) (1.8)
Education 16.7 12.4 13.6
(2.3) (3.7) (1.1)
Depressiont 3.8 9.7 6.5
Functional status:
(3.6)
8.1
(0.3)
(7.4)
14.0
(3.1)
(4.3)
- .
Verbal fluency' 36.9 21.5 40.5
Mini-Mental State Exam

(10.1)
27.5
(9.8)
24.7
(7.8)
- .
(2.1) (3.0)
•Young participants did not receive these measures due to anticipated ceiling and
floor effects.
'Depression was assessed with the Geriatric Depression Scale.
:Functional status was measured with the Functional Assessment Scale.
'Total number of words for three letters.
744 APPENDIX 11
Table A11.17. [VF.13] Cahn et al., 199.1: Data for Table A11.19. [VF.15a] Ruff et al., 1996: Data•
a Control Sample of Cognitively Intact Jf;lderly for the Controlled Oral Word Association Version
for a Sample of Healthy Adults 16-70 Years of Age
Letter•
Stratified by Three Educational Groups x Gender
n Age Education MIF Ratio : Fluency
Education Total
238 78.4 13.8 97/141 38.3 Groups Males Females Sample
(6.8) (2.6) (0.78)t
n 180 180 360
~12 36.9 35.9 36.5
tThe SD is considerably lower than expected.: (9.8) (9.6) (9.9)
+3 +4
13-15 40.5 39.4 40.0
(9.4) (10.1) (9.7)
-1 -1
~16 41.0 46.5 43.8
(9.3) (11.2) (10.6)
-1 -7
TableA11.18. [VF.14] Ivniketal., TottJl _,.,. 39.5 40.6 40.1

1996: Demographic Description of (9.8) (11.2) (10.5)
the Healthy Sample Partitioned into
Groups Used in Controlled Oral •Means, SDs, and correction factors to adjust raw scores
for level of education.
Word Association Testing
Age groups
56-59 55 Table A11.20. [VF.15b] Ruff et al., 1996: Percen-
60-64 90 tile Ranks, Normalized T Scores, and Interpretation
65-69 85 for the Education-Corrected Scores for the Con-
70-74 93
trolled Oral Word Association Test
75-79 146
80-84 149 Corrected Score Percentile T Score Interpretation
85-89 84
90-94 33 ~17 1 26.7
95+ 8 20 2 29.5 Seriously
deficient
Education 3 31.2 Deficient
21
~7 8 23 4 32.5
8-11 121 25 5 33.5 Deficient
12 239 26 8 35.8 Borderline
13-15 181 27 9 36.6
16-17 128 28 10 37.2 Borderline
~18 66 29 13 38.7 Low average
30 16 40.2
Gender
31 19 41.2
Males 286 32 21 41.9
Females 457 33 27 43.9 Low average
Baee 34 30 44.7 Average
35 34 45.9
Caucasians 741 36 38 46.9
Blacks 2 37 43 48.2
Handedneu 38 47 49.2
39 51 50.3
Right 682
40 58 52.0
Left 29
41 61 52.8
Mixed 32
42 64 53.6
Tott.al 743 43 67 54.4
44 69 55.0
APPENDIX 11 745

Corrected Score Percentile T Score Interpretation
45 72 55.8 Average
46 76 57.0 High average
47 78 57.7
48 80 58.5
49 82 59.1
50 85 60.4
51 87 61.3
52 89 62.3 High average
53 91 63.4 Superior
54 92 64.1
55 94 65.5
56 95 66.5
58 97 68.9 Superior
60 98 70.6 Very superior
~64 99 73.3
Table A11.21. [VF.16] Hoff et al., 1996: Data for

the Controlled Oral Word Association for the
Control All-Male Sample
n Age Education COWA Score
54 32.1 15.4 43.7

(9.7) (2.4) (10.0)
Table A11.22. [VF.17] Ponton et al., 1996: FAS Data• for a Sample of 300 Spanish-
Speaking Participants Stratified by Gender x Age x Education
Age Group
16-29 30-39 40-49 50-75
Education <10 >10 <10 >10 <10 >10 <10 >10
Malu
n 11 25 13 18 12 17 18 6
X 24.18 31.84 31.39 33.00 24.33 35.18 24.33 35.83
(SO) (9.95) (7.85) (9.92) (9.34) (6.91) (10.64) (12.66) (7.94)
Females
n 12 30 22 44 16 11 25 20
X 24.00 26.37 22.91 35.18 23.56 34.64 20.88 33.00
(SO) (8.79) (9.01) (9.95) (10.49) (8.79) (7.55) (11.98) (9.88)

746 APPENDIX 11
Table A11.23. [VF.l8] Crossley et al., 1997: Data for the FAS• and Animal Naming for a
Sample of Cognitively Intact Seniors Prulitioned by Age, Gender, and Educational Level
FAS! Animal Naming
M so: n M SD n
Age group
65-74 24.0 12.4 139 14.2 4.3 144
7/h'W 25.8 11.5 343 14.2 3.8 343
~85 24.0 10.8 146 12.5 3.8 148
Gender
Male 23.2 12.1l 258 14.2 4.2 258
Female 26.2 11.0! 370 13.6 3.9 377
.j
Education (years) I
I
0-6 16.2 6.9, 140 12.1 3.1 149
7-9 23.7 9.9 170 13.4 3.8 169
10-12 27.0 10.2 202 14.1 3.9 203
~13 34.2 12.6 115 16.3 4.1 113
TottJl aampk 25.0 11.61 628 13.8 4.3 635
•Total number of words for three letters. 1

I
I
!
I
I
Table A11.24. [VF.l9] Beatty et al., 199t: Data for the FAS and Animal Naming for the
Control Sample ~
Animal
n Age Education MMSE MIF Ratio FAS Score• Naming Score
38 73.7
(8.7)
13.4
(3.4)

28.7
(1.6) I 1 18120 36.8
(13.6)
18.0
(4.9)
MMSE, Mini-Mental State Exam.
Table A11.25. [VF.20] Nyberg et al., l997: FAS

Data for a Healthy Elderly Sample ·
Age Education FAS

n (Range) (Range) Score•
39 77.3 13.6 42.51

(66-87) (8-22) (9.77)

APPENDIX 11 747
Table A11.26. [VF.21] Salthouse et al., 1997: Data for a Sample of Healthy Participants Stratified into
Three Age Groups
Letter Fluency Category Fluency
Age Group Mean Age n Education %Male c F L Animals Furniture Vegetables
18-39 .29.0 40 15.5 4.2.5 16..2 14.9 15.0 .20.6 1.2.9 13.8
(4.8) (1.7) (4.7) (4 ..2) (4.1) (5.1) (.2.8) (3.3)
40-59 49.1 38 15..2 50.0 14.9 13.7 13.4 18.8 1.2.3 14.0
(5.1) (.2.5) (4.9) (4.6) (3.3) (5.7) (3.4) (3.9)
60-78 69..2 37 15.3 48.6 14.8 14.6 13.5 17.1 11.8 14.0
(5.1) (.2.6) (5 ..2) (4.7) (4.9) (4.6) (3.3) (3.3)
Table A11.27. [VF.22] Kempler et al., 1998: Data for Animal Naming for a Sample of Healthy
Adults Stratified by Age, Education, Gender, and Ethnicity
X Age X Years Age at %at Animals

Group n (SD) Education (SD) in us• Immigration Hornet (SD)
Age
54-74 195 70.0 10.3 16.0
(4 ..2) (5.1) (5.0)
75-99 1.2.2 80.8 10.1 14.4
(4.6) (5.0) (4.3)
Edaccdion
0-8 11.2 73.3 4.6 13.5
(7.5) (.2.5) (4..2)
~9 .205 7.2.7 13.3 16.4
(7.7) (3.0) (4.7)
Gender
Male 11.2 73.4 11.5 16.4
(7.3) (4.6) (4.4)
Female 205 7.2.7 9.6 14.7
(7.8) (5 ..2) (4.8)
Etlmiciey
African American 54 7.2.8 11.6 15..2
(9.1) (4.7) (4.4)
White 58 76.6 1.2.3 16.7
(7.6) (3.8) (4.2)
Chinese 67 72.5 10.9 11.8 60.9 98% 15.3
(7.3) (5.5) (9.3) (9.1) (5.1)
Hispanic 78 71.9 8.5 .27.6 44.0 82% 1.2.8
(7.1) (5.4) (19.6) (19.1) (3.9)
Vietnamese 60 71.6 8.6 7.0 64.3 98% 17.3
(5.8) (4..2) (6.4) (8 ..2) (5 ..2)
Total 317 73.0 10.3 15.5
(7.6) (5.0) (4.6)
•Number of years residing in United States.

tPercent who speak only their native language at home.
748 APPENDIX 11
Table A11.28. [VF.23] Stuss et al., 1998: Data for the Control Group• Stratified by Age and
Gender
Age
21-39 40--64 65--81
Tasks Males Females Males Females Males Females
n 10 10 9 16 9 8
Letter Ouency (FAS)t 53.2 48.4 44.1 42.7 37.1 47.5
(13.1) (10.3) (10.2) (11.0) (10.1) (14.8)
Animal Ouency 26.3 23.0 18.7 22.4 16.7 18.1
(2.6) (3.9) (3.3) (6.0) (1.7) (5.9)
Letter-based errors 2.1 1.4 2.3 2.2 2.3 3.0
(2.5) (1.8) (2.0) (1.6) (2.4) (2.2)
Semantic errors 0.7 0.1 0.7 1.1 0.6 0.4
(1.0) (0.3) (0.9) (1.4) (1.3) (0.7)
•Mean education for the sample is 13.9 (2.3) years; mean National Adult Reading Test IQ is 113.8 (6.1).
Table A11.29. [VF.24] Johnson-Selfridge et al., 1998: F AS and Animal Naming Data for a Sample of Male
Veterans Stratified into Three Ethnic Groups
FAS Animals
Ethnic WRAT-R FAS adjusted Animals adjusted
Group n Age Education Reading Score• scoret Score scoret
White 200 37.9 13.5 62.8 35.5 33.6 21.2 <JJJ.7

(2.6) (2.3) (14.3) (11.9) (4.6)
Black 200 37.8 12.9 50.4 32.5 35.1 18.3 18.8
(2.7) (2.1) (13.6) (lO.o) (4.8)
Hispanic 200 37.9 13.3 60.2 31.7 31.0 18.7 18.7
(2.6) (2.4) (13.9) (9.6) (5.2)

tscores after covarying for income, education, and Wide Range Achievement Test-Revised (WRAT-R) Reading scores.
Table A11.30. [VF.25] Dikmen et al., 1999: Test-Retest FAS Data for a Group of Normal,
Neurologically Stable Adults•
FAS Scoret
Test-Retest
n Age Education M/F Ratio WAlS FSJQ• Interval Time 1 Time2
81 28.5 12.2 60/40 108.8 11.1 43.25 44.47

(12.2) (1.9) (12.3) (0.6) (10.75) (10.36)
•Demographic information is provided for a larger sample of 138 participants; mean Wechsler Adult
Intelligence Scale full-scale IQ (Wechsler, 1955) is reported for the three groups used in this study combined;
20% of the sample had preexisting conditions that might affect test performance, the most significant being
alcohol abuse and a significant traumatic brain injury.
APPENDIX 11 749
TableA11.31. [VF.26] Manlyetal., 1999: Data for

Illiterate and Literate Elders (>65 years of Age)
with 0-3 Years of Education
Category
n Fluency•
Englisla-apealdng elden
Education-matched sample
Literate 43 11.79
(3.97)
Illiterate 43 11.55
(3.14)
Uneducated sample
Literate 26 11.34
(3.48)
Illiterate 47 12.15
(3.07)
Spaniah-8pflaldng elden
Education-matched sample
Literate 32 10.88
(3.90)
Illiterate 32 11.59
(3.21)
Uneducated sample
Literate 17 10.51
(3.06)
Illiterate 43 12.38
(3.02)
•score represents number of words averaged over three

conditions (animals, food, and clothing).
Table A11.32. [VF.27] Boone, 1999: FAS Data for a Healthy Elderly Sample•
Partitioned into Three IQ Levels
WAIS-R IQ Level
Average High Average Superior
n 53 39 59
FAS scoret 36.45 38.87 44.31
SD (9.26) (9.22) (11.88)
"Mean age for the sample is 63.07 (9.29), mean education is 14.57 (2.55), mean full-scale
IQ is 115.41 (14.11).
Table A11.33. [VF.28] Demakis, 1999: FAS Data

for the Control Sample
MIF FAS
n Age Education Ratio Score•
21 22.5 13.6 33167 37.8

(7.99) (1.46) (11.1)

750 APPENDIX 11
Table A11.34. [VF.29] Epker et al., 1999: Data for the FAS and Animal Naming for the
Control Sample
n Age Education MIF Ratio MMSE FAS Score• Animals Score
65 70.6 14.3 22143 28.45 45.31 19.49

(4.7) (2.9) (1.44) (12.67) (4.67)

MMSE, Mini-Mental State Exam.
Table A11.35. [VF.30a] Tombaugh et ·al., 1999:

Data for the FAS and Animal Naming forla Sample
of Healthy Adults Stratified by Ded10graphic
Groups
FAS• ~Naming
n M (SD) n M(SD)
Etluccation
0-8 163 24.9 140 13.9
(10.7) (3.9)
9-12 664 36.7 377 16.7
(12.2) (4.6)
13-16 392 42.6 173 19.0
(11.6) ' (5.2)
17-21 81 43.9 44 19.5
(12.3) (5.2)
Age
16-19 19 39.3 19 21.5
(12.0) (4.4)
20-29 106 41.2 41 19.9
(9.2) (5.0)
30-39 132 43.1 43 21.5
(11.4) (5.5)
40-49 121 43.5 45 20.7
(12.2) (4.2)
50-59 144 42.1 43 ' 20.1
(11.1) (4.9)
60-69 220 38.5 92 17.6
(13.7) (4.7)
70-79 334 34.8 228 16.1
(12.8) (4.0)
80-89 200 28.9 200 14.3
(11.7) (3.9)
90-95 24 28.2 24 13.0
(11.0) (3.8)
Gender
Male 559 37.0 310 17.4
(13.0) (5.1)
Female 741 37.8 425 16.5
(13.1) (5.0)
ToltJl 1,300 37.5 735 16.9
(13.1) (5.0)

APPENDIX 11 751
TableA11.36. [VF.30b] Tombaughetal.,1999: Data• for the FAS Stratified by Three Age Groups x Three
Education Groups
Age
16-59 ~79 80-95

Education
Percentile 0-8 9-12 13-21 0-8 9-12 13-21 0-8 9-12 13-21
Score (n=12) (n=268) (n=242) (n=76) (n=292) (n=185) (n=75) (n=102) (n=46)
90 48 56 61 39 54 59 33 42 56
80 45 50 55 36 47 53 29 38 47
70 42 47 51 31 43 49 26 34 43
60 39 43 49 27 39 45 24 31 39
50 36 40 45 25 35 41 22 29 36
40 35 38 42 22 32 38 21 27 33
30 34 35 38 20 28 36 19 24 30
20 30 32 35 17 24 34 17 22 28
10 27 28 30 13 21 27 13 18 23
M 38.5 40.5 44.7 25.3 35.6 42.0 22.4 29.8 37.0
(SD) (12.0) (10.7) (11.2) (11.1) (12.5) (12.1) (8.2) (11.4) (11.2)
Table A11.37. [VF.30c] Tombaugh et al., 1999: Data for Animal Naming Stratified by Three Age x Three
Education Groups
Age
16-59 ~79 80-95

Education
Percentile 0-8 9-12 13-21 0-8 9-12 13-21 0-8 9-12 13-21
Score (n=12) (n=268) (n=242) (n=76) (n=292) (n=185) (n=75) (n=102) (n=46)
90 26 30 20 22 25 18 19 24
75 23 25 17 19 22 16 17 20
50 20 23 14 17 19 13 14 16
25 17 18 12 14 16 11 12 14
10 15 16 11 12 13 9 11 12
M 19.8 21.9 14.4 16.4 18.2 13.1 13.9 16.3
(SD) (4.2) (5.4) (3.4) (4.3) (4.2) (3.8) (3.4) (4.3)
Table A11.38. [VF.31] Basso et al., 1999: Data for a Sample of Healthy Men on Two
Testing Probes over a 12-Month Interval
FAS Score•
n Age Education WAIS-R FSIQ Test Retest
50 32.50 14.98 109.30 47.68 48.42

(9.27) (1.93) (12.29) (10.82) (12.06)

752 APPENDIX 11
Table A11.39. [VF.32a] Gladsjo et al., lt99: Data for the FAS and Animal Naming for a
Healthy Sample• Stratified by Age and Education
Education
0-11 12-15 16-20
Age Range FASt Animal FAS Animal FAS Animal
n 103 415 250

20-34 38.21 17.74 40.30 21.11 44.38 22.88
(13.43) (5.52) (9.59) (5.90) (10.54) (4.73)
35-49 33.32 18.36 40.63 19.82 47.27 22.28
(11.93) (6.63) ! (11.43) (6.26) (13.33) (5.57)
50-101 31.47 15.28 . 38.63 18.05 41.81 19.35
(13.21) (3.80) i (11.98) (4.81) (12.75) (4.42)
•sample is 52% male, 45% African American, 55% Caucasian.

1'-rotal number of words for three letters.
Table A11.40. [VF.32b] Gladsjo et al., lt99: Data

for the FAS and Animal Naming for. Healthy
African-American and Caucasian Particip.ts, Stra-
tified by Age l
African Americans• Cau~ianst
Age Range FAS: Animal FAS; Animal
n 346 422
20-34 38.94 19.44 45.37 24.79
(10.49) (5.17) (9.45) (5.20)
35-49 38.61 19.11 47.00 22.96
(12.57) (6.16) (11.81) (5.64)
50-101 33.87 16.68 40.32 18.59
(12.96) (4.71) (12.39) (4.63)
"Meanage39.2(12.6)years,meaneducation13.4~.5)years.
tMean age 59.0 (19.6) years, mean education l4.5 (2.8)
years.
*Total number of words for three letters.
Table A11.41. [VF.33] Binder et al., 1999: Data for a Normal Elderly Sample
cos• Blessedt Animal
n Age Education %Male %Caucasian Score Score Naming
125 82.3 13.5 25 87 1.8 2.1 15.5

(4.4) (3.0) (1.8) (2.1) (4.5)
"Geriatric Depression Scale.

tshort Blessed Orientation-Memory-Concentra~n Test.
APPENDIX 11 753
Table A11.42. [VF.34] Fama et al., 2000: Data for the FAS, Animal Naming,
and Inanimate Objects Naming Conditions for the Control Sample
FAS Animal Naming Inanimate Objects
n Age Education Score• Score Naming Score
51 66.7 16.4 41.2 22.1 26.1
(7.4) (2.3) (12.9) (4.4) (7.3)
Table A11.43. [VF.35] Troyer, 2000: Data• for the Letter and Category Fluency
Conditions for a Sample of Healthy Adults
Version n Age Education M/F Ratio Test Score
FAS/CFL' 257/154 59.8 13.9 aono 42.5
(20.7) (2.9) (11.7)
Animals 407 19.5
(5.3)
Supermarket 156 22.9
(5.8)
"Demographic information is provided for a larger sample.

tTbe mean for FAS/CFL performance.
Table A11.44. [VF.36a] Acevedo et al., 2000: Data Table A11.4S. [VF.36b] Acevedo et al., 2000: Data
for the Category Fluency Test for Healthy English- for the Category Fluency Test for Healthy Spanish-
Speaking Participants Speaking Participants
Animal Animal
n Naming Vegetables Fruits Total n Naming Vegetables Fruits Total
Age Age
50-59 37 18.4 16.0 16.0 50.4 50-59 64 16.3 13.0 13.2 42.6
(4.9) (4.1) (4.1) (10.6) (3.9) (3.6) (3.3) (8.4)
60-69 107 17.1 14.4 13.7 45.2 60-09 97 17.2 13.1 13.4 43.6
(4.2) (3.9) (3.7) (9.6) (5.3) (4.0) (3.4) (10.0)
70-79 172 15.2 13.6 12.5 41.3 70-79 76 16.3 12.3 12.8 41.3
(4.3) (3.5) (3.1) (8.4) (4.4) (3.6) (3.6) (9.7)
.Etlueaticm EducCJtion
8-12 112 15.0 14.2 13.0 42.2 8-12 105 15.8 12.6 12.4 40.9
(4.3) (3.8) (3.1) (8.7) (4.6) (3.7) (3.3) (9.5)
13-16 154 16.3 14.0 13.3 43.6 13-16 94 17.1 13.1 13.8 44.0
(4.0) (3.7) (3.9) (9.5) (4.1) (3.9) (3.5) (9.6)
17+ 50 18.8 14.7 13.9 47.4 17+ 38 17.7 12.8 13.5 44.0
(5.4) (3.9) (3.6) (10.7) (5.7) (3.4) (3.2) (8.8)
Gender Gender
Male 82 16.2 11.9 11.9 40.0 Male 73 16.6 11.3 12.2 40.1
(4.6) (2.8) (3.3) (8.7) (5.7) (3.8) (3.5) (10.7)
Female 234 16.3 15.0 13.8 45.0 Female 164 16.7 13.5 13.5 43.7
(4.5) (3.7) (3.6) (9.5) (4.2) (3.6) (3.3) (8.8)
ToltJl 16.2 14.2 13.3 43.7 Total 16.7 12.8 13.1 42.6
(4.5) (3.8) (3.6) (9.6) (4.7) (3.6) (3.4) (9.5)
754 APPENDIX 11
Table A11.46. [VF.36c] Acevedo et al., 2000: Data for English Speakers Stratified by Gender x Age and
Gender x Education
Age
Men Women
50-59 60-69 70-79 50-59 60-69 70-79
n 7 30 45 30 77 J.i7
Animals 16.4 16.4 16.0 18.9 17.3 15.0
(3.3) (4.9) (4.7) (5.1) (3.9) (4.2)
Fruits 12.3 11.7 11.9 16.9 14.4 12.7
(2.3) (3.5) (3.4) (3.9) (3.5) (3.0)
Vegetables 11.7 11.8 12.0 17.0 15.4 14.2
(1.7) (2.8) (3.0) (3.8) (3.8) (3.5)
Total Huency 40.3 40.0 39.8 52.7 47.2 41.9
(4.5) (9.7) (8.6) (10.2) (8.8) (8.3)
Education
Men Women
8-12 13-16 17+ 8-l.i 13-17 17+
n 25 42 15 87 1l.i 35
Animals 15.6 16.1 17.4 14.8 16.4 19.4
(4.4) (4.4) (5.8) (4.3) (3.9) (5.2)
Fruits 11.9 11.7 12.3 13.3 13.9 14.6
(3.4) (3.3) (3.3) (2.9) (4.0) (3.6)
Vegetables 12.2 11.7 12.0 14.8 14.8 15.9
(2.3) (3.1) (3.0) (3.9) (3.6) (3.7)
Total Huency 39.8 39.4 41.7 42.9 45.2 49.9
(8.3) (8.8) (9.3) (8.8) (9.3) (10.4)
APPENDIX 11 755
Table A11.47. [VF.36d] Acevedo et al., 2000: Data for Spanish Speaken Stratified
by Gender x Age and Gender x Education
Age
Men Women
50-59 60-09 70-79 50-59 60-09 70-79
n 15 32 26 49 65 50
Animals 15.5 18.0 15.4 16.6 16.7 16.7
(3.4) (7.2) (4.2) (4.1) (4.0) (4.5)
Fruits ll.1 12.7 12.4 13.8 13.7 13.0
(3.0) (3.9) (3.2) (3.1) (3.2) (3.8)
Vegetables u.s 11.0 11.6 13.5 14.1 12.7
(3.4) (3.7) (4.2) (3.6) (3.7) (3.3)
Total fluency 38.3 417 39.3 43.9 44.6 42.2
(7.8) (12.3) (9.9) (8.2) (8.6) (9.5)
Education
Men Women
8-12 13-16 17+ 8-12 13-17 17+
n 39 21 13 66 73 25
Animals 16.3 16.8 17.1 15.6 17.2 18.0
(5.4) (5.2) (7.7) (4.2) (3.8) (4.5)
Fruits 12.2 12.8 11.5 12.6 14.1 14.5
(3.7) (3.2) (3.2) (3.1) (3.6) (2.7)
Vegetables u.s 10.6 10.9 13.1 13.8 13.8
(4.0) (3.7) (3.4) (3.6) (3.7) (3.0)
Total fluency 40.3 40.2 39.5 41.2 45.1 46.4
(10.9) (10.6) (10.9) (8.6) (9.1) (6.5)
Table A11.48. [VF.37] Chen et al., 2000: Data for

the Control Sample of Nondemented Elderly• Table A11.49. [VF.38a] Anstey et al., 2000: Data
Letter for a Sample of Australian Elderly•
Category
n Age %Male Fluencyt Flueno/ n Education % Male FAS scoret
Age
483 74.9 37.5 23.46 27.70 280 79.04 11.25 14 32.76
(4.4) (7.26) (6.31) (6.59) (2.79) (11.33)
"Lower than high school education, 31.9%. "Participants with Mini-Mental State Exam scores as low
tTotal number of words for letters P and S. as 17 were included.
:Total number of words for Animals and Fruits categories. 'Total number of words for three letters.
756 APPENDIX 11
Table A11.50. [VF.38b] Anstey et al.; 2000: Percentile Distribution for the Sample Stratified by
Age x Education
Age
62-69 70-79 80-89 90-95

Education
0-9 10-12 ~13 0-9: 10-12 ~13 0-9 10-12 ~13 0-9 10-12 ~13
n 7 7 6 43 60 29 36 29 26 8 4 5
p~
5 13 10 27 10 15 15 12 17 17 13 20 22
10 13 10 27 15 18 23 16 22 21 13 20 22
25 16 36 32 22· 22 31 21 26 29 25 21 24
50 20 41 47 28! 31 38 33 32 36 31 30 33
I
75 26 44 51 37. 42 47 37 40 43 36 42 56
90 46: 46 51 43 49 49
95 ssl 49 49 47 50 58
Table A11.51. [VF.39] Brady et al., 2001:tData for Table A11.53. [VF.41] Grady et al., 2002: Data for
the Initial Test and Retest 3 Years La~r for a the Animal Naming Test for a Sample• of Women
Sample of Healthy Males 1 with Established Coronary Disease, Stratified into
Estrogen/Progestin Replacement Treatment and
Placebo Groups
n Age Education Males Test 1 ~ Test 2 Animal
%White Naming
235 66.41 14.03 1009& 19.0 : 18.3
(6.73) (2.62) (4.8> I (4.9) Group n Age Education Race Score
Treabnent 517 66.3 12.7 90.9 15.9

(6.4) (2.7) (4.8)
Placebo 546 67.3 12.7 90.5 16.6
Table A11.52. [VF.40] Rosselli et -~L 2002a:
(6.3) (2.7) (4.8)
Performance on the FAS and Animal NlllllflgTests
for Three Linguistic Groups 1
•All participants were <80 years of age.
: Animal
FAS I Naming
n Score• Score
Table A11.54. [VF.42] Giovannetti et al., 2003:
Monolinguals
English 45 37.7 16.8 Data for the Animal Naming Test for the Control
(5.3) (5.2) Sample
Spanish 18 34.9 16.7
(4.1) I (3.8) Animal
WAIS-R Naming
Bilinguals n Age Education MIF Ratio FSIQ" Score
English 19 34.9 14.2
(4.8) (4.1) 31 25.2 15.0 21110 109.3 24.9
Spanish 19 35.2 14.5 (6.07) (1.48) (11.51) (5.5)
(4.8) (3.8)
"Wechsler Adult Intelligence Scale-Revised full-scale IQ.
APPENDIX 11 757
Table A11.55. [VF.43a] Lopez-Carlos et al., 2003: Table A11.56. [VF.43b] Lopez-Carlos et al., 2003:
Data for Monolingual Spanish Speakers with ::510 Data for Monolingual Spanish Speakers with ::510
Years of Education on the PMR Version of Pho- Years of Education on the PMR Version of Pho-
nemic Fluency and Animal Naming. Stratified by nemic Fluency and Animal Naming. Stratified by
Age Group Education Group
Animal Animal
Age PMR Naming Education PMR Naming
Group n Vocabulaly- Scorest Scores Group n Vocabulary• Scorest Scores
1~29 71 16.86 31.45 16.83 0-6 56 14.38 28.87 16.70

(9.37) (9.92) (4.28) (8.33) (10.14) (4.49)
30-49 44 19.41 33.95 19.21 7-10 59 21.12 35.71 18.71
(10.36) (12.67) (5.66) (10.02) (10.96) (5.24)
"Wechsler Adult Intelligence Scale-ni Vocabulary raw •Wechsler Adult Intelligence Scale-Ill Vocabulary raw
scores (Mexican version). scores (Mexican version).
1Total number of words for three letters. tTotal number of words for three letters.
Table A11.57. [VF.43c] Lopez-Carlos et al., 2003: Data for Monolingual Spanish
Speakers with ::510 Years of Education on the PMR Version of Phonemic Fluency and
Animal Naming. Stratified by Age x Education
Animal
PMR Naming
Age Education n Vocabulary• Scorest Scores
1~29 0-6 30 12.63 27.24 15.34
(6.49) (8.70) (4.54)
7-10 41 19.95 34.50 17.90
(10.04) (9.72) (3.79)
30-49 0-6 26 16.38 30.76 18.28
(9.85) (11.48) (3.95)
7-10 18 23.78 38.39 20.50
(9.74) (13.23) (7.36)
•Wechsler Adult Intelligence Scale-Ill Vocabulary raw scores (Mexican version).

1Total number of words for three letters.
758 APPENDIX 11
Table A11.58. [VF.43d] Lopez-Carlos et al., 2003: Data for Monolingual Spanish
Speakers with :::;10 Years of Education on the PMR Version of Phonemic Fluency
and Animal Naming, Stratified by Country Group
Countty Animal
Group n Vocabulary• PMR Scorest Naming Scores
Ins Angeles, USA 65 16.18 30.00 17.16

{8.52) {11.09) {4.46)
Jalisco, Mexico 50 19.98 35.62 18.52
{10.96) {10.30) {5.53)
"Wechsler Adult Intelligence Scale-III Vocabulary raw scores {Mexican venion).

tTotal number of words for three !etten.
Table A11.59. [VF.44] Miller, 2003 (An Update

on Seines et al., 1991): Data for a Sample of
Seronegative Homosexual/Bisexual Males Partici- Table A11.59. (Contd.)
pating in the Multi-Center AIDS Cohort Study,
Education Animal
Stratified by Age x Education
Age {Yean) FAS" Naming
Education Animal
45-59 <16
Age {Yean) FAS" Naming
Mean 41.05 20.27
25-34 <16 {SD) {10.41) {3.58)
Mean 40.35 21.10 n 39 11
{SD) {10.98) {4.14) 16
n 88 30 Mean 42.22 22.00
16 {SD) {12.06) (3.79)
Mean 48.61 22.52 n 32 7
{SD) {11.99) {4.77) >16
n 80 27 Mean 48.56 23.82
>16 {SD) {13.47) {5.42)
Mean 49.54 23.80 n 65 22
{SD) {12.43) {5.72) Total
n 81 25 Mean 44.91 22.53
Total {SD) {12.75) (4.87)
Mean 46.00 22.39 n 136 40
{SD) {12.47) {4.93) Total <16
n 249 82 Mean 41.10 21.44
35-44 <16 {SD) (11.23) (4.82)
Mean 41.78 22.16 n 227 73
(SD) {11.80) (5.73) 16
n 100 32 Mean 45.36 22.37
16 (SD) (12.62) (4.61)
Mean 43.52 22.32 n 199 59
{SD) (12.86) (4.81) >16
n 87 25 Mean 49.13 24.08
>16 (SD) (12.63) (5.54)
Mean 49.16 24.35 n 302 95
(SD) (12.44) (5.61) Total
n 156 48 Mean 45.60 22.79
Total {SD) (12.65) (5.19)
Mean 45.58 23.20 n 728 227
(SD) (12.77) (5.52)
n 343 105 •Total number of words for three letten.
APPENDIX 11 759
Table A11.60. [VF.45] Ravdin et al., 2003: Data

for a Sample of Healthy Elderly• Stratified by
Three Age Groups
Age Group
60-69 70-79 80-92
n 34 80 35
Age 66.26 74.56 83.71
(2.59) (2.67) (3.66)
Education 16.00 15.54 15.23
(2.17) (2.62) (3.21)
Estimated VIQt 120.00 121.24 119.14
(5.36) (5.32) (6.83)
Letter fluency total: 49.56 46.81 44.46
(11.57) (10.76) (13.28)
c 17.35 16.64 16.03
(4.97) (4.27) (5.51)
F 17.29 15.50 15.43
(3.66) (4.10) (4.34)
L 14.91 14.74 13.00
(4.46) (4.10) (5.09)
Semantic fluency totalf 51.35 46.91 41.60
(10.25) (10.03) (8.10)
Animals 20.68 19.09 16.34
(5.59) (5.38) (4.07)
Fruits 14.41 13.63 12.17
(3.39) (3.41) (3.43)
Vegetables 16.26 14.20 13.17
(4.00) (3.93) (2.86)
"The sample included 32 males and 117 females.

tverbal IQ is estimated with the AMNART.
~otal number of words for three letters.
1Total number of words for three categories.
Appendix 11m: Meta-Analysis Tables for
the Verbal Fluency Test
Table A11m.1. Results of the Meta-Analysis and Predicted Scores

for the FAS

Sample tize
Mean 30 79.20 86.81 19-411
Age
Mean 30 49.55 19.66 17.8-74.3
SD 30 6.42 4.51 1.4-20.7
Education
Mean 29 14.31 1.98 9.4-17.0
SD 29 2.33 0.82 0.8-4.2
IQ
Mean 7 113.91 7.72 102-121.9
SD 7 10.45 2.33 8.2-14.1
Percent male 17 54.33 28.41 0-100
Teat score meana
Combined mean 30 41.84 3.65 35.0-49.4
Combined SD 30 11.20 1.41 7.8-13.6
tWeigbted means and SDs.
760
APPENDIX 11M 761
Predicted number of words generated and SDs per age group•

(FAS)
95%CI
Age .Rmtge Predieted Score Lower Band Upper Band
18-19 41.13 40.11 44.16

J0-.14 43.10 41.47 44.92
JS-J9 44.!1 42.70 45.72
30-34 44.87 43.43 46.30 Standard deviation for all age
35-39 45.17 43.74 46.61 groups is 11.10.
40-44 45.13 43.68 46.58
4S-49 44.73 43.30 46.16
50-IU 43.99 42.63 45.35
SS-S9 41.89 41.66 44.12
60-64 41.44 40.36 42.51
65-69 39.64 38.69 40.59
10-14 37.48 36.49 38.48
"Based on the equation:

Predictetl tat acore = 34.29763 + 0.5537161• age - 0.0070315 • age2
Years of Correction
Education Factor
10 -2.00
11 -1.50
12 -1.00
13 -0.50
14 0
15 +0.50
16 +1.00
17 +1.50
With every year of education above or below 14, we suggest

correcting the obtained score by adding or subtracting 0.50 to
or from the predicted score given in the table for the relevant
age group. Extrapolation of this correction outside the bound-
aries of 10-17 years of education should be made with caution
as empirical data are not available beyond these education
ranges.
(continued)
762 APPENDIX 11M
Ordinary least-squares regression o£Wst means on age (quadratic)

Numberofcluden 18
R2 0.711
F<df)• p F<2.17l = 3!1.40, p < 0.000
Age 0.5537161 0.109 5.07" o.ooo• 0.323 to 0. 784

Age2 -0.0070315 0.001 -6.41 0.000 -0.009 to -0.005
Constant 34.29763 2.428 14.13 0.000 29.18 to 39.42
•significance test for age centered (sample m~ -aggregate mean): t =-6.62, p = 0.000.
Prediction
Predicted age range 1~74yefl's
SEe 0.71 :
95%CI 41.52-44.29
50
0
0
0 0
45
40 0
35 0
20 30 40 50 80 70 80
11118'
Figure A11 m.1. A scatterplot illustrating the dispenion of the data points around the regression line for the
FAS. The size of the bubbles reHects the w~ght of the data point, with larger bubbles indicating larger

Q<df>•P Q(29) = 270.34, p < 0.000
APPENDIX 11M 763
Tests for model &t--additioo of a quadratic term
Model BIC BIC'
Linear 0.478 0.459 47.038 -16.083

Quadratic 0.711 0.690 32.659 -30.462
Tests for parameter speci&catioos
Shapiro-Wilk W test W=0.983, p =0.891
Homoscedasticity
White's general test 1.672, p = 0.796
Significance tests for regression with the standard deviations
A regression of SDs on age yielded an R2 of 0.018 (F(l,l7l = 0.35, p = 0.561).

Therefore, the SD for the aggregate sample is suggested for use with
all age groups.
Education
Est. tau2 with education 0
R2 0.768
Education 0.4976037 0.202 2.47 0.025 0.070 to 0.925
IQ
R2 0.914
IQ 0.1135096 0.036 3.17 0.050 -0.000 to 0.227
Gender
t-test by gender:
n X male (SD) X female (SD) M-F cliff. p
6M,6F 43.783 (1.655) 43.872 (1.707) -0.088 -0.037 0.515

764 APPEND IX 11M
Table 11 m.2. Results of the Meta-Analysis and Predicted Scores for Animal Naming
Numbe r of studies included in the analysis 11

Sample size
Mean 25 86.67 81.58 31-411
Age
Mean 25 62.60 18.13 25 . ~7 . 5
SD 25 5.47 3.67 2.3-20.7
Education
Mean 24 14.42 1.89 10.5-17.0
SD 24 2.49 0.91 0.8-4.7
IQ
Mean 1 109.30
SD 1 11.51
Percent male 14 48.38 27.46 26.0-100
Test score means
Combined mean 25 18.94 3.15 13.3-24.9
Combined SD 25 4.65 0.54 3.7-5.
tw eighted means and SDs.
Predicted number of words generated and SD s per age group•

(Animal Naming)
95% CI
Age Predicted Lower Upper

Range Score Band Band
25-29 24.28 23.34 25.22

30-34 23.52 22.62 24.41
35-39 22.75 21.89 23.62
40-44 21.99 21.15 22.84 Standard deviations for
45-49 21.23 20.39 22.08 all age groups is 4.65.
50-54 20.47 19.62 21.33
55-59 19.71 18.83 20.59
60-64 18.95 18.03 19.87
65~9 18.19 17.22 19.16
70-74 17.43 16.40 18.46
75-79 16.67 15.57 17.77
80-84 15.91 14.73 17.08
85-87 15.38 14.14 16.61
• Based on the equation:

Predicted test score = 28.45972 -0.1521419 • age
APPENDIX 11M 765
Table 11 m.2. (Contd.)

25
20
15
0
0
10
20 30 so eo 10 eo
age
Figure A11 m.2. A scatterplot illustrating the dispersion of the data points around the regression line for
Animal Naming. The size of the bubbles reflects the weight of the data point, with larger bubbles indicating
larger standard error and smaller weight.
Signi&cance tests for regression with the test scores
OrdiDary least-squares regression of test meaos on age (linear)

R2 0.764
F<df)• p F<2.1o> = 177.27, p < 0.000
Age -0.1521419 0.011 -13.31 0.000 -0.178 to-0.127

Constant 28.45972 0.679 41.93 0.000 26.95 to 29.97
Predietion
SEe 0.49
95%CI 18.76-20.70
Tests for assumptions and model flt

Q<dO·P Q<24> = 1312.69, p<O.OOO
Tests for model &t--eddition of a quadratic term
Model BIC BIC'
Linear 0.764 0.754 17.198 -32.916

Quadratic 0.765 0.743 20.376 -29.739
766 APPENDIX 11M
Table 11m.2. (Contd.)
Tests for parameter speei&c:atioos

Shapiro-Wilk W test w = 0.955, p = 0.322
Homoscedasticity
I
Signiflcance tests for regression withithe SDs
A regression of SDs on age yielded an R2 _10.319 (Fu.to) = 10.62, p =0.009). Therefore, the SD
for the aggregate sample is suggested for~ with all age groups.
Education
Est. tau2 with education · 2.911
Number of observations 1 24
Number of clusters ' 11
R2 0.790
Education 0.2975954 0.154 1.93 0.083 -0.046 to 0.642
IQ .
Suflicient information for inclusion of IQ ~to regression analysis was not available.
Gender
Sufficient information for a t-test by gend~ was not available.
the Rey-Osterrieth Complex Figure (ROCF)
767
Table A12.1. Locator Table for the Rey-Osterrieth Complex Figure (ROCF)
Study Age• n Sample Composition IQ/ Education• Trials Reported Location i
ROCF.l Powell, 1979 41.0 Right-handed patients VIQ: %retention London
page 255 (14.05) 64 referred for neurological 107.70 on 40-minute
TableA12.2 screening, but brain (16.80) recall
damage was ruled out PIQ:
83.70
(21.55)
ROCF.2 King, 1981 39.6 71 Control group: healthy Education: Copy. Canada
page255 (21.4) volunteers or patients 11.4 40-minute recall,
TableA12.3 <30 with nonneurological (2.9) %recall,
30--60 or psychiatric conditions; FSIQ:
>60 divided into 3 age 104.5
subgroups (18.1)
ROCF.3 Huhtaniemi 19-29 25 Male college students College Copy, Maryland
et al., 1983 22 with good or poor students 3-minute recall
page256 attentional abilities (as
Table A12.4 determined by
performance on CPT)
were compared
ROCF.4 Bennett-Levy, 29.3 107 Volunteers: 76 M, 31 F: IQ: Copy, England
1984 (9.3) no history of head injury 104.9 40-minute recall
page256 17-49 or epilepsy (7.6) (quantified
TableA12.5 technique)
ROCF.S Speers & 35.00 40 20 M, 20 F normal Education: Copy, California
Hibbler, unpublished (10.79) volunteers, modified 16.15 immediate recall,
manuscript 23-70 ROCF was administered (2.77) 30-minute recall,
page257 10-22 24-hour recall
TableA12.6 Q.T.IQ: (Percent recall)
107.93
(8.73)
87-23
ROCF.6 Ardila et al., 200 Normal Spanisb-spealcing, Education: Copy Colombia >
"tt
1989 16-25 right-handed subjects: Dliterate "tt
m
page 258 26-35 data are stratified into vs. 2::10 z
Table A12.7 36-45 5 age ranges, by gender 0
46-55 and by education
-
X
N
ROCF.7 Van Gorp 57-85 156 Healthy elderly without Education: Copy, Los Angeles, >
"'tt
et al., 1990 history of neurological or 14.1 3-minute recall CA "'tt
m
page 259 57-65 28 psychiatric disorder; (2.9) z
TableA12.8 66-70 45 62 M, 94 F; FSIQ: 0
71-75 57 4 age groups 117.21 X
76-85 26 (12.59) ....
N
ROCF.8 Berry et al., 50-79 107 55 M, 52 F, elderly Education: Copy, Kentucky
1991 65 Caucasians without 15 immediate recall,
page 259 (2.9) cardiac, neurological, or (2.9) 30-minute recall
Table A12.9 psychiatric disease, or
psychoactive medication;
Rey and Taylor figures
were administered; test-
retest data for 1 year
interval are provided
for 41 subjects
ROCF.9 Tombaugh & 64 Study 1 compared 3rd-year Copy, Canada
Hubley, 1991 performance on ROCF and undergraduate immediate recall,
page260 Taylor figures; itemized students 4-minute recall,
Table A12.10, A12.11 scoring systems were used; 20-minute recall
67 study 2 addressed similar
issues with the addition of
2 scoring systems-itemized
and Osterrieth-Taylor
ROCF.IO Berry & 68 31M Healthy older Education: Copy, USA
Carpenter, 1992 (8.5) 29F volunteers with no history 15 immediate recall,
page 261 of neurological or (3.1) 4 delay intervals
Table A12.12 psychiatric illness;
divided into 4 equal groups,
each exposed to different delay
intervals (15, 30, 45, and 60 minutes)
(continued)
$
........
=
Study Age• n Sample Composition IQ/ Education• Trials Reported Location
ROCF.ll Delaney et al., 22-67 42 Study compared perfonnance 6-16 Copy, 7 V.A. Medical
1992 45.8 of nonnal adults on 12.8 immediate recall, Center
page 262 ROCF and Taylor ligures 20-minute recall facilities:
Table A12.13 in test-retest paradigm Cf,CA, FL,
with intervals of 1 month VA, MA, MN,
Ontario, Canada
ROCF.l2 Kuehn & Snow, 46.7 38 Study compared scores on Copy, Canada
1992 copy, recall, and percent recall 40-minute recall,
page 262 for ROCF and Taylor figures percent recall
Table A12.14 in a sample of patients Education:
referred for evaluation of 10.9
possible brain damage; FSIQ:
group 1 was presented 94.7
with ROCF first
Education:
Group 2 was presented with 13.5
Taylor figure first FSIQ:
93.4
ROCF.l3 Boone et al., 45-59 91 Fluent English-speaking, Education: Copy, Los Angeles,
1993b 6()...00 healthy older adults; 34 M, 14.5 3-minute recall, CA
page 263 70-83 57 F; 3 age groupings (2.5) %retention
Table A12.15 and 4 IQ levels FSIQ:
115.9
(13.0)
90-109
110-119
120-129
130-139
ROCF.l4 Chiulli et al., 70-93 153 Healthy elderly Education: Copy,
1995 without any serious medical immediate recall, >
page 264 70-74 46 illnesses and not taking 15.3 (2.4) 30-minute recall ""m
Table A12.16 75-79 58 medications; 3 age groups; 15.0 (3.6) z
80-91 49 data on proportion adopting 13.9 (3.0) 0
a conligural approach are provided X
_.
N
)>
ROCF.l5 Meyers & Group means: 30 in each of 4 groups Undergraduate students Copy, Iowa
"'tt
Meyers, 1995a 21.2-23.8 randomly assigned to 4 12.2-12.6 immediate or "'tt
m
page 264 experimental groups; 3-rninute recall, z
Table Al2.17 modified scoring procedure 30-minute recall 0
was used X
ROCF.l6 Ponton 38.4 300 Spanish-speaking 10.7 Copy, Los Angeles,
.....
N
et al., 1996 (13.5) healthy volunteers; (5.1) 10-minute recall CA
page 265 M1F ratio 40%/60%;
Table A12.18 16-29 data are partitioned by <10
30-39 gender (2) x age (4) x >10
40-49 education (2)
50-75
ROCF.l7 Rapport 18-84 318 Veterans referred to Copy,
et al., 1997 55.01 a V.A. hospital assessment 12.62 immediate recall
page 265 (14.31) service; majority were (2.77)
Table Al2.19 inpatients; 312 M, 6 F;
standard and Denman
scoring systems were
compared
ROCF.l8 Hartman 22.3 Two age groups were compared: 15.3 Copy, North Carolina
& Potter, 1998 18-32 students (13 M, 17 F) and healthy immediate recall
page 266 older adults (12 M, 18 F); BQSS and
Table A12.20 69.8 extended 36-point scoring system were 16.7
60-81 compared
ROCF.l9 Ostrosky- 20-29 15 A sample of 105 healthy Spanish- >6 Copy, Mexico City
Solis et al., 1998 30-39 15 speaking volunteers was partitioned immediate recall,
page 266 40-49 15 into 7 age groups 20-minute recall
Table A12.21 50-59 15
60-69 15
70-79 15
80-89 15
(continued)
...""
""
N

Study Age• n Sample Composition IQ/ Education• Trials Reported Location
ROCF.20 Fastenau 62.9 211 Healthy adults, 45% M, 95% 14.9 Indianapolis
et al., 1999 (14.2) Caucasian; Extended Complex (2.6)
page 267 Figure Test developed by the authors
Data are not reproduced was used; data for score conversion
in this book are presented in overlapping age groups
using midpoint interval technique
ROCF.21 Schreiber 29.5 18 Healthy controls (9 M, 9 F); BQSS and 15.1 Copy Boston,
et al., 1999 (11.5) 36-point scoring systems were (1.7) MA
page 267 compared
Table Al2.22
ROCF.22 Deckersbach 35.13 55 ontrol sample (38% M); 2 scoring 16.7 opy, Massachusetts
et al., 2000 (12.6) systems measuring organizational (2.3) immediate recall
page 268 approach and Meyers & Meyers' (1995 b)
Table Al2.23 system we re compared
ROCF.23 Miller, 40.4 729 Seronegative homosexual 16.2 opy, MACS
2003 (an update on (7.4) and bisexual males from (2.4) immediate recall, en te rs
Seines et al. , 1991) MACS; data 20-mioute recall
page 268 25-34 are partitioned by age x < 16
Table Al2.24 35-44 education 16
45-59 > 16
•Age column and IQ/education column contain information regarding range and/or mean and standard deviation for the whole sample and/or separate groups, whichever information is
provided by the authors.
)>
-o
-o
m
z
0
X
N
APPENDIX 12 773
Table A12.2. [ROCF.1] Powell, 1979: Percent

Retention of the ROCF Following a 40-Minute
Delay Compared to the Original Copy Score in
Individuals Referred for a Neuropsychological Eval-
uation but Cleared of Neurological Impairment
WAIS
Gender
n Age (MIF) VIQ PIQ Retention
64 41.0 43121 107.70 83.70 56.79

(14.05) (16.80) (21.55) (22.25)
Table A12.3. [ROCF.2] King, 1981: Data for the Total Sample and Three Age Groups of
Healthy Volunteers and Patients with Non-neurological or Psychiatric Conditions
ROCF
n Age Education WAIS FSIQ Copy 40-Minute Recall %Recall
Total 71 39.6 11.4 104.5 31.1 16.4 52.3

(21.4) (2.9) (18.1) (4.5) (7.1) (20.4)
36 <30 33.0 20.0 60.4
(2.8) (6.4) (18.1)
17 30-60 30.5 13.4 44.5
(4.7) (6.0) (19.9)
18 >60 27.8 12.2 44.3
(5.4) (5.9) (19.9)
Table A12.4. [ROCF.3] Huhtaniemi et al., 1983:

Data for Two Groups of Male College Students
ROCF
WAIS 3-Minute
FSIQ Copy Recall
Good attention 122 30.23 24.92

(5.2) (4.62) (5.62)
Poor attention 107 28.92 22.67
(12.0) (5.04) (6.81)
774 APPENDIX 12
Table A12.5. [ROCF.4] Bennett-Levy, 1984: Data for Copy Score, Strict and Lax 40-Minute De1ayed
Recall, Copy Time, Symmetry, Good Continuation, and Strategy Total for the Entire Sample (Consisting of
Medical Patients and Healthy Adults) and for Males and Females Separately
Good Strategy Copy Sbict Lax Copy
n Age IQ Symmetry Continuation Total Score Recall Recall Time
107 29.3 104.9 10.1 13.2 23.4 28.1 16.3 20.9 158.0
(9.3) (7.6) (3.2) (3.1) (5.0) (4.2) (5.2) (5.8) (51.5)
76 Males 29.3 104.0 10.5 13.7 24.2 28.6 17.1 21.9 159.1
(9.4) (7.9) (3.1) (3.1) (4.8) (3.9) (5.3) (5.6) (52.8)
31 Females 29.2 106.9 8.9 12.2 21.1 26.9 14.5 18.3 153.6
(9.1) (6.2) (3.2) (2.8) (5.0) (4.7) (4.7) (5.7) (47.8)
Table A12.6. [ROCF.5] Speers and Hibbler, Unpublished Manuscript: Data for a Sample of
Intact Participants on the Copy Condition Based on the Author's Unique Scoring System and
Percent Recalled on Three Recall Conditions
ROCF %retention
n Age Education Est. IQ" Copy Immediate 30 Minutes 24 Houn
40 35.00 16.15 107.93 99 87 84 84

(10.79) (2.77) (8.73)
(~70) (10-22) (87-123)
"IQ was estimated with the Quick Test.
Table A12.7. [ROCF.6] Ardila et al., 1989: Data

for the Copy Condition for a Sample• of Healthy
Colombians: Illiterates and Those Who Completed
at Least 10 Years of Education
Illiterate Educated
Age Men Women Men Women
16-25 21.7 19.7 35.5 35.1

26-35 25.2 16.7 35.6 35.3
36-45 25.3 13.1 34.3 34.1
46-55 22.6 14.5 34.9 35.2
56-65 12.0 10.6 34.7 34.8
•sample size is 10 for each cell.

APPENDIX 12 775
Table A12.8. [ROCF.7] Van Gorp et al., 1990:

Data for a Sample of Healthy Elderly
3-Minute
n Age VIQ PIQ Copy Recall
28 57~ 117.2 109.2 32.50 14.45

(11.3) (11.6) (4.7) (5.3)
45 ~70 114.8 111.5 32.93 14.13
(17.0) (16.8) (3.4) (7.8)
57 71-75 122.9 115.1 31.73 11.13
(11.4) (11.9) (3.4) (6.7)
26 76-85 110.6 101.0 30.14 8.41
(11.3) (8.8) (5.6) (5.9)
Table A12.9. [ROCF.8] Berry et al., 1991: Data for the ROCF and Taylor Figures for 54 Healthy Elderly
Subjects, as Well as for Baseline and 1-Year Follow-up ROCF Scores in a Subset of 41 Subjects•
ROCF (n=54) Taylor (n =54)
M/F 30-Minute 30-Minute

n Age Education Ratio Copy Immediate Recall Copy Immediate Recall
107 65 15 55152 33.2 23.4 22.5 32.9 24.8 23.3

(8.6) (2.9) (2.1) (6.1) (6.0) (2.3) (6.3) (7.1)
ROCF
Baseline (n =41) 1 Year (n=41)
30-Minute 30-Minute
Copy Immediate Recall Copy Immediate Recall
32.6 17.8 17.2 31.6 17.5 17.9

(2.4) (5.1) (5.1) (2.8) (5.1) (5.0)
•A modified scoring system was used in this study.
Table A12.10. [ROCF.9a] Tombaugh and Hubley, 1991: Data for the Sample of
Undergraduate Students on the Copy and Three Recall Conditions Based on the
Itemized Scoring Systems for the ROCF and Taylor Figures
Immediate 4-Minute 20-Minute

Figure n Copy Recall Delay Delay
ROCF 31 69.8 44.1 46.4 48.7

(2.1) (13.2) (13.4) (12.7)
Taylor 33 69.6 52.1 54.2 55.5
(1.5) (11.6) (10.5) (10.0)
776 APPENDIX 12
Table A12.11. [ROCF.9b] Tombaugh and Hubley, 1991: Data for a Sample of
Undergraduate Students on the Copy anfl Three Recall Conditions for the ROCF and
Taylor Figures Based on Two Scoring S)ttems
Immediate 20-Minute 30-Day
Figure n Copy Recall Delay Delay
Ifetlliud acoring .,.,_
ROCF 33 . 69.0 46.9 50.3 29.5
(1.8) (11.0) (12.2) (12.1)
Taylor 34 69.6 56.4 59.4 39.2
. (1.5) (11.6) (10.5) (10.0)
o.teniella-Taylor ICOring .,.,_

ROCF 33 34.9 23.5 25.5 14.6
(1.2) (5.1) (6.0) (6.1)
Taylor 34 35.1 28.6 30.3 19.8
(1.4) (5.9) (5.3) (8.1)
Table A12.12. [ROCF.10] Beny and C:ftnter, 1992: Data for a Sample of Healthy Elderly
on the Copy and Two Recall Conditions r Each Experimental Group (Based on Length of
Delay Interval)
Delay Sample Immediate Delayed
Period Size Age E~cation Copy Recall Recall
15 15 67.3 \5.1 30.8 19.3 19.2

(7.8) (2.6) (3.4) (3.9) (3.2)
30 15 69.2 15.2 31.0 19.1 18.4
(9.9) (3.6) (4.3) (7.6) (8.1)
45 15 67.5 15.2 32.5 22.6 22.1
(8.5) (2.7) (2.7) (6.3) (5.5)
60 15 67.4 15.3 33.4 20.1 18.9
(8.4) (3.4) (2.4) (7.5) (6.9)
Table A12.13. [ROCF.ll] Delaney et al., J992: Data for a Control Sample on
the Copy, Immediabl Recall, and 20-Minute Delayed Recall Conditions for the
ROCF and Taylor Figures
Age Education
n (Range) (Range) ROCF Taylor
Copy 42 45.8 12.8 33.8 33.6

~7 6-16 (2.1) (2.2)
Immediate recall 21.0 26.1
(7.8) (6.4)
20-minute delayed 20.8 25.7
recall (8.0) (7.2)
APPENDIX 12 777
Table A12.14. [ROCF.12] Kuehn and Snow, 1992: Data for Patients Referred for a
Neuropsychological Evaluation on Copy, 40-Minute Delayed Recall, and Percent Recall for the
Rey and Taylor Figures•
Groupt n Gender Education FSIQ ROCF Taylor
Copy 1 19 12M,7F 10.9 94.7 31.5 32.7

(4.0) (3.4)
2 19 1M, 12F 13.5 93.4 31.0 30.5
(6.5) (7.4)
Absolute recall 1 ll.2 9.2
(5.2) (7.5)
2 9.2 14.2
(6.3) (6.9)
Percent recall 1 35.2 29.5
(14.6) (22.1)
2 28.0 46.0
(16.8) (17.0)
"Mean age for the sample is 46.7 years.

tcroup 1, ROCF administered first; Group 2, Taylor administered first.
Table A12.15. [ROCF.13] Boone et al., 1993b: Data for a Sample of Healthy Elderly for
Three Age Groupings and Four Full-Scale IQ (FSIQ) Levels
n Age Education FSIQ Copy 3-Minute Recall %Retention
38 45-59 14.6 ll4.7 34.2 18.9 55.0

(2.6) (14.2) (1.8) (6.1) (17.1)
31 60-69 14.4 ll4.5 33.8 17.3 51.7
(2.1) (12.8) (2.8) (5.2) (13.8)
22 70-83 14.5 119.4 31.3 13.8 43.8
(2.9) (10.6) (4.7) (5.0) (14.8)
n FSIQ Age Education Copy 3-Minute Recall %Retention
32 ~109 60.3 13.4 32.6 15.2 46.3

(9.8) (2.2) (4.5) (4.9) (13.3)
23 110-119 62.1 14.2 33.5 16.7 49.6
(9.0) (2.2) (2.0) (5.6) (16.1)
21 120-129 63.0 15.1 33.7 18.9 56.0
(9.3) (2.4) (2.2) (5.4) (14.8)
15 130-139 62.9 16.4 34.3 19.4 56.0
(10.4) (2.4) (2.3) (7.6) (20.2)
778 APPENDIX 12
Table A12.16. [ROCF.l4] Chiulli et al., 1995:

Data for a Sample of Healthy Elderly for Three
ROCF Conditions
Age GrCiup
70-74 75-78 80-91
n 46 58 49
Age 72.7 82.4
(1.1) (3.0)
Educcdion 15.3 15.0 13.9

(2.4) (3.6) (3.0)
Gender(~) 52% 59% 49%

c.,
Accuracy 32.6 31.0 29.8
(2.8) (4.0)" (4.6)
Approach• 39% 36% 35%
IrruneditJte reccdl
Accuracy 17.2 14.2 12.9
(6.2) (6.6) (6.4)
Approach 55% 41% 40%
:JO..tninutedeltJyetlreccdl
Accuracy 16.9 14.2 12.4
(6.3) (6.2) (6.0)
Approach 55% 52% 41%
"Proportion of subjects adopting a configura! approach.
Table A12.17. [ROCF.l5] Meyers and Meyers, 1995a. Data for Undergraduate Students on the Copy
Condition and Different Combinations of Three Recall Conditions/Recognition Trial for Each Experimental
Group• (n=30 for Each Group)
Immediate 3-Minute 30-Minute
Group Age Education Gender Copy Recall Delay Delay Recognition
1 23.6 12.2 10M 34.7 26.7 26.6 21.9

(7.4) (0.6) 20F (1.7) (4.6) (4.4) (1.3)
2 21.2 12.4 17M 35.5 27.6 27.7 21.6
(4.2) (0.7) 13 F (0.9) (4.0) (3.9) (1.3)
3 23.8 12.6 11M 35.2 26.6 27.2 27.4 21.5
(5.4) (0.8) 19 F (1.0) (4.3) (3.6) (3.6) (1.5)
4 21.6 12.6 18M 35.5 25.3 20.9
(4.4) (0.9) 12 F (0.6) (3.7) (1.6)
"Scoring was based on the procedure developel by Meyers and Meyers (see text).
APPENDIX 12 779
Table A12.18. [ROCF.16] Ponton et al., 1996: Data for a Sample of 300 Spanish-Speaking Healthy
Participants Stratified by Gender, Age, and Education
Age Group
16-29 30-39 40-49 50-75
Education (Years)
<10 >10 <10 >10 <10 >10 <10 >10
Mala
n 11 25 13 18 12 17 18 6
Copy x 30.27 32.76 29.15 31.67 29.50 31.35 26.19 30.83
(SD) (4.13) (3.13) (5.68) (3.69) (4.21) (3.69) (4.96) (4.71)
Recall" x 18.32 21.34 14.77 22.17 16.58 21.71 14.06 18.67
(SD) (5.70) (6.31) (7.77) (6.54) (7.97) (5.62) (4.30) (8.85)
Femalea
n 12 30 22 44 16 11 25 20
Copy x 30.00 31.57 27.46 32.05 27.44 31.64 23.52 29.90
(SD) (4.09) (2.83) (6.24) (4.58) (4.77) (2.98) (7.97) (4.97)
Recall x 20.13 19.77 17.25 20.16 15.19 18.73 11.50 16.85
(SD) (7.47) (5.22) (6.28) (6.08) (5.46) (5.95) (6.26) (5.16)
"Ten-minute delayed recall.
Table A12.19. [ROCF.17] Rapport et al., 1997: Total Scores and

Individual Item Scores for a Sample of Patients Referred to the Veterans
Administration Hospital Assessment Service, Scored According to Denman
and Standard Systems
Denman Denman Standard Standard
Copy Recall• Copy Recall
Total acore
M 51.79 23.52 26.01 11.94
SD 15.57 15.35 7.89 7.64
Inditlitlual itema
M 2.16 0.98 1.44 0.66
SD 0.33 0.44 0.17 0.30
"Immediate recall.
Table A12.20. [ROCF.18] Hartman and Potter, 1998: Data• for Two Age Groups: Students
and Healthy Older Adults
Age MIF Immediate
Group n (Range) Education Ratio Copy Recall
Young 30 22.3 15.3 13/17 31.1 53.6

18-32 (3.6) (3.2)
Old 30 69.8 16.7 12/18 23.7 15.5
60-81 (5.2) (5.5)
•The extended 36-point scoring system was used.

780 APPENDIX 12
!
Table A12.21. [ROCF.19] Ostrosky-S~ et al., 1998: Data for Seven Age Groups• of
Healthy Spanish Speakers living in Mexif:o City
Age Mean MIF Immediate 20-Minute
Group Age Ratio Copy Recall Delayed Recall
20--29 24.4 1114 35.1 25.8 24.1
(2.9) (1.3) (4.9) (6.8)
30-39 32.8 5/10 35.0 24.1 24.6
(2.8) (1.1) (4.7) (4.4)
40-49 44.6 619 34.6 19.9 20.4
(3.1) (1.6) (4.8) (5.6)
50-59 54.2 4/11 34.2 19.2 16.8
(2.1) (1.5) (4.5) (6.9)
'
60-69 63.3 817 30.3 13.4 15.8
(2.8) (5.8) (7.4) (8.7)
70-79 74.8 619 29.4 12.2 10.8
(2.0) (1.1) (4.7) (4.8)
80-89 83.4 4/11 29.2 8.9 10.0
(3.1) (5.0) (4.3) (3.9)
"Each group included 15 participants.
I
Table A12.22. [ROCF.21] Schreiber e, al., 1999: Data for the
Control Group
n Age Education I MIF Ratio Copy
18 29.5
(11.5)
15.1
(1.7)
I 919 30.7
(3.4)
Table A12.23. [ROCF.22] Deckersbach fI al., 2000: Data• for the

Control Group I
% I Immediate
n Age Education Male ! Copy Recall
55 35.13 16.7 33.81 20.84
(12.6) (2.3) I (7.47)
(11
(2.71)
•scores are based on Meyers and Meyers' scoring system.
I
APPENDIX 12 781
Table A12.24. [ROCF.23] Miller, 2003 (An Up- Table A12.24. (Contd.)
date on Seines et al., 1991): Data for a Sample of
Education Immediate Delayed
Seronegative Homosexual/Bisexual Males Partici-
Age (Years) Copy Recall Recall
pating in the Multi-Center AIDS Cohort Study,
Stratified by Age x Education Total <16
Mean 34.14 21.50 21.32
Education Immediate Delayed (SD) (2.54) (6.59) (6.55)
Age (Years) Copy Recall Recall n 232 230 229
25-34 <16 16
Mean 34.54 22.65 22.07 Mean 34.55 22.24 22.13
(SD) (2.51) (6.80) (6.78) (SD) (2.07) (6.55) (6.63)
n 57 57 57 n 207 207 206
16 >16
Mean 34.81 22.83 23.03 Mean 34.84 23.23 22.69
(SD) (1.79) (6.71) (6.32) (SD) (1.99) (6.34) (6.34)
n 48 48 48 n 290 290 289
>16
Total
Mean 35.50 26.73 26.05
Mean 34.54 22.40 22.10
(SD) (0.81) (5.31) (5.29)
(SD) (2.22) (6.51) (6.50)
n 43 43 43
n 729 727 724
Total
Mean 34.91 23.90 23.54
(SD) (2.00) (6.59) (6.41)
n 148 148 148
35-44 <16
Mean 34.06 21.80 21.90
(SD) (2.49) (6.41) (6.15)
n lll 110 109
16
Mean 34.46 22.66 22.63
(SD) (2.00) (6.30) (6.47)
n Ill lll 110
>16
Mean 35.00 23.20 22.75
(SD) (1.72) (6.08) (6.07)
134 134 133
Total
Mean 35.00 22.60 22.45
(SD) (2.10) (6.62) (6.22)
n 356 355 352
4&-59 <16
Mean 33.92 19.93 19.63
(SD) (2.66) (6.51) (6.80)
n 64 63 63
16
Mean 34.47 20.68 20.10
(SD) (2.33) (6.86) (7.00)
n 48 48 48
>16
Mean 34.41 21.94 21.34
(SD) (2.47) (6.56) (6.57)
n l13 113 l13
Total
Mean 34.28 21.10 20.59
(SD) (2.59) (6.64) (6.74)
n 225 224 224
Appendix 12m: Meta-Analysis
Tables for the Rey-Osterrieth
Complex Figure (ROCF)
Table A12m.1. Results of the Meta-Analysis and Predicted Scores for the ROCF, Copy
Condition

Sample me
Mean 19 46.64 51.40 15-356
Age
Mean 19 62.73 19.27 21.2-82.4
SD 19 4.00 3.52 1.0-11.3
Education
Mean 19 14.33 0.98 12.2-16.2
SD 19 2.68 0.90 0.1--3.6
IQ
Mean 5 108.09 6.12 101.0-115.1
SD 5 11.74 3.06 8.~16.8
Percent male 11 53.92 21.04 33-100

Teat ~eore me1J118
Combined mean 19 32.20 1.79 29.~.5
Combined SD 19 3.59 1.46 0.6-5.6
tweighted means and SDs.
782
APPENDIX 12M 783
Predicted scores and SDs per age group (ROCF, Copy)•
95%CI 95%CI
Age Predicted Preclieted
.Range Seore Lower Band Upper band SD Lower Band Upper Band
JJ-J4 35.04 34.84 35.24 1.10 0.80 1.41

J5-J9 34.99 34.85 35.14 1.39 1.10 1.67
30-34 34.88 34.64 35.11 1.70 1.42 1.97
35-39 34.69 34.34 35.04 !.01 1.73 2.29
40-44 34.43 33.99 34.87 !.3! 2.02 2.63
45-49 34.11 33.60 34.61 !.64 2.29 2.98
50-84 33.71 33.17 34.26 !.95 2.56 3.34
5S-S9 33.!5 32.70 33.80 3.!6 2.82 3.70
60-64 3!.7! 32.19 33.25 3.57 3.08 4.07
~ 3!.11 31.63 32.60 3.89 3.33 4.44
10-14 31.44 31.03 31.86 4.!0 3.58 4.81
15-19 30.70 30.37 31.00 4.51 3.83 5.19

Predicted tat ecore=34.40434+0.0595862•age -0.0013855•age2
Predicted SD = - 0.333026 + 0.0625042 • age
Sigui&cance tests for regression with the test scores
Ordiaary least-squares regression of test meaos on age

(quaclratie)
R2 0.899
F<dO•P F<2.s> = 561.89, p < 0.000
Age 0.0595862 0.035 1.69° 0.130• -0.022 to 0.141

Age2 -0.0013855 0.000 -4.40 0.002 -0.002 to -0.001
Constant 34.40434 0.719 47.82 0.000 32.74 to 36.06
•significance test for age centered (sample means -aggregate mean): t = -25.20, p = 0.000.
Prediction
SEe 0.20
95%CI 33.11-33.90
(continued)
784 APPENDIX 12M
38
Rey-Osterrieth Complex Figure Copy. The size of the bubbles reflects the weight of the data point. with
larger bubbles indicating larger standard error and smaller weight.
Tests for heterogeneity in the 8oal data set

Q<dO·P Q(l8) = 354.90, p < 0.000
Moment-based estimate of between-study variance 0.977
Tests for model 8t-addition of a quadratic term
Linear 0.836 0.827 -9.399 -31.444

Quadratic 0.899 0.886 -15.563 -37.608
Tests for parameter speeiflcations
Shapiro-Wilk W test W=0.983, p=0.975
Homoscedasticity
White's general test 8.921, p < 0.063
APPENDIX 12M 785
Sigoi&cance tests for regression with the SD
Ordiaary least-squares regression of SDs on age (Unear)

R2 0.685
F<df), p Fo.s> = 79.10, p < 0.000
Age 0.0625042 0.007 8.89 0.000 0.046 to 0.079

Constant -0.333026 0.268 -1.24 0.249 -0.951 to 0.285
Prediction
SE., 0.22
95%CI 2.51-3.39
Education
Est tau2 with education 0.0000
R2 0.899
Education 0.0126504 0.163 0.08 0.940 -0.36 to 0.39
Gender
Information for the t-test by gender was not available.
Table A12m.2. Results of the Meta-Analysis and Predicted Scores for the ROCF,
Immediate Recall
Number or studies ineluded in the analysis 7

Yean of publieation 1991-2003
Number or data points used in the analysis 12
(a data point denotes a study or a ceO
(continued)
786 APPENDIX 12M

Variable n• so' Range
s-.pk.U.
Mean 12 60.63 73.51 15-356
Age
Mean 12 53.46 22.66 22.0-82.4
SD 12 ·5.61 4.09 1.0-11.3
Education
Mean 12 J4.47 1.34 12.2-16.2
SD 12 .2.36 1.21 0.1--'3.6
IQ
Mean 0
SD 0
Percent male 9 55.18 24.94 33-100
Tat ICOnl meana

Combined mean 12 *>.53 4.39 12.9-26.7
Combined SD 12 6.36 1.16 4.3-7.8
•Number of data points differs for different aqalyses due to missing data.
Predicted scores and SDs per a1e e't! (ROCF, Immediate Recall)•
95%CI 95%CI
Age Predicted Lower Upeer Predicted Lower Upper

Bl.mge Score Band Baad SD Band band
.2.2-24 14.91 22.34 27.$0 4.81 4.09 5.65

.25-.29 14.81 22.70 26.85 5.49 4.87 6.10
30-34 14.58 22.44 26.12 6.07 5.47 6.68
3S-39 14.18 21.76 26.81 6.55 5.85 7.26
40-44 13.64 20.92 26.36 6.93 6.11 7.74
4S-49 U.9$ 20.06 25.$4 7.19 6.30 8.09
50-lU u.n 19.21 25.• 1 7.35 6.43 8.27
5S-S9 11.11 18.39 23.$5 1.40 6.51 8.30
60-64 19.98 17.58 22.$8 7.35 6.54 8.16
65-69 18.69 16.70 20.&9 7.19 6.52 7.86
7~74 17.26 15.54 18.$8 6.91 6.42 7.42
75-79 15.67 13.67 17.&7 6.55 6.16 6.93
Predietetl le8t acore =23.5187 + 0.1292929 •t~Je - 0.0029745 • age2

Pretlkted SD = 0.34854 + 0.2456015 • age - 0.0021371 • age2
APPENDIX 12M 787
30
0
25
0 0
20
15
0
0
10
20 30 40 50 60 70 60
age
Rey-Osterrieth Complex Figure Immediate Recall. The size of the bubbles reflects the weight of the data
point, with larger bubbles indicating larger standard error and smaller weight.
Ordinary least square regression of test means on age (quadratic)

R2 0.822
F<dO· p F<2.6l = 17.69, p < 0.003
Age 0.1292929 0.246 0.53" 0.618" -0.472 to 0.731

Age2 -0.0029745 0.002 -1.24 0.260 -0.009 to 0.003
Constant 23.5187 5.229 4.50 0.004 10.72 to 36.31
"Significance test for age centered (sample means- aggregate mean): t = - 5.87, p = 0.001.
Prediction
SE. 1.22
95%CI 19.28-24.05
Tests for assumptions and model fit
Tests for heterogeneity in the 6nal dataset

Q<dO•P Qon=301.24, p < 0.000
Moment-based estimate of between-study variance 12.478
(continued)
788 APPENDIX 12M
Tests for model &t-rulitioo o£ a quadratic term
Linear 0.775 0.752 25.770 -15.397

Quadratic 0.822 0.782 25.440 -15.727
BIC' difference of 0.330 provides weak support for the quadratic model.
Tests £or parameter speci&catioos

Homoscedasticity
Significance tests for regression with the SD
Ordinary least-squares regression o£ SDs on age (quadratic)

R2 0.694
F<do.p F<2.6) = 7.58, p = 0.023
Age 0.2456015 0.074 3.31" 0.016" 0.064 to 0.427

Age2 -0.0021371 0.001 -3.09 0.021 -0.004 to-0.000
Constant 0.34854 1.621 0.22 0.837 -3.62 to 4.31
•significance test for age centered (sample means- aggregate mean): t = 2.39, p = 0.054.
Prediction
SE. 0.36
95%CI 5.94-7.37
Education
Number of obseiVations 12
Rz 0.823
Education -0.1463632 0.617 -0.24 .820 -1.66 to 1.36
Gender
Information for the t-test by gender was not available.
APPENDIX 12M 789
Table A12m.3. Results of the Meta-Analysis and Predicted Scores for the ROCF,
Long-Delayed Recall

Years of pubHcation 1991-2003
Sample size
Mean 11 62.07 75.73 15-356
Age
Mean 11 55.37 21.86 22.0-S2.4
SD 11 5.65 4.30 1.0-11.3
Education
Mean 11 14.64 1.26 12.2-16.2
SD 11 2.46 1.19 0.1--3.6
IQ
Mean 0
SD 0
Percent male 8 58.04 25.77 33-100
Test ecore means

Combined mean 11 19.95 4.38 12.4--26.6
Combined SD 11 6.67 1.17 4.4--8.1
790 APPENDIX 12M
Predicted scores and SDs, per age group• (ROCF,

Delayed Recallt)
95%CI
Age Range Predieted Score Lower Band Upper Band
JJ-J4 25.18 23.75 26.62

J5-J9 14.87 23.51 26.23
30-34 14.41 22.68 26.15
35-39 23.85 21.69 26.01
40-44 23.17 20.69 25.64 Standard cleviatioa for
45-49 22.38 19.75 25.01 all age groups is 6.67.
SO-S4 21.49 18.87 24.09
55-59 20.47 18.04 22.90
60-64 19.35 17.24 21.46
fJS.-69 18.12 16.38 19.86
10-14 16.78 15.22 18.33
15-19 15.33 13.40 17.25

Predicted test acore = 25.39903 + 0.0416485 • age -0.0022144 • age2
tThe predicted scores are relevant for the Copy-Immediate Recall-Delayed

Recall administration sequence (can be used with caution if 3-Minute Delayed
Recall is administered instead of Immediate Recall, but not both). The length of
the long-delay interval varies widely in the data reviewed (see text).
Ordinary least-squares regression of test means on age (quadratic)

R2 0.862
F<dO·P F<2.6) = 33.14, p < 0.0006
Age 0.0416485 0.207 0.20° 0.848• -0.466 to 0.549

Age2 -0.0022144 0.002 -1.05 0.332 -0.007 to 0.003
Constant 25.39903 3.982 6.38 0.001 15.65 to 35.14
•Significance test for age centered (sample means - aggregate mean): t = -5.96, p = 0.001.
Prediction
SEe 1.03
95%CI 19.27-23.30
APPENDIX 12M 791
30
25
20
15
0
10
20 30 40 80 70 80
age
Rey-Osterrieth Complex Figure Long-Delayed Recall. The size of the bubbles reflects the weight of the data
point, with larger bubbles indicating larger standard error and smaller weight.
Tests for heterogeneity in the fiDal data set

Q<dO•P Q(lo) = 302.99, p < 0.000
Tests for model 8t.-ddition of a quadratic term
Model BIC BIC'
Linear 0.836 0.817 21.213 -17.458

Quadratic 0.862 0.828 21.646 -17.025
BIC' difference of .433 provides weak support for the linear model.
Tests for parameter speei8eatioDs

Shapiro-Wilk W test w = 0.849, p = 0.041
Homoscedasticity
Signiflcance tests for regression with the SD
A regression of SDs on age yielded an R2 of 0.482 (F<2•6 ) = 3.38, p = 0.104). Therefore,

the SD for the aggregate sample is suggested for use with aH age groups.
(continued)
792 APPENDIX 12M
Education
Regression of test means on education ancl age
R2 0.863
Education -0.1076512 0.556 0.19 0.853 - 1.25 to 1.47
Gender ,
Information for the t-test by gender was nOt available.
Table A12m.4. Summary Table of Predifed Scores for the ROCF

Copy Immediate ReeaD Long-Delayed ReeaD
Age .Range Score SD Score SD Score SD
!2-.24 35.04 1.10 24.92 4.87 25.18 6.67

J5-J9 34.99 1.39 24.82 5.49 24.87 6.67
30-34 34.88 1.70 24.58 6.07 24.41 6.67
35-39 34.69 2.01 24.18 6.55 23.85 6.67
40-44 34.43 2.32 23.64 6.93 23.17 6.67
45-49 34.11 2.64 22.95 7.19 22.38 6.67
SO-S4 33.71 2.95 22.11 7.35 21.49 6.67
SS-59 33.25 3.26 21.12 7.40 20.47 6.67
6fJ.4J4 32.72 3.57 19.98 7.35 19.35 6.67
65-69 32.11 3.89 18.69 7.19 18.12 6.67
70-74 31.44 4.20 17.26 6.92 16.78 6.67
75-79 30.70 4.51 15.67 6.55 15.33 6.67
Appendix 13: Locator and Data
Tables for the Hooper Visual
Organization Test (HVOT)
scriptions ofthe studies in the text ofChapter 13. this appendix.
Table A13.1. Locator Table for the Hooper VISual Organization Test (HVOT)
HVOT.l Rao et al., 46.0 100 Control group of 13.3 Milwaukee, WI

1991a (11.6) rigorously screened (2.0)
page 275 participants (25 M, 75 F),
Table A13.2 paid for their participation
HVOT.I Libon 64-74 23 Healthy right-handed elderly 13.4 Philadelphia, PA
et al., 1994 participants (8 M, 15 F), (2.7)
page 276 75-94 14 (4 M, 10 F). 12.4
TableA13.3 (2.0)
HVOT.3 Richardson 81.5 101 Autonomously 11.0 New Haven, cr
&: Marottoli, 1996 (3.3) living, current drivers (3.7)
page 276 {53 M, 48 F); data are
Table Al3.4 76-80 provided for younger-old <12
81-91 and older-old by two 2:12
educational levels
HVOT.4 Walsh 73.2 32 Cognitively intact geriatric 11.7 Detroit, MI
et al., 1997 (7.7) rehabilitation inpatients (2.9)
page277 (10M, 22 F)
Table Al3.5
(continued)
793
794 APPENDIX 13
Study Age• n Sample Composition IQ/Education" Location
HVOT.5 Lichtenberg 76.9 74 Cognitively intact geriatric 10.8 Detroit, MI

et al., 1998 (5.9) rehabilitation inpatients (3.0)
page277 (19M, 55 F);
Table A13.6 38 African American,
36 European American
•Age column and IQ/education column contaqa information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, whicbe11er is provided by the authors.
Table A13.2. [HVOT.1] Rao et al., 1991a: Data for

a Control Sample
n Age Education MIF Ratio HVOTScore
100 46.0 13.3 25175 • 25.9

(11.6) (2.0) (2.4)
Table A13.3. [HVOT.2] Libon et al., 1994: Data for a Sample of Healthy Older Adults Stratified into Two
Age Groups
Age Group n Mean Age Educa~n MIF Ratio MMSE Score GDS Score HVOTScore
64-74 23 69.7 13.4 8/15 28.1 4.2 23.1

(3.3) (2.7) (1.1) (2.7) (4.1)
75-94 14 81.0 12.4 4/10 28.5 2.9 19.9
(4.3) (2.0) (1.0) (3.0) (3.4)
Table A13.4. [HVOT.3] Richardson and Marottoli, 1996: Data for a Sample of Healthy
Elderly Stratified by Two Age Groups x 1\vo Education Groups
Age/Education n Age Education %Male %Black HVOT Score
76-80 78.80 10.44 54.0% 18.0%

(1.07) (3.86)
<12 26 17.90
(4.01)
~12 24 21.69
(4.02)
81-91 84.08 1U>9 51.0% 2.0%
(2.56) (3.45)
<12 18 17.62
(6.17)
~12 33 19.71
(2.97)
APPENDIX 13 795
Table A13.S. [HVOT.4] Walsh et al., 1997: Data Table A13.6. [HVOT.5] Lichtenberg et al., 1998:
for a Sample of Cognitively Intact Geriatric Data for a Sample of Cognitively Intact Geriatric
Rehabilitation Patients Rehabilitation Patients
MIF HVOT African
n Age Education Ratio Score American/
European
32 73.2 11.7 10122 18.6 MIF American HVOT
(7.7) (2.9) (4.9) n Age Education Ratio Ratio Score
74 76.9 10.8 19155 38136 18.32

(5.9) (3.0) (4.03)
~
Tables for the Visual Form
Discrimination Test
these tables refer to study numben; and de- each study to a corresponding data table in
Table A14.1. Locator Table for the Vi$Ial Fonn Discrimination Test (VFDT)
Study Age n Sample Composition Education Location
VFDT.l Benton eta!., 19-54 58 Pati~ts without history Regarding total USA
1983b 55-74 27 oflbrain disease or sample of 85,
page 281 healthy subjects; 72 > 12 years,
Table A14.2 ala10st equal number 13 <12 years
of males and females
in'each age category;
~a are partitioned by
age and gender
VFDT.2 Campo & 18--39 222 Heakhy, unpaid volunteers Average education: Spain
Morales, 2003 40--59 175 li'f.ng in south and males 12.34 (south/southwest)
page 282 soothwest Spain; (4.14) years, females
Table A14.3 1$_ M, 206 F; all 11.85 (3.84) years;
inClependently functioning; no significant
ruaa are partitioned education difference
h)! age and education between men/women;
data reported by
age (18--39,
40--59 years)/education
(6-8, >9 years)
categories
796
APPENDIX 14 797
Table A14.2. [VFDT.l] Benton et al., 1983b: Mean, Median, and Range Per-
formance Data for a Sample of Adults Stratified by Two Age Groups by Gender
Age Group
19-54 55-74
Gender
Male Female Male Female

(n=28) (n=30) (n= 15) (n= 12)
VFDT
Mean 30.8 29.9 29.3 30.3
Median 31.0 30.0 30.0 31.0
Range 28-32 24--32 23-32 27--32
Benton et al.'s score interpretation is as follows: 26-32, within normal limits; 24 or 25,
borderline or mildly defective; ::::;23, severely defective.
Table A14.3. [VFDT.2] Campo and Morales, 2003: Means and SDs for each Variable of
the VFDT Stratified by Age and Years of Education
Age Group
18--39 40-59
Years of Education
6--8 >9 6--8 >9

(n=52) (n = 170) (n =63) (n=ll2)
Age
Mean 29.29 28.15 49.32 48.73
SD 5.70 6.25 5.26 5.39
l'eDrs of education
Mean 7.81 13.96 7.44 13.84
SD 0.49 3.33 0.84 3.20
Total score
Mean 30.27 30.72 28.81 30.16
SD 2.63 1.87 3.30 2.20
Correct respolllft
Mean 14.88 15.22 13.76 14.77
SD 1.64 1.15 2.40 1.46
Peripheral errors
Mean 0.50 0.37 1.29 0.62
SD 0.80 0.79 1.63 1.01
Diatortiota en-on
Mean 0.25 0.21 0.52 0.31
SD 0.52 0.45 0.69 0.57
Rotation errors
Mean 0.36 0.20 0.43 0.29
SD 0.79 0.45 0.71 0.51
Administered and scored according to Benton et al. (1983b); error types categorized according to
Kaskie and Storandt (1995).
the Judgment of Line Orientation Test
Table A15.1. Locator Table for the Judgment of Line Orientation Test (JLO)
Study Age• n Sample Composition IQIEducation• Location
JLO.l Benton 16-49 137 65 M, 72 F; general 12 Iowa

et al., 1983b 50-64 medical patients with no
page 288 65-74 evidence of brain disease
Table A15.2
JLO.J Eslinger 65-74 179 25 M, 162 F; healthy controls 13.7 Iowa
&: Benton, 1983 75-84 recruited from senior or retirement 12.5
page289 85-94 organizations with no history of 12.1
Table A15.3 neurological disease or psychiatric
hospitalization
JL0.3 Eslinger 73.1 53 25M, 28 F; healthy controls 12.0 Iowa
et al., 1985 60-88 recruited from senior-citizen and
page 289 community organizations; no
Table A15.4 history of neurological or
psychiatric disorder
JL0.4 Rao 42.8 40 M 10, F 30; volunteers recruited 14.0
et al., 1989 (8.1) from newspaper ads and (2.3)
page 289 screened for hypertension, cardiac or
Table A15.5 cerebrovascular disease, neurological WAIS-R
illness, head injury, substance abuse, Verbal IQ:
and psychiatric illness 108.1
(6.3)
798
APPENDIX 15 799

Study Age• n Sample Composition IQ/Education • Location
JLO.S Ska 55-64 95 M 19, F 76; healthy volunteers with 10.13 Canada
et al., 1990 65-74 no history of alcoholism, drug (3.38)
page 290 75-84 abuse, neurologic-al or psychiatric 9.46
Table A15.6 illness (3.40)
8.06
(2.77)
JL0.6 Rao 46.0 100 M 75, F 25; paid volunteers were 13.3 Wisconsin
et al., 1991 (11.6) screened with neurological evaluation (2.0);
page 290 and MRI; exclusion criteria
Table A15.7 included history of substance abuse, WAIS-R
psychiatric illness, head injury, and Verbal IQ:
other neurological disorders; 99 106.5
participants were Caucasian (6.9)
JLO. 7 Meader 20--42 12 M 8, F 4; paid volunteers were 12-20 Georgia
et al., 1993 (31) members of the staff at Medical (16)
page 291 College of Georgia; none had
Table Al5.8 history of neurological, psychiatric, or
"major" medical disease and none
used psychoactive drugs; study
used repeated measures design in
which participants were administered
either saline or scopolamine at least
72 hours apart
JL0.8 Kempen 65.2 13 M 3, F 10; volunteers with normal 16.5 California
et al., 1994 (5.9) vision were recruited during a (2.9)
page 291 routine ophthalmic exam by one of
Table A15.9 the authors at the UCSD School of
Medicine; the only exclusion criteria
was Snellen distance acuity of 20/50
JL0.9York & 61.89 15 M 6, F 9; volunteers were 15.07 Maine
Cermak, 1995 (8.67) orthopedic patients from one of two (3.41)
page 291 rehabilitation hospitals;
Table A15.10 screened for history of
cerebrovascular accidents and
other neurological deficits
JLO.IO Ivnik 50--59 2 M 71, F 145; volunteers were 5,7 Minnesota
et al., 1996 60--64 5 part of the MOANS project; 8-11
page 292 65-69 12 exclusion criteria were 12
Table Al5.11 70-74 24 psychiatric or neurological illness 13--15
75-79 24 1~17
80-84 69 ~18
85-89 40
90-94 16
95+ 3
JLO.ll Fleming 26.1 27 M 16, F 11; paid volunteers were 15.4 Washington,
et al., 1997 (7.4) recruited from the community and (3.2) D.C.
page 293 screened for substance use,
Table A15.12 psychiatric illness, neurological WAIS-R
disease, and other medical Prorated IQ:
diagnoses 101.0
(13.8)
JLO.l2 Finton 70.4 24 M 13, F 11; elderly control 15.7
et al., 1998 (6.0) participants were screened via (2.8)
page 293 physical and neurological exams for
Table A15.13 cognitive deficits;
(continued)
800 APPENDIX 15
JLO.l3 40.83 12 M 3, F 9; paid healthy volunteers NART: United Kingdom

Obonsawin et al., (12.55) were injected with a saline solution 35.09
1998 after abstaining from alcohol, tea, (7.23)
page 293 and coffee; all participants were
Table A15.14 screened via medical examination
JLO.l4 Meyers 36.70 30 M 14, F 16; healthy volunteers were 13.67
et al., 1999 (20.5) screened for a history of a number of (3.47)
page 294 conditions, including neurological
Table A15.15 disease, closed head injUI)', motor WAIS FSIQ:
vehicle accidents, 113.97
learning disabilities, and loss of (13.51)
consciousness
JLO.l5 Basso M: Total M 26, F 26; volunteers were M: 13.92
et al., 2000 22.04 52 undergraduate students screened via {1.01)
page 294 (3.53) interview for history of learning
Table A15.16 disability, neurological illness, F: 13.85
F: psychiatric disease, and head {1.05)
22.62 trauma; all were right-handed, and
(7.24) while most of the sample was
Caucasian, a small percentage were
non-white
JLO.l6 Bell 34.4 29 72% F; volunteers were 13.0
et al., 2001 (12.5) friends and family members of {1.7)
page 294 16--60 early-onset temporal lobe
Table A15.17 epilepsy patients; FSIQ:
exclusion criteria were 97.7
substance abuse, medical or {6.4)
psychiatric history affecting
cognition, use of psychotropic
medication, loss of eonsciousness
longer than 5 minutes, developmental
learning disorder, and repetition of a grade
JLO.l7 Montse 63.84 76 M 38, F 38; volunteers were friends 9.64 Barcelona,
et al., 2001 (9.93) and spouses of Parkinson's disease {4.17) Spain
page 295 39-85 patients who were free of neurological
Table A15.18 and psychiatric illness
JLO.l8 32.58 240 M 120, F 120; healthy hetero- and 16.27 London, UK
Rahman & (.'5.87) homosexual volunteers were (3 ..'54)
Wilson, 2003 recruited from King's College,
page 29.'5 local community, and
Table A1S.l9 social networks;
no exelusion criteria provided
JLO. 19 18-39 40 M 47%; healthy nonstudent 15..'53
Salthouse 40-59 38 volunteers with at least 11 {2.30)
et al., 1997 60-78 37 years of education; no other
page 296 exclusion criteria provided; short
Table A15.20 form was used
JL0.20 65.8 82 Volunteers were 14.0 New York
Woodward {6.7) geriatric individuals who were {2.3)
et al., 1998 55-84 "community dwellers;" sample was 9-20
page 296 healthy and not diagnosable for
Table A15.21 depression; 97% were
Caucasian; short form was used
APPENDIX 15 801
Table A15.2. [JL0.1] Benton et al., 1983a: Data• Table A15.3. [JL0.2] Eslinger and Benton, 1983:
for the Sample of General Medical Patients with Mean JLO T Scores for a Sample of Healthy
No Evidence of Brain Disease, Stratified by Elderly, Stratified into Three Age Groups
Gender and Three Age Groupings
Age MJF• JLO
Age Groups Group n Ratio Education T Scores
16-49 50-64 65-74 65-74 87 22165 13.7 53.0

(7.4)
Mala 75-84 67 10n5 12.5 48.0
Mean 25.6 24.3 22.7 (9.6)
n 27 17 21
85-94 25 3/22 12.1 44.3
Females (14.6)
Mean 23.3 22.2 20.8 •Average educational levels for males and females are
n 31 26 15 12.6 and 13.1, respectively.
•sos are not reported.
Table A15.4. [JL0.3] Eslinger et al., 1985: Mean

JLO Correct Responses for Normal Controls
n MIF Age Education JLO

Ratio (Correct)
53 25/28 73.1 12.0 24.8

(3.7)
Table A15.5. [JL0.4] Rao et al., 1989: Mean JLO Correct Responses for Normal Controls
WAIS-R JLO
n MIF Ratio Age Education Verbal IQ MMSE• (Correct)
40 10/30 42.8 14.0 108.9 29.8 26.9

(8.1) (2.3) (11.9) (0.5) (4.7)
•Mini-Mental State Exam (Folstein et al., 1975).
Table A15.6. [JL0.5] Ska et al., 1990: Mean JLO Correct Responses for a Sample of Healthy
Elderly Partitioned into Three Age Groups
Range of
Age Group Mean Age n MIF Ratio Education JLO (Correct) JLO scores
55-64 59.20 38 7/31 10.13 24.39 16--30

(2.97) (3.38) (3.98)
65-74 68.85 40 6/34 9.46 23.75 15--30
(2.90) (3.40) (3.64)
75-84 77.88 17 6/11 8.06 21.71 14-30
(2.62) (2.77) (5.02)
802 APPENDIX 15
Table A15.7. (JW.6] Rao et al., 1991: iData for a Table A15.11. (JW.10] Ivnik et al.,
Normal Control Sample 1996: Demographic Description of the
Sample Partitioned into Groups Used
MIF WAIS-R I JLO
in JW Testing
n Ratio Age Education Verbal IQI (Correct)
n
100 75125 46.0 13.3 106.5 27.2
(11.6) (2.0) (6.9) 1 (4.1) Age groupe
56-59 2
60-64 5
65-$ 12
Table A15.8. (JW.7] Meader et al., 1~: Mean 70-74 24
JW Correct Responses for Healthy Staff Employ- 75-79 45
80-84 69
ees of a Medical College '
85-89 40
MIF JLO 90-94 16
n Ratio Age Education (Correct) 95+ 3
12 814 31 16 27.33 Education

(range 20-42) (range 12-20) (2.35) ~7 1
8-11 48
12 60
13-15 51
16-17 33
Table A15.9. (JW.8] Kempen et al., 1~: Mean ;?:18 23
JW Correct Responses for Non-C~'vely Im- Gender
paired Patients o£ the University of C · rnia San
Males 71
Diego School of Medicine · Females 145
MIF JLO .RGce
n Ratio Age Education (Correct) Caucasians 215
-------------------------------- B~b
13 3110 65.2 16.5 25.0
(5.9) (2.9) (8.0) Bantleclneu
------------------------------- ru~t 201
Left 6
Mixed 9
Table A15.10. (JW.9] York and Ce~ 1995: TotGl 216

Data for Mean JW Correct Respon*s for a
Sample of Orthopedic Rehabilitation Pa~nts
n MIF Age Education JLO
Ratio (Correct)
15 619 61.89 15.07 24.6

(8.67) (3.41) (5.14)
Table A15.12. (JW.ll] Fleming et al., f)97: Data for Mean JW Correct Responses
for a Control Group :
WAIS-R WRAT-R JLO
n MIF Ratio Age Educatio1 Prorated IQ Reading (Correct)
27 16/11 26.1 15.4 . 101 101 24.1

(7.4) (3.2) . (13.8) (14.8) (4.9)
APPENDIX 15 803
Table A15.13. [JL0.12] Finton et al., 1998: Mean Table A15.17. [JL0.16] Bell et al., 2001: Mean
JLO Correct Responses for a Sample of Healthy JLO Correct Responses for a Control Sample
Participants
JLO
MIF DRS" JLO n Age Education %Male FSIQ" (Correct)
n Age Ratio Education Score (Correct)
29 34.4 13.0 28 97.7 24.7
24 70.4 13111 15.7 138.1 24.5 (12.5) (1.7) (6.4) (3.8)
(6.0) (2.8) (3.5) (3.7)
"WAIS-IU FSIQ seven-subtest short form.
"Dementia Rating Scale.
Table A15.14. [JL0.13] Obonsawin et al., 1998: Table A15.18. [JL0.17] Montse et al., 2001: Mean
Mean JLO Correct Responses for a Sample of JLO Error Scores for a Sample of Healthy
Healthy Participants from the United Kingdom Participants from Barcelona, Spain
Raven's MIF JLO
MIF NART- Matrix JLO n Ratio Age Education Errors
n Ratio Age Score Score (Correct)
76 38138 63.84 9.64 15.85
12 319 40.83 35.09 51.17 24.92 (9.93) (4.17) (7.30)
(12.55) (7.23) (7.81) (2.54) (range 39-85) (range 1-23)
"National Adult Reading Test.
Table A15.15. [JL0.14] Meyers et al., 1999: Mean

JLO Correct Responses for a Sample of Healthy
Participants
MIF WAIS JLO
n Ratio Age Education FSIQ (Correct)
30 14/16 36.70 13.67 113.97 26.70

(20.5) (3.47) (13.51) (3.16)
Table A15.16. [JW.15] Basso et al., 2000: Mean JW Correct Responses Stratified by
Gender for a Sample of Undergraduate Students
JLOt
n Age Education Estimated FSIQ" (Correct)
Males 26 22.04 13.92 112.67 25.27

(3.53) (1.01) (3.78) (4.08)
Females 26 22.62 13.85 112.08 22.12
(7.24) (1.05) (4.20) (4.13)
"Regression method developed by Barona et al. (1984).

tsDs were calculated from the 95% confidence intervals provided by the authors.
804 APPENDIX 15
Table A15.19. [JLO.l8] Rahman and Wilson, 2003: Mean JLO Correct Responses for
the Control Sample Stratified by Sexual Prientation and Gender
n Age ~ucation Raven's Matrix JLO (Correct)
H~
Males 60 29.91 '15.96 47.05 28.40
(6.60} . (3.29) (7.41) (1.84)
Females 60 26.80 :16.65 46.86 24.81
(5.87) i (3.29) (6.73) (3.43}
HOfiiONIICUtll
Males 60 32.08 16.51 44.83 24.18
(5.66) (3.86) (6.68) (3.75)
Females 60 29.61 15.95 45.55 25.56
(5.35) (3.71) (6.46} (3.75)
Short Form
Table A15.20. [JLO.l9] Salthouse et al., 1997

(Short Form): Mean JLO Correct Respqnses for a
Healthy Sample
Age Group~
18-39 40-59 60-78
n 40 38 37
%Male 42.5 50 48.6
Age 29.0 49.1 69.2
(4.8) (5.1) (5.1)
Education 15.5 15.2 15.3
(1.7} (2.5) (2.6}
JLO (Correct) 12.7 12.0 12.1
(1.7} (2.2) (2.3)
Table A15.21. [JL0.20] Woodward et' al., 1998 (Short Form): Mean JLO Correct
Responses for a Healthy, Nondepressed Geriatric Sample
JLO (Correct)
n Age Education MMSE" Score FormO FormE
82 65.8 14.0 27.5 11.6 11.5

(6.7) (2.3) (2.0} (2.4) (2.1)
(range 51H14) (range 9-20} ; (range 21-30) (range ~15) (range 5-15)
"Mini-Mental State Exam (Folstein et al., 19f5).

the Design Fluency Tests
these tables refer to study numbers and de- each study to a corresponding data table in this
scriptions of the study in the text of Chapter 16. appendix.
Table A16.1. Locator Table for the Design Fluency Tests

RFFl'.l Ruff et al., 28.2 50 Normal control volunteers 13.2 California

1986b (8.8) (1.7)
page 303
Table Al6.2
RFF1'.2 Ruff et al., 16-24 358 161 M, 197 F; recruited participants <12 California,
1987 25-39 were excluded if they had a history 13-15 Michigan
page303 40-54 of psychiabic hospitalization, chronic >16
Table Al6.3 55-70 polydrug abuse, or neurological disorder,
data are stratified by 4 age groups
and 3 educational levels
RFF1'.3 Demakis &: 19.1 134 61 M, 73 F; controls had no College
Harrison, 1997 (2.0) learning disabilities and were screened students
page 304 for current or past neurological or
Table Al6.4 psychiabic disease
RFF1'.4 Fama et al., 66.7 51 Normal controls; participants were 16.4 California
1998 (7.4) excluded from the study if they (2.3)
page 304 had a significant history of
Table Al6.5 psychiatric disorders, neurological PIQ:
illness, past or present history of alcohol 115.6
or substance abuse, or other serious (5.9)
medical conditions
(continued)
805
806 APPENDIX 16

RFFT.5 Berning 20 124 34 M, 90 F; undergraduate students were College Mississippi

eta!., 1998 recruited from psychology courses students
page 305 at the University of Mississippi and
Table A16.6 given course credit for
participation
RFFT.6 Demakis, 22.50 21 Undergraduate students (67% F); 13.60 Midwestern
1999 (7.99) data are partitioned by finn and (1.46) United States
page 305 mixed-handedness groups
Table A16.7
RFFT.7 Ross eta!., 23.90 90 College students (55% F) recruited from FSIQ: Midwestern
2003 (7.32) introductory psychology courses, 108.1 United States
page 305 received course credit for (9.2)
Table A16.8 participation; 44% Caucasian, 30%
Mrican American, 9% Hispanic, and
6% other; exclusion criteria were history
of neurological disorder, learning disability,
or psychiatric conditions involving
medication usage; 48 subjects were retested
an average of 35.2 days later
Design Fluency.l 35.8 16 7 M, 9 F; controls recruited; exclusion 15.2 Southern
Boone eta!., 1991 (13.7) criteria included history of alcohol or (2.8) California
page 306 drug abuse, head injury, seizure
Table A16.9 disorder, cerebral vascular disease or FSIQ:
stroke, psychiatric history, or any renal, 109.1
hepatic, or pulmonary disease (10.9)
Design Fluency.2 69.4 80 35 M, 45 F; volunteers were screened 14.6
Woodard eta!., (10.6) for dementia, current or past (3.4)
1992 neurological illness or injury, substance
page 306 abuse, drug use, and history of
Table A16.10 psychiatric disorder
Design Fluency.3 20-35 70 Subjects were divided into 2 age groups: 12.36 Canada
Daigneault et a!., 38M, 32 F (younger group); 30M, 28 F (2.09)
1992 45-65 58 (older group); French-speaking; exclusion 12.11
page 306 criteria included consumption of more (3.63)
Table A16.11 than 24 beers, 5 bottles of wine,
or 15 oz of spirits per week; consumption
of cocaine, LSD, or psychostimulants; any
neurological or psychiatric consultation,
psychoactive medication, head trauma
with hospitalization, or major surgery;
Design Fluency.4 34.7 20 9 M, 11 F; participants were screened for 13.6
Beatty et a!., 1993 (7.7) a history of central nervous system (1.4)
page 307 disease or injury, major medical illness,
Table A16.12 major psychiatric disorder, or current
alcohol or drug abuse; one of the
subjects apparently had a past history of
substance abuse
Design Fluency.5 27.7 87 28 M, 59 F; volunteers had no history of 14.4
Varney et a!., 1996 (13.1) neurological or psychiatric illness, loss (2.0)
page 308 of consciousness due to head trauma,
Table A16.13 or severe febrile illness
Design Fluency.6 19.1 134 61 M, 73 F; controls had no College
Demakis & (2.0) learning disabilities and were screened students
Harrison, 1997 for current or past neurological or
page 308 psychiatric disease
Table A16.14
APPENDIX 16 807

Design Fluency.7 25.06 66 19 M, 47 F; undergraduates were primarily 15.21 Ontario,

Carter et al., 1998 (7.83) recruited; inclusion criteria included (1.60) Canada
page 308 age 18-60, right-handed, English as
Table A16.15 6rst or main language. FSIQ >79, and FSIQ:
no signi6cant neurological, systemic, 100.85
or psychiatric illness (11.07)
Design Fluency.S 20 52 64 college students (91% F) enrolled at College Texas
Harter et al., 1999 Texas Tech University; 81% reported students
page 309 a history of neurological disorder; means
Table A16.16 and SDs were provided both for the
sample as a whole and with the 12
students with possible neurological
dysfunction excluded
Design Fluency.9 19.1 27 4 M, 23 F; undergraduates; one subject College Barcelona,
Mataix-Cols et al., (1.3} was left-handed; exclusion criteria students Spain
1999 included history of psychiatric disorder
page 309
Table A16.17
Design Fluency.IO 55.8 49 16 M, 9 F; right-handed controls 14.1 London, UK
Abrahams et al., (11.6} were screened for neurological disorder (3.1)
2000 or signi6cant head injury FSIQ:
page 309 114.6
Table A16.18 (9.9)
Design Fluency.ll 34.32 22 22 all-male, paid participants were recruited 13.36 Southern
Boone et al., 2001 (14.81) and screened for a history of learning (2.15) California
page 310 disability, major psychiatric disorder,
Table A16.19 substance abuse, or neurological FSIQ:
disorder 107.14
(15.89)
Table A16.2. [RFFf.l] Ruff et al., 1986b: Data for All Five RFFf Parts
Combined for Healthy Individuals from the San Diego, California, Region
RFFT Total RFFT
Unique Perseverative Error
n Age Education Designs Errors Ratio
50 28.2 13.2 103.0 5.8 0.06•
(8.8) (1.7) (23.0) (7.3}
·so is not provided.

808 APPENDIX 16
Table A16.3. [RFIT.2] Ruff et al., 19~: Unique Designs and Perseverations for All Five Parts of the
RFIT Stratified by Age Group and Eduyation Level for Healthy Individuals
1 Education (Years)
<12 12-15 >16

Age Unique Perseverative• ' Unique Perseverative" Unique Perseverative"
Group n Designs Errors n Designs Errors n Designs Errors
16-24 30 103.3 4.6 29 108.2 5.3 21 113.1 6.8

(23.8) (6.2) (18.6) (4.4) (21.2) (3.9)
25-39 31 82.1 6.9 36 99.9 7.1 34 104.1 5.7
(18.7) (7.5) (23.6) (5.8) (25.4) (5.9)
40--45 28 86.0 7.9 32 93.9 7.3 32 106.3 5.1
(22.4) (6.2) (17.8) (6.1) (16.2) (4.5)
55-70 27 67.5 7.4 31 77.2 9.0 27 84.4 9.3
(19.9) (6.2) (17.7) (7.6) (20.8) (6.4)
"The authors did not include 20 outlier score~ (~2 SD from the mean) for the entire sample in these mean values.
I
Table A16.4. [RFIT.3] Demakis and Harrison,

1997: Data for All Five Parts of the RFfl' Com-
bined for a Sample of College Students •
RFFf
Male/Female Unique
n Ratio Age Designs
134 61173 19.1 86.92

(2.0) (21.91)
Table A16.5. [RFIT.4] Fama et al., 1 : Data for All Five Parts of the RFIT
Combined for a Group of Healthy Indivi uals
EstnUted RFFf Uni~ue

n Age Education NARt• IQ Designs
51 66.7 16.4 115.6 28.8 76.1

(7.4) (2.3) (3.9) (1.1) (12.1)
"National Adult Reading Test (Nelson, 1982).

tMini-Mental State Exam (Folstein et al., 197Sf.
~Data are based on 50 participants. )
APPENDIX 16 809
Table A16.6. [RFFT.5] Berning et al., 1998: Data For Each Part of the RFFT for a
Sample of Undergraduate Students•
Median RFFf
n Age Measure Part I Part II Part III Part IV PartV Total
124 20
Unique Designs 17.8 19.6 19.6 20.2 21.7 99.0
(4.7) (4.5) (4.2) (4.4) (4.5) (18.9)
Perseverations 0.9 1.0 1.3 1.4 1.6 6.2
(1.7) (1.3) (1.7) (1.9) (1.8) (5.6)
Error Ratio 0.05 0.05 0.07 0.07 0.08 0.06
(0.13) (0.07) (0.08) (0.12) (0.09) (0.05)
•The sample includes 34 males and 90 females.
Table A16.7. [RFFT.6] Demakis, 1999: Test-Retest Data for Five RFFT Parts
Combined for a Sample of Undergraduate Students
RFFf Total Unique Designs
n Age Education %Female Initial Test Retest (3 Weeks)
21 22.50 13.60 67 100.9 117.7

(7.99) (1.46) (24.5) (26.9)
Table A16.8. [RFFT.7] Ross et al., 2003: Initial Test and !-Month Retest Data from College
Undergraduates (55% Female) in the Midwest
RFFf Scores
Testing Session n Age FSIQ Total Designs Perseverative Error Ratio
Initial testing 90 23.90 108.10 106.3 8.4 0.0790

(7.32) (9.20) (23.1) (8.8) (0.0521)
Retest 48 114.5 8.2 0.0753
(24.6) (8.1) (0.0695)
Table A16.9. [Design Fluency.!] Boone et al.,

1991: Mean Scores for Free Condition in Controls
from Southern California (9 Women, 7 Men)
Total
Unique
n Age Education FSIQ Designs
16 35.8 15.2 109.1 26.0

(13.7) (2.8) (10.9) (11.0)
810 APPENDIX 16
Table A16.10. [Design Fluency.2] Woodard et al., 1992: Mean Scores for Free and
Fixed Conditions in 80 Older Normals (35 Men, 45 Women)
Free condition Fixed condition
Wrong Number
Novel Nameable Perseverative Novel Nameable Perseverative of Lines
17.0 1.9 8.6 12.0 0.8 4.6 1.4

(9.3) (2.2) (7.5) (7.1) (1.3) (4.2) (2.1)
Mean age= 69.4 (10.6), mean education= 14.6 (3.4), mean W AIS-R Vocabullii)' scaled score= 10.0.
Table A16.11. [Design Fluency. 3] Daigneault et al., 1992: Data for French-
Speaking Healthy Canadian Volunteers
Age Free Perseverative
Group n Age Education Condition Errors
20-35 70 27.71 12.36 24.33 0.01

(4.05) (2.09) (12.40) (0.08)
44-65 58 56.62 12.11 28.48 0.01
(5.29) (3.63) (15.00) (0.02)
Table A16.12. [Design Fluency.4] Beatty et al.,

1993: Data for Fixed Condition in Controls {9 Males,
11 Females)
Fixed Rule
n Age Education Condition Violations
20 34.7 13.6 25.4 0.6

(7.7) (1.4) (12.6) (2.0)
Table A16.13. [Design Fluency.5] Varney et al.,

1996: Data for the Free Condition in Controls
(28 Males, 59 Females)
n Age Education Free Condition
87 27.7 14.4 16.1

(13.1) (2.0) (9.13)
1S-77 12-21 4--51
Table A16.14. [Design Fluency.6] Demakis and Harrison, 1997: Data for the Free
and Fixed Conditions in College Students
Male/Female Free Fixed Total
n Ratio Age Condition Condition Designs
134 61173 19.1 18.58 18.24 36.82

(2.0) (9.52) (5.94) (13.42)
APPENDIX 16 811
Table A16.15. [Design Fluency.7] Carteret al., 1998: Data for the Free and Fixed Conditions
in College Undergraduates in Canada•
Free Condition Fixed Condition
n Novel Perseverations Nameable Novel Perseverations Nameable Not 4 Lines
66 13.9 7.1 0.2 16.7 6.4 0.4 2.5

(6.3) (7.8) (0.4) (6.1) (5.5) (0.8) (2.3)
•sample of 19 males, 47 females; age= 25.06 (7.83); education= 15.21 (1.60); FSIQ = 100.85 (11.07).
Table A16.16. [Design Fluency.8] Harter et al., Table A16.18. [Design Fluency.10] Abrahams et al.,
1999: Data for the Free Condition in College 2000: Data for the Free and Fixed Conditions in
Undergraduates (91% Female) Right-handed Older Controls (16 Males, 9 Females)
Total Free Fixed
Unique Scribbled n Age Education FSIQ Condition Condition
n Age Designs Perseverations Responses
25 55.8 14.1 114.6 25.0 22.2
52 20 8.92 0.00 0.06 (11.6) (3.1) (9.9) (7.8) (7.7)
(5.03) (0.00) (0.23)
Table A16.17. [Design Fluency.9] Mataix- Table A16.19. [Design Fluency.ll] Boone et al.,
2001: Data for the Free Condition in Male Controls
Cols et al., 1999: Data for College Under-
from Southern California
graduates (23 Women, 4 Men) in Spain
Total
Free Fixed
Unique
n Age Condition Condition
n Age Education FSIQ Designs
27 19.1 25.1 10.8
22 34.32 13.36 107.14 29.00
(1.3) (13.9) (2.8)
(14.81) (2.15) (15.89) (15.14)
Appendix 1.7: Locator and Data Tables for
the Tactual [Performance Test
scriptions of studies in the text of Chitpter 17. this appendix.
Table A17.1. Locator Table for the Ta~al Performance Test (TPT)
TPT.I Halstead, 1947 15-50 28 14 subjects without psychiatric Education: Chicago

page 318 diagnosis or history of 7-18
Table A17.2 brain injury; 14 subjects with IQ:
psychiatric diagnosis 70--140
TPT.2 Reitan, 1955b, 32.36 50 35 M, 15 F volunteers Education: Indiana
1959 (10.78) hospitalized with paraplegia 11.58 (2.85)
page 318 and neurosis FSIQ:
Table A17.3 112.6 (14.3)
TPT.3 Reitan & Wolfson, No information provided USA
1985 regarding the normative
page 319 sample; cutoffs for "severity
Data are not reproduced ranges" (perfectly normal,
in this book normal, mildly impaired,
seriously impaired) are
presented
TPT.4 Klove & Lochen American, and Norwegian Education: Wisconsin,
(in Klove, 1974) 31.6 22, controls; no exclusion 11.1 Norway
page 320 32.1 22. criteria reported 12.2
Table A17.4 FSIQ:
109.3
111.9
812
APPENDIX 17 813

Study Age• n Sample Composition IQ!Education• Location
TPI'.5 Wiens & Matarazzo, 48 All males, neurologically nonnal; Education: Portland,
1977 23.6 divided into 2 groups; 13.7 OR
page 320 21-27 random sample of 29 were 14.0
Table A17.5 24.8 retested 14-24 weeks later FSIQ:
21-28 117.5
105-139
118.3
108-131
TPI'.6 Cauthen, 1978a Sample: 117 35 M, 82 F subjects recruited from WAIS IQ: Canada
page 320 20-60 hospital volunteers and 91-111
Table A17.6 Groups: service clubs; all but 3 112-122
20-29 were right-handed; data 123-139
30-39 represented in age x IQ ceUs
40-49
50-60
TPI'.7 Harley et al., 1980 Sample: 193 V.A. hospitalized patients; Education: Wisconsin
page 321 55-79 T-score equivalents 8.8
Table Al7.7 Groups: reported IQ:>80
55-59 56
60-64 45
~9 35
70-74 37
75-79 20
TPI'.S Pauker, 1980 Sample: 363 Volunteers fluent in English; WAIS IQ: Toronto,
page 322 19-71 152 M, 211 F; no 89-102 Canada
Table A17.8 Groups: physical disability, sensory 103-112
19-34 de6cit, current medical 113-122
35-52 illness, brain disorder or 123-143
53-71 alcoholism; data
presented in age x IQ cells
TPI'.9 Anthony et al., 38.88 100 Nonnal volunteers, no history of Education: Colorado
1980 (15.80) medical or psychiatric 13.33 (2.56)
page323 problems, head injury, brain FSIQ:
Table A17.9 disease or substance abuse 113.5 (10.8)
TPI'.IO Bak & Greene, Participants were equally divided Education: Texas
1980 50-62 15 in 2 age groupings; subjects were 13.7 (1.91)
page323 55.6 (4.44) fluent in English and denied
Table A17.10 67-86 15 history of neurological 14.9 (2.99)
74.9 (6.04) problems; 1st group had 9 F,
2nd group had 10 F
TPI'.ll Fromm-Auch & Sample: 190 111 M, 82 F; participants Education: Canada
Yeudall, 1983 15-64 described as nonpsychiatric 8-26
page323 25.4 and nonneurological; 83% 14.8 (3.0)
Table A17.11 (8.2) right-handed; 5 age groupings FSIQ:
Groups: 119.1 (8.8)
15-17 32
18-23 74
24-32 56
33-40 18
41-64 10
(continued)
814 APPENDIX 17

TPT.l2 Moore 284 Data for healthy adults recruited FSIQ: Canada
et al., 1984 19-27 56 through newspaper ads 115
page 324 28-36 64 are partitioned into 6 age groups; 112
Table A17.12 37-45 59 various performance measures and 111
46-55 60 time to completion are reported 116
56-65 20 115
66-76 25 115
TPT.l3 Schear, 1984 Sample: 556 Neuropsychiatric sample; 35% Education: Kansas
page 325 20-69 had evidence of various signs 2-22
Tables A17.13, Al7.14 Groups: of organic brain syndromes,
20--29 111 alcohol encaphalopathy,
3()...39 112 epilepsy, etc; 49% exhibited
-ID-49 111 nonorganic psychotic
50-59 155 disorders, schizophrenia,
60-69 67 alcoholism, etc; 5 age
decades are represented
TPT.l4 Russell, 1985 43.5 19 Caucasian controls admitted to Education: Miami
page 325 (13.6) a neurology ward for suspected 14.8
Table A17.15 neurological condition but (6.4)
showed no evidence of brain
damage; all but 2 were
male; 6-block version used;
17M, 2 F
TPT.l5 Heaton et al., 15-81 553 356M, 197 F; exclusion criteria Education: Colorado,
1986 39.3 included history of neurological ~20 California,
page 326 (17.5) illness, significant head trauma, 13.3 (3.4) Wisconsin
Table Al7.16 <40 319 and substance abuse; sample <12 (132)
~9 134 was divided into 3 age groups 12-15 (249)
;::60 100 and 2 education groups; % ;::16 (172)
classification as normal
provided
TPT.l6 Alekoumbides 19-82 Ill Medical and psychiatric V.A. Education: s. California
et al. 1987 46.85 inpatients & outpatients without 1-20
page 326 (17.17) cerebral lesion or history 11.43 (3.20)
Table Al7.17 of alcoholism or cerebral FSIQ:
contusions; all except for
one were male 105.9 (13.5)
TPT.17 Bomstein et al., 17-52 23 Volunteers: 9 M, 14 F; no history Verbal IQ:
1987a 32.3 of neurological or psychiatric 88--128
page 327 (10.3) illness; test-retest data for a 105.8 (10.8)
Table Al7.18 3-week interval are Performance IQ:
provided 85--121
105.0 (10.5)
TPT.l8 EI-Sheikh et al., 17-24 32 Undergraduate and Cairo,
1987 20.6 graduate students with no Egypt
page 327 (1.4) history of brain damage;
Table A17.19 test-retest data are provided
TPT.l9 Dodrill, 1987 27.73 120 60 M, 60 F volunteers; data Education: Washington
page 328 (11.04) for various intelligence 12.28 (2.18)
Table A17.20 levels are presented FSIQ:
100 (14.4)
APPENDIX 17 815

Study Age" n Sample Composition IQ!Education• Location
TPI'.JO Yeudall Sample: 225 Volunteers classifled in 4 Education: Canada

et al., 1987 15-40 age groupings; 8891> were 14.55 (2.78)
page 328 Groups: right-handed; 127 M, 98 F FSIQ:
Table Al7.21 15-20 112.25 (10,25)
21-25
26-30
31-40
TPI'.Il Ernst, 1987 65-75 110 51 M, 59 F volunteers Education: Brisbane,
page329 69.6 (2.7) 10.3 Australia
Table Al7.22
TPI'.U Clark &: Klonoff, 35-68 79 All male, right-handed, WAIS-R FSIQ: Canada
1988 55.5 (8.0) coronary bypass surgery 105.9 (12.2)
page330 patients; test-retest data;
Table Al7.23 6-block version was used
TPI'.I3 Elias et al., 138 Healthy participants; data are Education: Maine
1990 20-31 partitioned into 3 age 15.4
page 331 37-49 groups 15.7
Table Al7.24 55-67 14.9
TPI'.M Thompson &: 39.43 489 Healthy volunteers Education: California,
Heaton, 1991 (17.76) 13.19 (3.46) Colorado,
page 331 FSIQ: Ohio,
Table Al7.25 113.09 (12.07) Michigan
TPI'.IIS Heaton et al., 486 Volunteers: urban and rural; Education: California,
1991,2004 42.1 data collected over 15 years 13.6 (3.5) Washington,
page 332 (16.8) through multicenter collabo- FSIQ: Teras,
Data are not Groups: rative efforts; strict exclusion 113.8 (12.3) Oklahoma,
reproduced in this book 20-34 criteria; 6591> M; data are Education groups: Wisconsin,
35-39 presented in T-score 6-8 Illinois,
40-44 equivalents for M and F 9-11 Michigan,
45-49 separately in 10 age 12 New York,
50-54 groupings by 6 education 13-15 Virginia,
55-59 groupings; in the 2004 16-17 Massachusetts,
60-64 edition, age range is 18+ Canada
65-69 expanded to 85 years, and the
70-74 data are presented for
75-80 African-American and
Caucasian participants
separately
TPI'.16 Elias et al., Groups: 427 Healthy participants; data are Education: Maine
1993 15-24 partitioned into 6 age groups 12-19
page 332 25-34 by gender
Table Al7.26 35-44
45-54
55-64
;::65
TPI'.I7 Barrett 43.9 1052 Air Force veteran controls; High school
et al., 2001 (7.6) presumably all male; and college
page 333 SDs for the test data
Table A17.27 are not provided
816 APPENDIX 17
Table A17.2. [TPT.1] Halstead, 1947: Data for the Control Group
(Including Patients with Psychiatric Diamoses): Mean Total Time for
the Three Trials and Mean Scores For ~mory and Localization for
the Total Group and for Each Subgroup'
TPr
n Total I
Memory U>calization
Total 30" 10.8 8.2 5.7

Civilian 14 9.5 8.4 6.1
{5.2-19.6) {6-10) {2-10)
Military 10 12.0 7.8 5.6
{6.3-15.9) {6-9) {1-8)
Miscellaneous 6 11.7 8.2 4.9
{5.~13.3) {7-10) {1-7)
"Two participants were tested twice.
I
Table A17.3. [TPT.2] Reitan, 1955b, ~959: Data for Individuals Referred for Neuropsychological
Evaluation, with Negative Neurological Findings: Mean Total Time for the Three Trials and Mean Scores
for Memory and Localization !
WAIS TPr
n Age Education VIQ PIQ FSIQ Total Memory Localizatioo
50 32.36 11.58 110.82 112.18 112.64 12.59 7.65 5.29

{10.78) {2.85) {14,46) I {14.23) {14.28) {5.20) {1.41) {2.12)
Table A17.4. [TPT.4] Klove and Lochen tiD

Klove, 1974): Data for American and Norwegian
Controls: Means for Total Time in Minutef. Memory, and Localization Scores for Each Group
TPr
n Age Education WAIS IQ Total Memory U>calization
Americans 22 31.6 11.1 109.3 14.0 7.2 4.3

Norwegians 22 32.1 12.2 111.9 13.7 7.5 5.2
APPENDIX 17 817
Table A17.S. [TPT.S] Wrens and Matarazzo, 1977: Data for Male Applicants to Patrolman Program: Mean
Time in Minutes and SD to Complete the TPT with the Preferred Hand, Nonpreferred Hand, Both Hands,
and Total, as Well as Mean and SD for Memory and Localization Scores for Two Equal Groups of Subjects•
TPT
n Age Education WAIS FSIQ Preferred Nonpreferred Both Total Memory Localization
24 23.6 13.7 117.5 4.85 3.02 1.87 9.74 8.46 5.67
(21-27) (12-16) (8.3) (1.92) (1.20) (0.89) (3.16) (0.98) (1.74)
24 24.8 14.0 118.3 4.38 2.97 1.83 9.18 8.67 6.13
(21-28) (12-16) (6.8) (1.24) (1.03) (0.84) (2.42) (0.76) (2.61)
Test Retest
n Age Education FSIQ Total Memory Localization Total Memory Localization
29 24 14 118 9.36 8.38 5.34 8.19 8.72 7.10
(21-28) (12-16) (2.73) (0.82) (2.41) (2.70) (0.88) (1.82)
•The data for a subset of 29 subjects include means and SDs for total time in minutes, memory, and localization scores for
boththe original testing and retest.
Table A17.6. [TPT.6] Cauthen, 1978a: Data for Canadian Volunteers Presented in Four Age
by Three IQ Groupings: Mean Times in Minutes and SD for Preferred Hand, Nonpreferred
Hand, Both Hands, and Total, as Well as Means and SD for Memory and Localization
n Age IQ Preferred Nonpreferred Both Total Memory Localization
18 20-29 91-111 5.1 4.3 3.0 12.4 6.8 3.8
(1.4) (1.3) (1.2) (3.2) (1.4) (2.1)
11 112-122 4.3 3.6 2.2 10.1 8.1 5.5
(0.9) (1.4) (0.9) (2.0) (1.6) (2.1)
14 123-139 4.7 3.0 1.8 9.6 8.2 6.4
(1.2) (1.1) (0.7) (2.1) (0.9) (1.6)
5 30-39 91-111 7.4 4.0 2.5 3.9 5.8 4.2
(2.5) (2.3) (0.9) (5.3) (1.8) (0.8)
17 112-122 6.4 4.4 2.5 13.2 7.8 5.5
(2.8) (2.5) (1.5) (5.9) (1.1) (2.0)
6 123-139 4.5 3.4 1.9 9.5 8.5 5.2
(1.1) (0.9) (0.3) (2.3) (1.0) (3.2)
9 40-49 91-111 6.9 5.5 2.9 15.3 6.1 3.3
(2.2) (1.8) (0.8) (3.5) (1.3) (1.9)
8 112-122 5.4 4.6 3.0 12.9 7.6 4.1
(1.9) (1.7) (2.1) (5.2) (1.1) (2.2)
10 123-139 5.7 4.0 1.6 11.4 6.7 4.5
(1.8) (1.5) (0.4) (3.4) (1.5) (2.1)
7 50-60 91-111 8.5 6.2 3.9 18.6 5.3 2.1
(2.6) (1.8) (2.1) (3.3) (1.4) (1.1)
4 112-122 7.1 5.3 3.4 16.0 6.5 3.8
(3.1) (2.8) (2.0) (7.0) (3.1) (3.1)
8 123-139 6.3 7.4 3.5 17.2 5.5 2.8
(1.9) (3.1) (1.2) (5.0) (1.1) (1.8)
818 APPENDIX 17
Table A17.7. [TPT.7] Harley et al., 1980: Data for Veterans Administration-Hospitalized Patients: Means
and SDs for Time in Minutes per Block for Dominant and Nondominant Hands, Both Hands as Well as
Total TIDle per Block for the Three Trial!; Combined by Age Groupings
n Age WAIS FSIQ Education Dominant Nondominant Both Total
Totclliii.UIIple
56 55-59 98.57 10.1 2.53 2.30 1.83 2.21
(11..43) (2.42) (2.54) (2.70) (3.31)
~129
45 60-64 98.68 9.8 2.27 1.94 1.77 2.70
(9.93) (1.73) (1.35) (1.93) (3.81)
~121
35 65-69 97.51 8.7 3.56 2.37 2.38 2.17
(11.18) (3.26) (1.67) (2.16) (1.86)
~laG
37 70-74 100.41 8.8 3.68 4.24 2.43 3.25
(9.!E) (3.24) (3.77) (2.59) (3.41)
82-125
20 75-79 IOU5 6.5 4.30 2.15 2.06 2.63
(IO.i8) (2.64) (2.51) (1.43) (1.42)
81-119
Alcolwl-equaled .....,Ze
47 55-59 99.00 10.1 2.72 2.47 1.99 2.40
(ll.'IJ) (2.58) (2.72) (2.89) (3.57)
~~9
33 60-64 96.00 9.3 2.41 2.10 2.03 2.94
(9.43) (1.81) (1.41) (2.00) (3.92)
~117
23 65-69 99.00 8.8 3.45 1.93 2.10 1.72
(12.06) (3.16) (1.29) (2.14) (1.47)
~ISO
37 70-74 100.00 8.8 3.68 4.24 2.43 3.25
(9.99) (3.24) (3.77) (2.59) (3.41)
82-1t5
20 7~79 102.00 6.5 4.30 3.15 2.06 2.63
(10.18) (2.64) (2.51) (1.43) (1.42)
81-U9
APPENDIX 17 819
Table A17.8. [TPf.8] Pauker, 1980: Data for Canadian Volunteers: Means and SDs for
Total Time in Seconds, as Well as Memory and Localization Scores for the Sample as a
Whole and for Three Age Groupings by Four WAIS IQ Levels
WAIS IQ
Age 89-102 103-112 113-122 123-143 89-143
19-34 n=21 n=53 n=60 n=28 n=162

Total 932.57 830.79 692.95 616.71 755.93
(294.57) (298.21) (267.62) (174.93) (285.73)
Memory 7.10 7.70 8.33 8.61 8.01
(1.48) (1.41) (1.13) (1.07) (1.35)
Localization 4.71 4.66 5.68 6.36 5.34
(2.19) (2.46) (2.39) (2.09) (2.41)
35-52 n=20 n=34 n=56 n=25 n=135
Total 1139.80 961.79 888.11 788.64 925.53
(257.07) (320.49) (288.75) (344.50) (318.45)
Memory 6.30 6.74 7.43 8.16 7.22
(1.66) (1.56) (1.31) (1.43) (1.56)
Localization 2.40 3.38 4.05 4.08 3.64
(1.60) (2.16) (2.12) (2.31) (2.16)
53-11 n=4 n=15 n=27 n=20 n=66
Total 1591.25 1342.60 1108.70 1189.90 1215.71
(417.50) (346.88) (317.15) (409.95) (375.07)
Memory 5.00 5.33 6.22 6.55 6.05
(0.82) (1.63) (1.48) (1.73) (1.62)
Localization 0.75 2.07 3.07 2.00 2.38
(0.96) (1.71) (1.66) (2.25) (1.92)
19-11 n=45 n=102 n=143 n=73 n=363
Total 1083.22 949.73 847.87 823.63 902.60
(340.02) (355.55) (322.77) (386.83) (356.10)
Memory 6.56 7.03 7.58 7.89 7.36
(1.62) (1.70) (1.48) (1.62) (1.64)
Localization 3.33 3.85 4.55 4.38 4.17
(2.30) (2.43) (2.38) (2.81) (2.50)
Table A17.9. [TPf.9] Anthony et al., 1980: Data for Normal Volunteers: Mean Time in
Minutes and SD Divided by Number of Blocks Placed and Means and SDs for Memory and
Localization Scores
WAIS TPT
n Age Education FSIQ VIQ PIQ Time Memory Localization
100 38.88 13.33 113.54 113.24 112.26 0.44 7.80 4.64

(15.80) (2.56) (10.83) (11.59) (10.88) (0.25) (1.49) (2.15)
820 APPENDIX 17
Table A17.10. [TPT.10] Bak and Greene, 1980: Data for a Healthy Sample: Mean T'une in
Seconds and SDs for the Right Hand, Left Hand, Both Hands, and Total Time, as Well as
Means and SDs for Memory and Localization Scores
TPT
n Age Education Right Left Both Total Memory Locali-tion
15 50-62
55.6 13.7 365.20 306.80 169.80 841.80 5.27 2.07
(4.44) (1.91) (151.49) (161.51) (71.03) (326.01) (1.49) (1.53)
15 67-86
74.9 14.9 571.60 514.00 301.20 1,386.80 5.07 1.60
(6.04) (2.99) (289.36) (276.64) (116.14) (627.34) (2.02) (1.55)
Table A17.11. [TPT.ll] Fromm-Auch and Yeudall, 1983: Data for Canadian Volunteers:
Mean Time in Minutes, SDs, and Range for Preferred Hand, Nonpreferred Hand, Both
Hands Combined, and Total Time for Each Age Grouping•
TPT
n Age Preferred Nonpreferred Both Total Localization Memory
32 15-17 4.6 3.3 1.7 9.5 6.8 8.9

(1.2) (1.2) (0.5) (2.1) (2.5) (1.0)
2.6-6.8 1.1-0.4 0.8-3.3 4.7-14.1 1-10 ~10
74 18-23 5.1 3.5 2.1 10.6 5.7 8.2

(2.2) (1.6) (1.3) (4.5) (2.1) (1.3)
1.9-13.5 1.1-10.8 0.4-9.3 4.2-29.1 1-10 4-10
56 24-32 4.5 3.1 1.8 9.4 5.5 8.3
(1.8) (1.1) (0.8) (3.0) (1.8) (1.1)
1.7-9.5 1.5-7.1 0.5-4.6 3.8-18.8 2-9 ~10
18 33-40 4.9 3.7 2.3 10.9 5.6 8.6

(1.7) (1.0) (0.8) (2.9) (2.2) (1.1)
1.9-9.0 2.2-5.9 1.4-4.4 5.9-19.4 1-9 ~10
10 41-04 5.6 4.2 2.5 12.2 4.9 7.7

(1.5) (1.6) (1.2) (3.6) (1.8) (1.3)
4.0-9.0 2.4-8.1 1.4-5.5 8.3-20.6 2-7 6-9
"Mean correct blocks, SD, and range are reported for localization and memory bials.
Table A17.12. [TPT.12] Moore et al., 1984: Data for 183 Normal Volunteers Stratified into Six Age Groups
n Mean Recall
Age Mean Mean Location Completioo

Group Age Male Female Total FSIQ Shape Location proportion time (seconds)
19-27 23.1 24 32 56 115 8.3 5.7 0.67 737

28-36 31.8 36 28 64 112 7.6 4.6 0.59 805
37-45 40.8 31 28 59 111 7.2 3.4 0.46 893
~ 50.8 26 34 60 116 6.8 3.6 0.50 1,032
56-05 61.2 8 12 20 115 5.8 1.9 0.33 1,281
~76 69.5 8 17 25 115 5.0 1.7 0.32 1,377
APPENDIX 17 821
Table A17.13. [TPT.13a] Schear, 1984: Data for Patients with Neurological and Psychiatric Disturbance:
Means, SDs, and Ranges for Time in Minutes Required for Completion and Number of Blocks for the Right
Hand. Left Hand, and Both Hands for Five Age Decades
Right Left Both
n Age Education Minutes Blocks Minutes Blocks Minutes Blocks
111 20-29 11.72 6.98 8.95 5.39 9.36 3.63 9.85

(1.50) (2.30) (2.14) (2.62) (1.78) (2.29) (1.08)
6-16 1.~10 1-10 1.3-10 2-10 0.7-10 1-10
112 30-39 12.11 7.04 8.66 6.30 8.81 4.51 9.53
(2.21) (2.43) (2.49) (2.75) (2.39) (2.80) (1.52)
6-18 2.4-10 0-10 1.6-10 0-10 0.9-10 1-10
111 40-49 11.71 7.69 7.81 6.96 8.41 5.07 9.32
(2.75) (2.37) (3.11) (2.50) (2.89) (2.74) (2.05)
5-21 2.2-10 0-10 2.4-10 0-10 1-10 1-10
156 50-59 11.16 8.70 7.35 8.20 7.45 6.51 8.40
(3.61) (1.85) (3.03) (2.21) (3.24) (2.92) (2.86)
2-22 3.3-10 0-10 2.3-10 0-10 1.2-10 0-10
67 11.13 8.75 6.85 8.35 6.39 7.36 7.54
(3.60) (1.81) (3.25) (2.24) (3.51) (3.00) (3.06)
3-20 3.5-10 0-10 2.6-10 0-10 2.1-10 0-10
Table A17.14. [TPT.13b] Schear, 1984: Data for Patients with Neurological and
Psychiatric Disturbance: Means, SDs, and Ranges for Time in Minutes and
Number of Blocks•
Total
n Age Minutes Blocks Memory Localization
111 20-29 16.01 28.15 7.14 3.43

(6.57) (4.34) (1.99) (2.79)
3.~ 4-30 2-10 0-10
112 30-39 17.86 27.00 6.40 2.54
(7.25) (5.29) (1.85) (2.20)
6.5-30 4-30 0-10 0-8
111 40-49 19.73 25.54 6.02 1.94
(6.58) (7.09) (2.01) (2.01)
5.6-30 2-30 2-10 0-9
156 50-59 23.41 23.18 5.49 1.47
(6.24) (8.28) (1.92) (1.73)
8.4-30 0-30 1-10 0-7
67 60-69 24.46 20.78 4.96 1.45
(6.56) (9.05) (1.96) (1.46)
8.6-30 0-30 1-9 0-5
"Data are also reported for memory and localization.

822 APPENDIX 17
Table A17.15. [TPT.14] Russell, 1985: Data for Veterans Administration Patients: Mean Time in Minutes
and SDs for the Dominant Hand, Nondominant Hand, Both Hands, and Total Time, as Well as Mean and
SD for Memory and Localization Scores for the 10-Block and 6-Block Versions
TPT
n Age Education Dominant Nondominant Both Total Memory Localization

10 blocks 19 43.5 14.8 6.88 5.67 13.33 15.68 7.42 4.10
(13.6) (6.4) (3.36) (3.25) (2.03) (6.94) (1.64) (2.57)
6 blocks 1.88 1.41 0.87 4.17 4.84 3.84
(0.99) (1.12) (0.41) (2.16) (0.83) (1.54)
Table A17.16. [TPT.15] Heaton et al., 1986: Data for a Sample of Normal Controls: Mean Total Time in
Minutes per Block and Mean Memory and Localization Scores for the Six Subgroups, as Well as Percent
Classified as Normal Using Russell et al.'s (1970) Criteria
TPT % Classified Normal
WAIS Total Total

n Age Education Mean ss• Time Memory Localization T"IUie Memory Localization
319 <40 11.9 0.39 8.1 5.3 87.5 97.8 65.5
134 40-59 11.2 0.50 7.5 4.0 69.4 90.3 41.8
100 ~60 9.7 0.85 6.2 2.0 23.0 69.0 9.0
132 <12 9.5 0.64 6.9 3.6 53.7 78.8 35.6
249 12-15 11.2 0.47 7.7 4.4 75.9 93.2 49.0
172 ~16 12.9 0.43 8.0 5.0 78.5 96.5 61.1
•Mean WAIS scaled scores are also reported for each group.
>
"tJ
"tJ
m
z
0
Table A17.17. [TPT.l6] Alekoumbides et al., 1987: Data for Veterans Administration Inpatients: Mean Time in Minutes and SD to Correctly Place all the Blocks X
for the Preferred Hand, Nonpreferred Hand, Both Hands, and Total Time• ......
WAIS T1me Blocks per Minute
n Age Education FSIQ VIQ PIQ Preferred Nonpreferred Both Total Preferred Nonpreferred Both Total Memory Localization
111 46.85 11.43 105.89 107.03 103.1 7.80 6.18 4.31 18.29 1.58 2.10 3.42 2.05 6.28 2.67
(17.17) (3.20) (13.47) (14.38) (13.02) (3.73) (3.51) (3.43) (9.87) (0.90) (1.20) (2.08) (1.06) (2.01) (2.29)
•Mean number of blocks correctly placed per minute and SD are sUDlmarized for the preferred hand, nonpreferred hands, both hands, and total time. Memory and localization scores as
well as demographic information are also provided.
Table A17.18. [TPT.l7] Bomstein et al., 1987a: Data for Healthy Volunteers: Means and SDs for Total Time, Memory, and Localization Scores for Both Testing
Sessions, as Well as Raw Score Change and SD, Median Raw Score Change, and Mean Percent of Change
Raw Score Median Raw Mean%
Test Retest Change Score Change of Change
n Age VIQ PIQ Time Memory Localization Time Memory Localization Time Memory Localization Time Memory Localization Time Memory Localization
23 32.3 105.8 105.0 10.7 8.4 5.1 7.4 8.9 6.3 3.25 0.65 0.88 2.2 1.0 1.0 27 6 34
(10.3) (10.8) (10.5) (4.2) (0.9) (2.3) (2.6) (1.0) (2.9) (3.3) (0.61) (1.9)
~
w
824 APPENDIX 17
Table A17.19. [TPT.l8] EI-Sheikh et al., 1987:

Data for Egyptian Students: Mean Times in
Minutes and SDs for the Dominant Hand, Non-
dominant Hand, Both Hands, and Total, as Well as
Means and SDs for Memory and Localization
Scores for Both Test and Retest•
Test Retest
Dominant hand 4.95 3.32
(1.96) (1.43)
Nondominant hand 3.68 2.83
(1.28) (1.43)
Both hands 2.35 1.74
(0.84) (0.71)
Total 11.01 7.71
(3.38) (3.03)
Memory 8.34 8.59
(1.41) (1.93)
Localization 5.53 6.88
(2.59) (2.46)
n=32; Mean age, 20.6 (1.4).
Table A17.20. [TPT.19] Dodrill, 1987: Data for a Sample of Volunteers: Mean T1me in Minutes for TPT
Total Time and Mean Scores for Memory and Localization for the Total Sample and for Various Levels of
Intelligence
WAIS-R
n Age Education FSIQ VIQ PIQ Total Time Memory Localization
120 27.73 12.28 100.00 100.92 98.25 13.65 7.86 4.97
(11.04) (2.18) (14.35) (14.73) (13.39) (7.21) (1.26) (2.36)
n FSIQ Total Time Memory Localization

7 130 10.8 8 5
18 125 10.9 8 5
34 120 10.9 8 6
64 ll5 10.8 8 6
93 110 11.6 8 6
101 105 12.1 8 6
75 100 12.4 8 6
60 95 12.9 8 5
48 90 14.2 8 5
33 85 17.7 7 4
19 80 21.0 7 4
10 75 26.5 7 3
70 6 3
APPENDIX 17 825
Table A17.21. [TPT.20] Yendall et al., 1987: Data for Canadian Volunteers: Means and SDs for Time in
Seconds to Execute the Task With the Preferred and Nonpreferred Hands Separately and Together, as Well
as Means and SDs for Memory and Localization Scores, for Each Age Grouping and for the Total Sample
%Right
n Age Education Hand FSIQ Preferred Nonpreferred Combined Memory Localization
62 15--20 12.16 79.03 111.75 286.80 195.49 106.90 8.73 6.47

(1.75) (10.16) (101.90) (84.93) (43.71) (l.Oi) (2.44)
73 21-25 14.82 86.30 109.79 312.96 209.16 128.36 8.11 5.51
(1.88) (9.97) (131.61) (86.44) (74.42) (1.29) (2.10)
48 26-30 15.50 89.58 113.95 265.66 181.86 103.35 8.13 5.42
(2.65) (10.61) (92.45) (67.06) (42.34) (1.55) (2.23)
42 31-40 16.50 90.48 116.09 278.11 206.61 134.01 8.19 5.24
(3.11) (9.51) (101.94) (69.08) (53.13) (1.35) (1.97)
225 15--40 14.55 85.78 112.25 288.94 199.00 118.07 8.30 5.70
(2.78) (10.25) (111.22) (79.26) (5i.75) (1.33) (2.24)
Table A17.22. [TPT.21] Ernst, 1987: Data for Australian Volunteers: Mean Times in Minutes and SDs for
the Preferred Hand, Nonpreferred Hand, Both Hands, and Total Time•
Dominant Nondominant Both Total
n Gender Time Blocks Time Blocks Time Blocks Time Blocks Memory Localization
51 M 9.1 9.5 7.4 9.7 5.1 9.9 21.4 29.2 6.6 2.8
(4.1) (1.5) (3.4) (1.7) (2.9) (1.0) (9.0) (3.0) (1.6) (1.9)
59 F 10.3 9.0 10.1 9.1 6.6 9.8 26.9 28.0 5.9 2.0
(4.1) (2.2) (4.2) (2.0) (3.8) (0.9) (10.9) (3.8) (1.6) (1.8)
•In addition, mean number of blocks and SDs are presented for each time measure and for Memory and Localization.
Table A17.23. [TPT.22] Clark and Klonoff, 1988: Data for Male Coronary Bypass Surgery Patients in
Canada: Mean Time in Minutes and SDs for Time to Complete the Task with the Right Hand, Left Hand,
Both Hands, and Total, as Well as Mean and SD for Memory and Localization for Each of the Four Testing
Probes for the 6-Biock TPT Version
TPT
WAIS-R
Age FSIQ Right Left Both Total Memory Localization
Presurgery 55.5 105.9 2.37 1.74 1.05 5.16 4.46 3.33

(8.01) (12.2) (1.16) (0.86) (0.59) (2.29) (1.11) (1.59)
Postsurgery
3 months 2.07 1.83 0.97 4.87 4.64 3.47
(0.86) (1.10) (0.49) (2.16) (1.09) (1.67)
12 months 2.06 1.63 1.05 4.73 4.i8 3.69
(0.82) (0.81) (0.60) (1.87) (1.05) (1.48)
24 months 2.19 1.65 0.92 4.i7 4.i2 3.44
(1.01) (0.91) (0.42) (1.87) (1.10) (1.65)
826 APPENDIX 17
Table A17.24. [TPT.23] Elias et al., 1990: Data for 183 Healthy Volunteers Partitioned into 11tree Age
Groups
n WAIS TPT
Age
Group Male Female Education VIQ PIQ Total Time (Minutes) Memory Localization
20-31 41 47 15.7 119 116 9.40 8.00 6.10

(3.52) (1.20) (2.10)
37-49 23 38 15.4 122 122 11.90 7.80 4.90
(5.20) (1.30) (2.20)
55-67 12 22 14.9 124 121 15.70 7.10 3.20
(8.90) (1.20) (1.90)
Table A17.25. [TPT.24] Thompson and Heaton, 1991: Data for Healthy Volunteers: Mean Tunes in
Minutes and SDs for Dominant Hand, Nondominant Hand, Both Hands, and Total Tune, as well as
Memory and Localization Scores
n Age Education WAIS IQ Dominant Nondominant Both Total Memory I.ncalizatlon
489 39.43 13.9 113.09 6.78 5.72 3.51 14.85 7.59 4.43
(17.76) (3.46) (12.07) (7.56) (6.13) (5.72) (10.20) (1.58) (2.45)
Table A17.26. [TPT.26] Elias et al., 1993: Data for 427 Healthy Volunteers Partitioned into Six Age Groups
by Gender•
TPT
n Total TIDle (Minute$) Memory Localization
Age Group Male Female Male Female Male Female Male Female
15-24 37 24 9.60 9.55 8.27 7.83 6.51 6.17

(0.83) (0.78) (0.21) (0.31) (0.36) (0.50)
25-34 40 56 10.20 11.03 7.77 7.70 5.92 4.93
(0.58) (0.53) (0.26) (0.19) (0.37) (0.28)
35-44 36 56 10.88 11.81 7.86 7.04 4.64 4.55
(0.61) (0.56) (0.24) (0.25) (0.34) (0.34)
45-54 25 46 12.38 13.62 7.24 6.87 3.76 3.76
(0.89) (0.81) (0.23) (0.19) (0.32) (0.29)
55-64 25 35 17.29 14.70 6.80 6.86 3.40 2.91
(2.15) (0.98) (0.27) (0.23) (0.40) (0.33)
~65 24 23 16.49 18.66 6.62 6.30 3.12 3.00
(1.18) (1.70) (0.22) (0.38) (0.40) (0.53)
"Education range for the sample is 12-19 years.

APPENDIX 17 827
Table A17.27. [TPT.27] Barrett et al., 2001: Data for

Air Force Veteran Controls:
Localization and Memory Scores and Mean Times in Minutes for Dominant Hand,
Nondominant Hand, Both Hands, and Total Time•
n Age Localization Memory Dominant Nondominant Both Total
1,052 43.9 2.59 6.16 7.98 7.06 3.85 18.87

(7.6)
"SDs are not provided for the test scores.

Appendix 1~: Locator and Data Tables
for the Weqhsler Memory Scale
(WMS-R, \N,MS-111, and WMS-IIIA)
I
I
Study numbers and page numbers pr4'ided ' in Locator table also provides a reference for
these tables refer to study numbers f1nd de- each study to a corresponding data table in
scriptions of studies in the text of Ch;>ter 18. this appendix.
828
Table A18.1. Locator Table for the Wechsler Memory Scale (WMS-R, WMS-III, and WMS-IIIA) >
-a
-a
Sample IQ/ Length of Scoring Subtests Special m
Study Age• n Composition Education• Delay Method Administered Notes Location z
0
WMS-8.1
Wechsler, 1987
16-17
18-19
53
0
Sample designed to
represent tbe nonnal
Mean
WAIS-R FSIQ
30 minutes Wechsler
(1987)
FuDWMS-R Normative data for 18-
19, 25-34, and ~
USA
(standardization) ...
X
co
page 346 20-24 50 population of tbe U.S.; was 110 (15); 54 age groups were
Data are not 25-34 0 stratified sample however, tbe statistically interpo-
reproduced 35-44 54 based on age, sex, race, fuDWAIS-R lated (i.e., no data
in tbis book ~ 0 and geographic region was adminls- were actually
55-64 54 tered only to collected)
65-69 55 tbe 35-44 and
70-74 50 55-69 groups;
a 4-test short
form was given
to tbe otber 4
age groups on
whom data
were collected
WMS-B.J 50-70 47 Group of healthy, well- 14.4 (2.7) years 30 minutes Wechsler FuDWMS-R Provides WMS-R San Diego, CA
CuDumet al., 75-95 32 educated, community- 14.6 (3.0) years (1987) "preliminary norms"
1990 dwelling older adult for "nonnal elderly
page 347 volunteers recruited subjects;" forgetting
Tables Al8.2, A18.3 via flyers and newspaper rates for verbal and
and subsequently nonverbal material
screened via telephone also provided
"for neuropsychological
risk factors, history of
neurological disorder,
learning disability.
major psychiatric
disorder, major medical
illness or substance
abuse;" also excluded
if taking any
medications which
might negatively affect
perfonnance on
memory tests; low-
dose antihypertensives
were allowed
(continued)
=
N
\0
Table A18.1. (Contd.) CIO
w
Sample IQI Length of Scoring Subtests Special
=
Study Age• n Composition Education• Delay Method Administered Notes Location
WMS..R.3 25-34 50 Sample designed to match Prorated 30 minutes Wechsler Full WMS-R Florida
Mittenberg et al., 1980 u.s. Census data WAIS-R FSIQ (1987)
1992 stratified on age, gender, based on
page 348 ethnicity, and education; Vocabulmy and
Tables A18.4, A18.5 differs from WMS-R Block Design
standardization in that subtests only
all subjects reside in Mean=
Florida; recruited from 101.3 (14.6)
"local businesses, Range= 72-131
weekend/evening adult Median=100
education and
vocational!tecbnical
classes"
WMS-11.4 70-74 25 Consecutive admission No information; 30 minutes Wechsler WMS-RLogical Provides clinical Detroit,
Lichtenberg & 75-79 23 sample of cognitively however, (1987) Memory sub- comparison data (not Ml
Christensen. 80-99 18 intact geriatric medical Mattis test only normative data) for
1992 patients from an urban Dementia gtmamcmedical
page 348 hospital; about 113 had Rating Scale patients seeking
Table A18.6 hip fractures, 113 had cutoff was treatment in an urban
knee replacement due 129 to insure medical setting
to arthritis, 113 showed intact
"deconditioning from cognition
lengthy illness;" sample
comprised of 43 F,
23 M; 35 Caucastan,
31 black
WMS-8.5 Ivnik Total sample: Sample represents a For total sample 30 minutes WMS-R Full WMS-R; Data are reported using Rochester and
et al., 1992b 56-74 274 combination of (n=441) converted however, only the midpoint interval Olmsted
page 349 75-94 167 (1) patients who had a to Mayo allowed 3 technique (Pauker, County, MN
Data are not Data tables medical tlllllll1 at the WAIS-R IQ: summary learning trials 1988) to
reproduced we Mayo Clinic and were deemed VlQ = 105.5 (10.0) scores during both "maximize available
in this book midpoint "normal" because they the V'ISIIal and information"
interval lacked adlve neurological PIQ = 107.3 (11.4) Verbal Paired >
'"tl
method: or psychiatric conditions Associates 1 Provides '"tl
56-66 154 that would compromise FSIQ = 106.6 (10.5) subtests; no a statistically derived m
~ 161 cognitive l1mctioning additional trials estimate of probable z
62-72 169 (chronic medical illness were admini- WMS..R wlues for 0
65-75 168 was not an exclusion Years of stered if individuals >14 years old X
68-18 178 criterion, all were able education: criterion .....
00
71-81 160 to function independ- ~7(0.7%) reached by NOTE: Mayo and
74-84 151 ently) and (2) "normal 8-11 (13.6%) third trial WMS-R summary >
"tt
77-87 123 controls" from a 12 (39%) scores are not "tt
80-90 84 research project at 13-15 (22%) Interchangeable m
83-94 53 Mayo's Alzheimer's 18-17 (15%) z
Disease Patient Registry; 18 (9.8%) 0
criterion for normality
determination was as
above
Sample primarily
-B-educated,
-
X
Q)
Caucasian older
adults from an
agricultural
region
WMS..B.8 81.5 (3.3) 101 All autonomously living Education: 30 minutes Wechsler Logical Memory Data are presented In New Haven, CT
Richardson & elderly. current (1987) and V"JSual age by education cells:
Marottoli,1996 78-80 driven; 48 F, 53 M; 11.0 (3.7) Reproduction <12. ~12 yean
page 351 free from neurological subtests
TabJeA18.7 81-91 and psychlatric illness
WMS..IL7 76.48 131 Healthy adults over Education: 30 minutes Presumably 1.ogical Memory The study aimed at Rural
Marcopulos (7.87) S5 yean of age. living In Mean= Wechsler and Visual development of central
et al., 1997 a rural setting 6.65 (2.14) (1987) Reproduction snormative data for V"uginia
page 352 subtests rural community-
Tables A18.8-A18.11 55-64 Range= dwelling older adults
65-74 0-10 with no more than 10
15-M yean of formal educa-
~85 lion; data are reported
by age and age by
education; percent
retention was
calculated
WMS..B.8 18-34 399 Healthy. Three education 30 minutes Wechsler FullWMS-R Provides nonnative data Sydney,
Shores o'k generally well- levels reported- (1987) for urban Australians; data Australia
Cantain, 2000 educated adults <12 (n=91), with Logical reported by age/education
page 353 (193 M, 206 F) 12 (n=91), Memory categories for V"JSual
Data are >12 (n=217) years- subtest scored ~roduction and Visual
not reproduced for V"uual Reproduction according to Memory Span subtests;
In this book and Visual lvison, 1993 performance for men
Memory Span vs. women noted for
subtests Logical Memory,
V"JSUal Paired Associates,
and Verbal Paired
Associates subtests;
tables convert
performance raw to
scaled scores
a...
(continued)
e3
~
Sample IQ/ Length of Scoring Subtests Special

Study Age• n Composition Education• Delay Method Administered Notes Location
:w.MS.-.DJ.l 15-79 1,250 ~~ ~ Wecb.iler E.ull w.ws.Jl1 .Baw.-.are. ... USA

Wechsler, 1997 partitioned standardization mately (1997) reported which can be (standardization)
page 353 into 11 age sample 30 minutes converted to scaled
Data are not groupings scores
reproduced in
this book
WMS-ffi-A.l 15-79 1,250 Nationally collected Approxi- Wechsler Full WMS-III-A Raw scores are USA
Wechsler, 2002 partitioned standardization sample; mately (1997) (Logical Memory reported which can be (standardization)
page 354 into 11 age data are from WMS-111 30 minutes & Family converted to scaled
Data are not groupings standardization sample Pictures) scores
reproduced in
this book
• Age column and IQ/education column contain information regarding range and/or mean and standard deviation for the whole sample and/or separate groups, whichever information is
>
"tl
"tl
m
z
0
X
())
APPENDIX 18 833
Table A18.2. [WMS-R.2a] Cullum et al., 1990: Data for a Sample of Healthy
Older Adults Partitioned into Two Age Groups
Age W-70 (n=47) Age 75-95 (n = 32)
M SD M SD
WMS-ll subtm
Digit Span 14.9 3.7 15.3 3.7
Visual Span 14.7 2.4 14.1 2.5
Logical Memory I 29.2 7.7 25.0 7.5
Logical Memory II 25.6 7.9 20.9 8.4
Verbal Paired
Associates I 19.7 2.9 18.3 2.8
Verbal Paired
Associates II 7.6 0.7 6.9 1.2
VISual Reproduction I 34.3 3.9 29.1 7.3
Visual Reproduction II 29.1 6.1 20.1 9.1
Visual Paired
Associates I 14.1 3.7 12.7 3.5
VISual Paired
Associates II 5.6 0.9 4.8 1.4
Figural Memory 6.5 1.7 6.5 1.5
WMS-ll raw eummary •corea

Attention/Concentration 64.7 9.3 64.5 11.2
Verbal Memory 78.0 16.3 68.5 16.8
VISual Memory 55.0 7.1 48.3 9.7
General Memory 133.0 20.7 116.6 22.0
Delayed Memory 81.1 12.8 64.5 17.4
Table A18.3. [WMS-R.2b] Cullum et al., 1990: Table A18.4. [WMS-R.3a] Mittenberg et al., 1992:
Savings Scores for the Two Age Groups Data for a Sample of 50 Healthy Adults 25-34 Years
of Age: Raw Score Means and SDs for the WMS-R
Age W-70 Age 75-95
(n=47) (n=32) Subtest M SD Range
Savings Score M% SD M% SD Information/Orientation 13.88 0.33 13-14

Mental Control 4.98 1.13 2-6
Logical Memory 87 13 83 18 Figural Memory 7.20 1.49 4-10
Verbal Paired Logical Memory I 26.04 6.81 11-40
Associates 96 8 88 13 Visual Paired
VISual Reproduction 85 15 68 25 Associates I 15.12 3.39 5-18
Verbal Paired
Visual Paired
Associates I 20.80 3.16 12-24
Associates 97 20 84 23 Visual Reproduction I 34.52 5.39 16-41
Digit Span 14.96 3.71 7-22
Visual Memory Span 17.66 3.62 10--25
Logical Memory II 22.38 7.38 9--37
Visual Paired
Associates II 5.60 0.95 2-6
Verbal Paired
Associates II 7.62 0.86 4-8
Visual Reproduction II 32.88 6.59 13-41
834 APPENDIX 18
Table A18.5. [WMS-R.3b] Mittenberg et al., 1992:

WMS-R Weighted Raw Score Composites
Subtest Composites
Verbal Memory 72.58 15.30

Visual Memory 56.46 8.91
General Memory 129.04 20.25
Delayed Recall 80.74 15.16
Attention/Connection 69.58 13.40
Table A18.6. WMS-R.4] Lichtenberg and Chris-

tensen, 1992: Data for a Sample of Cognitively
Intact Geriatric Medical Patients: Means and SDs
for Logical Memory I and II from the WMS-R
Logical Memory Scores
Group Age n Mr so, Mn SDn
Overall 66 17.8 5.8 13.7 6.6

Group 1 70-74 25 19.4 5.4 15.7 6.7
Group 2 75-79 23 16.5 6.3 12.9 6.3
Group 3 80-99 18 17.3 5.4 12.1 6.6
Table A18.7. [WMS-R.6] Richardson and Marottoli, 1996: Data for a Healthy
Elderly Sample (n = 101), Partitioned by Two Age Groups and Two Educational
levels: Means and SDs for Logical Memory I and II and Visual Reproduction I
and II
Age Group
76-80 81-91
Education
<12 ~12 <12 ~12

(n=26) (n=24) (n=18) (n=33)
Logical Memory I 14.17 19.24 14.29 19.57

(6.48) (5.62) (4.70) (7.13)
Logical Memory II 8.88 12.70 9.12 12.69
(5.24) (5.49) (4.41) (8.44)
Visual Reproduction I 20.29 28.24 20.31 23.70
(6.91) (4.77) (7.96) (6.80)
Visual Reproduction II 8.63 16.15 11.56 11.97
(6.57) (7.33) (10.16) (8.57)
APPENDIX 18 835
Table A18.8. [WMS-R.7a] Marcopulos et al., 1997: Data for a Sample of Rural Older Adults with 10 Years
of Education or Less: Means and SDs for WMS-R Logical Memory I by Age Educational Level•
Years of Education Total by Age
0-4 lHi 7~ 9-10
Age n M (SD) n M(SD) n M (SD) n M(SD) n M (SD)
55-64 1 4.0 2 19.0 2 16.5 2 14.0 7 14.7

(-) (11.3) (3.5) (1.4) (7.1)
65-74 5 11.2 8 14.0 20 16.9 10 16.7 43 15.7
(5.9) (7.9) (5.8) (5.3) (6.2)
7~ 12 11.8 15 13.4 28 14.1 8 19.4 63 14.2
(6.4) (7.4) (6.6) (9.3) (7.3)
85+ 5 7.4 1 0.0 10 7.9 2 12.5 18 7.8
(2.4) (-) (6.4) (12.0) (6.1)
Total by education 23 10.4 26 13.5 60 14.1 22 17.0
(5.7) (8.0) (6.8) (7.3)
•Total sample n = 131; mean score= 13.8, SD = 7.2, range 0-32.
Table A18.9. [WMS-R.7b] Marcopulos et al., 1997: Data for a Sample of Rural Older Adults with 10 Years
of Education or Less: Means and SDs for WMS-R Logical Memory II by Age and Educational Level•
0-4 lHi 7~ 9-10
Age n M(SD) n M (SD) n M (SD) n M(SD) n M(SD)
55-64 1 2.0 2 14.0 2 11.0 2 11.0 7 10.6

(-) (4.2) (1.4) (0.0) (4.4)
65-74 5 4.4 8 8.9 20 11.4 10 12.7 43 10.4
(3.0) (7.5) (6.7) (5.4) (6.6)
7~ 12 7.4 15 8.5 28 9.4 8 14.1 63 9.4
(5.0) (6.1) (7.1) (9.2) (6.9)
85+ 4 1.8 1 0.0 10 3.8 2 10.5 17 3.9
(2.1) (-) (3.7) (13.4) (5.2)
(4.6) (6.5) (6.8) (7.1)
"Total sample n = 130; mean score= 9.1, SD = 6.5, range 0-28.

836 APPENDIX 18
Table A18.10. [WMS-R.7c] Marcopulos et al., 1997: Data for a Sample of Rural Older Adults with 10Years
of Education or Less: Means and SDs for WMS-R Visual Reproduction I by Age and Education Level•
0-4 5-6 7-S 9-10
Age n M (SD) n . M(SD) n M (SD) n M(SD) n M (SD)
55-64 12.0 2 25.0 2 25.5 2 31.5 7 25.1

(-) (1.4) (4.9) (4.9) (7.1)
65-74 5 8.0 8 15.3 20 18.6 10 21.6 43 17.4
(5.4) (5.7) (7.7) (9.2) (8.5)
75-S4 13 12.8 15 : 12.7 28 18.4 8 14.9 64 15.5
(5.9) (5.7) (8.8) (8.1) (7.9)
85+ 5 10.2 2· 3.5 10 9.2 2 15.0 19 9.5
(4.2) ! (2.1) (6.1) (5.7) (5.7)
I
Total by education 24 11.2 27 . 13.7 60 17.2 22 19.5
(5.5) (7.2) (8.7) (9.3)
"Total sample n = 133; mean score= 15.8, SD,., 8.4, range 0-35.
Table A18.11. [WMS-R.7d] Marcopulos et al., 1997: Data fora Sample of Rural Older Adults with 10Years
of Education or Less: Means and SDs for ~S-R VISual Reproduction II by Age and Education Level•
0-4 ·5-6 7-S 9-10
Age n M(SD) n M(SD) n M(SD) n M(SD) n M (SD)
55-64 2.0 2 22.5 2 18.0 2 12.5 7 15.4

(-) (4.9) (7.1) (17.7) (10.8)
65-74 5 4.6 8 6.1 20 11.3 10 11.7 43 9.7
(5.9) (4.8) (7.9) (7.5) (7.4)
75-S4 13 5.5 15 6.2 28 9.9 8 6.9 64 7.7
(5.3) (5.9) (10.9) (4.3) (8.4)
85+ 4 5.8 2 0.0 10 2.4 2 2.5 18 2.9
(4.3) (0.0) (4.4) (3.5) (4.2)
(5.0) (7.0) (9.5) (7.5)
•Total sample n = 132; mean score=8.1, SD=S:2, range 0-34.

for the List-Learning Tests
837
Table A19.1. Locator Table for the List-Learning Tests CIO
~
CIO
Study Age• n Sample Composition IQ/Education • Trials Reported Location
RAVLT.l Rey, 1941, - 132 French-speaking - 1-V Switzerland

1964 Swiss stratified
page 375 into 5 occupational
Table A19.2 groups
RAVLT.2 Query & 19-81 677 The study provides Education: I & V, postinter- North Dakota
Megran, 1983 19-24 54 norms for male 11.44 ference recall,
page 376 25-29 88 ambulatory inpatients recognition,
Table Al9.3 30-34 109 treated for a variety WAIS IQ: V-1 difference
35-39 54 of physical 93.83
40-44 50 complaints at a
45-49 52 VAMC
50-54 83
55-59 81
60-64 57
65-69 26
70-81 23
RAVLT.3 48.62 47 VAMC psychiatric Education: 1-V, postinte r- lllinois
Rosenbe rg et al., (16.60) and neurological 10.81 ference recall,
1984 47.51 45 inpatients classified (3.01) recognition
page 376 (13.59) as memory- 11.87
Table A19.4 impaired and non- (2.58)
memory-impaired
RAVLT.4 Cohen, 60-64 81 Not·mative data are Education: 1-V, Peoria,
et al ., pe rsonal 65-69 provided for elderly 13.8 interference, IL
communication 70-74 volunteers per age postinterferenc
page 377 75-89 group, for males and recall, immediate
Table A19.5 fe males separately recognition,
(53 F, 28 M) 30-minute
delayed recall,
delayed )>
recognition -o
-o
m
RAVLT.5 Hyan et al ., 45.86 85 VAMC inpatients; Education: I- V, postinter- Kansas z
1986 (14.05) alte rnate form reliability 11 .85 ference recall,
page 377 for AVLT assessed (2.5 l) and recognition -0X
Table AJ 9.6 for alte rnate forms
1.0
RAVLT.6 Bleecker 40-49 196 The study presents Means 1-V, Maryland >
"tt
et al., 1988 5Ch59 norms for healthy range recognition "tt
m
page 378 60-69 subjects broken 13-18 z
Table A19.7 70-79 down by age group years for 0
80-89 and gender different X
(87 M, 109 F) groups .....
loD
RAVLT.7Wiens 19-51 222 The study presents 2::12 1-V, postinter- Oregon
et al., 1988 29.1 nonnative data for for all ference recall,
page 379 (6.0) healthy job applicants, groups recognition,
Tables A19.8-A19.10 representing an derived indices
occupational cross-section
of the community;
193 M, 29 F; data
are stratilled by FSIQ,
age, and education
RAVLT.8 Crawford - 60 The study compared test- 1-V, UK
et al., 1989 retest performance over interference,
page 379 27 days with the same postinterference
Tables A19.11, A19.12 form of RAVLT and with a recall,
parallel version developed recognition
by the authors, in
healthy adults
RAVLT.9 Nielsen 20-54 101 53 M, 48 F; majority VIQ: 1-V, total for Denmark
et al., 1989 Groups: had undergone minor 98.6 trials 1-V,
page 380 20-29 35 surgery and were tested (12.2) 15-minute delayed
Table A19.13 30-39 27 several weeks postsurgery recall
40-54 39
RAVLT.IO Roth 27.5 61 45 M, 16 F; controls in a Education: 1-V, interference, Detroit, Ann
et al., 1989 (SE=l.O) study on neuropsychological 12.8 postinter- Arbor, Ml
page 380 dellcits in acute spinal ference recall,
Table A19.14 cord injury recognition,
(means and SEs
reported)
(continued)
~
10
=
....
Study Age• n Sample Composition IQ/Education• Trials Reported Location
=
RAVLT.ll Geffen 16--86 153 Nonns provided for Education: Standard recall, Australia
et al., 1990 16-19 25 adults aged 11.2 delayed recall,
page 381 20-29 20 16--86, by age and gender; (2.2) and recognition
Tables A19.15-A19.17 3Ch39 23 variety of perfonnance trials, plus serial
40-49 23 indices explore position and
50-59 20 different memory mechanisms; functional
60-69 22 equal number of males indices
70--86 20 and females
RAVLT.l2 lvnik 55-97 394 The study provides raw data Education: 1-V, inter- Olmsted
et al., 1990 55-59 45 and summary scores for an <8 ference, postinter- County,
page 382 60--64 53 elderly sample stratified years ference recall, MN
Tables A19.18-A19.20 65-69 64 into 7 age groups; to>17 30-minute
70-74 67 145M, 249 F years delayed recall,
75-79 69 recognition, errors,
80-84 49 and 4 summary
85-97 47 scores
RAVLT.l3 Miller Age The study compared V, total for MACS
et al., 1990 range: perfonnance of 3 trials 1-V centers at
page 382 21-72 groups of homosexual! Baltimore,
Table A19.21 bisexual men: Chicago,
37.20 769 1. Seronegative 16.36 L.A.,&
(7.52) (2.34) Pittsburgh
35.66 727 2. Asymptomatic 15.70
(6.47) seropositive (2.44)
36.90 84 3. Symptomatic 16.06
(7.04) seropositive (2.50)
RAVLT.l4 Shapiro Four alternate fonns of 1-V, inter- VIrginia
& Harrison, 1990 AVLT were compared; ference, post-
page383 2 were generated interference
Tables Al9.22-A19.24 according to the criteria recall
developed by the authors; >
~
2 subject samples were ~
66 17 used: VAMC patients and - m
z
19 25 undergraduate students - Cl
X
_.
\0
BAVLT.l5 57-65 28 Norms are provided for 14.2 1-V, postinterfer- California >
Mitrushina et al., 6&-70 45 highly functioning elderly 14.0 ence recall, "tl
"tl
1991 71-75 57 stratified into 14.6 recognition, m
page 384 76-85 26 4 age groups; 13.3 false-positives, z
0
Tables A19.~A19.27 62 M, 94F rates of acquisition
X
and forgetting, .....
primacy/recency effect 1.0
BAVLT.l6 57-65 19 Perfonnance of highly 14.4 I, V, postinter- California

Mitrushina & Satz, 6&-70 40 functioning elderly sample 13.7 ference recall
1991a 71-75 47 is compared over 3 14.5
page 384 76-85 16 longitudinal annual 14.0
Tables A19.28, Al9.29 probes; 49 M, 73 F
BAVLT.l7 Seines 25-34 733 The study reports nonnative Education: V, total for MACS
et al., 1991 35-44 data collected on a large <College trials 1-V, centers at
page 385 45-54 sample of seronegative college postinterference Baltimore,
Table A19.30 homosexual/bisexual >college recall, delayed Chicago,
men; data are stratified recall, delayed L.A.,&
by age and education recognition Pittsburgh
BAVLT.l8 Delaney 45.8 42 Controls without 12.8 I, III, V, Connecticut,
et al., 1992 22-67 histories of neurological 6-16 postinterference California,
page385 or psychiatric problems recall, delayed Florida, Virginia,
Table A19.31 recall, delayed Massachusetts,
recognition New York,
Minnesota
BAVLT.l9 Ivnik Age range 530 The study provides age- Education: Raw scores for Minnesota
et al., 1992c 5fHl7; specific nonns for elderly; various recall
page 386 divided scoring procedures were $;7 and recognition
Table A19.32 into developed, which convert to 2:18 trials are
groups raw and computed years reported in
based on scores into scaled scores; the earlier
midpoint tables are not reproduced article
interval in this book
(continued)
:....
~
N
Study Age• n Sample Composition IQ/Education• Trials Reported Location
RAVLT.20 Savage 1&-19 134 The study provides ;:::12 I-V, inter- Louisiana
& Gouvier, 1992 20-29 normative data for years ference, postinter-
page 387 30-39 healthy adults, ference recall,
Tables Al9.33, Al9.34 40-49 stratified by age group recognition, 30-
50-59 and gender; 66 M, 68 F minute delayed
60-69 recall, delayed
70-76 recognition
RAVLT.21 Crossen 29.9 60 Performance on RAVLT Education: I-V, total,
& Wiens, 1994 (6.2) and CVLT was compared 14.7 (1.6) interference, postin-
page 387 on a group of healthy FSIQ: terference recall,
Table Al9.35 job applicants; 52 M, 8 F 106.3 recognition
RAVLT.22 Geffen 31.3 51 The study explored equi- Education: I-V, inter- Brisbane,
et al., 1994 (12.7) valence between the 12.2 ference, postin- Australia
page 388 original form of the RAVLT (2.4) terference recall,
Tables Al9.3&-Al9.39 and a new form on a 20-minute
sample of healthy delayed recall,
volunteers; 25 M, 26 F recognition
RAVLT.23 Torres 29.0M 84 Data for 160 healthy 14.4 M I, V, inter- USA
et al., 2001 (10.2) controls, stratified by (2.1) ference trial,
page 388 30.5 F 76 gender 14.8 F postinterference
Table A19.40 (11.7) (1.9) recall
RAVLT.24 Miller, 38.2 920 Seronegative homosexual 16.3 I, V, total MACS
2003 (an update on (7.4) and bisexual males from (2.4) for I-V, interference centers
Seines et al., 1991) MACS; data trial, 20-minute delayed
page 389 25-34 are partitioned by age x Education: recall, delayed recog-
Table A19.41 35-44 education <16 nition, number of false-
45-59 16 positive errors
>16
HVLT-R.1 Friedman 60-71 237 Healthy African-American <12 Standard indices, plus Florida
et al., 2002 72-84 sample; 108M, 129 F; 12 delayed cued recall >
""C
page 389 data are stratified by 2 >12 and learning index ""C
m
Table Al9.42 age levels, 3 educational z
levels, and gender; 0
tables for conversion into X
percentiles are also provided ....
ID
WHO-UCLA AVLT.l 38.4 300 Spanish-speaking healthy 10.7 V, recall after California >
Ponton et al., 1996
page 390
(13.5) volunteers;
MIF ratio 40%160%;
(5.1) interference and
20-minute delayed ""
m
z
Table A19.43 16-29 data are partitioned by <10 recall 0
30--39 gender x 4 age groups x >10
40-49 2 education groups ....><
\D
50-75
• Age column and I Q/education column contain information regarding range and/or mean and standard deviation for the whole sample and/or separate groups, whichever information is
e
844 APPENDIX 19
Table A19.2. [RAVLT.l] Rey, 1941, 1$64: Data for French-Speaking Swiss
Participants
Trial
Subject Groups JI III IV v

Manual laborers
(n=25)
Mean 7.0 10.5 12.9 13.4 13.9
SD 2.1 l.9 1.6 2.0 1.2
Professionals
(n=30)
Mean 8.6 1i.8 13.4 13.8 14.0
SD
Students
1.5 f·O 1.4 1.1 1.0
(n =47)
'
Mean 8.9 1•. 7 12.8 13.5 14.5
SD 1.9 ~.8 1.5 1.3 0.7
Elderly laborers
(70-90 years, n = 15)
Mean 3.7 •.6 8.4 8.7 9.5
SD 1.4 1.4 2.4 2.3 2.2
Elderly professionals :
(70-88 years, n = 15)
Mean 4.0 7.2 8.5 10.0 10.9
SD 2.9 •. 9 2.5 3.3 2.9
Table A19.3. [RAVLT.2] Query and MeWan, 1983: Data for Male Ambulatory Inpatients
Treated at North Dakota Veterans A~tration Medical Center for Physical Complaints"
Postinterference
Trial I Trial V Recall Recognition
Age n M SD M SD : M SD M SD V-1 Difference
19-24 54 6.15 1.46 11.50 0.63 . 9.80 1.66 12.81 2.26 5.35
25-29 88 5.98 1.43 11.27 1.87 ' 9.91 2.36 12.16 2.51 5.29
30-34 109 5.68 1.17 10.71 4.19 I 9.os 2.94 13.03 1.57 5.03
35-39 54 5.49 1.77 11.80 1.94 9.55 2.05 13.45 1.12 6.31
40-44 50 5.51 1.44 11.14 1.59 9.37 2.60 12.86 2.25 5.63
45-49 52 5.10 1.27 10.43 1.92 ' 8.18 2.91 12.23 1.70 5.33
50-54 83 5.01 1.62 9.38 2.66 I 7.12 3.40 11.48 2.67 4.37
55-59 81 4.53 2.50 8.80 5.04 . 5.96 4.28 10.75 6.41 4.27
60-64 57 4.09 1.61 7.54 2.57 5.81 2.64 9.96 2.62 3.45
65-69 26 4.12 1.26 7.29 6.12 ' 5.21 2.58 9.50 3.33 3.17
70-81 23 3.14 1.50 5.86 2.04 3.45 2.92 8.91 3.64 2.72
•Mean education for the sample is 11.44 y~ mean IQ is 93.83.

APPENDIX 19 845
Table A19.4. [RAVLT.3] Rosenberg et al., 1984: Data for Male Psychiatric and Neurological Inpatients
from the Veterans Administration Medical Center in Chicago•
Years RAVLTTrial
Recall after
n Age Education II III IV v Interference Recognition
Non-memory- 45 47.51 11.87 4.96 6.81 8.66 9.40 9.71 7.81 11.53
impaired (13.59) (2.58) (1.78) (2.31) (2.33) (2.53) (3.04) (3.71) (3.06)
Memory- 47 48.62 10.81 3.91 5.00 5.71 6.13 6.89 4.07 8.18
impaired (16.60) (3.01) (1.93) (2.06) (2.27) (2.77) (2.91) (2.79) (4.05)
"Mean full-scale IQ for the entire sample is 93.11 (13.43).
Table A19.5. [RAVLT.4] Cohen et al., Personal Communication: Data for Elderly
Volunteers•
Age Group
60-64 65-69 70-74 7&-89
n 23 13 12 9
Trial I 5.70 (1.48) 5.46 (1.39) 5.67 (1.23) 4.94 (1.63)
Trial II 8.13 (1.69) 8.88 (2.08) 8.42 (1.38) 8.17 (1.80)
Trial III 10.00 (1.84) 10.31 (1.81) 10.04 (1.48) 9.33 (1.92)
Trial IV 11.58 (2.35) 11.00 (1.97) 10.67 (1.66) 10.72 (2.28)
Trial V 11.83 (2.39) 11.54 (2.61) 10.71 (1.82) 11.33 (2.12)
ListB 6.13 (2.20) 5.81 (1.93) 5.17 (1.48) 5.50 (1.17)
List B errors 0.00 (0.00) 0.23 (0.44) 0.33 (0.65) 0.44 (0.73)
Trial VI (PostiR) 10.15 (2.94) 9.23 (3.21) 8.21 (2.23) 7.78 (3.87)
Intrusions from list B 0.09 (0.29) 0.08 (0.28) 0.25 (0.45) 0.33 (0.71)
Trial V-I 6.04 (1.60) 6.08 (2.17) 5.04 (1.32) 6.39 (2.00)
Trial V-VI 1.67 (1.87) 2.31 (1.75) 2.50 (1.17) 3.56 (2.60)
Total errors 0.70 (1.40) 1.38 (1.80) 2.08 (2.61) 2.00 (2.24)
Immediate Recognition 12.65 (2.10) 12.46 (2.40) 12.25 (2.05) 12.22 (3.67)
Delayed (30-minute) recall 9.98 (2.66) 9.38 (2.95) 9.00 (3.08) 8.61 (3.00)
Delayed (30-minute) recognition 12.61 (1.70) 12.69 (1.75) 11.58 (2.50) 11.33 (3.88)
Mala
n 8 7 9 4
Trial I 5.24 (1.50) 6.00 (1.00) 4.22 (1.32) 3.75 (2.50)
Trial II 6.88 (1.46) 8.36 (1.49) 6.44 (1.13) 6.50 (0.58)
Trial III 8.25 (2.38) 9.79 (1.78) 7.22 (1.80) 8.25 (1.71)
Trial IV 9.38 (1.92) 10.71 (2.36) 8.22 (1.64) 9.75 (2.22)
Trial V 10.50 (1.69) 12.14 (1.86) 8.72 (0.90) 9.25 (1.89)
ListB 5.38 (1.22) 5.57 (1.40) 4.06 (1.55) 4.25 (1.71)
List B errors 0.25 (0.46) 0.43 (0.54) 0.11 (0.33) 0.25 (0.50)
Trial VI (PostiR) 6.75 (2.71) 8.57 (3.50) 5.94 (1.98) 8.50 (1.73)
Intrusions from list B 0.00 (0.00) 0.14 (0.38) 0.11 (0.33) 0.25 (0.50)
Trial V-I 5.25 (2.19) 6.14 (1.95) 4.61 (1.87) 5.50 (1.73)
Trial V-VI 3.75 (2.12) 3.57 (1.99) 2.78 (1.50) 0.75 (0.50)
Total errors 4.38 (6.68) 1.29 (0.95) 1.67 (3.08) 1.25 (1.89)
Immediate recognition 9.71 (1.89) 11.43 (2.22) 11.89 (1.27) 11.75 (2.50)
Delayed(30-minute)recall 7.71 (1.80) 7.71 (2.98) 4.89 (1.54) 8.25 (2.63)
Delayed (30-minute) recognition 10.43 (1.72) 11.14 (3.08) 11.89 (2.03) 11.50 (2.08)
"Mean education for the sample is 13.8 years.

PostiR, postinterference recall.
846 APPENDIX 19
Table A19.6. [RAVLT.S] Ryan et al.,l9$6: Data for Inpatients from the Veterans Administration
Medical Center in Kansas Referred for fsychological and/or Neuropsychological Assessment
n Gender Race VVAI5-RFSIQ ~ Education
85 82M 61 White 19.74 45.86 11.85

3F 24 Nonwhite (11.11) (14.05) (2.51)
Trial
Form I II III IV v Postinterference Recognition" Total I-V
Original
M 4.69 6.34 7.29 8.35 8.87 6.18 9.82 35.55
SD 1.70 2.08 2.55 3.03 3.20 3.75 3.64 11.08
Alternate
M 4.46 6.05 6.93 7.39 7.99 5.44 10.39 32.81
SD 1.78 2.39 2.59 2.87 2.81 3.27 2.91 11.18
r 0.63 0.62 0.60 0.62 0.71 0.66 0.65 0.77
"For this variable, n=84.
Table A19.7. [RAVLT.6] Bleecker et al, 1988: Data for Healthy Volunteers from the Johns Hopkins
Teaching Nursing Home Study of Nornutl Aging
AVLT
n ~ Gender Education v~· BOlt I II III IV v Recognition
15 40-49 M 15 ~ 3 7.2 9.5 10.7 11.6 12.3 14.1

(3.1) (!l3) (2.5) (1.6) (2.3) (3.0) (2.6) (2.6) (1.3)
16 40-49 F 15 55 5 7.7 10.9 12.3 12.6 13.6 14.5
(3.0) (10.2) (4.5) (1.6) (2.1) (2.0) (1.8) (1.7) (0.9)
20 50-59 M 13 51 5 6.5 9.8 11.5 11.8 12.8 14.4
(3.2) (10.2) (5.4) (1.6) (2.0) (2.0) (2.0) (2.2) (0.9)
22 50-59 F 14 54 5 7.4 10.0 12.0 12.8 13.6 14.5
(2.9) (10.2) (3.6) (1.3) (1.5) (1.5) (1.6) (1.7) (0.8)
23 60-69 M 13 sa 5 5.6 8.0 9.9 10.3 11.3 13.9
(2.8) (7'.0) (5.4) (1.8) (1.9) (1.9) (2.5) (1.8) (1.7)
29 60-69 F 14 5S 6 6.6 9.6 11.1 12.2 12.6 14.4
(2.6) (9.0) (4.7) (1.8) (2.0) (2.1) (2.2) (2.3) (1.0)
18 70-79 M 15 56 3 6.3 8.2 9.6 10.0 10.8 14.2
(2.8) (7,7) (3.0) (1.6) (2.2) (2.2) (2.3) (2.2) (0.6)
29 70-79 F 15 59 5 6.3 9.4 10.4 11.8 12.5 14.5
(3.4) (9.8) (5.1) (1.7) (2.2) (2.2) (2.3) (2.1) (0.8)
11 80-89 M 18 61 5 5.0 7.1 8.0 8.7 9.2 13.0
(1.9) (M) (3.4) (1.2) (1.4) (2.1) (1.9) (2.1) (4.1)
13 80-89 F 15 58 7 5.6 8.2 9.7 11.0 11.1 13.9
(2.0) (7J)) (3.5) (1.6) (1.5) (2.3) (1.8) (1.4) (1.4)
•vvechsler Adult Intelligence Scale-Revised VoeabuLuy raw scores.

tBeck Depression Inventory.
Table A19.8. [RAVLT.7a] Wiens et al., 1988: Data for Job Applicants by Wechsler Adult Intelligence Scale-Revised (WAIS-R) Full-Scale IQ (FSIQ) >
"0
"0
Trial m
WAIS-R Postinterference Distractor Trial Words Learned Percentage Total z
FSIQ n I II III IV v Recall Recognition Ust (B) (Trial V-I) Recall Errors Repetitions (Trials I-V) 0
><
80--89 5 8.0 10.4 10.8 11.0 11.0 10.6 14.0 6.6 3.0 99.5 2.2 2.6 51.2
.....
1.0
(2.5) (1.7) (2.2) (2.1) (3.0) (2.4) (0.7) (2.6) (1.7) (22.0) (1.8) (3.2) (10.9)
90-99 29 7.1 9.7 11.4 12.2 13.0 11.2 14.0 6.0 5.9 86.7 3.2 5.6 53.4
(1.6) (1.8) (2.1) (1.8) (2.0) (2.2) (1.1) (1.6) (2.2) (14.5) (3.9) (5.7) (7.4)
100--109 81 7.2 9.9 11.8 12.4 12.9 11.6 14.2 6.5 5.7 90.1 2.2 5.2 54.2
(1.8) (2.5) (2.0) (1.9) (1.8) (2.3) (0.9) (1.5) (1.9) (12.0) (2.9) (5.8) (8.2)
110-119 55 7.5 10.4 11.9 13.1 13.2 12.1 14.0 6.8 5.7 91.8 2.1 7.0 56.1
(1.7) (2.2) (1.9) (1.4) (1.6) (2.3) (1.2) (1.5) (2.1) (12.1) (2.4) (6.9) (7.0)
120--129 38 7.7 10.7 12.7 13.3 13.7 12.6 14.4 7.2 6.0 92.5 2.0 5.4 58.1
(1.8) (2.2) (1.7) (1.5) (1.7) (1.9) (0.8) (1.9) (1.8) (9.9) (2.3) (6.3) (7.2)
130-139 3 10.0 12.3 13.7 15.0 14.7 14.3 15.0 7.7 4.7 97.9 2.0 0.7 65.7
(2.6) (2.5) (1.5) (0) (0.6) (1.2) (O) (1.5) (2.5) (10.4) (1.7) (1.2) (6.7)
© Swets & Zeitlinger (1988).
Table A19.9. [RAVLT.7b] Wiens et al., 1988: Data for Job Applicants by Age
Trial
Postinterference Distractor Trial Words Learned Percentage Total
Age n I II III IV v Recall Recognition UstB (Trial V-I) Recall Errors Repetitions (Trials I-V)
20--29 126 7.4 10.4 12.2 13.0 13.4 12.1 14.2 6.8 6.0 90.4 2.2 5.8 56.3
(1.7) (2.2) (1.9) (1.7) (1.7) (2.2) (1.0) (1.6) (2.0) (12.4) (3.0) (6.1) (7.4)
30-39 71 7.4 9.9 11.7 12.4 12.7 11.7 14.2 6.5 5.3 92.0 2.3 5.0 54.2
(1.9) (2.5) (2.0) (1.8) (1.8) (2.2) (1.1) (1.7) (1.9) (12.7) (2.5) (5.5) (8.3)
40--49 12 7.3 9.8 11.4 12.3 12.5 11.2 13.8 6.6 5.5 88.9 2.7 7.3 53.3
(2.2) (2.7) (2.6) (1.8) (2.5) (3.1) (0.9) (1.8) (2.6) (10.8) (2.3) (10.1) (10.3)
© Swets & Zeitlinger (1988). ~

.....
!
Table A19.10. [RAVLT.7c] Wiens et al., 1988: Data for Job Applicants by Education
Trial
Years of Postinterference DistractorTrial Words Learned Percentage Total
Education n I II III IV v Recall Recognition ListB (Trial V-1) Recall Errors Repetitions (Trials 1-V)
12 34 7.0 9.9 11.7 11.9 12.4 11.4 13.9 6.6 5.3 93.1 1.9 6.6 52.9
(1.6) (3.5) (2.0) (3.5) (2.3) (2.41 (.1.2) (1.8} (.2.0) (J.l..5) {l.JI) ~6.7) ~7~)
13 25 7.5 10.1 11.9 12.7 13.2 12.1 13.9 6.0 5.7 91.4 1.4 7.1 55.3
(1.2) (2.4) (2.4) (1.7) (1.6) (2.1) (1.2) (1.6) (1.6) (10.4) (1.7) (7.3) (7.7)
14 50 7.2 9.9 ll.8 13.0 13.2 12.3 14.4 6.7 6.0 93.3 2.5 6.5 55.2
(1.9) (2.3) (2.1) (1.9) (2.0) (2.2) (0.8) (1.8) (2.2) (ll.O) (3.2) (6.9) (8.4)
15 19 7.4 10.3 12.4 12.6 13.2 11.4 14.3 6.6 5.7 86.0 2.3 2.6 55.8
(2.2) (2.9) (2.0) (2.1) (1.9) (2.7) (0.7) (0.9) (2.1) (50.7) (3.5) (2.7) (9.5)
16 80 7.6 10.5 12.1 13.0 13.3 12.0 14.2 6.8 5.7 90.0 2.6 5.0 56.5
(1.9) (2.2) (1.8) (1.4) (1.5) (2.1) (0.9) (1.7) (2.0) (11.8) (3.0) (5.6) (7.2)
~17 5 7.8 10.4 12.4 13.8 13.4 ll.2 13.4 57.8 5.6 83.0 1.2 4.2 57.8
(2.6) (3.0) (1.1) (1.3) (1.7) (3.1) (1.8) (1.1) (3.0) (16.6) (1.6) (1.6) (7.7)
>
"tt
"tt
m
z
0
X
ID
APPENDIX 19 849
Table A19.11. [RAVLT.8a] Crawford et al., 1989: Data for Healthy Participants• (n =60), Recruited from
Nonmedical Health Service Personnel and Fire Service in the UK; Scores for Matched Groups Receiving
the Original AVLT and the Parallel Version
Trial
Postinterference
I II m IV v ListB Recall Recognition
Original 8.30 11.()() 11.80 12.73 13.20 6.90 11.90 25.37

(1.80) (2.11) (2.36) (2.02) (1.61) (2.22) (2.55) (2.68)
Parallel 7.37 10.50 11.70 12.63 13.00 6.43 11.43 25.13
(1.67) (2.52) (2.25) (1.50) (1.48) (1.89) (2.00) (2.96)
"Demographic information for the sample Is not available.
Table A19.12. [RAVLT.Sb] Crawford et al., 1989: Data for Healthy Participants (n =60); Scores for
Groups Receiving Either the Same or a Different AVLT Version at 27-Day Retest
Trial
Postinterference
I II m IV v ListB Recall Recognition
Test 7.87 11.10 11.93 13.03 13.33 6.70 11.93 25.30

(1.76) (2.19) (2.00) (1.59) (1.56) (2.40) (1.95) (2.47)
Same. version retest 10.53 12.87 13.67 13.90 14.13 7.70 13.43 26.67
(2.39) (1.81) (1.40) (1.35) (1.14) (2.29) (1.68) (2.47)
Test 7.53 10.40 11.57 12.33 12.87 6.63 11.40 25.20
(1.76) (2.43) (2.56) (1.88) (1.80) (1.70) (2.58) (3.15)
Different version retest 7.50 10.27 11.87 12.70 12.90 6.23 11.93 24.57
(2.13) (.2.16) (2.00) (1.82) (2.11) (1.91) (2.77) (3.67)
Table A19.13. [RAVLT.9] Nielsen et al., 1989: Data for Danish Participants; Majority Were Tested Several
Weeks Post-Minor Surgery
Trial
15-Minute Delayed
n Age I II m IV v Total Recall
35 20-29 6.31 9.77 11.31 12.31 12.74 52.46 11.91

(1.53) (2.00) (1.97) (1.79) (1.46) (7.21) (1.76)
4-10 6-14 7-14 8-15 9-15 37~ 9-15
27 5.85 9.04 10.26 11.59 12.22 48.96 11.22
(1.15) (1.55) (1.60) (1.71) (1.76) (6.19) (2.56)
4-8 6-12 6-14 7-15 8-15 33-64 5-15
39 5.67 8.18 10.08 10.77 11.41 46.10 9.92
(1.31) (1.89) (2.29) (2.02) (.2.14) (7.99) (2.73)
4-9 4-12 5-15 6-15 7-15 27-63 4-15
Table A19.14. [RAVLT.10] Roth et al., 1989: Data for the Control Group
Trial
Postinterference
n Age Education I II III IV v 8 RecaR Recogoitioo
61 27.5 12.8 7.4 10.4 11.9 12.8 13.7 6.7 12.4 14.0
(l.O)• (0.2) (0.2) (0.3) (0.3) (0.2) (0.2) (0.3) (0.3) (0.2)
•standard error.
Table A19.15. [RAVLT.lla] Geffen et al., 1990: Data for Healthy Australian Adults Stratifled by Age
Group: Males•
Age Group
16-19 20-29 ~ 40-49 50-59 60-69 70-86

Trials (n= 13) (n=10) (n=10) (n=11) (n=11) (n=10) (n=10)
6.9 8.4 6.0 6.4 6.5 4.9 3.6

(1.8) (1.2) (1.8) (1.8) (2.0) (1.1) (0.8)
II 9.7 10.8 8.0 9.0 8.6 6.4 5.7
(1.7) (1.9) (2.4) (2.3) (2.0) (1.2) (1.7)
III 11.5 11.3 9.7 9.8 10.1 8.0 6.8
(1.2) (1.6) (2.7) (2.0) (1.6) (2.6) (1.6)
IV 12.8 12.2 10.9 11.5 10.7 8.5 8.3
(1.5) (1.8) (2.8) (1.9) (1.9) (2.7) (2.7)
v 12.5 12.2 11.4 10.9 11.8 8.9 8.2
(1.3) (2.2) (2.6) (2.0) (2.6) (2.0) (2.5)
Total 53.4 54.9 46.0 47.5 47.6 36.7 32.6
(5.4) (7.0) (10.9) (8.3) (8.5) (8.4) (8.3)
Total repeats 5.9 8.0 3.0 4.1 7.3 5.0 5.1
(5.6) (4.6) (3.6) (2.9) (7.5) (3.6) (8.6)
Extra list intrusions 0.39 0.90 1.20 0.55 0.73 0.30 0.90
(0.65) (1.29) (3.12) (0.82) (1.19) (0.68) (1.67)
ListB 6.9 6.5 5.3 6.1 5.0 4.9 3.5
(1.9) (1.8) (1.6) (2.1) (2.3) (1.6) (1.3)
Postinterference recall 11.2 11.1 9.7 9.7 9.6 7.2 6.4
(1.6) (1.7) (2.3) (2.5) (2.9) (2.8) (1.7)
20-minute delayed recall 11.3 10.6 10.4 10.5 10.0 7.1 5.6
(1.7) (2.4) (2.3) (2.7) (2.6) (3.8) (lL6)
&copidon
List A 14.4 14.2 13.5 14.2 13.9 12.4 11.5
(0.9) (0.8) (1.5) (1.0) (0.9) (2.8) (2.6)
ListB 8.4 8.2 4.4 6.9 4.7 4.9 3.0
(2.8) (2.7) (2.0) (2.6) (2.9) (2.7) (2.5)
p(A) 0.95 0.90 0.92 0.92 0.90 0.82 0.81
List A (0.04) (0.05) (0.04) (0.06) (0.06) (0.13) (0.10)
p(A) 0.77 0.76 0.64 0.71 0.65 0.65 0.59
ListB (0.09) (0.09) (0.01) (0.09) (0.10) (0.09) (0.08)
Miaaaipmenla
A to B 0.77 1.00 0.70 1.18 1.18 2.20 0.80
(1.01) (1.89) (1.25) (1.33) (1.54) (1.32) (1.03)
BtoA 0.08 0.40 0.60) 0.18 0.27 1.40 1.00
(0.28) (0.84) (0.70) (0.60) (0.65) (1.78) (1.25)
Total false-positives 1.39 3.20 1.50 2.64 2.91 4.30 3.10
(1.76) (1.99) (1.35) (2.66) (2.55) (3.95) (2.96)
•Data for additional indices developed by the authors are included.

APPENDIX 19 851
Table A19.16. [RAVLT.llb] Geffen et al., 1990: Data for Healthy Australian Adults Stratified by Age
Group: Females•
Age Group
16-19 20-29 ~ 40-49 50-59 60-69 70-86

Trials (n=l2) (n=10) (n=13) (n=11) (n=9) (n=12) (n=10)
7.8 7.7 8.0 6.8 6.4 6.0 5.6

(1.9) (1.0) (2.0) (1.5) (1.5) (2.2) (1.4)
11 10.5 10.5 10.8 9.4 8.2 9.0 6.9
(2.0) (2.0) (2.1) (1.5) (2.4) (2.0) (2.1)
III 12.3 12.2 11.5 11.4 10.2 10.8 8.9
(1.2) (2.3) (1.7) (1.7) (2.1) (2.0) (1.9)
IV 12.5 12.0 12.9 11.7 11.1 11.3 10.1
(1.7) (1.6) (1.3) (2.1) (1.9) (1.4) (1.9)
v 13.3 12.9 12.7 12.8 11.6 11.9 10.1
(1.5) (1.5) (1.3) (1.4) (2.1) (1.6) (1.2)
Total 56.5 55.3 55.9 52.1 47.6 49.0 41.6
(6.0) (6.6) (6.3) (7.1) (7.7) (7.1) (6.6)
Total repeats 5.5 10.6 5.0 8.0 4.9 4.8 3.5
(6.5) (14.3) (5.8) (4.8) (3.7) (2.8) (4.8)
Extra list inbusions 0.92 1.20 1.23 0.83 0.78 0.67 0.50
(1.38) (1.40) (1.74) (1.19) (1.30) (1.07) (0.97)
ListB 7.7 7.9 6.5 5.2 4.6 5.3 4.2
(1.3) (2.0) (1.5) (1.3) (1.9) (1.1) (1.9)
Postinterference recall 11.9 11.6 12.1 11.1 9.9 9.8 7.8
(2.5) (2.5) (1.9) (2.4) (2.8) (1.6) (1.8)
20-minute delayed recall 11.4 11.0 12.2 11.1 10.2 10.3 8.3
(2.5) (2.0) (2.5) (2.3) (2.7) (2.3) (2.1)
~
List A 13.8 14.4 14.2 14.4 13.7 13.8 13.6
(2.0) (0.8) (1.7) (0.8) (1.1) (1.1) (2.0)
ListB 7.8 8.0 8.9 7.4 5.7 7.5 7.5
(3.1) (2.9) (4.1) (2.8) (2.4) (3.6) (3.7)
p(A) 0.92 0.91 0.89 0.88 0.88 0.90 0.84
List A (0.08) (0.09) (0.08) (0.07) (0.08) (0.06) (0.11)
p(A) 0.74 0.75 0.78 0.73 0.68 0.74 0.73
ListB (0.10) (0.10) (0.13) (0.09) (0.07) (0.11) (0.10)
M~
A to B 0.33 0.40 0.31 1.17 1.56 1.42 0.90
(0.49) (0.97) (0.63) (1.64) (1.94) (1.44) (1.37)
BtoA 0.25 0.30 0.00 0.17 0.33 0.58 1.10
(0.45) (0.48) (0.00) (0.58) (0.71) (0.67) (0.88)
Total false-positives 2.33 3.60 4.23 4.58 3.67 2.92 5.60
(2.96) (3.92) (3.37) (2.68) (3.71) (3.12) (5.72)
"Data for additional indices developed by the authors are included.

852 APPENDIX 19
Table A19.17. [RAVLT.llc] Geffen et al. 1990: Mean (SO) Number of Words Recalled in Five Grouped
Serial Positions of Words in List A Averased Over the Five Acquisition Trials
Age
Serial Position 16-19 20-29 ' ~9 40-49 50-59 60-69 70-86
Malea
I (1-3) 12.7 12.3 10.3 12.9 11.2 9.0 9.3
(2.2) (2.1) (2.9) (1.9) (3.5) (2.7) (2.3)
II (4-6) 10.3 11.1 8.6 9.0 10.5 6.2 5.2
(1.9) (1.8) (3.1) (2.7) (2.2) (3.7) (2.6)
III (7-9) 7.8 7.7 6.7 6.0 7.2 3.5 2.4
(2.7) (3.0) (3.4) (3.0) (3.2) (2.4) (2.5)
IV (1~12) 10.6 12.0 10.0 8.8 8.7 8.6 7.6
(1.9) (2.4) (2.1) (2.6) (3.0) (3.2) (3.8)
v (1~15) 12.0 11.8 10.4 10.8 10.1 9.4 8.1
(1.9) (2.1) (3.3) (2.3) (2.1) (1.4) (2.8)
F...,.,_
I (1-3) 12.5 12.4 13.2 12.6 11.7 11.9 10.3
(1.9) (2.3) (1.6) (2.7) (3.5) (1.7) (2.6)
II (4-6) 11.6 11.0 11.2 9.8 8.9 10.1 8.1
(1.7) (1.8) (2.1) (2.4) (2.3) (1.8) (2.3)
III (7-9) 10.2 8.4 8.7 8.2 7.3 8.0 4.9
(2.3) (3.8) (2.8) (2.6) (3.4) (2.9) (2.0)
IV (1~12) 11.3 11.5 11.2 10.9 9.0 9.1 8.4
(1.5) (1.8) (2.7) (2.5) (2.9) (1.6) (2.8)
V(1~15) 11.0 12.0 11.6 10.6 10.7 9.9 9.9
(2.9) (2.7) (2.4) (2.1) (1.9) (3.1) (3.3)
© Swets ~ Zeitlinger (1990).
Table A19.18. [RAVLT.12a] Ivnik Table A19.18. (Contd.)

et al., 1990: Demographic Description
n
of the Cognitively Intact Sample
Partitioned into Groups Used in the Mtwifal .,.,.,.
RAVLT Testing Single 51
n Married 235
Divorced 7
Age Widowed 100
No response 1
55-59 45
60-64 53 .Edueation (,-.)
65-69 64 <8 9
7~74 67 8-11 60
7>79 69 12-15 226
60-64 49 16-17 60
85-97 47 > 17
B,.,.,.._ 39
Right-handed 369
Left-handed 14
Mixedlboth 11
Gend.r
Men 145
Women 249
APPENDIX 19 853
Table A19.19. [RAVLT.12b] Ivnik et al., 1990: Data for the Cognitively Intact Sample Stratified into Seven
Age Groups
Trials
Age Recall after 30-Minute
Group n II III IV v List B Interference Delayed Recall Recognition Errors
lJ5.-.S9 45
M 6.8 9.5 11.4 12.4 13.1 5.3 11.2 10.4 14.0 0.6
so 1.6 2.2 2.0 1.9 1.9 1.7 2.5 3.1 1.3 0.9
Range 4-10 6-14 6-15 7-15 7-15 2-9 >15 0-15 10-15 0-3
60-64 53
M 6.4 9.0 10.6 11.7 11.9 4.9 10.0 9.9 13.9 0.8
SD 1.9 2.3 2.3 2.7 2.0 1.5 3.1 3.1 1.5 1.2
Range ~13 ~14 6-14 7-15 7-15 ~ 4-15 ~15 8-15 0-5
65-69 64
M 5.7 8.6 9.7 10.6 11.2 4.7 9.1 8.3 13.3 0.9
SD 1.6 2.1 2.3 2.4 2.4 1.5 3.2 3.5 2.0 0.9
Range 1-10 5-13 4-15 6-15 6-15 1-9 0-15 0-15 8-15 0-3
7~74 67
M 5.5 7.8 9.1 10.2 10.5 4.1 8.3 7.4 12.7 1.0
SD 1.5 1.8 2.1 2.3 2.6 1.5 2.9 3.1 2.1 1.2
Range 2-9 ~12 4-13 ~14 5-15 1-8 1-14 0-13 6-15 0-5
75-79 65
M 5.0 7.0 8.2 9.2 10.1 4.2 7.8 6.9 12.5 1.5
so 1.5 1.9 2.2 2.2 2.2 2.0 2.7 2.9 2.4 1.6
Range 1-8 ~12 ~15 4-15 5-15 1-10 2-15 0-14 6-15 0-7
80-84 49
M 4.4 6.5 7.7 8.6 9.0 3.5 6.7 5.5 12.3 1.2
SD 1.2 1.5 2.1 2.5 2.5 1.6 2.5 3.3 2.4 1.4
Range 2-7 ~10 ~12 1-14 4-15 0-8 1-14 0-12 2-15 0-i
85-97 47
M 4.0 6.0 7.4 7.9 9.1 3.1 6.2 5.4 12.3 1.5
so 1.5 1.8 2.2 2.4 2.3 1.4 2.6 2.7 2.3 1.6
Range 0-7 2-10 2-12 ~15 ~14 0-7 2-14 0-13 6-15 0-7
854 APPENDIX 19
Table A19.20. [RAVLT.l2c] Ivnik et al., 1990: Summary Scores for the Cognitively
Intact Sample Stratified into Seven Age Groups
Total Learning Short-Term Long-Term

Age n Learning Over Trials Percent Retention Percent Retention
55-59 45
M 53.2 19.3 85.0 79.1
SD 8.2 5.8 12.9 18.7
Range 33-67 3-34 45-100 0-108
60-64 53
M 49.7 17.8 82.8 81.7
SD 9.1 7.0 18.3 18.0
Range 32-68 .2-30 44-118 30-118
6S-69 64
M 45.8 17.2 80.5 72.3
SD 9.5 6.1 21.2 23.6
Range 22-66 (-3)-28 0-133 0-125
7~74 67
M 43.1 15.6 78.7 68.4
SD 9.1 6.9 18.4 23.6
Range 1~1 (-1)-33 12-120 0-111
7fl-79 69
M 39.6 14.5 76.5 67.1
SD 8.7 6.6 18.9 21.6
Range 21-63 2-31 25-133 0-110
80-84 49
M 36.2 14.3 74.0 60.1
SD 7.4 7.3 21.7 33.4
Range 21-51 1-30 25-133 0-150
85-97 47
M 34.4 14.4 69.1 58.7
SD 8.6 6.6 22.7 23.4
Range 1-56 0-31 25-133 0-100
Table A19.21. [RAVLT.l3] Miller et al., 1990: Data for Homosexual/Bisexual Males Participating in the
Multi-Center AIDS Cohort Study
Race•
CES Depression
w 8 H 0 Scale CD4 Age Education Trial V Trials 1-V Total
Seronegative 92 2 4 2 9.08 97o.42 37.20 16.36 12.83 52.75

(9.03) (332.46) (7.52) (2.34) (1.85) (8.05)
Asymptomatic, 91 2 6 2 9.44 561.90 35.66 15.70 12.64 52.18
seropositive (9.27) (277.98) (6.47) (2.44) (1.88) (8.30)
Symptomatic, 90 2 5 3 15.21 277.22 36.90 16.06 12.40 50.51
seropositive (11.19) (269.45) (7.04) (2.50) (2.20) (9.52)
•w, white; 8, black; H, hispanic; 0, others (percentages).

APPENDIX 19 855
Table A19.22. [RAVLT.l4a] Shapiro and Harrison, 1990: The Original AVLT (List AB), the Alternate List
(CD), and Two New Lists (EF and GH)
List AB" List cot ListEF List GH
Drum Desk Book Bowl Street Baby Tower Sky

Curtain Ranger Flower Dawn Grass Ocean Wheat Dollar
Ben Bird Train Judge Door Palace Queen Valley
Coffee Shoe Rug Grant Arm Up Sugar Butter
School Stove Meadow Insect Star Bar Home Han
Parent Mountain Harp Plane Wd'e Dress Boy Diamond
Moon Glasses Salt County Wmdow Steam Doctor Winter
Garden Towel Finger Pool City Coin Camp Mother
Hat Cloud Apple Seed Pupil Rock Flag Christmas
Farmer Boat Chimney Sheep Cabin Army Letter Meat
Nose Lamb Button Meal Lake Building Com Forest
Turkey Gum Key Coat Pipe Friend Nail Gold
Color Pencil Dog Bottle Skin Storm Cattle Plant
House Church Glass Peach Fire Village Shore Money
River Fish Rattle Chair Clock een Body Hotel
"Rey (1964).
tLezak (1983, 1995, 2004).
Table A19.23. [RAVLT.l4b] Shapiro and Harrison, 1990: Data for the
Original and Three Alternate Forms for a Sample of 25 University
Undergraduates•
Trial List AB List CD List EF List GH
I 7.00 7.40 6.84 7.28

(1.63) (1.63) (1.93) (2.39)
II 10.20 10.08 9.76 10.00
(2.24) (1.87) (1.90) (2.25)
III 11.76 11.40 11.12 11.80
(2.45) (1.71) (2.26) (2.08)
IV 12.40 12.40 12.16 12.52
(2.20) (1.68) (1.82) (1.78)
v 13.04 12.80 12.92 13.40
(2.09) (1.55) (2.14) (1.35)
ListB 7.20 6.76 7.04 7.52
(1.89) (1.59) (1.62) (1.50)
Postinterference 12.00 11.64 11.68 12.24
recall (2.61) (2.12) (2.64) (2.31)
"Mean age 19 years.

856 APPENDIX 19
Table A19.24. [RAVLT.14c] Shapiro and Harrison, 1990: Data for the
Original and Three Alternate Fonns for a Sample of 17 VAMC Medical
Patients•
Trial UstAB UstCD UstEF UstGH
4.06 3.29 3.52 3.41

(1.43) (1.96) (1.55) (1.37)
II 5.52 4.94 4.76 4.71
(1.66) (2.08) (2.61) (1.57)
III 6.12 5.59 5.76 5.76
(2.00) (2.09) (2.31) (2.44)
IV 6.41 5.71 6.47 5.65
(2.12) (2.26) (2.65) (2.18)
v 6.47 6.47 6.88 6.47
(2.72) (2.83) (3.16) (3.24)
UstB 3.35 3.41 3.18 3.41
(1.66) (1.66) (1.55) (1.97)
Postinterference 4.06 3.29 3.71 4.17
recall (3.65) (2.87) (3.67) (2.96)
• Mean age 66 years.
Table A19.25. [RAVLT.15a] Mitrushina et al., 1991: Demographic

Characteristics for the Sample of Normal Elderly•
Age Groups
All 57~ 66-70 71-75 76-85
Age 70.7 62.8 68.2 72.9 78.7

(5.4) (2.3) (1.3) (1.4) (2.7)
Education 14.1 14.2 14.0 14.6 13.3
(2.9) (2.0) (2.0) (3.4) (3.6)
WAIS-R FSIQ 117.2 115.8 119.3 118.7 112.0
(12.6) (12.6) (14.5) (10.8) (11.7)
n 156 28 45 57 26
•The sample includes 62 males and 94 females.

APPENDIX 19 857
Table A19.26. [RAVLT.15b] Mitrushina et al., 1991: Average Recall for Four Age Groups of Normal
Elderly
Trials
Postinterference
Age Group I II III IV v Recall Recognition FP" V-VI V-1
57~ 6.4 8.8 10.4 11.4 12.1 10.3 13.2 0.8 1.9 5.7
(1.5) (2.4) (2.5) (2.3) (2.4) (3.0) (1.3) (1.4) (1.7) (2.0)
66-70 5.9 8.5 9.8 11.3 11.5 9.1 13.0 1.0 2.4 5.6
(1.6) (2.3) (2.3) (2.3) (3.0) (3.3) (1.1) (1.6) (2.5) (2.9)
71-75 5.1 7.5 8.7 9.7 10.3 8.4 12.7 1.1 1.9 5.2
(1.8) (2.2) (2.4) (2.7) (2.9) (3.5) (1.9) (1.7) (2.2) (2.5)
76-85 5.1 6.8 8.3 9.5 9.7 7.7 12.6 0.8 2.0 4.7
(1.6) (2.1) (2.3) (2.8) (2.8) (3.4) (1.9) (1.2) (2.6) (2.7)
"False-positive errors.
Table A19.27. [RAVLT.15c] Mitrushina et al.,

1991: Primacy/Recency Effects for the Entire
Sample•
Trial
n III IV v
Beginning 1.9 3.0 3.4 3.9 3.9
Middle 1.1 1.7 2.3 2.7 3.1
End 2.6 3.0 3.3 3.6 3.6
•Provides mean number of words recalled within each

segment of the list across 6w acquisition trials.
Table A19.28. [RAVLT.16a] Mitrushina and Satz,

1991a: Demographic Characteristics for a Sample
of Normal Elderly Stratified into Four Age Groups•
Age Groups
57~ 66-70 71-75 76-85
Mean age 62.2 68.2 72.9 78.3

(2.5) (1.2) (1.4) (2.5)
Mean education 14.4 13.7 14.5 14.0
(2.0) (1.8) (3.1) (3.6)
Male/female(%) 10190 15185 30170 W78
n 19 40 47 16
"Mean Wechsler Adult InteUigence Scale-Revised full.

scale IQ for the sample is 118.2 (13.0).
858 APPENDIX 19
Table A19.29. [RAVLT.16b] Mitrushina and Satz, 1991a: Performance over 'Three Longitudinal Probes for
Four Age Groups of Normal Elderly
Age Groups
57-65 66-70 71-75 76-85

(n=19) (n=40) (n=47) (n=16)
Tl T2 1'3· Tl T2 T3 T1 T2 T3 T1 T2 T3
Trial I 6.7 6.4 7.9 6.0 6.2 7.3 5.1 5.4 6.4 5.1 5.8 6.0
(1.6) (1.3) (2.3) (1.6) (1.8) (1.8) (2.0) (1.7) (1.8) (1.5) (1.2) (1.9)
Trial V 12.4 12.3 11.9 11.8 11.7 12.1 10.4 10.1 10.4 10.3 10.6 9.8
(2.6) (2.5) (2.8) (2.5) (2.5) (2.3) (2.7) (3.3) (3.2) (2.4) (3.2) (4.0)
Postinterference 10.7 10.8 10.5 9.5 9.9 10.4 8.7 8.4 8.9 8.4 8.5 7.9
recall (3.2) (3.2) (3.4) (3.0) (3.2) (2.8) (3.5) (3.6) (3.6) (3.5) (3.8) (4.7)
•Three annual probes.
Table A19.30. [RAVLT.17] Seines et al., 1991:

Data for Seronegative Homosexual and Bisexual
Males who Participated in the Multi-Center AIDS
Cohort Study Table A19.30. (Contd.)
Age Group Education
25-34 35-44 45-54 <College College >College
n 309 290 97 n 229 202 302
Education 16.1 16.4 16.7 Mean age 36.1 35.6 38.4
(2.2) (2.3) (2.6) (7.4) (7.2) (7.8)
Race 96.4%C• 96.6%C 95.9%C Race 94.8%C 96.0%C 96.7%C
Total score 54.4 51.4 49.5 Total score 50.7 53.2 53.3
(7.8) (8.1) (7.9) (7.5) (8.1) (8.3)
Trial V 13.0 12.6 12.3 Trial V 12.6 12.8 12.9
(1.8) (1.8) (1.8) (1.7) (1.9) (1.8)
Recall after 11.3 10.7 10.6 Recall after 10.5 10.9 11.2
interference (2.4) (2.6) (2.8) interference (2.5) (2.5) (2.6)
Delayed recall 11.1 10.4 10.2 Delayed recall 10.1 10.8 10.9
(2.5) (2.9) (3.2) (2.7) (2.9) (2.8)
Delayed recognition 14.4 14.1 14.0 Delayed recognition 14.1 14.4 14.2
(0.9) (1.2) (1.2) (1.2) (0.9) (1.1)
•c. Caucasian.
APPENDIX 19 859
Table A19.31. [RAVLT.18] Delaney et al., 1992: Data for the Control Sample
Trial
Postinterference 20-Minute
Form n Age Education I III v Recall Delayed Recall Recognition
A 42 45.8 12.8 6.0 10.1 11.6 9.9 9.9 13.6

(22-67) (6-16) (2.1) (2.4) (2.5) (3.2) (3.1) (1.8)
c 6.1 10.0 11.8 9.9 9.2 14.0
(2.2) (2.4) (2.8) (3.3) (3.5) (1.2)
Table A19.32. [RAVLT.19] Ivnik Table A19.33. [RAVLT.20a] Savage and Gouvier,
et al., 1992c: Demographic Descrip- 1992: Data for Healthy Adults: Trials 1-V
tion of the Cognitively Intact Sample
Partitioned into Groups Used in the Trials
RAVLT Testing Age n I II III IV v
n 16-19
Males 10 6.2 7.7 9.9 11.1 12.3
Age lf'OUJ'JII (0.8) (1.8) (1.5) (2.3) (1.6)
56-59 41 Females 10 6.5 7.7 10.6 11.7 12.9
60-64 72 (1.7) (2.3) (2.4) (1.2) (0.2)
65-69 83 20-29
7~74 82 Males 10 6.4 8.4 9.6 10.1 10.5
75-79 105 (2.0) (2.4) (2.3) (3.1) (1.9)
~ 76
Females 9 5.7 7.3 8.0 9.6 10.3
85-89 49
(1.0) (1.3) (2.7) (2.3) (2.0)
90-97 22
B..,.,_ 30-39
Males 9 5.5 7.6 9.0 9.5 9.8
Right-handed 501 (1.1) (1.0) (2.3) (2.7) (2.1)
Left-handed 16 Females 10 6.2 8.6 10.8 10.7 11.8
Mixed/both 13 (3.2) (3.8) (2.3) (2.2) (1.9)
Gender 40-49
Males 200
Males 9 6.0 7.3 9.1 9.8 10.4
(1.2) (2.0) (1.8) (1.2) (2.1)
Females 330
Jlariltd Sltdu Females 10 5.7 7.3 9.1 9.4 10.4
(1.5) (2.2) (2.7) (3.2) (3.1)
Single 69 50-59
Married 318 Males 9 5.7 8.1 9.1 9.4 9.3
Divorced 12 (0.8) (1.0) (1.8) (2.5) (2.2)
Widowed 131
Females 10 5.6 7.9 8.8 11.3 11.6
.8Gc:e (1.6) (1.5) (2.2) (2.4) (2.2)
Caucasian 528 ~
Black 1 Males 9 5.0 6.0 7.4 8.2 8.4
Hispanic 1 (1.0) (1.1) (2.5) (2.3) (2.3)
.Et.fucadon (,_..) Females 10 5.6 7.4 7.0 9.0 9.3
$.7 8 (1.2) (1.5) (2.7) (2.5) (1.9)
8-11 84 7~76
12 192 Males 10 5.3 6.3 7.5 7.9 8.1
13-15 117 (1.4) (1.2) (1.6) (2.4) (2.4)
16-17 82 Females 9 5.6 6.5 6.5 6.7 7.4
<::18 47 (1.1) (1.7) (1.8) (1.6) (1.6)
860 APPENDIX 19
Table A19.34. [RAVLT.20b] Savage and Gouvier, 1992: Data for Healthy Adults: List B
Through Delayed Recognition
30-Minute
Postinterference Immediate Delayed Commission Delayed
Age ListB Recall Recognition Recall Errors Recognition
16-19
Males 5.9 10.6 14.1 9.9 0.60 13.9
(0.74) (2.5) (1.2) (2.5) (0.96) (1.2)
Females 5.4 12.1 14.4 11.4 0.60 14.3
(1.2) (2.6) (0.51) (2.6) (0.96) (0.67)
20-29
Males 4.7 10.1 14.2 10.0 0.30 14.0
(1.6) (3.0) (1.6) (3.4) (0.67) (1.6)
Females 5.3 8.7 14.0 8.6 0.55 13.8
(2.1) (2.3) (1.1) (2.1) (0.88) (1.6)
30-39
Males 4.6 8.3 13.1 9.0 0.11 12.6
(2.1) (2.7) (2.0) (3.3) (0.33) (2.4)
Females 5.0 10.7 14.0 11.7 0.40 13.6
(2.2) (2.9) (0.94) (2.8) (0.52) (2.8)
40-49
Males 5.0 8.1 13.1 7.6 0.11 13.5
(1.0) (1.9) (1.9) (2.0) (0.33) (2.1)
Females 4.7 8.2 13.3 7.6 0.90 12.9
(1.4) (3.4) (2.2) (3.5) (1.4) (2.8)
50-59
Males 4.3 8.3 12.8 7.5 0.88 12.5
(2.0) (2.4) (1.7) (2.7) (1.4) (2.1)
Females 4.5 9.2 12.0 9.4 0.60 13.3
(1.6) (1.9) (2.8) (2.3) (1.1) (2.4)
APPENDIX 19 861
Table A19.35. [RAVLT.2l] Crossen and Wiens, Table A19.36. [RAVLT.22a] Geffen et al., 1994:
1994: Data for Job Applicants (n =60):° Compar- Data for the Rey AVLT Forms land 4 for a Sample
ison of AVLT and CVLT Scores of 51 Healthy Australian Volunteers•
AVLT CVLT Fonn 1 Fonn4
Trial I 7.0 7.5 Trial I 6.82 6.82

(1.6) (1.6) (1.47) (1.58)
Trail II 9.7 10.5 Trial II 9.35 8.90
(2.0) (1.9) (1.98) (1.98)
Trial III 11.1 11.7 Trial III 10.92 10.76
(2.0) (2.1) (1.97) (1.99)
Trial IV 11.7 12.5 Trial IV 11.55 11.53
(2.1) (2.0) (2.12) (2.05)
Trial V 12.2 13.0 Trial V 12.47 12.00
(1.8) (1.8) (1.91) (1.99)
Total Words 51.7 55.1 Total (I-V) 51.12 50.02
(7.5) (7.7) (7.42) (7.68)
ListB 7.0 7.9 List B 6.02 5.68
(2.0) (1.9) (1.73) (1.68)
Postinterference Recall 10.6 11.7 Postinterference 10.88 10.65
(3.1) (2.3) recall (2.91) (2.71)
Recognition 13.4 14.7 20-minute delayed 10.82 10.33
(1.2) (1.4) recall (2.99) (2.83)
Recognition A total 13.71 13.65
"The sample includes 52 men and 8 women with mean
(1.60) (1.48)
age of29.9 (6.2) years, mean education of 14.7 (1.6) years,
and mean WAIS-R FSIQ of 106.3. Recognition A p[A] 0.90 0.90
(0.08) (0.07)
Recognition B total 6.94 7.59
(2.92) (2.77)
Recognition B p[A] 0.71 0.73
(0.10) (0.09)
"The sample includes 25 males and 26 females with a

mean age of 31.3 (12.7) years, mean education of 12.2
(2.4) years, and mean National Adult Reading Test-
estimated IQ of 115.6 (6.26).
862 APPENDIX 19
Table A19.37. [RAVLT.22b] Geffen et al., 1994: Frequency (Approximate Occurrence

per Million) and Length (Number of Letters) of Words, Comparing Rey AVLT Form 1
with Form 4, List A
Form 1 Fonn4
List A Frequency Length List A Frequency Length
Drum 11 4 Pipe 20 4
Curtain 13 7 Wall 160 4
Bell 18 4 Alarm 16 5
Coffee 78 6 Sugar 34 5
School 492 6 Student 131 7
Parent 15 6 Mother 216 6
Moon 60 4 Star 25 4
Garden 60 6 Painting 59 8
Hat 56 3 Bag 42 3
Fanner 23 6 Wheat 9 5
Nose 60 4 Mouth 103 5
Turkey 9 6 Chicken 37 7
Color 141 6(5) Sound 204 5
House 591 5 Door 312 4
River 165 5 Stream 51 6
Mean 119.47 5.2 94.6 5.2
Range 9-591 3-7 ~12 3-8
Table A19.38. [RAVLT.22c] Geffen et al., 1994: Frequency (Approximate

Occurrence per Million) and Length (~mber of Letters) of Words, Comparing
Rey AVLT Form i witllForm 4, List B .
Form 1 Fonn4
ListB Frequency Length ListB Frequency Length
Desk 65 4 Bench 35 5
Ranger 2 6 Officer 101 7
Bird 31 4 Cage 9 4
Shoe 14 4 Sock 4 4
Stove 15 5 Fridge• 23 6
Mountain 33 8 Cliff 11 5
Glasses 29 6 Bottle 76 6
Towel 6 5 Soap 22 4
Cloud 28 5 Sky 58 3
Boat 72 4 Ship 83 4
Lamb 7 4 Goat 6 4
Gun 118 3 Bullet 28 6
Pencil 34 6 Paper 157 5
Church 348 6 Chapel 20 6
Fish 35 4 Crab• 2 4
Mean 56.47 4.93 42.33 4.87
Range J.-.348 3-8 2-157 3-7
-The actual words are not present in the tab!~ of word frequency.
APPENDIX 19 863
Table A19.39. [RAVLT.22d] Geffen et al., 1994: Word List for Testing Rey AVLT
Recognition (Form 4)•
Alarm (A) Eye (SA) Soap (B) Ship (B) Bottle (B)
Aunt (SA) Crab (B) Wall (A) Car (PA) Seat (SB)
Bag (A) Star (A) Clock (SA) Mother (A) Sock (B)
Creek (SA) Rag (PA) Sound (A) Duck(SA) Tone (SA)
Officer (B) Bun (PA) Bench (B) Wheat(A) Fridge (B)
Mouth (A) Cage (B) Bullet (B) Floor (SPA) Rock (SPB)
Arrow (SB) Cliff (B) Night (SA) Sky (B) Bread (SA)
Student (A) Sugar (A) Chapel (B) Door(A) Pipe (A)
Hail (PA) Cream (PA) Chicken (A) Bridge (PB) Ball (PA)
Paper (B) Stream (A) Coat (PB) Painting (A) Goat (B)
•A, words from list A; B, words from list B; P, phonemic associate of words on lists A and B;
S, semantic associate of words on lists A and B.
Table A19.40. [RAVLT.23] Torres et al., 2001: Data for the Control Group
WAIS-R
Interference Recall after
n Age Education VIQ FSIQ Trial I Trial V Trial Interference
Males 84 29.0 14.4 113 114 6.6 12.7 6.3 11.6

(10.2) (2.1) (14) (13) (2.1) (2.1) (2.1) (2.9)
Females 76 30.5 14.8 109 112 7.2 13.2 6.7 12.3
(11.7) (1.9) (12) (12) (2.1) (1.6) (2.4) (2.3)
864 APPENDIX 19
Table A19.41. [RAVLT.24) Miller, 2003 (Update on Seines et al.,1991): Data for a Sample of Seronegative
Homosexual/Bisexual Males Participating in the Multi-Center AIDS Cohort Study, Stratified by
Age x Education
Education Interference Delayed Delayed False-
Age (years) Trial I Trial V Total 1-V Trial Recall Recognition Positives
25-34 <16
Mean 6.67 12.63 51.36 7.01 10.63 14.0 0.58
(SO) (1.72) (1.67) (7.02) (2.03) (2.62) (1.04) (1.13)
n 102 102 102 67 67 67 67
16
Mean 7.18 13.13 55.39 6.80 11.62 14.36 0.39
(SO) (1.72) (1.84) (7.47) (1.94) (2.24) (0.94) (0.72)
n 96 96 96 66 66 66
>16
Mean 7.41 13.49 56.91 7.26 11.97 14.30 0.36
(SO) (1.74) (1.41) (6.59) (2.03) (1.90) (0.93) (0.65)
n 35 110 98 66 66 66 66
Total
Mean 7.08 13.07 54.50 7.03 11.40 14.22 0.45
(SO) (1.75) (1.68) (7.40) (2.00) (2.33) (0.98) (0.86)
n 296 296 296 199 199 199 199
35-44 <16
Mean 6.50 12.40 50.35 6.11 10.12 13.49 0.55
(SO) (1.72) (1.80) (7.51) (1.87) (2.87) (1.76) (0.94)
n 128 128 128 81 80 80 80
16
Mean 6.69 12.69 52.11 6.24 10.29 13.68 0.48
(SO) (1.59) (1.78) (7.50) (2.13) (3.04) (1.53) (0.77)
n 112 112 112 79 79 79 79
>16
Mean 6.85 12.80 52.71 7.17 10.72 13.88 0.35
(SO) (1.76) (1.88) (8.21) (2.17) (2.73) (1.46) (0.57)
n 177 177 177 121 121 121 1.21
Total
Mean 6.70 12.65 51.82 6.60 1G.43 13.71 0.44
(SO) (1.71) (1.83) (7.86) (2.13) (.2.87) (1.57) (0.75)
n 417 417 417 281 280 280 280
45-59 <16
Mean 6.21 11.87 47.72 5.85 9.95 13.41 0.72
(SO) (1.97) (1.74} (7.40} (2.36} (2.28) (1.71) (1.12)
n 61 61 61 40 39 39 39
16
Mean 6.27 11.93 48.89 5.94 9.27 13.88 0.67
(SO) (1.40) (1.96} (8.68) (1.75) (3.51) (1.75) (1.34)
n 44 44 44 33 33 33 33
>16
Mean 6.30 12.54 50.79 5.92 10.47 13.86 0.56
(SO} (1.69) (1.89} (8.04) (1.69} (2.89} (1.34) (1.11)
n 100 100 100 71 70 70 70
Total
Mean 6.27 12.21 49.47 5.90 10.05 13.74 0.63
(SO) (1.72} (1.88) (8.08) (1.90) (2.92) (1.55) (1.16)
n 205 205 205 144 142 142 142
APPENDIX 19 865

Education Interference Delayed Delayed False-
Age (years) Trial I TnalV Total 1-V Trial Recall Recognition Positives
Total <16
Mean 6.50 12.37 50.15 6.38 10.27 13.66 0.60
(SD) (1.78) (1.76) (7.41) (2.09) (2.67) (1.55) (1.05)
n 291 291 291 188 186 186 186
16
Mean 6.80 12.72 52.79 6.39 10.60 13.97 0.48
(SD) (1.64) (1.87) (8.03) (2.01) (2.98) (1.42) (0.88)
n 252 252 252 178 178 178 178
> 16
Mean 6.85 12.91 53.30 6.85 10.97 13.98 0.41
(SD) (1.78) (1.81) (8.08) (2.09) (2.65) (1.32) (0.78)
n 375 375 375 258 257 257 257
Total
Mean 6.72 12.69 52.16 6.58 10.65 13.88 0.49
(SD) (1.75) (1.82) (7.97) (2.08) (2.77) (1.42) (0.90)
n 918 918 918 624 621 621 621
Table A19.42. [HVLT-R.l] Friedman et al., 2002: Table A19.42. (Contd.)
Performance on the Hopkins Verbal Learning Test-
Performance Years of M
Revised for a Sample of Healthy African-American
Index Gender Education (SO) n
E lderly Stratified by Gender and Education
Percent Retention Male < 12 91.05 37
Performance Years of M
(10.61 )
Index Gender Education (SO ) n
12 5.45 11
Ages 60-71 (14.11)
Total Recall Male <12 16.89 37 > 12 1.00 6
(3.40) (24.62)
12 17.09 11 Total .80 54
(2.21) (13.56)
> 12 18.33 6 Female <12 92.97 30
(3.98) (13.18)
Total 17.09 54 12 95.63 16
(3.23) (1 .53)
Female < 12 16.93 30 > 12 .64 11
(2.96} (19.21 )
12 19.56 16 Total 92.11 57
(2.22) (16.21)
> 12 21.64 11 Total < 12 91.91 67
(4.88} (11.77)
Total 18.58 57 12 91.4 27
(3.70) (17.33)
Total < 12 16.91 67 > 12 3.35 17
(3.18) (20.57)
12 18.56 27 Recognition Male < 12 9.03 37
(2.50) Discrimination (1.74)
> 12 20.47 17 Index 12 9.36 11
(4.74) (2.06}
Total 17.86 ll1 > 12 9.67 6
(3.54) (1.97)
Delayed Recall Male < 12 6.43 37 Total 9.17 54
(1.54) (1. 1)
12 5.82 11 Fe mal < 12 9.37 30
(1.08) (2.13}
> 12 5.67 6 12 9.63 16
(2.16) (1.54)
Total 6.22 54 > 12 10.73 11
(1.54) (1.74)
Female < 12 6.47 30 Total 9.70 57
1.38 (1.95}
12 7.37 16 Total < 12 9.18 67
(1.31) {1.91)
> 12 7.45 11 12 9.52 27
(2.50) (1.74)
Total 6.91 57 > 12 10.35 17
(1.67) (1.84)
Total < 12 6.45 67 Total 9.44 111
(1.46} (1.89)
12 6.74 27 (continued )
(1.43)
> 12 6.82 17
(2.48)
Total 6.58 111
(1.64)
APPENDIX 19 867
Table A19.42. (Contd.) Table A19.42. (Contd.)
Perfonnance Years of M Perfonnance Years of M

Index Gender Education (SD) n Index Gender Education (SD) n
Aga7~ 12 80.75 4
Total Recall Male <12 13.71 49 (28.32)
(4.09) >12 50.00
12 14.25 4 (-)
(2.06) Total 87.80 54
>12 15.00 1 (17.99)
(-) Female <12 89.47 57
Total 13.78 54 (17.45)
(3.93) 12 96.45 11
Female <12 15.96 57 (12.18)
(4.05) >12 72.25 4
12 20.00 11 (29.02)
(2.65) Total 89.58 72
>12 18.50 4 (17.92)
(6.35) Total <12 89.32 106
Total 16.72 72 (16.93)
(4.23) 12 92.27 15
Total <12 14.92 106 (18.16)
(4.20) >12 67.80 5
12 18.47 15 (27.03)
(3.58) Total 88.82 126
>12 17.80 5 (17.90)
(5.72) Recognition Male <12 7.53 49
Total 15.46 126 Discrimination (2.72)
(4.34) Index 12 8.00 4
Delayed Recall Male <12 5.08 49 (1.83)
(1.74) >12 9.00 1
12 4.75 4 (-)
(2.06) Total 7.59 54
>12 3.00 1 (2.64)
(-) Female <12 8.44 57
Total 5.02 54 (2.61)
(1.75) 12 11.27 11
Female <12 6.05 57 (0.90)
(2.03) >12 10.00 4
12 7.82 11 (2.45)
(1.17) Total 8.96 72
>12 5.75 4 (2.61)
(3.30) Total <12 8.02 106
Total 6.31 72 (2.69)
(2.08) 12 10.40 15
Total <12 5.60 106 (1.88)
(1.96) >12 9.80 5
12 7.00 15 (2.17)
(1.96) Total 8.37 126
>12 5.20 5 (2.70)
(3.11)
Total 5.75 126
(2.04)
Percent Retention Male <12 89.14 49
(16.47)
868 APPENDIX 19
Table A19.43. [WHO-UCLA AVLT.l] Ponton et al., 1996: Data for a Sample of 300 Spanish-Speaking
Healthy Participants Stratified by Gender, Age, and Education
Age Group
1~29 30-39 40-49 50-75
Education (Yean)
<10 >10 <10 >10 <10 >10 <10 >10
Mala n 11 25 13 18 12 17 18 6
Trial V x 12.73 13.12 12.23 13.33 12.92 13.53 12.11 12.67
recall (SD) (1.56) (1.90) (1.64) (1.33) (1.78) (1.42) (1.68) (1.51)
Recall after x 11.73 12.24 11.46 11.56 10.50 13.00 10.50 11.00
interference (SD) (1.35) (2.67) (2.15) (2.09) (2.97) (2.03) (2.09) (1.67)
20-minute x 12.36 12.52 11.23 12.61 11.42 13.18 10.83 11.83
delayed recall (SD) (1.91) (2.08) (2.42) (1.61) (2.35) (1.78) (2.15) (1.60)
Fetn4Jlee n 12 30 22 44 16 11 25 20
Trial V x 13.33 13.53 12.77 13.77 12.56 13.27 11.52 13.20
recall (SD) (1.56) (1.94) (2.22) (1.41) (0.96) (2.05) (1.94) (1.32)
Recall after x 11.58 12.37 11.59 12.11 10.56 12.09 10.24 10.75
interference (SD) (1.73) (2.31) (2.72) (2.10) (1.63) (1.87) (2.62) (2.61)
20-minute x 11.75 12.93 11.86 12.89 11.06 12.50 10.63 12.45
delayed recall (SD) (2.18) (2.45) (2.59) (2.01) (1.61) (1.90) (2.36) (1.96)
for the Rey Auditory-Verbal
Learning Test (Rey AVLT)
Rey AVLT, Trial I
Number o£ studies iDcluded in the analysis 8

Yean of publication 1988-2003
Number o£ data points used in the analysis 24
Total number o£ participants 1,910
Sample size
Mean 24 50.99 64.78 9-417
Age
Mean 24 57.37 19.36 19.0-82.0
SD 24 2.98 3.26 1.~11.7
Education
Mean 18 14.07 0.76 12.8-16.0
SD 14 1.99 1.06 0.5-3.6
IQ
Mean 2 112.98 1.41 112-114
SD 2 12.49 0.71 12-13
Percent fiiGle 16 39.73 41.37 ~100
TatKOtWmeana
Combined mean 24 6.09 0.89 4.4-7.4
Combined SD 24 1.71 0.29 1.2-2.2
(continued)
869
870 APPENDIX 19M
Predicted number of words recaDed

age group• (Bey AVLT, Trial I) .
,..d SDs, per
95%CI
Age Predicted I..Dwer Upper

RtJnge Seore Band Band
J0-.!4 '7.10 6.78 7.41
.25-.29 '7.10 6.78 7.43
30-34 '7.08 6.71 7.44
35-39 '7.01 6.59 7.42 Standard deviation for all age groups is 1.'71.
40-44 6.90 6.46 7.35
45-49 6.'76 6.31 7.21
50-S4 6.58 6.15 7.01
tlS-S9 6.36 5.98 6.74
60-64 6.10 5.80 6.41
65-69 15.81 5.61 6.00
70-74 5.4'7 5.41 5.54
75-19 5.10 4.97 5.24
•Based on the equation: 1
Predicted tat ICOnJ = 6.581533 + 0.0399874 • !rge- 0.0007624 • oge2

I
Signiflcanee tests for regression with +e test seores
I
OrdiDary leut-11J118"18 regresaion or teat - Oil age (quadratic)
Number of obsemations ' 24
Number of clusters : 8
R2 . 0.842
FcdO• p Fc2.7) = ~.38, p < 0.000
Age 0.0399874 0.031 1.28• 0.241° -0.034 to 0.114

Age2 -0.0007624 0.000323 -2.36 0.050 -0.002 to 0.000
Constant 6.581533 0.580 11.34 0.000 5.21 to 7.95
•signiflcance test for age centered (sample m . - aggregate mean): t = -7.16, p = 0.000.
Predictioa
SEe 0.16
95%CI 6.13-6.77
APPENDIX 19M 871
0 0
7
0
6 0 0
0 0
0
5
0
4
20 30 40 50 60 70 60
age
Rey AVLT trial I. The size of the bubbles reflects the weight of the data point, with larger bubbles indicating
Tests for assumptions and model fit
Tests for heterogeneity in the 6aal data set

Q(dl), p Q(23) =445.01, p < 0.000
Tests for model 8t-4ddition of a quadratic term
Model BIC BIC'
Linear 0.762 0.751 -42.698 -31.258

Quadratic 0.842 0.827 -49.378 -37.938
BIC' difference of 6.680 provides strong support for the quadratic model.
Tests for parameter specifications

Homoscedasticity
White's general test 8.055, p =0.090
Significance tests for regression with the SDs
A regression of SDs on age yielded an R2 of 0.335 (F(l.7>=8.57, p =0.022).

Therefore, the SD for the aggregate sample is suggested for use with all age groups.
(continued)
872 APPENDIX 19M
EtTeets of delllOJI'8Phic variables
Education
Est. tau2 without education 0
B_2 0.850
'f
Education 0.188231 0.179 0.334 -0.250 to 0.627
Gender
t-test by gender
n X male (SO) X female (SO) M-F difference t p
7M,5F 6.964 (1.869) 5.794 (1.566) 1.170 3.148 0.005
Table A19m.2. Results of the Meta-Analy ~and Predicted Scores for the
Bey AVLT, Trial V
(Relevant values are weighted on the stan( lard error for the test mean)
Description o£ the aggregate sample
Number of studies inelucled in the ~ 8

Yean of pubUcatloo , 1988-2003
Number of data points ued in the ~ 23
(a data point denotes a study or a cell t
Total number of partieipmts 1,901
Variable n• xt so' Range
s....,..-
Mean 23 49.34 59.06 12-417
Age
Mean 23 58.54 17.97 19.0-82.0
so 23 2.51 3.00 1.0-11.7
Edueation
Mean 17 14.02 0.74 12.8-16.0
so 14 2.09 1.08 0.~.6
IQ
Mean 2 113.11 1.41 112-114
so 2 12.56 0.70 12-13
APPENDIX 19M 873
Percent mtJk 15 38.86 38.69 ~100
Tat•core_,..
Combined mean 23 11.55 1.18 9.~13.4
Combined SD 23 2.36 0.39 1.6-3.0
Weighted means and SDs.
Predicted number of words recaDed and SDs, per age group• (Bey AVLT, Trial V)
95%CI
Age Predictecl I..Dwer Upper

Bt.ange Score Band Band
JO-J4 12.85 12.48 13.23

JS-J9 12.96 12.66 13.26
30-34 12.99 12.66 13.34
35-39 12.96 12.55 13.37
40-44 12.85 12.38 13.32
Stanclanl deviation for all age groups is 2.36.
45-49 12.66 12.17 13.16
50-S4 12.41 11.93 12.89
5S-S9 12.08 11.65 12.50
60-64 11.67 11.34 12.00
65-69 11.20 10.99 11.39
70-74 10.64 10.53 10.76
75-79 10.02 9.74 10.30

Predieted tat 8conl = 11.46148 + 0.0948657 • age- 0.0014639 • age2
Ordioary least-squares regressioo of test meaos oo age (quadratic)

R2 0.877
F(dl), p F<2.1l = 98.10, p < 0.000
Age 0.0948657 0.045 2.10° 0.074• -0.012 to 0.202

Age2 -0.0014639 0.000 -3.20 0.015 -0.003 to -0.000
Constant 11.46148 0.906 12.65 0.000 9.32 to 13.60
"Signi6cance test for age centered (sample means- aggregate mean): t = -8.27, p = 0.000.
Predietioa
Predicted age range ~79years
SEe 0.18
95%CI 11.76-12.46
(continued)
874 APPENDIX 19M
14
12
10
0
8
20 30 40 50 80 80
age
Rey AVLT trial V. The size of the bubbles reflects the weight of the data point, with larger bubbles indicating

Q(dfl• p Q(22)=421.02, p < 0.000
Tests for model 8t~ of a quadratic term
Linear 0.733 0.720 -24.482 -27.218

Quadratic 0.877 0.864 -39.110 -41.846
Tests for parameter speci8eatioas
Homoscedasticity
Signi8canee tests for regression with the SDs
A regression of SDs on age yielded an R2 of0.263 (F0 •7,=7.18, p=0.032).

Therefore, the SD for the aggregate sample is suggested for use with all age groups.
Educatioa
Est. tau2 without education 0
R2 0.875
Term Coefiicient SE t p 95%CI
Education 0.1582632 0.223 0.71 0.504 -0.387 to 0.703
Gender
t-test by gender
n Xmale (SD) Xfemale (SD) M-F difference t p

7M, 4F 1.2.747 (0.386) 11.820 (1.035) 0.927 2.189 0.028
Bey AVLT, RecaD after Interference
or
Number studies iDduded In the aoalysU 7
Yean of publicatioa 1988-2001
or
Number data points ued In the aoalysU 20
In education/gender-stratified data)
Total number or participaats 983
xt
s,.,. . .
Variable
Mean
n•
20 39.62
so'
25.69
Range
12-126
Age
Mean 20 58.61 17.99 19.0-82.0
SD 20 2.58 3.17 1.0-11.7
&luc:acion
Mean 14 13.92 0.59 12.8-14.8
SD 11 2.02 1.10 0.5-3.6
lQ
Mean 2 113.09 1.41 112-114
SD 2 12.55 0.70 12-13
Pereent male 12 35.34 37.21 0-100
Tatacore-
Combined mean 20 9.72 1.55 6.7-12.3
Combined SD 20 2.93 0.39 2.2-3.5
"Number of data points dJffers for dJfferent analyses due to missing data.
(continued)
876 APPENDIX 19M
Predicted number of words reealled and SDs, per age

group• (Bey AVLT, RecaD after Interference)
95%CI
Age Predieted Lower Upper

&nge Score Band Band
JO-U 11.88 11.65 12.12

JS-29 11.88 11.60 12.15 Standard dmation for aD age groups Is 1.83.
30-34 11.79 11.42 12.16
35-39 11.64 11.19 12.09
40-44 11.41 10.91 11.91
4S-49 11.11 10.59 11.62
50-54 10.73 10.24 11.22
55-59 10.28 9.85 10.71
60-64 9.76 9.42 10.09
6S-69 9.16 8.95 9.37
1~14 8.49 8.34 8.64
15-19 7.75 7.44 8.05

Predkled tat ICON= 11.01093 + 0.0718987 • age- 0.0014714 • age2
Sigoiflcance tests for regression with the test scores
Ordinary least-squares regression of test means on age (quadratic)

R2 0.923
Fed!). p Fe2.&)=309.74, p < 0.000
Term Coefficient SE p 959&CI
Age 0.0718987 0.040 1.79" 0.123" -0.026 to 0.170

Age2 -0.0014714 0.000 -3.48 0.013 -0.003 to -0.000
Constant 11.01093 0.715 15.40 0.000 9.26 to 12.76
"Significance test for age centered (sample means - aggregate mean): t = -10.15, p =0.000.
Predietion
SE.. 0.18
95%CI 10.13-10.84

Qed!). p Qe 19) = 451.56, p < 0.000
APPENDIX 19M 877
12
10 0
0
6
20 30 40 50 60 70 60
age
Rey AVLT Recall After Interference. The size of the bubbles reflects the weight of the data point, with larger
bubbles indicating larger standard error and smaller weight.
Model BIC BIC'
Linear 0.839 0.830 -17.296 -33.511

Quadratic 0.922 0.913 -28.934 -45.149
Tests for parameter speci8cations

Shapiro-Wilk W test W =0.902, p = 0.045
Homoscedasticity
A regression of SDs on age yielded an R2 of 0.111 (F(l.6 ) = 2.18, p =0.190). Therefore, the
SO for the aggregate sample is suggested for use with all age groups.
Education
R2 0.943
(continued)
878 APPENDIX 19M
Education 0.4297837 0.262 1.64 0.162 -0.243 to 1.103
Gender
t-test by gender
n X male (SD) X female (SD) M-F difference p
9M,8F 10.222 (0.535) 9.959 (0.591) 0.263 0.331 0.373
Table A19m.4. Results of the Meta-Analysis and Predicted Scores for

the Bey AVLT, Recognition
Number of studies included iD the analysis 4

Number of data points used iD the analysis 14
(a data point denotes a study or a ceU
Total number of participants 453
Sample me
Mean 14 22.22 21.86 10-126
Age
Mean 14 52.64 22.23 17.5-78.7
SD 14 1.64 0.79 0.~2.7
Education
Mean 10 13.95 0.40 13.3-14.6
SD 7 2.31 1.17 1-3.6
IQ
Mean 0
SD 0
Percmtmak 14 36.11 42.61 0-100
Test acat"e meant

Combined mean 14 13.19 0.81 12.3-14.4
Combined SD 14 1.65 0.56 0.5-2.4
APPENDIX 19M 879
Predicted number of words recogalzed and SDs, per

age group• (Bey AVLT, Recognition)t
95%CI

.Rattp SClore Band Band
20-!4 14.23 14.07 14.40

25-29 14.06 13.91 14.22
30-34 13.89 13.73 14.04
35-39 13.72 13.54 13.89 Standard deviation for all age groups is 1.65.
40-44 13.54 13.34 13.75
4lS-49 13.37 13.13 13.61
50-S4 13.10 12.92 13.48
SS-lJ9 13.03 12.70 13.35
(JQ...$4 12.85 12.49 13.22
65-69 12.68 12.27 13.10
1~14 12.51 12.05 12.97
15-19 11.34 11.82 12.85

Predkted tal acore = 15.00957- 0.0344756 • age
t'I'he predicted scores are relevant for the following administration
sequence: five acquisition trials, interference trial, recall after
interference, recognition (immediate or after a short delay).
Signiflcance tests for regression with the test scores
Ordiaary least-squares regression of test means on age (linear)

R2 0.892
F<dO• p Fo.3)=43.69, p<0.007
Age -0.0344756 0.005 -6.61 0.007 -0.051 to -0.018

Constant 15.00957 0.173 86.83 0.000 14.46 to 15.56
Predietion
SEe 0.15
95%CI 13.00-13.57
(continued)
880 APPENDIX 19M
20 30 40 50 60 70 60
age
Rey AVLT Recognition. The size of the bubbles reflects the weight of the data point, with larger bubbles
indicating larger standard error and smaller weight.

Q(d0• p Q( 13) = 120.53, p < 0.000
Tests for model &t-mlition of a quadratic term
Model BIC BIC'
Linear 0.892 0.883 -30.011 -28.546

Quadratic 0.896 0.877 -27.914 -26.448
Tests for parameter specl&cations
Shapiro-Wllk W test W = 0.871, p = 0.043
Homoscedasticity
A regression of SDs on age yielded an R2 of 0.471 (F(I.Jl = 9.71, p = 0.053). Therefore, the
SD for the aggregate sample is suggested for use with all age groups.
Education
APPENDIX 19M 881
R2 0.788
Education 0.2712587 0.265 1.02 0.414 -0.870 to 1.413
Gender
t-test by gender
n X male (SO) X female (SO) M-F difference p
5M,5F 14.100 (0.077) 13.152 (0.437) 0.948 2.136 0.033
Table A19m.S. Results of the Meta-Analysis and Predicted Scores for the
Bey AVLT, Total Recall, Trials 1-V
Number of studies included iD the analysis 6

Number of data points used iD the analysis 20
StJtrtpJ. me
Mean 20 50.37 66.28 12-417
Age
Mean 20 59.69 17.63 19.0--82.0
SO 20 1.59 0.78 1.0--0.7
Education
Mean 14 14.10 0.73 13.0-16.0
SO 11 1.94 1.16 0.5-3.6
lQ
Mean 0
SO 0
Percent male 13 40.33 38.53 0-100
Test meant
IIConl
Combined mean 20 47.47 5.51 36.2--56.3
Combined SO 12 8.85 1.06 7.4-10.3
Estimated so* 20 8.85 0.97 7.0-10.3
tweighted means and SOs.
*Estimated SOs for the total score means were estimated as follows:
2(SO trial I)+ 2(SO trial V) + 0.9

882 APPENDIX 19M
Table A19m.S. (Contd.)
Predicted number of words recalled and SDs, per age

group• (Bey AVLT, Total Recall)
95%CI
Age Predicted Lower Upper

&nge Score Band band
JQ.-.24 5US 53.87 55.23
J5-J9 55.03 54.36 55.70
30-34 55.15 54.23 56.06
Standard deviation for all age groups is 8.85.
35-39 54.91 53.77 56.06
40-44 54.33 53.04 55.6.2
45-49 53.38 52:06 54.70
50-S4 52.08 50;83 53.3.2
SS-S9 50.42 4lt37 51.47
6(}...$4 48.40 4i.64 49.17
65-69 46.03 45:55 46.5.2
70-74 43.31 42;67 43.94
75-79 40.9 38,99 41.45

Predicted tal acorw =47.98303+ 0.4520533 •age- 0.007125 •age2
Sigoiftcance tests for regression with the test scores

I
Onlioary least ~quares regression of left me11111 on age (quadratic)

R2 0.948
F<dO• p F<2.5> = 319.79, p <0.000
Age 0.45.20533 0.1.20 3.77• 0.013" 0.144 to 0.760

Age2 -0.007125 0.001 -5.58 0.003 -0.010 to -0.004
Constant 47.98303 2.209 .21.72 0.000 42.31 to 53.66
"Significance test for age centered (sample means-aggregate mean): t = -11.69, p = 0.000.
Prediction
Predicted age range 20-79 yellll
SEe 0.49
95%CI 49.70-51.60
APPENDIX 19M 883
80
50
~ ~----~r-----r-----.-----.-----,------r-
20 80 0 80
age
Fipre A19m.S. A scatteiplot illustrating the dispenion of the data points around the regression line for the
Rey AVLT Total RecaU. The size of the bubbles reflects the weight of the data point, with larger bubbles
indicating larger standard error and smaller weight.
Tests for assumptioos and model 8.t
Tests for heterogeneity iD the 8Dal dataset

Q(df)• p Q(l9) = 611.27, p <0.000
Tests for model &t--tUidltion of a quadratic term
Model BIC BIC'
Linear 0.785 0.773 39.348 -27.744

Quadratic 0.948 0.942 13.910 -53.182
Tests for parameter speei6cations

Shapiro-Wilk W test W = 0.983, p = 0.969
Homoscedasticity
SigDi&eanee tests for regression with the SDs
A regression of SDs on age yielded an a2 ofO.OlO (Fu.s)=0.17, p=0.699). Therefore,
(continued)
884 APPENDIX 19M
Education
R2 0.957
Education -0.8448568 0.546 .-1.55 0.197 -2.361 to 0.671
Gender
t-test by gender
n X male (SD) X female (SO) M-F difference p
11M, 9F 50.117 (1.63) 49.578 (1.520) 0.539 0.238 0.408
Table A19m.6. Summary Table of Predi~ed Scores for the Bey AVLT
~after Total
Age nJnge Trial I Trial V Irf;erference Recognition• Recallt Leamingt Forgetting'
20-!U 7.10 12.85 ; 11.88 14.23 54.55 6.00 0.99

JS-.29 7.10 12.96 11.88 14.06 55.03 1.11
30-34 7.08 12.99 11.79 13.89 55.15 1.23
35-39 7.01 12.96 11.64 13.72 54.92 1.36
40-44 6.90 12.85 11.41 13.54 54.33 1.48
41$-49 6.76 12.66 11.11 13.37 53.38 5.60 1.60
50-S4 6.58 12.41 10.73 13.20 52.08 1.72
5S-S9 6.36 12.08 10.28 13.03 50.42 1.84
(JQ...$4 6.10 11.67 9.76 12.85 48.40 1.96
6lHi9 5.81 11.20 9.16 12.68 46.03 2.08
70-74 5.47 10.64 8.49 12.51 43.31 2.21
75-19 5.10 10.02 7.75 12.34 40.22 4.60 2.33
SD 1.71 2.36 2.93 1.65 8.85
Weighted means 6.31 11.89 10.21 13.20 49.58 1.72
(0.82) (1.24) (1.68) (0.67) (6.21) (0.47)
"The predicted scores are relevant for the following administration sequence: five acquisition trials, interference trial,
recall after interference, and recognition (immediate or after a short delay).
tTrials 1-V.
*Learning= trial V -trial I. Data in the aggregate sample were inconsistent and did not allow predictions. Original data
corresponding to 3 points along the age continuiun are presented in order to demonstrate the slope of the age-related
decline in learning capacity.
'Forgetting= Trial V- Recall after Interference~ Based on a linear regression (R2 = 0. 743, Fu.s> = 52.38, p < 0.0004).
for the Benton Visual Retention Test
885
Table A20.1. Locator Table for the Benton Visual Retention Test (BVRT) I0\
Study Age• n Sample Composition IQ/Education• FonnlAdministration Location
BVRT.l Benton- 15-69 600 Hospital inpatients IQ: FormC,D,E Iowa

Sivan,1992 769 or outpatients; data are ~110 to ~59 Adm. A,B,C,D
page 402 120 partitioned by age groups and 8-item
Data are not and IQ levels; expected abbreviated
reproduced number of correct responses version
in this book and errors are presented
(some means and SDs)
BVRT.2 Benton, 41 100 Patients on medical and Education: FormC,E Iowa
1962 16-60 neurological wards with 10 Adm.C,B
page403 no evidence of cerebral
Table A20.2 disease; number of errors
reported
BVRT.3 Klonoff 80-92 172 War veterans; data presented 7.04 Forme Canada
& Kennedy, 1965 in 6 age groupings; number Adm.A
page 403 correct and errors reported
Table A20.3
BVRT.4 Arenberg. 18-102 857 Male volunteers, mostly Caucasian; Well-educated Forme Baltimore/
1978 sample is divided by 7 age groups sample Adm.A Washington DC
page 404 and date of testing; number of
Table A20.4 errors reported
BVRT.5 Eslinger 73.1 53 Healthy volunteers, 25 M, 28 F; 12.0 Forme Iowa
et al., 1985 60-88 total correct and errors Adm.A
page404 reported
Table A20.5
BVRT.6 Larrabee 72.9 78 Healthy volunteers; 17M, 61 F; 12.2 Forme Florida
et al., 1986 (6.9) number of errors reported (3.3) Adm.A
page404
TableA20.6
BVRT.7 Randall Mean for 120 Volunteers; 51 M, 69 F; IQ: Forme Mississippi >
et al., 1988 IQ groups: data partitioned into 60 to ;:;20 Adm. A,C "tl
"tl
page 405 23.87- 6 IQ groups; total FormD m
TableA20.7 26.20 correct and errors reported Adm.D z
c
><
N
0
BVRT.8 Robertson- 20-29 1,643 Volunteers; data partitioned into gender, <College degree. Forme Baltimore/ DC >
"tt
Tchabo & Arenberg. 30-39 2 education groups, and 7 College degree Adm.A Washington "tt
m
1989 40-49 age groups; total errors z
page405 50-59 reported 0
Table A20.8 60-69 ><
70-79 IV
80-89 0
BVRT.9 Alder 25-34 277 Male volunteers; sample is Highly Forme Baltimore/
et al., 1990 35-44 divided into 5 age groups; educated Adm.A Washington DC
page 406 45-54 number of errors reported
Table A20.9 55-&4
65+
BVRT.IO Prakash 15-19 90 331 M, 329 F; exclusion Higher Forme,D,E India
& Bhogle, 1992 20-24 86 criteria ..evident physical secondary (collapsed)
page 406 25-29 84 or psychological disorders;" education Adm.A
Table A20.10 30-34 56 sample is stratified into
35-39 53 10 age groups; number
40-44 62 correct reported
45-49 73
50-54 55
55-59 62
60-64 39
BVRT.ll Robinson- 72.23 122 Healthy elderly; number 13.61 Forme Missouri
Whelen, 1992 (9.0) correct, errors, and error (3.4) Adm.A
page 407 types FormD
Table A20.11 Adm.C
BVRT.l2 Lee & 34.7 81 Male manual workers, guards, 12.9 NCTB Korea
Lee, 1993 (8.16) clerks, and technicians; (2.5) battery,
page 407 number correct for Recognition
Table A20.12 Recognition trial reported trial
BVRT.l3 Youngjohn 11h39 1,128 Healthy volunteers; 464 M, 12-14 Adm.A USA
et al., 1993 40-49 664 F; data are stratified 15-17
page 407 50-59 into 5 age groups by 18-25
Table A20.13 60-69 3 education groups;
70-84 number correct and errors
reported
(continued)
I.....
Table A20.1. (Contd.) ICD
Study Age• n Sample Composition IQ/Education • FonnlAdministration Location
BVRT.l4 Palmer 71.4 1,149 Community-dwelling older WeD- Form C Western

et al., 1994 (4.69) males; number correct educated Adm. A USA
page 408 reported
Table A20.14
BVRT.l5 Escalona 30 67 Control participants; 56 M 8 NCTB Venezuela
et al., 1995 16--45 11 F; number correct for battery,
page 408 Recognition trial reported Recognition
Table A20.15 trial
BVRT.l6 Giambra 28--33 1,721 Volunteers; 1,163 M, 558 F; Highly Form C Baltimore/
et al., 1995 34-39 mostly Caucasian; sample educated Adm.A Washington DC
page 408 40--45 is partitioned into gender x
Table A20.16 46-51 10 age groups; number of
52-57 errors reported
58-63
64--69
70-75
76-81
82--87
BVRT.l7 Resnick 20-29 2,000 Volunteers; 1,365 M, 635 F; 1-IigWy Form C Baltimore/
et al ., 1995 30-39 mostly Caucasian; sample educated Adm. A Washington DC
page 409 40-49 is partitioned into gender x
Table A20.17 50---59 7 age groups; data for
60-69 7 types of e rrors reported
70--79
80--102
BVRT.l8 Steenhuis 78.5 591 Elderly participants with no 9.8 Multiple- Canada
& Ostbye, 1995 (6.7) cognitive impairment; (4.0) choice
page 409 61 % F; number correct administration
Table A20.18 reported
)>
BVRT.l9 Dealberto 60--64 1,389 Elderly participants; 574 M, <6 Recognition France "'0
et al., 1996 65-70 815 F; no exclusion criteria; 6-12 trial "'0
m
page 410 sample partitioned by gender, > 12 z
Table A20.19 2 age groups, and
3 educational levels; numbe r
-0X
0"'
correct reported
BVRT.!IO Jacobs 75.07 118 Older English speakers and 8.85 Recognition New York >
""0
et al., 1997 (5.90) Spanish speakers; 75% and (3.78) for matching ""0
m
page 410 74.91 118 72% F, respectively; 8.41 form C; z
Table A20.20 (5.71) cognitive impairment (3.98) Memory c
or dementia among exclusion form D ><
criteria; number correct 1\J
0
reported
BVRT..11 Carmelli 5~ 589 White male WWII veterans; Education: Version Massachusetts,
et al., 1999 no exclusion criteria; ~12 not Indiana,
page411 sample partitioned by (n=341) specified California
Table A20.21 4 smoking groups and >12
4 alcohol intake groups; (n=254)
number correct reported
BVRT.U Coman 77.7 156 Healthy participants; mostly 12.67 Adm.A Iowa
et al., 1999 (7.89) Caucasian; 31M, 125 F; (3.46)
page 412 sample is partitioned into
Table A20.22 9 age groups x 11 educational
levels; expected number
correct reported
BVRT..I3 Manly 76.2 43 Nondemented participants in 0-3 Recognition New York
et al., 1999 (6.1) the Columbia Aging Project trial,
page 412 74.8 43 (>65 years old); Medicare recipients; matching
Table A21.23 (5.7) 74% F; sample is stratified into
21iteracy groups; number
correct reported
BVRT.lW Mathiesen 45.5 52 Male control participants; 9.5 Forme N01way
et al., 1999 (10.8) number correct and errors (1.8) Adm.A
page 412 reported
Table A20.24
BVRT.25 Amir, 15-44 260 124M, 136 F; ll.6M FormD United
2001 no exclusion criteria; (2.71) Adm.A Arab
page 413 sample partitioned into FormE Emirates
Table A20.25 4 IQ groups by gender 12.0 F Adm.A
and 2 education groups (2.80)
(continued)
co
~
!

Study Age• n Sample Composition IQ/Education• Form/Administration Location
BVRT.26 Touradji 75.7 106 Nondemented participants in 12.9 Recognition New York
et al., 2001 (7.2) the Columbia Aging Project (3.5) trial,
page 413 77.9 87 (>65 years old); Medicare recipients; 12.0 matching
Table A21.26 (7.3) sample is stratified into U.S.-bom (3.7)
vs. foreign-hom groups; number
correct reported
BVRT.27 Coman 55-64 156 Same as BVRT.22; 12.67 Adm.A Iowa
et al., 2002 65-74 number correct reported (3.46)
page 414 75-84
Table A20.27 85+
BVRT.28 Manly 73.9 192 Nondemented participants in 12.8 Recognition New York
et al., 2002 (5.8) the Columbia Aging Project (2.8) trial,
page 414 74.6 192 (>65 years old); Medicare recipients; 13.0 matching
Table A21.28 (5.9) 68% F; data are presented for (3.0)
African-American and white
participants separately; number
correct reported
BVRT.29 Farabat 42.28 50 Male control participants; 2::12 Version Egypt
et al., 2003 (5.54) number correct reported not
page 414 specified
Table A20.29
BVRT.30 Kawas <50 1,425 Volunteers; 1,004 M, 421 F; Highly Forme Baltimore/
>
"'0
"'0
et al., 2003 50-59 sample is partitioned into educated Adm.A Washington DC m
page 415 60-69 6 age groupings; number z
Table A20.30 70-79 of errors reported 0
80-89 X
N
90+ 0
BVRT.31 Reinprecht 81 141 Male participants born in 1914; Adm.A Sweden >
""0
et al., 2003 sample is partitioned into ""0
m
page 415 3 groups based on z
Table A20.31 hypertension status; 0
number correct reported X
BVRT.32 Ruggieri 30.08 50 Control participants; 24M, 26 F; 11.76 Version Italy N
0
et al., 2003 (8.37) number correct reported (3.04) not
page 416 specified
Table A20.32
BVRT.33 Witjes-Ane 42 88 Volunteers; 40 M, 48 F; Mostly Version Netherlands
2003 18-64 number correct reported ~high not
page 416 school specified
Table A20.33
• Age column and IQ/education column contain information regarding range and!or mean and standard deviation for the whole sample and/or separate groups, whichever is provided by
the authors.
...co
\D
892 APPENDIX 20
Table A20.2. [BVRT.2] Benton, 1962: D,lta for 100

Inpatients (mean age=41, mean years of edu-
cation= 10) in Iowa with Medical or N•urological
Disorders but without Cerebral Disease;or Injury
BVRT Errors
Copy (Administration C, 0.8

Form Cor E)
Memory (Administration B, 6.6
Form Cor E)
Table A20.3. [BVRT.3] Klonoff and Kennedy, 1965: Data for

Community-Dwelling Canadian VeteraJI; Tested in 1963--1964
BVRT"
Age Mean
n Range Education Correct Errors
172 80-92 7.04 3.97 11.71

(1.70) (4.64)
·range 0-8 range 1-26
•Mean number correct and errors for administration

r
A, form C.
Table A20.4. [BVRT.4] Arenberg, 1978: Data• for 857 Male, Mostly Caucasian, Well-
Educated, High Socioeconomic Status 1\esidents of the Baltimore-Washington DC Area:
Sample Is Divided According to Date of Testing
Age Group Tested 1960-1964 T~ 1965-1968 (1st half) Tested 1968-1973 (2nd half)
<30 1.25 (n=8) 2.75 (n=12) 2.55 (n=42)

30s 2.61 (n=67) 2.73 (n= 15) 3.13 (n=61
40s 2.88 (n=98) 3.15 (n=40) 3.76 (n=41)
50s 3.50 (n = 100) 4.51 (n=37) 4.48 (n=48)
60s 4.58 (n=66) 5.09 (n=35) 5.31 (n=39)
70s 6.33 (n=SS) 6.05 (n=20) 6.64 (n=50)
~80 11.75 (n=8) 12.00 (n=3) 8.33 (n=12)
•Mean number of errors for administration A, form C.
Table A20.5. [BVRT.S] Eslinger et al.. 1985: Data for Normal

Elderly Volunteers Recruited in Iowa
BVRT"
n Age Education MIFI\af.o Correct Errors
53 73.1 12.0 25128: 5.6 7.4

60-88 ( 1.6) (3.3)
-Total correct and total errors for administtatiOJA, form C.

APPENDIX 20 893
Table A20.6. [BVRT.6] Larrabee et al., 1986: Data on Healthy

Elderly Participants (73 Caucasian, 5 African American) Recruited in
Florida
BVRT
n Age Education MIF Ratio VIQ PIQ Errors•
78 72.9 12.2 17/61 112 114 7.8

(6.9) (3.3) (12.8) (11.5) (3.2)
"Number of errors for administration A, form C.
Table A20.7. [BVRT.7] Randall et al., 1988: Data• for 120 Participants in Mississippi (69 Females,
=
51 Males) in six IQ Ranges (n 20 per Group) with Mean Ages across Groups of 23.87-26.20
Administration A Administration D Administration C
IQ Group Correct Errors Correct Errors Correct Errors
60-69 2.50 15.50 1.80 19.10 6.10 5.55

(1.64) (5.05) (1.32) (6.62) (3.03) (5.63)
70-79 4.30 10.15 4.50 9.30 8.00 2.90
(1.26) (2.00) (2.21) (6.15) (2.20) (3.45)
80-89 6.50 5.40 6.50 4.80 8.80 1.90
(1.90) (2.99) (2.01) (3.16) (1.32) (1.92)
90-109 8.30 2.10 8.75 1.35 9.85 0.15
(1.34) (1.83) (0.85) (0.93) (0.37) (0.36)
110-119 8.05 2.50 8.50 1.60 9.75 0.25
(1.19) (1.79) (1.10) (1.31) (0.44) (0.44)
120+ 8.25 2.00 8.40 1.90 9.70 0.30
(0.79) (1.07) (0.94) (1.21) (0.47) (0.47)
"Number correct and errors for administrations A (form C), D (form D), and C (form C).
Table A20.8. [BVRT.8] Robertson-Tchabo and Arenberg, 1989: Data• for 1,643 Participants in Baltimore-
Washington DC Divided by Gender, Education, and Seven Age Groupings
Age
20-29 30-39 40-49 50-59 60-69 70-79 80-89
Males, no college degree 2.02 3.56 3.75 5.07 6.25 7.72 11.13
(1.53) (3.04) (2.46) (2.79) (2.41) (3.92) (4.29)
n=45 n=57 n=51 n=42 n=28 n=47 n=15
Males, college degree 2.47 2.49 2.94 3.67 4.66 6.03 8.15
(1.87) (1.87) (1.89) (2.57) (2.81) (3.20) (4.25)
n=92 n=181 n=155 n=156 n=123 n=139 n=40
Females, no college degree 2.90 3.04 3.87 5.11 6.02 7.53 9.44
(2.08) (2.21) (1.82) (1.95) (2.58) (3.37) (3.84)
n=19 n=27 n=23 n=35 n=42 n=32 n=9
Females, college degree 2.70 2.67 2.61 3.68 4.57 6.79 7.62
(1.82) (2.11) (2.17) (1.99) (2.17) (2.97) (4.43)
n=37 n=70 n=28 n=38 n=56 n=43 n=13
"Number errors for administration A (form C).

894 APPENDIX 20
Table A20.9. [BVRT.9] Alder et al., 1990: Data for Table A20.10. [BVRT.10] Prakash and Bho-
271 Males in Baltimore-Washington DC J'resented gle, 1992: Data for a Sample of 660 Indian
in Five Age Groupings ' Participants (331 Male, 329 Female) with
Higher Secondary Education, Divided into
BVRT WAIS Vocabulary
10 Age Groupings
Age n Errors• Ra" Score
Age BVRT
25-34 27 2.59 ~.26
Grouping n Correct•
(2.36) (9.42)
35-44 74 2.77 115.68 15-19 90 7.94
(2.06) (10.48) (1.32)
45-54 101 3.36 67.01 20-.24 86 7.48
(2.47) ~.89) (1.88)
55-64 42 4.19 ~.62 25-29 84 7.77
(2.96) (\)..05) (1.69)
65+ 33 5.12 ~.91 30-34 56 7.88
(3.32) ,.96) (1.72)
35-39 53 7.70
•Mean errors for administration A, form C. (1.77)
40-44 62 7.65
(1.49)
45-49 73 7.73
(1.67)
S0-54 55 6.93
(2.09)
55-59 62 7.34
(1.70)
60-64 39 6.36
(1.80)
•Mean correct for administration A, Forms C, D,

and E (apparently collapsed).
Table A20.11. [BVRT.ll] Robinson-~len, 1992: Data• for 122 Caucasian Participants Recruited in
Missouri with a Mean Age of 72.23 (9.0) and Mean Education of 13.61 (3.4)
Total Right Left
Form Correct Errors Omission Distortion Perseveration Rotation Misplacement Size Side Side
c 5.55 7.38 0.86 2.87 1.07 1.48 0.80 0.30 4.08 3.00
(1.69) (3.67) (1.36) (1.990 (1..21) (1.34) (1.05) (0.79) (2.27) (1.72)
D 9.38 0.65 0.05 0.24 0.01 0.06 0.18 0.12 0.24 0.31
(1.20) (1.31) (0..22) (0., (0.09) (0..23) (0.44) (0.51) (0.53) (0.86)
•Number correct, number of errors, and error ~ for administration A, form C followed by administration C, form D.
APPENDIX 20 895
Table A20.12. [BVRT.12] Lee and Lee, 1993:

Data for Male Korean Manual Workers, Guards,
Clerks, and Technicians
BVRT
n Age Education Correct•
81 34.7 12.9 8.1

(8.6) (2.5) (1.5)
"Mean correct for the Recognition trial.
Table A20.13. [BVRT.13] Youngjohn et al., 1993: Data• for a Sample of 1,128 Volunteers (464 Male, 664
Female), Divided into Five Age Groups and Three Educational Levels
Age Groupings
18-39 40-49 50-59 60-69 70-84
Education Correct Errors Correct Errors Correct Errors Correct Errors Correct Errors
12-14 7.59 3.38 7.11 4.22 6.66 4.90 6.18 5.55 5.62 7.28
(1.52) (2.37) (1.53) (2.62) (1.47) (2.42) (1.67) (2. 74) (1.73) (3.55)
n=29 n=18 n=130 n=129 n=53
15-17 8.04 2.52 7.78 3.48 7.08 4.21 6.70 4.99 6.06 6.74
(1.19) (1.70) (1.54) (2.78) (1.70) (2.85) (1.47) (2. 78) (1.84) (4.34)
n=27 n=23 n=146 n=159 n=54
18-25 8.11 2.67 7.42 3.74 7.55 3.64 6.80 4.93 6.22 6.33
(1.28) (1.78) (1.22) (2.47) (1.53) (2.76) (1.55) (2.87) (1.57) (3.63)
n=18 n=19 n=133 n=134 n=49
"Mean correct and errors for administration A (form not specified).
Table A20.14. [BVRT.14] Palmer et al., Table A20.15. [BVRT.15] Escalona et al., 1995:
1994: Data for Primarily Caucasian, Well- Data for Venezuelan Control Participants
Educated Males in High-Level Occupations,
who Were Tested between 1986 and 1989 BVRT
n Age M/F Ratio Education Correct•
BVRT
n Age Correct• 67 30 56111 8 6.2
(16-45) (2)
1,149 71.4 6.61
(4.69) (1.51) "Mean correct for Recognition trial.
"Mean correct for administration A, form C.

896 APPENDIX 20
Table A20.16. [BVRT.16] Giambra et ~·· 1995: Data for a

Sample of 1,721 Mostly Caucasian, Wtll-Educated Partici-
pants (1,163 Males, 558 Females), DividED into Gender x Age
Groupings
BVRT Errors•
Age Grouping Men Women
28-33 2.22
(1.93)
34-39 2.56 2.96
(1.94) (2.10)
40-45 3.05 3.11
(2.56) (2.28)
46-51 2.89 2.43
(2.44) (2.23)
52-57 3.24 3.48
(2.38) (2.02)
58-63 3.76 4.47
(2.61) (2.80)
64-69 4.52 5.39
(2.82) (3.31)
7~75 5.93 6.59
(2.96) (3.05)
76-81 7.55 7.50
(3.62) (3.50)
82-87 8.49
(4.36)
"Mean errors for administration A, form C.
Table A20.17. [BVRT.17] Resnick et lllt• 1995: Data for a Sample of 2,000, Mostly Caucasian, Well-
Educated Participants (1,365 Men, 6351 Women), Recruited in the Baltimore-Washington DC Area,
Divided into Seven Age Groups x Gendet
Error Type•
Age Group/Gender n Omission Addfion Distortion Perseveration Rotation Misplacement Size
~29
Men 166 0.02 0.00 1.07 0.39 0.40 0.38 0.05
(0.1) (0.9) (1.0) (0.7) (0.7) (0.7) (0.2)
I
Women 102 0.04 O.fH 1.00 0.43 0.38 0.51 0.05
(0.2) (0.~) (1.2) (0.7) (0.6) (0.8) (0.3)
I
~9
Men 243 0.01 0.~ 1.07 0.53 0.53 0.47 0.06

(0.1) (0.,) (1.1) (0.8) (0.8) (0.9) (0.3)
Women 110 0.09 O.Oo 1.26 0.43 0.65 0.39 0.02
(0.4) (O.f) (1.2) (0.8) (0.9) (0.6) (0.1)
t
APPENDIX 20 897

Error Type•
Age Group/Gender n Omission Addition Distortion Perseveration Rotation Misplacement Size
40-49
Men 214 0.07 0.00 1.46 0.56 0.50 0.48 0.06
(0.4) (0.0) (1.3) (0.7) (0.7) (0.7) (0.3)
Women 56 0.09 0.00 1.54 0.46 0.46 0.36 0.04
(0.3) (0.0) (1.3) (0.7) (0.7) (0.7) (0.2)
50-59
Men 211 0.13 0.01 1.73 0.64 0.64 0.63 0.08
(0.5) (0.1) (1.6) (0.8) (0.8) (0.9) (0.3)
Women 95 0.33 0.00 1.98 0.71 0.76 0.52 0.06
(0.8) (0.0) (1.3) (0.9) (0.9) (0.7) (0.3)
60-69
Men 222 0.21 0.01 2.22 0.79 0.73 0.70 0.14
(0.6) (0.1) (1.6) (1.0) (0.9) (0.9) (0.4)
Women 112 0.54 0.00 2.08 0.90 1.01 0.56 0.12
(1.0) (0.0) (1.4) (1.1) (0.9) (0.8) (0.4)
70-79
Men 226 0.54 0.02 2.72 1.14 0.91 0.79 0.19
(1.0) (0.1) (1.8) (1.2) (1.1) (1.0) (0.5)
Women 114 0.58 0.00 2.57 1.28 1.21 0.88 0.16
(1.5) (0.0) (1.7) (1.2) (1.1) (1.0) (0.5)
80-102
Men 83 0.82 0.05 3.81 1.00 1.30 1.05 0.39
(1.4) (0.3) (2.4) (1.2) (1.0) (1.1) (0.8)
Women 46 1.09 0.11 3.46 1.33 1.28 1.00 0.39
(1.8) (0.3) (1.9) (1.3) (1.0) (1.2) (0.8)
•Means for seven types of error from administration A, form C.
Table A20.18. [BVRT.l8] Steenhuis and Ostbye, 1995: Data for

Canadian Participants with no Cognitive Impainnent
BVRT
n Age Education %Female Correct•
591 78.5 9.8 61 11.84

(6.7) (4.0) (2.34)
•Mean correct for multiple-choice administration.
Table A20.19. [BVRT.l9] Dealberto et al., 1996: Data for a Sample o£1,389 French Participants (574 Male,
815 Female), Aged 60--70 Years, Partitioned by Gender, Two Age Groups, and Three Educational Levels:
No Exclusion Criteria Were Used
Gender Age Education (Years)
Male Female 60-64 65-70 <6 ~12 >12
Correct• 11.7 11.4 11.6 11.4 11.1 11.8 12.3

(2.0) (2.0) (1.9) (2.0) (2.0) (1.9) (1.6)
"Mean for number correct (apparently for Recognition trial).

898 APPENDIX 20
Table A20.20. [BVRT.20] Jacobs et al., 1997: Data• for 118 English Speakers and 118 Spanish
Speakers Recruited in New York
Age Education Gender Matching Memory
English 75.07 8.85 75% female 8.30 6.79

(5.90) (3.78) (1.60) (1.95)
Spanish 74.91 8.41 72% female 7.58 5.74
(5.71) (3.98) (1.93) (1.99)
"A recognition trial for visual perceptual assessment (matching, form C) and for visual memory (form D)
was administered. Means are for total correct.
Table A20.21. [BVRT.21] Cannelli et al., 1999: Data for 589 Caucasian Male Veterans Aged 59-69 Years,
Tested in 1985-1986, who Were Recruited from Massachusetts, Indiana, and California: Sample Is DMded
into Four Smoking Groups and Four Alcohol Intake Groups
Smoking Status Daily Intake of Alcohol
Never Former Former Current
Smoked (Quit <::10 Years) (Quit <10 Years) Smoker No Drinlcs ~1 1-3 <::3
n 199 222 72 102 158 204 150 83

BVRT correct• 6.7 6.5 6.2 6.3 6.1 6.6 6.4 6.5
"Mean for correct (test version not specified).
Table A20.22. [BVRT.22] Coman et al., 1999: Data• for 156 Primarily Caucasian Participants (31 Male,
125 Female) Recruited in Iowa, who Averaged 77.7 (7.89) Years of Age, with a Range of 61-97, and
12.67 (3.46) Years of Education, with a Range of 4-20: Sample Is Partitioned into Nine Age Groups by
11 Educational Levelst
Age
Years of Education 55 60 65 70 75 80 85 90 95
8 6.54 6.09 5.65 5.21 4.77 4.32 3.88 3.44 3.00

9 6.72 6.28 5.83 5.39 4.95 4.50 4.06 3.61 3.17
10 6.90 6.46 6.01 5.57 5.12 4.68 4.23 3.79 3.34
11 7.09 6.64 6.20 5.75 5.30 4.86 4.41 3.96 3.52
12 7.27 6.82 6.38 5.93 5.48 5.03 4.59 4.14 3.69
13 7.46 7.01 6.56 6.11 5.66 5.21 4.76 4.31 3.87
14 7.64 7.19 6.74 6.29 5.84 5.39 4.94 4.49 4.04
15 7.82 7.37 6.92 6.47 6.02 5.57 5.12 4.67 4.21
16 8.01 7.56 7.10 6.65 6.20 5.75 5.29 4.84 4.39
18 8.38 7.92 7.47 7.01 6.56 6.10 5.65 5.19 4.74
20 8.74 8.29 7.83 7.37 6.91 6.46 6.00 5.54 5.08
"Expected number correct scores.

tMean correct for administration A (form not specified) was 5.37 (1.92), with a range of0-10.
APPENDIX 20 899
Table A20.23. [BVRT.23] Manly et al., 1999: Data for Older

Participants (74% Female) with 0-3 Years of Education, Recruited
in New York
BVRT Correct
%Spanish Recognition
n Age speakers Trial Matching
Literate 43 76.2 72 5.12 6.58

(6.1) (2.35) (2.16)
Illiterate 43 74.8 86 3.75 5.35
(5.7) (1.75) (2.25)
Table A20.24. [BVRT.24] Mathiesen et al., 1999: Data for Male

Participants, Younger than 65 Years, Recruited in Norway
BVRT"
W AIS-R Vocabulary
n Age Education Scaled Score Correct Errors
52 45.5 9.5 8.7 7.5 4.0

(10.8) (1.8) (1.2) (1.4) (2.3)
"Means for number correct and number of errors for administration A, form C.
Table A20.25. [BVRT.25] Amir, 2001: Data• for 260 Participants Recruited in United Arab Emirates
FormD FormE
n Age Education Correct Errors Correct Errors
Gender
Males 124 21.7 11.6 7.65 2.48 7.57 2.80
(5.89) (2.71) (1.97) (2.58) (1.78) (2.46)
Females 136 21.6 12.0 7.33 3.33 7.91 2.31
(4.66) (2.80) (1.87) (3.15) (1.64) (2.13)
IQ'-l
Superior 17 8.53 1.18 8.88 1.06
(2.26) (1.59) (0.99) (1.08)
Above average 51 8.20 2.51 8.12 2.06
(1.20) (3.19) (1.32) (1.56)
Average 175 7.41 3.05 7.75 2.48
(1.73) (2.55) (1.62) (2.00)
Below average 17 5.06 7.24 5.53 6.12
(1.74) (2.84) (2.34) (4.09)
Eductrtion
Low 48 6.98 3.58 7.58 2.73
(~9 years) (2.17) (3.15) (1.67) (2.24)
High 44 7.98 2.48 8.09 2.20
(<::university) (1.42) (2.13) (1.61) (2.12)
"Mean number correct and errors for administration A (form D followed by formE 2 weeks later).
900 APPENDIX 20
Table A20.26. [BVRT.26] Touradji et al, 2001: Data for Older Non-Hispanic, Fluent
English-speaking, White Participants (74111 female) Recruited in New York
BVRTCorrect
n Age Edueation Recognition Trial Matching Trial
U.S.-bom 106 75.7

(7.2) ~I:) 7.67
(1.73)
9.07
(1.26)
Foreign-hom 87 77.9
(7.3)
Ito
(7)
7.81
(1.58)
8.71
(1.52)
Table A20.27. [BVRT.27] Coman et al., ~:Data for a Sample of

156 Primarily Caucasian Participants, Rectwted in Iowa, Partitioned
by Four Age Groups !
Age Croup n BVRT Correct"
55-64 6 6.83
(1.17)
65-74 54 6.30
(1.59)
75-84 67 4.90
(1.75)
85+ 29 4.45
(2.13)
"Mean correct for administration A (form not specified).
Table A20.28. [BVRT.28] Manly et al., ~2002: Data for Older Participants (689& Female)
Recruited in New York
Benton Correct
WRAT-3 Recognition Matching

Ethnicity n Age E4ucation Reading" Trial Trial
African American 192 73.9 12.8 44.2 7.4 8.9

(5.8) ·(2.8) (7.2) (1.8) (1.4)
White 192 74.6 !13.0 49.3 8.1 9.4
(5.9) :(3.0) (4.1) (1.5) (1.2)
•WRAT, Wide Range Achievement Test.

APPENDIX 20 901
Table A20.29. [BVRT.29] Farahat et al., 2003:

Data for Male Control Participants Recruited in
Egypt: 42 with Secondary Education and Eight
with University Degrees
n Age BVRT Correct•
50 42.28 5.48
(5.54)
"Mean for number correct (administration and form not

specified).
Table A20.30. [BVRT.30] Kawas et al., 2003: Data for a Sample of

1,425 Participants (1,004 Men, 421 Women), 72% of whom had
College Degrees, Recruited in Baltimore-Washington DC, Parti-
tioned by Six Age Groups
BVRT WAIS Vocabulary
Age Group n Errors• n Raw Score
<50 298 3.11 300 64.08

(2.28) (10.18)
00-59 546 3.72 546 64.97
(2.60) (9.54)
60-69 815 4.67 681 65.00
(3.05) (10.04)
70--79 760 6.78 608 65.61
(4.01) (8.80)
80--89 380 9.09 267 63.26
(4.50) (10.83)
90+ 40 9.73 19 58.00
(5.72) (15.39)
•Mean for total number of errors for administration A (presumably form C).
Table A20.31. [BVRT.31] Reinprecht et al., 2003: Data for a Sample

of 14181-Year-Old Men Recruited in Sweden, Tested in 1995-1996,
Classified into Three Groups According to Level of Hypertension
No Hypertension Hypertension Hypertension
at Age 68 and 81 at 81 but not 68 at 68 and 81
n 22 11 108
BVRT
correct• 4.7 4.5 4.3
(1.6) (1.9) (1.6)
•Mean for number correct for administration A (form not specified).

I
~2 i APPENDIX 20
!
Table A20.32. [BVRT.32] Ruggieri et ~-. 2003: Data for Italian
Control Participants ~
i BVRT
n Age Education IM/F Ratio Correct•
50 30.08 11.76 24126 7.76

(8.37) (3.04) (1.89)
"Mean correct (test administration and form not specified).
Table A20.33. [BVRT.33] Witjes-Ane et rj., 2003: Data for a Sample

of Control Participants Recruited in the Netherlands, Tested in
1993-1998
n Age BVRT Correct•
88 42 40148 I 7.5
18-64 (1.6)
"Mean for number oorrect (administration and form not specified).

Appendix 21 : Locator and Data Tables
for the Finger Tapping Test (FTT)
Table A21.1. Locator Table for the Finger Tapping Test (FIT)
IQ/
Study Age• n Sample Composition Education" Location
Fl'.l Vega & Parsons, 40.8 50 Control group of patients 11.1 Oklahoma
1967 (13.1) with no CNS dysfunction; (3.2)
page 424 37M, 13 F
Table A21.2
Fl'.! Goldstein & 156 Control group of general Topeka,
Shelly, 1972 medical and psychiatric KS
page 424 V.A. patients, mostly males
Table A21.3
Fl'.3 Goldstein & 20-29 29 Sample of neurologically
Braun, 1974 30-39 75 intact participants, divided
page 425 40-49 47 into 6 age groups
Table A21.4 50-59 30
60-69 16
70-79 12
Fl'.4 Finlayson 34.12 17 Control group of healthy adults Grade school
et al., 1977 (8.72) and hospitalized medical and
page425 34.37 17 psychiatric patients, all males; High school
Table A21.5 (7.36) data are presented in T scores
35.22 17 University
(7.78)
(continued)
903
904 APPENDIX 21

IQ/
Study Age• n Sample Composition Education• Location
Fl'.5 Wiens & 23.6 24 Nonnal young men divided into FSIQ: Oregon
Matarazzo, 1977 24.8 24 2 groups 117.5
page426 118.3
TableA21.6
Fl'.6 Dodrill, 27.3 50 Healthy control group; 12.0 Washington
1978a (8.4) 30M, 20F (2.0)
page426
Table A21.7
Fr.7 Dodrill, 41.1 25 Healthy control group; 10.7 Washington
1978b 20M, 5 F
page426
Table A21.8
Fl'.8 Morrison Mode=19 60 Students from introductory College Idaho
et al., 1979 psychology courses; students
page 427 30M, 30 F; data are presented
Table A21.9 for dominant hand, averaged
over test-retest and
interexaminer trials
Fl'.9 Anthony 38.9 100 Healthy control group FSIQ: Colorado
et al., 1980 (15.8) 113.5
page 427 (10.8)
Table A21.10
Fl'.10 Bak & ~2 15 6M,9F 13.7 Texas
Greene, 1980 (1.91)
page 427 67-86 15 5 M,10 F; 14.9
Table A21.11 healthy participants (2.99)
Fl'.ll Eckardt 45.6 20 Control group of V.A. 60% had California
& Matarazzo, 1981 (11.1) medical inpatients; data some college
page428 for dominant hand education
Table A21.12 are provided for test and
retest over 12-22 days
Fl'.1! Pirozzolo Mean 60 Healthy control group, Mean USA
et al., 1982 early 60s mostly males 11-12 years
page 428
Table A21.13
Fl'.13 Rounsaville 24.9 29 CETA workers, 59% M 11.2 Massachusetts
et al., 1982
page 428
Table A21.14
Fl'.14 Yeudall et al., FSIQ: Alberta,
1982 14.8 99 Delinquent adolescents, 95.3 Canada
page 428 14.5 47 nondelinquent adolescents 117.1
Table A21.15
Fl'.15 O'Donnell 20.0 30 Student volunteers; 13.7
et al., 1983 (1.9) 21M, 9 F (1.4)
page 429
Table A21.16
Fl'.16 Prigatano 59.6 25 Controlgroupofhealthy 10.5 Oklahoma City
et al., 1983 (9.0) adults; 84% M (3.3) and Canada
page 429
Table A21.17
APPENDIX 21 905

IQ/
Study Age• n Sample Composition Education" Location
Fr.l7 Fromm-Auch 15-17 193 Normal volunteers; 14.8 Alberta

&Yeudall, 1983 18-23 111 M, 82 F; data are (3.0) Canada
page 429 24-32 partitioned by 5 age
Table A21.18 33-40 groups x gender
41-64
Fl'.l8 Bomstein, 1985 20-39 365 178M, 187 F; 12.3 Western
page 430 40-59 paid volunteers (2.7) Canada
Tables A21.19, A21.20 60-69 free of neurological or <HS
psychiatric illness; <::HS
data are presented by
3 age x 2 education x gender
groups
Fl'.l9 Villardita <25 10 Healthy volunteers; 8-13 Catania,
et al., 1985 45-54 10 data are partitioned into years Italy
page 430 55-64 10 4 age groups
Table A21.21 65-74 10
Fl'.IO Heaton et al., 32.7 100 79 M, 21 F; controls with 14.5 Denver
1985 (13.5) no neurological illness, head (2.84)
page 431 trauma, or substance abuse
Table A21.22
Fl'.lU Kane et al., 1985 38.9 46 Control sample of 12.3 Oklahoma City,
page 431 (11.3) medical and nonschizo- (2.6) Pittsburgh
Table A21.23 phrenic V.A. psychiatric patients;
data are reported in T scores
Fl'.U Heaton 15-81 553 356 M, 197 F; normal with ~20 Colorado,
et al., 1986 39.3 no history of neurological 13.3 California,
page 431 (17.5) illness, head trauma, or (3.4) Wisconsin
Table A21.24 <40 substance abuse; <12
40-59 data are presented in 3 12-15
2::60 age and 3 education groups 2::16
Fl'.23 Bomstein, 18-39 365 178M, 187 F; paid volunteers free 12.3 Western
1986a 40-59 of neurological or psychiatric (2.7) Canada
page 432 ~9 illness; data are stratified by <HS
Tables A21.25, A21.26 3 age x 2 education x gender groups; <::HS
proportion of participants classified
as impaired is presented
Fl'.l4 Polubinski & 18-24 120 Undergraduate students; 60 M, 60 F; College Ohio
Melamed, 1986 data are partitioned by firm and students
page 432 mixed handedness groups
Table A21.27
Fl'.25 Trahan et al., 18-32 713 Healthy subjects; 382 M, 331 F; data are
1987 33-47 presented by age x gender
page 432 48-62
Table A21.28 63-91
Fr.!& Yeudall et al., 15-20 62 Normal adults; 127 M, FSIQ: Alberta,
1987 21-25 73 98 F; data are stratified 111.75 Canada
page433 26-30 48 by 4 age groups x gender 109.79
Table A21.29 31-40 42 113.95
116.09
(continued)
906 APPENDIX 21

IQ/
Study Age" n ' Sample Composition Education• Location
Fl'.27 Alekoumbides 46.9 123 · Mostly inpatients of a large 11.4

et al., 1987 {17.2) general hospital without {3.2)
page 433 a history of neurological
Table A21.30 disorder, mostly males
Fl'.28 Bomestein et al., 62.7 134 Healthy adults with no history 11.7
1987b {4.3) of neurological or psychiatric {2.9)
page 434 illness; 49 M, 85 F; matched
Table A21.31, A21.32 with neurological patients;
classification rates based on
conventional and optimal
cutoff scores are presented
Fl'.29 Bomstein 32.3 23 !Volunteers from a university VIQ=105.8
et al., 1987a {10.3) community; 9 M, 14 F; (10.8)
page 434 test-retest data over 3- PIQ=105.0
Table A21.33 week period are presented {10.5)
Fl'.30 Russell, 46.19 155 ·V.A. patients suspected of having 12.29 Miami,
1987 (12.86) neurological condition but with (3.0) Cincinnati
page 435 negative neurological findings;
Table A21.34 148M, 7 F
Fl'.31 Thompson et al., 40.59 426 f279 M, 147 F; normal subjects; 13.15
1987 {18.27) Percent falling in {3.49)
page 435 lateralized dysfunction
Table A21.35 range is presented
Fl'. 31 van den Burg et al., 37.4 40 !Control group of healthy subjects; Northern
1987 {11.9) 16M, 24 F Holland
page 436
Table A21.36
Fl'.33 Bomstein & Suga, 55-70 134 49 M, 85 F; paid volunteers screened 11.7 Western
1988 62.7 for a history of neurological or {2.9) Canada
page 436 {4.3) psychiatric disorders;
Table A21.37 divided into 3 education groups: Range, mean:
' 17M, 29 F 5-10,8.5
16M, 28 F 11-12, 11.7
16M, 28 F >12,15.0
Fl'.34 Ardila & Rosselli, ~55 346 ;Healthy older adults with MMSE 0-5 Bogota.
1989 scores ~23; data are stratified 6-12 Colombia
page 436 by 5 age x 3 education groups >12
Table A21.38
FI'.3S Heaton et al., 42.1 486 Urban and rural volunteers; 13.6 California,
1991,2004 {16.8) data collected over 15 {3.5) Washington,
page 437 Groups: years through multicenter FSIQ: Texas,
Data are not 20-34 collaborative efforts; strict 113.8 Oklahoma,
reproduced 35-39 exclusion criteria; 65% M; {12.3) WISCOnsin,
in this book 40-44 data are presented in 6-8 Illinois,
45-49 T-score equivalents for 9-11 Michigan,
50-54 M and F separately in 10 12 New York,
55-59 age groupings by 6 13-15 Virginia,
00-64 education groupings; 16-17 Massachusetts,
65-69 in the 2004 edition, age range ~18 Canada
7~74 is expanded to 85 years and
75-80 the data are presented for
African-American and Caucasian
participants separately
APPENDIX 21 907

IQ/
Study Age· n Sample Composition Education• Location
Fl'.36 Ruff & 1~70 358 Normal volunteers screened for 7-22 years California,
Parker,l993 psychiatric hospitalization, Michigan,
page 437 16-24 chronic poly-drug abuse, or Eastern
Tables A21.39, A21.40 25-39 neurological disorders; seaboard
40-54 data are stratified by 4 age x
55-70 gender groups; data for
left band-dominant sample are
also presented; 179M, 179 F
Fl'.37 Russell & 45.1 176M Norms are collected from 12.5 Cincinnati,
Starkey, 1993 (13.0) standardization sample for (2.8) Miami
page438 the HRNES manual;
Table A21.41 40.7 24F Controls are V.A. 14.5
(15.3) patients without CNS (2.6)
pathology
Fl'.38 Dikmen 34.2 384 Normal and neurologically stable 12.1 Washington,
et al., 1999 (16.7) adults; some had neurological (2.6) Colorado,
page 439 conditions; 66% M; data on California
Table A21.42 test-retest reliabilities and practice
effect are provided
Fl'.39 McCurry 74.6 120 Nondemented Japanese 11.7 Seattle
et al., 2001 (2.7) American elderly; 43.8% M; (2.9)
page 439 87.0 81 weighted test scores are 10.0
Table A21.43 (5.1) stratified into 2 age groups (2.7)
Fl'.40 Sackellares & 1~ 40 Control group; performance rates Florida
Sackellares, 2001 X=33.2 for both bands and asymmetry
page 440 indices are provided for right-
Table A21.44 banders and left-banders separately
Fl'.41 Prigatano & 33.6 15 Control group; 46.7% M 12.86 Phoenix,
Borgaro, 2003 (9.61) (1.19) AZ
page 441
Table A21.45
Table A21.2. [FT.1] Vega and Parsons, 1967: Data

for the Control Group, which Included 43 Patients
Hospitalized for Causes Other than Central Ner- Table A21.3. [FT.2] Goldstein and Shelly,
vous System Dysfunction and Seven Nonhospita- 1972: Data for the Control Sample of
lized Participants• General Medical and Psychiatric Veterans
Administration Patients, Mostly Males
WAIS Dominant
n Gender Age Education FSIQ Hand Dominant Nondominant
n Hand Hand
50 37M 40.8 11.1 99.4 44.6
13 F (13.1) (3.2) (12.9) (9.2) 156 46.8 41.2
(10.4) (8.8)
"Data are available only for the dominant hand.
908 APPENDIX 21
Table A21.4. [Fl'.3] Goldstein and Braqn, 1974: Data for a Healthy
Sample Consisting of 201 Men and Eight Women Divided into Six
Age Groups
Age Mean Preferred Nonpreferred Percent

Group Age n Hand Hand Reversal•
~29 24.5 29 54.1 49.0 17
(4.4) (4.6)
34.9 75 53.4 47.9 11
(5.4) (5.2)
40-49 44.7 47 53.3 47.4 9
(4.8) (4.5)
50-59 53.5 30 50.5 45.2 10
(5.4) (4.4)
60-69 64.2 16 47.4 43.0 19
(8.6) (7.0)
70-79 72.2 12 44.5 41.1 25
(7.4) (7.5)
"Percent of subjects who were tapping faster wlth the nonpreferred hand than
with the preferred hand. I
Table A21.S. [FI'.4] Finlayson et al., Hf77: Data Presented in T Scores for the Control
Group, which Included Healthy Indivicfuals and Hospitalized Medical and Psychiatric
Patients, All Males ;
T Score
Educational Dominant Nondominant
Level n Age Edu<$tion FSIQ Hand Hand
Grade school 17 34.12 7.H 101.88 51.06 55.77

(8.72) (I.t1J (10.23)
High school 17 34.37 12.00 112.71 54.00 55.81
(7.36) 0 (12.21)
University 17 35.22 17.35 129.53 55.00 56.10
(7.78) (1.116) (7.58)
Table A21.6. [Ff.5] Wiens and Ma~. 1977: Data for Male Applicants to a
Patrolman Program I
WAIS Preferred Nonpreferred
Group n FSIQ Age EiucaUon Hand Hand
1 24 117.5 23.6 ;13.7 54.0 48.4

(4.6) (4.4)
2 24 118.3 24.8 '14.0 54.5 50.0
(4.0) (4.1)
APPENDIX 21 909
Table A21.7. [FT.6] Dodrill, 1978a: Data for the Control Group with no
Evidence of Neurological Disorder
Preferred Nonpreferred
n Age Education %Male Race Hand Hand
Total 50 27.3 12.0 60 98% White

sample (8.4) (2.0)
Males 30 56.0 51.43
(5.6) (5.7)
Females 20 51.2 48.0
(4.0) (3.7)
Table A21.8. [FT.7] Dodrill, 1978b: Data for the Table A21.11. [FT.10] Bale and Greene, 1980:
Control Group with no Evidence of Neurological Data for Healthy Right-Handed Older Adults
Disorder
Age Right Left
%Right- Right Left Group n Gender Age Education Hand Hand
n Age Education Gender Handed Hand Hand
50-62 15 6M 55.6 13.7 44.53 40.80
25 41.1 10.7 20M 100 53.4 49.6 9F (4.44) (1.91) (6.71) (4.77)
SF (6.2) (5.4) 67-86 15 5M 74.9 14.9 38.73 36.33
10 F (6.04) (2.99) (4.13) (5.93)
Table A21.9. [FT.8] Morrison et al., 1979: Data Table A21.12. [FT.ll] Eckardt and Matarazzo,
for the Dominant Hand on Test-Retest and 1981: Test-Retest Data for the Dominant Hand for
Interexaminer Trials Averaged over Two Probes the Control Group of Medical Inpatients from the
for Students from Introductory Psychology Courses Veterans Administration Hospital
with a Modal Age of 19
Dominant Hand
Test-Retest Interexaminer
n Age %Male Test Retest•
n Tapping n Tapping
20 45.6 100 42.1 43.6
Males 30 51.37 30 54.79 (11.1) (9.4) (10.4)
(4.45) (3.71)
"Test-retest interval was 12-22 days.
Females 30 48.52 30 50.37
(4.97) (4.63)
Table A21.10. [FT.9] Anthony et al., 1980: Data Table A21.13. [FT.12] Pirozzolo et al., 1982: Data
for the Control Group of Healthy Adults for the Control Group of Healthy Elderly (Mean
Age Early 60s, Mean Education 11-12 Years,
WAIS Dominant Nondominant Mostly Males)
n Age Education FSIQ Hand Hand
n Right Hand Left Hand
100 38.9 13.3 113.5 52.6 48.2
(15.8) (2.6) (10.8) (9.1) (7.6) 60 50.07 46.73
(8.75) (8.43)
910 APPE DIX 21
Table A21.14. [FT.13] Rounsaville et al ., 1982: D ata for a Sample of Comprehensive

Employment Training Act Workers
% Right- Dominant ondominant

% Male Education Handed Age Hand Hand
29 59 11.2 90 24.9 48.48 42.5
Table A21 .15. [FT.l4] Yeudall et al., 1982: Data for Delinquent and ondelinquent Canadian Adolescents
WAlS Number Preferred onpreferred

n Age FSIQ of Males Femal Hand Hand
Delinquent 99 14.8 95.3 64 35 40.0 37.0

(6.7) (5.8)
ondelinquent 47 14.5 117.1 29 18 42.9 40.2
(7.7) (6.9)
Table A21 .16. [FT.15] O ' Donnell et al., 1983: Data for a Control Group of tudent
Volunteers
% Right- WAI Dominant

n Age Education Gender Handed F IQ Hand
30 20.0 13.7 21M 90 117.1 51.3

(1.9) (1.4) 9F (10.3) (5.5)
Table A21 .17. [FT.16] Prigatano et al., 1983: Data for the Control Group
%Right- Dominant ondominant

n Age Education % Male Handed WAlS FSIQ Hand Hand
25 59.6 10.5 84 96 112.0 48.4 42.6

(9.0) (3.3) (11.0) (7.7) (6.0)
APPENDIX 21 911
Table A21.18. [IT.l7] Fromm-Auch and Yeudall,

1983: Data for a Sample of Healthy Canadian
Adults Stratified by Age and Gender•
Age n Hand (SD) Hand (SD)
Mala
15-17 17 47.6 43.6
(5.8) (4.9)
18-23 44 49.5 45.4
(6.9) (6.9)
24-32 31 50.6 46.0
(6.6) (6.1)
33-40 12 53.4 49.8
(5.9) (4.7)
41-64 4 44.4 41.4
(5.8) (3.5)
Fcmaalu
15-17 15 42.7 41.1
(7.9) (6.2)
18-23 30 43.6 41.2
(7.5) (6.5)
24-32 25 45.2 40.9
(6.7) (5.7)
33-40 6 45.8 44.3
(5.5) (4.6)
41-64 6 40.4 38.6
(4.8) (4.8)
"Mean education 14.8 years, mean full-scale IQ 119.1.
Table A21.19. [IT.l8a] Bomstein, 1985: Data for a Sample of Healthy Canadian Adults
Stratified by Age, Education, and Gender
Number Number Preferred Nonpreferred

Education of Males of Females Hand Hand
By age
20-39 13.0 107 64 47.2 43.5
(2.3) (6.5) (5.4)
40-59 11.9 31 66 40.3 37.8
(2.8) (7.5) (5.8)
60-69 11.8 40 57 35.4 33.8
(2.9) (7.7) (6.4)
Byedueadon
48.5 <HS" 51 57 39.8 37.6
(16.6) (9.2) (7.8)
41.1 ~HS 127 130 43.3 40.3
(16.8) (8.2) (7.2)
(continued)
912 APPENDIX 21

Education of Males; of Females Hand Hand
Bfgender
39.2 12.4 178 46.1 42.6
(17.2) (2.9) (7.0) (6.9)
47.3 12.2 187 38.6 36.5
(16.1) (2.5) (8.4) (6.8)
•Hs, high school.
Table A21.20. [Ff.18b] Bomstein, 198$: Data Presented in Age x Education (<High School, ;:::High
School) x Gender Groupings '
~ale Female
<HS ~HS <HS ~HS
Age Group n M (SD) j n M (SD) n M (SD) n M (SD)
Ptefon ed 1umd
20-39 21 49.7 86 48.5 13 45.2 49 44.3
(6.0) (6.5) (6.0) (5.8)
40-59 13 42.3 17 43.4 22 36.3 43 40.5
(5.2) (7.9) (7.8) (7.1)
60-69 16 39.1 23 43.0 22 29.7 34 32.2
(5.7) (4.7) (6.2) (6.0)
Nonpnf•rretllumd
20-39 47.0 44.8 40.7 40.6
(5.5) (6.4) (5.0) (5.6)
40-59 39.8 39.5 35.2 37.8
(3.6) (5.8) (5.8) (6.0)
60-69 35.2 39.3 29.8 32.0
(5.2) (6.2) (5.6) (4.9)
Table A21.21. [Fr.19] Villardita et 19&5: al.

Data• for Healthy Italian Adults with 8-~3 Years
of Schooling Partitioned into Four Age Glk>ups
Age Preferred NoJt>referred
Group n Hand fland
<25 10 93.4 :90.1

(7.9) (8.0)
45-54 10 82.8 80.4
(11.0) :(9.2)
55-64 10 70.3 69.5
(11.9) (9.4)
65-74 10 67.9 69.3
(9.9) ~1.3)
•score Is the total number of taps recorded for !feb hand

on two trials.
APPENDIX 21 913
Table A21.22. [FT.20] Heaton et al., 1985: Data Table A21.24. [FT.22] Heaton et al., 1986: Data
for the Neurologically Healthy Control Group for a Sample of Healthy Adults Stratified by Age
and Education
MIF Total for
n Age Education Ratio Both Hands Dominant Nondominant
n Hand Hand
100 32.7 14.15 79/21 104.27
(13.5) (2.84) (12.03) Age erou,.
<40 319 53.6 48.8
40-59 134 52.8 47.5
2:60 100 47.9 43.5
Education erou,.
<12 132 48.7 44.0
12-15 249 53.3 48.1
Table A21.23. [FT.21] Kane et al., 1985: Data• for 2:16 172 53.7 49.0
the Control Group Consisting of Medical and
Nonschizophrenic Veterans Administration Psy-
chiatric Patients
Dominant Nondominant
n Age Education Hand Hand
46 38.9 12.3 51.5 51.5

(11.3) (2.6) (10.1) (8.5) Table A21.25. [FT.23a] Bomstein, 1986a: Propor-
tion of Participants Classified as Impaired in a
"Data are reported in T scores.
Sample of Healthy Canadian Adults (n = 365)
% Classified as
Mean Median Mode Impaired
Preferred 42.3 43 47 79.9

hand (taps)
Nonpreferred 39.5 39.5 39 70.2
hand (taps)
Table A21.26. [FT.23b] Bomstein, 1986a: Proportion of Participants Classified as Impaired in a Sample
Stratified by Age x Education x Gender
Preferred Hand Nonpreferred Hand

Males (<50)" Females (<46) Males (<44) Females ( <40)
Age <HS 2:HS <HS 2:HS <HS 2:HS <HS 2:HS

18-39 35 58.6 61.5 50 20 36.8 69.2 43.4
(7120) (51187) (8113) (23/46) (4120) (32/87) (9.13) (20146)
40-59 100 75 95.5 79.5 84.6 81.3 77.3 56.8
(13113) (12116) (21122) (35144) (ll/13) (13116) (17/22) (25144)
60-69 100 91.3 100 100 100 73.9 95.4 9619
(16/16) (21123) (17122) (33133) (16/16) (17123) (21122) (32133)
•cutoff criteria for impairment, which differed for males and females, are presented in parentheses.
HS, high school.
914 APPENDIX 21
Table A21.27. [Ff.24] Polubinski and Melamed, 1986: Data• for Undergraduate
Students Partitioned into Firm and Mixed Right-Handedness Group
n Age Education Right Hand Left Hand
Men
Finn 30 19.7 13.3 93.6 77.6
(1.4) (0.7) (10.1) (10.0)
Mixed 30 20.1 13.6 93.5 79.6
(1.6) (0.9) (13.1) (12.1)
Women
Finn 38 19.8 13.4 91.1 76.0
(1.2) (0.7) (12.4) (13.8)
Mixed 22 19.4 13.3 88.8 74.5
(0.8) (0.6) (9.9) (11.6)
• Scores are presented for 15-second trials.
Table A21.28. [Ff.25] Trahan et al., 1987: Data•

for a Sample of Neurologically Intact Adults
Age/ Preferred Nonpreferred
Gender n Hand Hand
18-32
Males 175 54.58 50.31
(5.19) (5.38)
Females 221 48.71 44.75
(5.52) (5.60)
33-47
Males 145 53.28 48.52
(6.29) (6.64)
Females 56 49.12 45.07
(5.91) (5.12)
48-{;2
Males 29 52.80 47.98
(5.24) (5.78)
Females 35 46.84 43.59
(6.05) (5.92)
63-91
Males 33 49.23 45.31
(8.94) (8.43)
Females 19 43.13 41.03
(6.69) (7.38)
•score is the mean of three trials for each hand.

APPENDIX 21 915
Table A21.29. [FI'.26] Yendall et al., 1987: Data for Healthy Canadian Adults Stratified by Age for the
Entire Sample and for Males and Females Separately
WAIS-R Preferred Nonpreferred
Age Group n Age Education FSIQ '*'
Right-Handed Hand Hand
Entire•• (n =225)
15-20 62 17.76 12.16 111.75 79.03 46.59 42.51
(1.96) (1.75) (10.16) (6.60) (5.81)
21-25 73 22.70 14.82 109.79 86.30 47.28 44.38
(1.40) (1.88) (9.97) (8.13) (7.20)
26-30 48 28.06 15.50 113.95 89.58 50.47 45.64
(1.52) (2.65) (10.61) (7.28) (6.68)
31-40 42 34.38 16.50 116.09 90.48 50.52 46.54
(2.46) (3.11) (9.51) (8.37) (6.75)
15-40 225 24.66 14.55 112.25 85.78 48.38 44.54
(6.16) (2.78) (10.25) (7.76) (6.76)
F-'" (n=98)
15-20 30 17.73 12.10 110.32 73.33 44.77 41.63
(1.84) (1.52) (10.64) (7.37) (5.69)
21-25 36 22.83 14.53 107.28 83.33 44.36 41.62
(1.54) (1.99) (9.14) (7.48) (6.82)
26-30 16 28.69 14.94 113.10 93.75 48.14 43.24
(1.25) (2.32) (11.37) (6.99) (4.44)
31-40 16 33.88 16.19 114.27 87.50 44.35 41.98
(2.53) (2.29) (11.32) (5.64) (4.65)
15-40 98 24.03 14.12 110.19 82.65 45.10 41.95
(5.95) (2.43) (10.46) (7.12) (5.76)
Jlalea (n = 127)
15-20 32 17.78 12.22 113.00 84.38 48.36 43.36
(2.09) (1.96) (9.72) (5.31) (5.89)
21-25 37 22.57 15.11 112.30 89.19 50.20 47.14
(1.26) (1.74) (10.27) (7.78) (6.55)
26-30 32 27.75 15.78 114.38 87.50 51.68 46.88
(1.57) (2.79) (10.43) (7.24) (7.34)
31-40 26 34.69 16.69 117.31 92.31 54.32 49.35
(2.41) (3.55) (8.21) (7.52) (6.34)
15-40 127 25.15 14.87 113.87 88.19 50.97 46.60
(6.29) (2.99) (9.83) (7.26) (6.81)
Table A21.30. [Ff.27] Alekoumbides et al., 1987:

Data for a Sample of Medical and Psychiatric
Veterans Administration Patients without a History
of Neurological Disorder, Mostly Males and Mostly
Inpatients
123 46.9 11.4 43.4 38.5

(17.2) (3.2) (10.2) (9.1)
916 APPENDIX 21
Table A21.31. [Fr.28a] Bornstein et at., 1987b: Data for a Sample of Healthy
Canadian Adults (Subsample of Bornsteii,
I
1985)
n Age Edu~on Hand Hand
Total 134 62.7 ll7

sample (4.3) (2,.9)
Males 49 41.9 37.3
(5.8) (5.4)
Females 85 33.0 32.7
(6.9) (5.9)
Table A21.32. [Fr.28b] Bornstein et iai., 1987b: Classification Rates Based on

Conventional and Optimal Cutoff Scores or Two Hands for Males and Females
Percent Correctly Classified
Sample Cutoff Scorei Control Brain-Damaged
Conventional!
Males preferred <50 8 95
Males nonpreferred <44 10 79
Females preferred <46 2 95
Females nonpreferred <40 12 93
Optimal
Males preferred $32 92 29
Males nonpreferred $31 86 45
Females preferred $21 95 26
Females nonpreferred $26 88 48
Table A21.33. [Fr.29] Bomstein et al., 1987a: Data on Test-Retest

Performance after a 3-Week Interval for~a Sample of 23 Healthy
Adults (9 Males, 14 Females)
Test Dominant Nondominant
1 44.8 42.5
(6.3) (6.7)
2 43.5 43.0
(7.1) (7.1)
Da~ for Patients Seen in Veterans Administration Medical Centers

Table A21.34. [Fr.30] Russell, 1987:
who Were Suspected of Having a Neurol 'cal Condition but who had Negative Neurological Findings
Average Number
WAIS of Taps for
n Gender Age ~ucation Race FSIQ Both Hands
155 148M 46.19 12.29 147C• 111.9 48.48
7F (12.86) (3.00) SA (5.70)
ON
•c, Caucasian; A, African American; N, other.
APPENDIX 21 917
Table A21.35. [Fr.31] Thompson et al., 1987: Percent of Normal Participants•

Scoring in the Lateralized Dysfunction Range
Dominant Nondominant Intennanual
Hemisphere Hemisphere Percent
Groups n Dysfunction Dysfunction Difference Scores
All right 167 10.18 17.96 10.8

(9.4)
Mixed right 226 12.05 10.27 8.3
(8.3)
Left 33 18.18 0.00 5.2
(5.9)
Total 426 11.79 12.50 9.0
(8.7)
"Mean age= 40.59 (18.27), mean education= 13.15 (3.49) years.
Table A21.36. [Fr.32] van den Burg et al., 1987:

Data for Control Group Collected in Northern
Holland
n Gender Age Total Score•
40 16M 37.4 339

24F (11.9) (38)
"Total score represents number of taps over three 10-

second trials for both hands.
Table A21.37. [Fr.33] Bornstein and Suga, 1988: Data for a Sample of Healthy Older
Canadian Volunteers Stratified by Education
Education Number of Number of Preferred Nonpreferred
(years) Age" Males Females Hand Hand
5-10 62.3 17 29 34.1 32.8

(7.1) (5.9)
11-12 62.9 16 28 36.9 35.2
(7.9) (5.2)
>12 63.0 16 28 37.4 35.4
(8.1) (7.0)
• Age range for the sample is 55-70 years.

918 APPENDIX 21
Table A21.38. [Ff.34] Ardila and Ros~lli. 1989: Table A21.40. [Fr.36b] Ruff and Parker, 1993:
Data for Healthy Older Adults Collected ~ Bogota, Data for a Sample of Healthy Adults• Stratified by
Colombia Age and Gender
Age Group! Preferred N4tpreferred Gender/ Dominant Nondominant
Education Hand ;Hand Age Hand Hand
Group n M (SD) M (SD)
SS-60
0-5 40.9 37.2 Women
I
6-12 44.4 i 39.2 16-24 45 49.5 45.6
>12 48.1 :46.3 (5.1) (5.1)
25-39 45 49.0 44.6
61~
(4.1) (4.6)
0-5 39.7 36.2 40-54 44 47.0 43.5
6-12 43.3 39.9 (5.6) (5.2)
>12 41.7 39.6 ~70 45 45.7 40.4
(5.5) (5.2)
66-10
Total 179 47.8 43.5
0-5 32.7 31.6 (5.3) (5.4)
6-12 39.8 37.0
>12 40.0 37.5 Men
11-15 I 16-24 45 52.9

(5.1)
48.2
(4.4)
0-5 32.4 29.7 25-39 44 52.7 48.7
6-12 36.2 35.2 (6.8) (5.7)
>12 39.4 36.0 40-54 45 54.3 48.9
(5.7) (5.8)
>15
~70 45 53.5 48.3
0-5 26.2 ~.6 (6.4) (5.0)
6-12 30.0 7.7 Total 179 53.4 48.5
>12 33.5 ~1.3 (6.0) (5.2)
I
MoleiCitllple
I Total 358 50.6 46.0
I. (6.3) (5.9)
"Education range 7-22 years.
Table A21.39. [FT.36a] Ruff and Parke,, 1993:

Data for the Left Hand-Dominant Satfple of
Healthy Participants ·
Men !Women
n 17 18
Age 37.9 38.7
(18.0) . (16.1)
Education 13.7 14.1
(2.8) ' (2.6)
(5.8) (4.5)
Nondominant hand 46.7 : 41.9
(13.2) (6.2)
APPENDIX 21 919
Table A21.41. [Fr.37] Russell and Starkey, 1993: Data for a Sample of Veterans Administration Patients
Without Central Nervous System Pathology, Stratified by Gender
Gender n Age Education Race· Hand Hand
Male 176 45.1 12.5 165W 45.2 39.5

(13.0) (2.8) 118 (8.9) (7.9)
Female 24 40.7 14.5 23W 44.2 38.8
(15.3) (2.6) 18 (9.2) (8.3)
•w, white; 8, black.

Material from the manual for the Halstead-RusseQ Neuropsychological Evaluation System-Revised (HRNES-R) copyright
© 1993, 2001 by Western Psychological Services. Reprinted by permission of the publisher, Western Psychological
Services, 12031 Wtlshire Boulevard, los Angeles, California, 90025, U.S.A. Not to be reprinted in whole or in part for any
additional purpose without the expressed, written permission of the publisher. All rights reserved.
Table A21.42. [Fr.38] Dikmen et al., 1999: Test-Retest Data for Normal and Neurologically Stable
Adults•
Dominant Nondorninant
Hand Hand
WAIS Test-Retest
n Age Education M/F Ratio FSIQt Interval (Months) Time 1 nme2 Time 1 Time2
384 34.2 12.1 66134% 108.8 9.1 50.88 51.36 47.02 47.87
(16.7) (2.6) (12.3) (3.0) (6.59) (6.46) (6.39) (6.47)
• A number of participants had preexisting conditions that might affect test perfonnance, the most significant being alcohol
abuse and a significant traumatic brain injury.
twAIS FSIQ, (Wechsler, 1955).
Table A21.43. [Fr.39] McCurry et al., 2001: Data• for a Sample of Nondemented Japanese American
Elderly
%Right- Dominant Nondorninant
Age Group n Mean Age Education %Male Handed Hand Hand
70-79 120 74.6 11.7 54.2 94.4 42.9 39.4

(2.7) (2.9) (8.1) (6.1)
80-101 81 87.0 10.0 28.4 93.4 36.4 34.2
(5.1) (2.7) (9.2) (7.7)
•Test scores are weighted.

920 APPENDIX 21
Table A21.44. [Ff.40] Sackellares and Sackellares, 2001: Data for the
Control Group•
Dominant Nondominant Asymmeby
Handedness n Hand Hand Index
Right 28 53.83 47.67 10.70
(7.55) (5.56) (8.97)
Left 12 53.43 48.71 8.76
(4.94) (6.31) (9.45)
•Mean age=33.2, range 18-50 years.
Table A21.45. [FI'.41] Prigatano and Bo~. 2003: Data for the Control Group
fl. J, Right- Dominant Nondominant
n Age Male Education lfanded Hand Hand
15 33.6 46.7 12.86 100 51.49 47.32

(9.61) (1.19) (4.33) (3.75)
for the Finger Tapping Test (FIT)
Table A21 m.1. Results of the Meta-Analysis and Predicted Scores for the
FIT: Males, Dominant Band
Only those studies reporting data for males and females separately were included in
the analyses.
Total number of partieipants 963
xt sot
s,.,.
Variable n• Range
me
Mean 20 37.76 30.69 12-175
Age
Mean 20 49.33 19.55 19-77
SD 20 3.99 2.58 0.5-9
Eductmon
Mean 6 12.08 1.48 10.5-14
SD 4 2.48 1.14 0.5-3.3
Tatscoremecma
Combined mean 20 51.09 3.47 44.5-56.0
Combined SD 20 6.50 1.68 3.7-8.9
(continued)
921
922 APPENDIX 21M
Predicted number of taps averaged 01fl'r 8ve trials and SDs, per age group
(FIT, males, dominant band)• ·
95%CI 95%CI
Age Predieted Lower Predicted Lower Upper

BGnge Score Band SD Band Band
!0-U 54.41 53.20 55. 4.60 3.98 5.23

!5-!9 5U7 53.55 54.1il 4.96 4.41 5.50
30-34 53.83 53.16 5.31 4.84 5.78
35-39 53.38 52.41 54.!
54. 5.66 5.26 6.07
40-44 51.81 51.61 54. 6.01 5.67 6.36
4S-49 51.15 50.79 53.5J 6.37 6.06 6.68
50-IU 51.37 49.88 52.8f; 6.71 6.43 7.01
5lS-S9 50.48 48.76 52.2b 7.07 6.77 7.38
60-64 49.49 47.32 51.6J 7.43 7.08 7.77
fi5.4J9 48.38 45.50 51.2f" 7.78 7.38 8.18
10-14 47.17 43.29 51.0f 8.13 7.66 8.60
Predicted te.e ICOre=54.11232 +0.0621745 •~e- 0.0021787 •age2

Predicted SD=3.013817 +0.0706054•age
I
Signiflcance tests for regression with ~e test scores

I
Ordinary least-14JUB"'S regression of tes4 meaas on age (quadratic)

Number of clusters j~--
R2 0',6H6
F<dO• p F<2.7)=6.1S, p <0.023
Term CoefBcient SE p 95%CI
Age 0.0621745 0.186 11.33• 0.748" -0.378 to 0.503

Age2 -0.0021787 0.002 -1.02 0.341 -0.007 to-0.003
Constant 54.11232 3.654 li&.81 0.000 45.47 to 62.75
"Significance test for age centered (sample mealls- aggregate mean): t = -3.41, p = 0.011.
Prediction
Predicted age range ~74years
SEe 0.84
95%CI 49.95-53.26
APPENDIX 21M 923
0
55
50
45
20 30 40 50 80 70 80
age
Fipre A21 m.1. A scatterplot illustrating the dispersion of the data points around the regression line for
the FTI', Male, Dominant Hand. The size of the bubbles reflects the weight of the data point, with larger
bubbles indicating larger standard error and smaller weight

Q<dO• P Qusl = 85.19, p < 0.000
Linear 0.656 0.637 30.215 -18.353

Quadratic 0.686 0.649 31.374 -17.194
BIC' difference of 1.159 provides weak support for the linear model.
Tests for parameter speeUleations
Shapiro-Wdk W test W = 0.964, p = 0.629
Homoscedasticity
(continued)
924 APPENDIX 21M
Signiflcance tests for regression with the SDs
OrdiDary least-squares regression of SDs on age (linear)

R2 0.677
F<dO·P F< t.7l = 56.39, p < 0.0001
Term Coefficient SE p 95,., Cl
Age 0.0706054 0.009 7.51 0.000 0.048 to 0.093

Constant 3.013817 0.516 5.84 0.001 1.794 to 4.233
Predietion
SEe 0.21
95%CI 5.96-6.78
EfFects of demographic variables
Education
Amount of data on education available in the literature was not sufficient for the analyses.
FIT: Males, Nondominant Hand
Only those studies reporting data for males and females separately were included in
the analyses.
Number of studies induded In the analysis 7

Yean ofpublieation 1974-1993
Number of data points used In the analysis 19
Sa,.,. ...
Mean 19 38.74 32.60 12-175
Age
Mean 19 49.92 19.25 20-77
SD 19 4.07 2.51 1.5-9
Eduecmon
Mean 5 12.06 1.55 10.5-14
SD 3 2.80 0.67 2.0-3.3
Test score_,.,
Combined mean 19 46.44 3.17 41.1-51.4
Combined SD 19 6.08 1.42 4.1-8.4
APPENDIX 21M 925
Predicted number of taps averaged over 6ve trials and SDs, per age group
(FIT, males, nondominant band)•
95%CI 95%CI

&nge Score Band Band SD Band Band
J~U 50.00 49.12 50.88 4.84 4.20 5.48

J5-J9 49.35 48.61 50.10 4.86 4.35 5.37
30-34 48.71 47.97 49.44 4.94 4.44 5.45
35-39 48.06 47.19 48.93 5.08 4.52 5.65
40-44 47.41 46.32 48.50 5.28 4.65 5.91
45-49 46.76 45.40 48.11 5.53 4.86 6.20
50-IU 46.11 44.47 47.75 5.85 5.17 6.52
55-S9 45.46 43.52 47.40 6.!2 5.57 6.86
60-64 44.81 42.56 47.06 6.65 6.07 7.22
65-69 44.16 41.60 46.73 7.13 6.65 7.62
1~14 43.51 40.63 46.39 7.68 7.27 8.09
Predicwd tat acore =52.92403- 0.1298114 •age

Predicted SD = 5.45396- 0.0534967 • age+ 0.0011612 • age2
Ordinary least-squares regression of test means on age (linear)

R2 0.622
F<dl)• p Fo.sl = 14.95, p < 0.008
Age -0.1298114 0.034 -3.87 0.008 -0.212 to -0.048

Constant 52.92403 1.076 49.18 0.000 50.29 to 55.56
Prediction
SE., 0.79
95%CI 45.22-48.30
(continued)
926 APPENDIX 21M
55
0
40
~ ~ 40 ~ i ~ ro ~
age I
Figure A21 m.2. A scatterplot illustrating the ~rsion of the data points around the regression line for the
FIT, Male, Nondominant Hand. The size of e bubbles reflects the weight of the data point, with larger
bubbles indicating larger standard error and s aller weight.
Tests for assumptions and model 8t ;

..
Tests for heterogeneity In the fbud ..... set
Pooled estimates for 6xed effect
Pooled estimates for random effect
I 48.370
47.679
Qcdf). p : Qns> = 105.34, p < 0.000
Tests for model&t-.lditioo of a ~tic term

I
Model BIC BIC'
Linear 0.622 0.600! 28.159 -15.550

Quadratic 0.630 0.5841 30.687 -13.022
BIC' difference of 2.528 provides positive su rt for the linear model.
Tests for parllllleter speei&cations

Shapiro-Wille W test W = 0.~ p = 0.538
Homoscedasticity
White's general test 6.616, v,=0.037
Ordinary least-squares regression of SDs on age (quadratic)

R2 0.665
F(df)o p F<2.6) =47.21, p < 0.0002
Age -0.0534967 0.055 -0.96" 0.375" -0.190 to 0.083

Age2 0.0011612 0.001 2.17 0.073 -0.000 to 0.002
Constant 5.45396 1.204 4.53 0.004 2.506 to 8.402
"Significance test for age centered (sample means- aggregate mean): t = 8.58, p < 0.000.
Prediction
SE. 0.29
95%CI 5.25--6.40
Education
Table A21 m.3. Results of the Meta-Analysis and Predicted Scores for the
FIT: Females, Dominant Hand
Only those studies reporting data for males and females separately
were included in the analyses.

Sample .U.
Mean 10 43.59 41.98 19-221
Age
Mean 10 44.91 21.48 19-77
SD 10 4.24 2.43 0.5-8.4
Education
Mean 2 12.49 0.71 12-13
SD 2 1.27 1.06 0.5-2
Test score means
Combined mean 10 47.57 2.77 43.1-51.2
Combined SD 10 5.45 0.93 4.0-6.7
tWeighted means and SDs.
928 APPENDIX 21M
Predicted number of taps averaged pver &ve trials and SDs, per age group
(FIT, females, dominant hand)• ~
95% CI 1 95%CI
Age Precllcted Lower Upper Predicted Lower Upper

&mge Score Band fand SD Band Band
!0-14 50.07 49.06 $1.08 4.66 4.12 5.20
!5-!9 49.74 48.79 4o.68 4.84 4.33 5.34
30-34 49.34 48.45 $0.22 5.01 4.54 5.49
35-39 48.88 48.05 49.71 5.19 4.74 5.64
40-44 48.36 47.58 49.14 5.36 4.93 5.80
45-49 47.78 47.07 ~.49 5.54 5.11 5.97
SO-S4 47.13 46.51 ~.76 15.71 5.29 6.14
SS-59 46.43 45.90 1fl.96 5.89 5.46 6.33
60-64 45.66 45.25 :.01 8.07 5.62 6.52
~ 44.83 44.56 .11 8.14 5.77 6.71
10-14 43.84 43.79 ~.09 8.41 5.92 6.92
*Based on the equations:

Predicted tal ecore =50.80925- 0.00513091" age - 0.0012358 • age2
Predicted SD =3.869681 + 0.0351682 • ;age 1
Sfgni6cance tests for the regression . . test scores

r
Ordinary least-1quares regression of le1l means on age (quadratic)
R2 0.937
F<dl)• p F<2.3J ="·78, p < 0.003
Age -0.0051309 0.031 :.....o.11• 0.878• -0.103 to 0.093

Age2 -0.0012358 0.000 }--3.98 0.028 -0.002 to -0.000
Constant 50.80925 0.846 60.03 0.000 48.12 to 53.50
• Significance test for age centered (sample llf8llS -aggregate mean): t = -10.64, p < 0.002.
Preclietion
Predicted age range 20-74 years .
Mean predicted score 47.47 (2.05) .
SEe 0.33
95%CI 46.82-48.11 .
APPENDIX 21M 929
0
50
45
~ ~------.-----~-----.-----.-----.------~
20 30 50 80 80
•
Figure A21 m.l. A scatterplot illustrating the dispersion of the data points around the regression line for the
FIT, Female, Dominant Hanel The size of the bubbles reflects the weight of the data point, with larger
Tests for _...,ticms and model 8t
Tests for heterogene~ in the 8nal data set

Pooled estimates for &xed effect 48.571
Qccll).p Q(9) = 46.94, p < 0.000
Tests for model 8t--eddltioa of a quadratic term
Model Adjusted~ BIC BIC'
Unear 0.912 0.901 4.998 -22.009

Quadratic 0.937 0.919 3.934 -23.072
BIC' difference of 1.()64 provides weak support for the quadratic model.
Tests for parameter speei8catioas
Shapiro-Wdk W test W =0.943, p = 0.590
Homoscedasticity
(continued)
930 APPENDIX 21M
Significance tests for the regression ~n SDs
Ordinary least-squares regression of SDs on age (linear)

Number of observations · 10
~ 0.663
=
Fo.3) 34.74, p < 0.009
Age 0.0351682 0.006 ' 5.89 0.010 O.ol6 to 0.054

Constant 3.869681 0.371 1G.42 0.002 2.688 to 5.052
Prediction
Mean predicted SD 5.54 (0.58) ;
SEe 0.24
95%CI 5.08-6.00
Education I
Amount of data on education available in ~e literature was not sufficient for the analyses.
Table A21m.4. Results of the Meta-~ysis and Predicted Scores for the
FIT: Females, Nondominant Hand ! .
(Relevant values are weighted on the s~dard error for the test mean)
Description of the aggregate sampl~
Only those studies reporting data for males fDd females separately were included in the analyses.
Number of studies included in the ~ 3
Number of data points used in the ~ 9
s,.,.
Variable
Mean
.ue
n•
9
xt
44.78
sot
44.76
Range
19-221
Age
Mean 9 ~.40 21.11 20-77
SD 9 ...69 2.17 2-8.4
Eductmon
Mean 1 1•.00
SD 1 •.00
Ted ecore metma
Combined mean 9 43.65 2.46 40.4-48.0
Combined SD 9 5.57 1.20 3.7-7.4
"Number of data points differs for different: analyses due to missing data.
APPENDIX 21M 931
Predicted number of taps averaged over 8ve trials and SDs, per age group (FIT, females, non-
dominant hand)•
95%CI 95%CI
Age Bmlge Predicted Score Lower Band Upper Band Predicted SD Lower Band Upper Band
!0-U 46.31 44.42 48.21 4.74 3.72 5.76

~~J9 45.80 44.07 47.53 4.69 3.86 5.52
30-34 45.28 43.71 46.86 4.70 3.98 5.43
35-39 44.77 43.33 46.21 4.77 4.07 5.46
40-44 44.26 42.94 45.57 4.89 4.18 5.60
4S-49 43.74 42.53 44.96 5.07 4.33 5.81
50-IU 43.22 42.08 44.37 5.31 4.56 6.05
SS-$9 42.71 41.60 43.82 5.61 4.88 6.33
60-64 42.20 41.09 43.31 5.96 5.29 6.63
65-69 41.68 40.53 42.83 6.37 5.79 6.95
10-14 41.17 39.95 42.39 6.84 6.40 7.28
Predicted fat ecore = 48.62897- 0.1029087 • age

Predicted SD=5.68279-0.0677816•age+O.OOU554•agc
Signi&cance tests for the regression on test scores
Ordmary least-squares regression of test means on age (Jinear)

R2 0.779
F<dO· p Fn.2> = 23.84, p < 0.039
Age -0.1029087 0.021 -4.88 0.039 -0.194 to-0.012

Constant 48.62897 1.380 35.23 0.001 42.69 to 54.57
Predictioo
SEe 0.69
95%CI 42.39-45.10
(continued)
932 APPENDIX 21M
48 0
48
44
42
0
0
40
20 30 40 50 so 70 so
age
Fipre A21 m.4. A scatterplot illustrating the dispersion of the data points around the regression line for the
FIT, Female, Nondominant Hand. The size of the bubbles reflects the weight of the data point. with larger
Tests for heterogeneity in the 6oal data set

Q(dO. p Q(8l = 55.77, p < 0.000
Tests for model 8t-ddition of a quadratic term
Model Adjusted If BIC BIC'
Linear 0.779 0.748 11.709 -11.401

Quadratic 0.779 0.706 13.905 -11.402
BIC' difference of 0.001 provides weak support for the quadratic model.
Tests for parameter speclflcations

Shapiro-Wilk W test W =0.959, p =0.793
Homoscedasticity
APPENDIX 21M 933

Signi6cance tests for the regression on SDs
OrdiDary least-squares regression of SDs on age (quadratic)

R2 0.800
F<dfl• p
Age -0.0677816 0.056 -1.21" 0.351" -0.309 to 0.174

Age2 0.00ll554 0.000 2.39 0.139 -0.001 to 0.003
Constant 5.68279 1.416861 4.01 0.057 -0.413 to 11.78
• Significance test for Age centered (sample means -aggregate mean): t =3.78, p =0.063
Prediction
SE., 0.37
95%CI 4.64-6.08
EfFects of demographic variables
Edueation
Tables for the Grip Strength Test
(Hand Dynamometer)
Table A22.1. Locator Table for the Hand Dynamometer

D. I Matarazzo et.al., 21-28 29 Normal young men; 12-16 Oregon

1974 X=24 patrolmen applicants FSIQ:
page 447 118
Table A22.2
D.l Wiens & Matarazzo, 23.6 24 Normal young men divided FSIQ: Oregon
1977 24.8 24 into 2 groups 117.5
page 447 118.3
Table A22.3
D.3 Dodrill, 1978b X=41.1 25 Control group; 20 M, 5 F X=10.7 Washington
page 447
Table A22.4
D.4 Dodrill, 1979 Nonneurologicalsample: Washington
page448 27.51 47 M 12.47
Table A22.5 27.49 47 F 12.36
D.S Rounsaville et.al., 24.9 29 CETA workers, 59% M 11.2 Massachusetts
1982
page 448
Table A22.6
D.6 Yeudall et al., 1982 FSIQ: Alberta,
page 448 14.8 99 Delinquent and 95.3 Canada
Table A22.7 14.5 47 nondelinquentadoresoents 117.1
934
APPENDIX 22 935

Study Age" n Sample Composition IQ/Education" Location
0.7 Prigatano et al., 59.6 25 Control group of healthy 10.5 Oklahoma City
1983 (9.0) adults; 84% M (3.3) and Canada
page 449
TableA22.8
0.8. Fromm-Auch at 1~17 193 Normal volunteers; Education: Alberta,
Yendall, 1983 18-23 111 M, 82 F; data are 14.8 (3.0) Canada
page449 24-32 partitioned by 5 age WAIS-R FSIQ:
TableA22.9 33-40 groups x gender 119.1 (8.8)
41-64
0.9 Bornstein, 1985 20-39 365 178M, 187 F; 12.3 Western
page 450 40-59 paid volunteers (2.7) Canada
Tables A22.10, A22.11 60-69 free of neurological or <HS
psychiatric illness; 2:HS
data are presented by
age x education x gender
0.10 KofBer & Zehler, 20-29 206 Normal sample; 100 M
1985 30-39 106 F; stratified
page 450 40-49 by 5 age groups and
Table A22.12 50-59 gender
ro-77
0.11 Heaton et al., 32.7 100 79 M, 21 F; controls with 14.5 Denver
1985 (13.5) no neurological illness, head (2.84)
page450 trauma, or substance abuse
Table A22.13
0.12 Kane et al., 1985 38.9 46 Control sample of 12.3 Oklahoma City,
page 451 (11.3) medical and nonschizo- (2.6) Pittsburgh
Table A22.14 phrenic V.A. psychiatric patients;
data are reported in T scores
0.13 Heaton et al., 1986 1~1 553 356 M, 197 F; normal subjects with 0-20 Colorado,
page 451 39.3 no history of neurological 13.3 California,
Table A22.15 (17.5) illness, head trauma, or (3.4) Wisconsin
<40 substance abuse; 7.2% <12
40-59 left-handed; data are 12-15
2:60 presented in 3 age and 2:16
3 education groups
0.14 Yendall et al., 1987 1~20 62 Normal adults; 127 M, FSIQ: Alberta,
page 451 21-25 73 98 F, data are stratiBed 111.75 Canada
Table A22.16 26-30 48 by 4 age groups x gender 109.79
31-40 42 113.95
116.09
0.13 Thompson et al., 40.59 426 279 M, 147 F; normal subjects; 13.15
1987 (18.27) percent falling in (3.49)
page 452 lateralized dysfunction
Table A22.17 range is presented
0.16 Ernst, 1988 65-75 85 Normal elderly; 39 M, 46F 10.4 Queensland,
page 453 (3.1) Australia
Table A22.18
(continued)
936 APPENDIX 22

Study Age• n Sample Composition IQ/ Education• Location
0.17 Heaton et al., 42.1 486 Urban and rural volunteers; 13.6 California,
1991,2004 (16.8) data collected over 15 (3.5) Washington,
page 453 Groups: years through multicenter FSIQ: Texas,
Data are not 20-34 collaborative efforts; strict 113.8 Oklahoma,
reproduced 35-39 exclusion criteria; 65% M; (12.3) WISCOnsin,
in this book 40-44 data are presented in 6-8 IDinois,
45-49 T-score equivalents for 9-11 Michigan,
50-54 M and F separately in 10 12 New York,
55-59 age groupings by 6 13-15 Vrrginia,
60-64 education groupings; 16-17 Massachusetts,
65-69 in 2004 edition, age ~18 Canada
70-74 range is expanded to 85
75-80 years and data are
presented for African-
American and Caucasian
participants separately
0.18 Russell &: 45.0 175M Norms are collected from 12.5 Cincinnati,
Starkey, 1993 (12.9) standardization sample for (2.8) Miami
page 454 HRNES manual;
Table A22.19 40.7 24F controls are V.A. 14.5
(15.3) patients without CNS (2.6)
i pathology
I
0.19 Tremont 16-74 157 : Patients referred for evaluation 11.53 Oklahoma
et al., 1998 : but determined to be (2.76)
page 455 I neurologically normal; 12.62
Table A22.20 71 M, 86 F; data for (2.76)
dominant hand partitioned 15.63
by FSIQ (3 levels) are (3.37)
presented for 3 IQ
levels
0.20 Dikmen 34.2 384 Normal and neurologically 12.1 Washington,
et al., 1999 (16.7) stable adults; some had (2.6) Colorado,
page 455 neurological conditions; California
Table A22.21 66% M; data on tests-
retest reliabilities and
practice effect are provided
0.21 Triggs et al., 37 60 130 right-handed, 30 left- At least HS
2000 (9) \ handed healthy volunteers;
page456 21-57 r data are presented for
I
Table A22.22 · 2 hands by 2 handedness
groups
O.U Christensen 70-74 199 i Healthy elderly, 37%-52% M; 11.78 Australia
et al., 2001 75-79 120 ! data are presented as the 11.34
page 456 80-84 41 ·, mean for 2 hands for 4 11.00
Table A22.23 ~85 14 age groups 10.79
•Age column and IQ'education column con · information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, whichevtr information is provided by the authors.
I
i
APPENDIX 22 937
Table A22.2. [D.1] Matarazzo et al., 1974: Data for Men who Met Selection Criteria for the
Portland Police Department
Hand Hand
n Age Education WAIS FSIQ Test Retest Test Retest
29 24 14 118 56.84 55.16 53.59 51.74

(7.66) (8.48) (6.14) (7.20)
Table A22.3. [D.2] Wiens and Matarazzo, 1977: Data for Male Applicants to a Patrolman
Program
WAIS Preferred Nonpreferred
Group n FSIQ Age Education Hand Hand
1 24 117.5 23.6 13.7 58.1 53.4

(7.3) (5.5)
2 24 118.3 24.8 14.0 57.5 53.9
(6.3) (6.2)
Table A22.4. [D.3] Dodrill, 1978b: Data for Control Participants
..Right- Right Left

n Age Education Gender Handed Hand Hand
25 41.1 10.7 20M 100 48.1 44.9

SF (13.4) (12.2)
Table A22.5. [D.4] Dodrill,1979: Data for Control

Participants
Dominant
n Age Education SES" Hand
Males 47 27.51 12.47 49.45 54.13

(9.95)
Females 47 27.49 12.36 47.41 34.00
(5.96)
"SES, socioeconomic status.

938 APPENDIX 22
Table A22.6. [D.5] Rounsaville et al., 1982: Data for a Sample of Comprehensive Employ-
ment Training Act (CETA) Workers
%Right- Dominant Nondominant
n %Male Education Handed Age Hand Hand
29 59 11.2 90 24.9 41.05 38.18
Table A22.7. [D.6] Yeudall et al., 1982: Data for Delinquent and Nondelinquent Canadian Adolescents
WAlS Number Number of Preferred Nonpreferred
n Age FSIQ of Males Females Hand Hand
Delinquent 99 14.8 95.3 64 35 37.0 35.5

(11.0) (10.0)
Nondelinquent 47 14.5 117.1 29 18 33.1 30.3
(7.8) (6.9)
Table A22.8. [D.7] Prigatano et al., 1983: Data for the Control Group
%Right- WAlS Dominant
n Age Education %Male Handed FSIQ Hand
25 59.6 10.5 84 96 112.0 45.1

(9.0) (3.3) (11.0) (11.4)
APPENDIX 22 939
Table A22.9. [0.8] Fromm-Auch and Yeudall,

1983: Data for a Sample of Healthy Canadian
Adults Stratified by Age x Gender•
n Hand (SD) Hand (SD)
Jlala
15-17 17 38.0 35.8
(8.4) (9.6)
18-23 43 49.7 46.6
(9.7) (9.9)
24-32 31 51.8 49.6
(8.1) (7.2)
33-40 12 52.9 51.2
(8.3) (7.9)
41-64 4 44.5 47.9
(10.9) (11.9)
F-'-
15-17 15 28.1 26.3
(5.0) (5.2)
18-23 29 28.8 26.4
(7.8) (6.2)
24-32 24 34.4 30.2
(9.2) (6.8)
33-40 6 27.7 28.6
(3.2) (3.1)
41-64 6 28.0 24.1
(6.2) (6.8)
•Mean education 14.8 years, mean full-scale IQ 119.1.
Table A22.10. [D.9a] Bornstein, 1985: Data for a Sample of Healthy Canadian Adults
Number of Number of Preferred Nonpreferred
Education Males Females Hand Hand
Bfiage
20-39 13.0 107 64 43.1 40.1
(2.3) (12.1) (11.4)
40-59 11.9 31 66 34.0 31.5
(2.8) (9.7) (10.4)
60-69 11.8 40 57 32.0 29.4
(2.9) (10.0) (9.2)
Bfi.daecdion
48.5 <HS 51 57 35.7 33.7
(16.6) (11.9) (11.9)
41.1 ~HS 127 130 38.8 35.8
(16.8) (12.1) (11.5)
'Bfl gend«-
39.2 12.4 178 47.5 44.3
(17.2) (2.9) (9.0) (8.7)
47.3 12.2 187 28.8 26.5
(16.1) (2.5) (6.2) (6.2)
APPENDIX 22
Table A22.11. [D.9b] Bornstein, 19851 Data Presented in Age x Education (<High School,
~High School) x Gender Groupings I
~e Female
<HS ~HS <HS ~HS
Age Group n M (SD) n M (SD) n M (SD) n M (SD)
Pt-fo• red hand

20-39 21 50.8 86 49.9 13 32.7 50 31.0
(11.5) (8.4) (8.7) (5.4)
40-59 13 39.8 17 48.2 22 27.7 43 29.8
(6.0) I (7.3) (5.9) (5.8)
60-69 16 38.7 22 44.5 22 25.6 34 25.0
(5.9) (5.6) (5.3) (4.9)
Nonpreferred hand
20-39 47.7 46.4 31.2 28.7
(11.7) (7.6) (8.0) (5.0)
40-59 38.2 46.4 24.9 26.9
(6.5) (9.1) (6.7) (5.4)
60-69 37.2 II 39.3 24.0 22.8
(5.4) (5.5) (6.0) (4.8)
I
I
I
Table A22.12 •. [D.lO] Koffier and Ztfder, 1985:

Data for Healthy Adults Stratified Iby Age x
I
Gender I
Dominant $ondominant
Age/Gender n Hand Hand
.IIQ-.19
Men 41 53.8 50.3
(7.8) (7.4)
Women 39 33.3 30.5
I
(4.7) (4.4)
30-39
Men 23 55.4 53.3
(7.1) (7.4)
Women 25 33.7 31.1
(6.2) (5.6)
40-49
Men 13 50.2 49.2
(5.3) (7.8)
Women 14 30.7 28.7
(5.5) (4.3)
50-S9
Men 12 44.3 44.8
(5.4) (5.8)
Women 13 28.8 25.3
(3.6) (3.8)
60-77
Men 11 45.5 41.3
(5.4) (6.7)
Women 15 28.3 23.5
(6.3) (5.2)
APPENDIX 22 941
Table A22.13. [D.ll] Heaton et al., 1985: Data for a Control Sample
MIF Kilograms
n Age Education Ratio for 8oth Hands
100 32.7 14.15 79121 88.99
(13.5) (2.84) (21.4.2)
Table A22.14. [D.12] Kane et al., 1985: Data• for the Control Group
consisting of Medical and Nonschizophrenic Veterans Administration
Psychiatric Patients
46 38.9 12.3 53.7 55.2
(11.3) (2.6) (6.5) (7.1)
•Data are reported in T sconiS.
Table A22.1 S. [D.13] Heaton et al., 1986: Data for a Sample of Healthy
Adults Stratified by Age and Education
n Hand Hand
Apgroupe
<40 319 51.4 47.8
40-59 134 51.7 46.8
:2:60 100 44.3 40.5
Ecft.c:alion groupl
<12 132 47.1 43.4
12-15 249 51.1 47.1
:2:16 172 51.5 47.1
942 APPENDIX 22
Table A22.16. [0.14] Yeudall et al., 1987: Data for Healthy Canadian Adults Stratified by Age for the
Entire Sample and for Males and Females Separately
WAIS-R IllRight- Preferred
Age Croup n Age Education FSIQ Handed Hand
Endre.....,. (n=JJ5)
15-20 62 17.76 12.16 111.75 79.03 37.22 :W.57
(1.96) (1.75) (10.16) (10.68) (10.20)
21-25 73 22.70 14.82 109.79 86.30 40.33 37.74
(1.40) (1.88) (9.97) (14.33) (13.81)
26-30 48 28.06 15.50 113.95 89.58 45.20 42.19
(1.52) (2.65) (10.61) (11.59) (11.91)
31-40 42 :W.38 16.50 116.09 90.48 45.22 42.81
(2.46) (3.11) (9.51) (13.47) (11.60)
15-40 225 24.66 14.55 112.25 85.78 41.42 38.76
(6.16) (2.78) (10.25) (13.01) (12.61)
F - ' a (n = 98)
15-20 30 17.73 12.10 110.32 73.33 30.20 28.07
(1.84) (1.52) (10.64) (5.56) (4.52)
21-25 36 22.83 14.53 107.28 83.33 29.79 27.22
(1.54) (1.99) (9.14) (6.91) (6.CH)
26-30 16 28.69 14.94 113.10 93.75 33.88 30.25
(1.25) (2.32) (11.37) (7.65) (6.35)
31-40 16 33.88 16.19 114.27 87.50 32.98 30.10
(2.53) (2.29) (11.32) (10.10) (6.62)
15-40 98 24.03 14.12 110.19 82.65 31.09 28.Q
(5.95) (2.43) (10.46) (7.33) (5.81)
Jlala (II= JJ7)

15-20 32 17.78 12.22 113.00 84.38 43.58 40.46
(2.09) (1.96) (9.72) (10.25) (10.37)
21-25 37 22.57 15.11 112.30 89.19 51.49 48.56
(1.26) (1.74) (10.27) (11.36) (10.83)
26-30 32 27.75 15.78 114.38 87.50 51.05 48.35
(1.57) (2.79) (10.43) (8.55) (9.07)
31-40 26 :W.69 16.69 117.31 92.31 52.28 50,15
(2.41) (3.55) (8.21) (9.57) (8.74)
15-40 127 25.15 14.87 113.87 88.19 49.49 46.75
(6.29) (2.99) (9.83) (10.53) (10.46)
APPENDIX 22 943
Table A22.17. [D.15] Thompson et al., 1987: Percent of

Healthy Participants Scoring in the Lateralized Dysfunction
Range
Intermanual
Dominant Nondominant Percent
Hemisphere Hemisphere Difference
Groups n Dysfunction Dysfunction Scores
All right 167 19.16 8.98 7.7

(9.8)
Mixed right 226 17.70 19.47 9.1
(12.5)
Left 33 48.48 6.06 -0.2
(12.0)
Total 426 20.66 14.32 7.8
(11.7)
Table A22.18. [D.16] Ernst, 1988: Data for Healthy Elderly Australian Volunteers
Dominant/
.. Right- Dominant Nondominant Nondominant
n Age Education Handed Hand Hand Ratio
Total 85 70.0 10.4 99 1.1
(2.6) (3.1) (0.2)
Males 39 41.7 38.5
(6.2) (5.1)
Females 46 26.9 23.1
(4.0) (4.9)
Table A22.19. [D.18] Russell and Starkey, 1993: Data for a Sample of Veterans Administration
Patients Without Central Nervous System Pathology, Stratified by Gender
Gender n Age Education Race· Hand Hand
Males 175 45.0 12.5 164W 45.4 41.9
(12.9) (2.8) 118 (10.6) (10.3)
ON
Females 24 40.7 14.5 23W 27.1 25.5
(15.3) (2.6) 18 (8.1) (8.1)
•w, white, 8, black, N, other.

Material from the manual fur the Halstetul-Bussell Neuropsychologfcal Evaluaflon System-ReWed
(HRNES-R) copyright @ 1993, 2001 by Westero Psychological Services. Reprinted by permission of the
publisher, Western Psychological Services, 12031 Wilshire Boulevard, Los Angeles, California, 90025,
U.S.A. Not to be reprinted in whole or in part fur any additional purpose without the expressed, written
permission of the publisher. All rights reserved.
944 APPENDIX 22
Table A22.20. [D.19] Tremont et al., 1998: Data for Dominant Hand in Patients Referred for
Neuropsychological Evaluation who Were Determined to be Neurologically Normal
Grip Strength
Performance Group n Age Education FSIQ Male Female
Below average 35 34.03 11.53 84.89 44.54 27.49

(13.8) (2.76) (4.84) (ll.45) (7.03)
Average 84 40.55 12.62 99.15 43.92 26.37
(16.73) (2.76) (8.05) (10.51) (7.64)
Above average 38 41.71 15.63 119.92 50.31 26.65
(14.65) (3.37) (7.55) (10.10) (5.29)
Table A22.21. [D.20] Dikmen et al., 1999: Test-Retest Data for Normal and Neurologically Stable Adults•
Dominant Hand Nondominant Hand
MIF WAIS Test-Retest
n Age Education Ratio FSIQt Interval Tune 1 Tune2 11me 1 11me2
384 34.2 12.1 66134% 108.8 9.1 43.34 42.41 40.60 39.65
(16.7) (2.6) (12.3) (3.0) (13.33) (13.44) (12.89) (13.19)
•A number of participants had preexisting conditions that might affect test performance, the most significant being aleohol
abuse and a significant traumatic brain Injury.
twAIS FSIQ, Wechsler Adult Intelligence Scale full-scale IQ. (Wechsler, 1955).
Table A22.22. [D.21] Triggs et al., 2000: Data for Left-Handed and
Right-Handed Healthy Volunteers•
Left-Handers Left-Banders
(n=30) (n=30)
Left Hand Right Hand Left Hand Right Hand
23 22 22 23
(6) (6) (6) (6)
•Mean age 37 (9) years, equal number of males and females in two groups.
APPENDIX 22 945
Table A22.23. [D.22] Christensen et al., 2001: Data for a Sample of Healthy Australian
Elderly, Expressed as Mean of the Scores for Two Hands Averaged over Four Trials for
Four Age Groups for Males and Females Separately
Grip Strength
Age Group Age n %Males Education Male Female
70-74 79.82 199 52 11.78 29.7 15.3
(6.3) (5.2)
75-79 84.46 120 43 11.34 25.5 13.9
(7.3) (5.0)
80-84 88.04 41 37 11.00 21.5 12.6
(8.2) (4.0)
2';85 93.71 14 50 10.79 21.5 9.7
(4.4) (3.1)
for the GriJP Strength Test (Hand
Dynamom~ter)
Table A22m.1. B.esults of the Meta-IWIIIIIYlilli and Predicted Scores for the
Hand DyaamOJDter Test: Males, Do t Hand
Relevant values are we ted on the
Description or the agrepte sample;
Ooly those studies reporting data for males aJtd females separately were included
in the analyses. I
Number of studies Included In the aaalysU 9
Yean of pablbtion · 1974-1993
Number of data points used In the .....,.. 15
(a data point denotes a study or a cell ·
in educatkmlpder-stratified data)
Total number el puticiplmts 713
Stltttpk . .
Mean 15 36.59 39.35 11-178
Age
Mean 15 41.2q 15.62 22.57-70
SD 13 4.08 4.10 1.3-17.2
:
Eclueadota
Mean 10 13.o3 2.43 10.4-16.7
SD 8 2.66 0.91 1-3.6
r..,_,.._
Combined mean 15 49.~ 4.34 41.7-56.8
Combined SD 15 8.e:t 2.72 5.3-13.4
"Number of data points differs for chfferent ~ due to missing data.

'Weighted means and SDs.
946
APPENDIX 22M 947
Predicted grip strength in kilograms averaged over trials and SDs

per age group• (Dynamometer, males, dominant hand)
95%CI

Bange Score Band Band
Jlf-J9 suo 51.21 54.60

30-34 51.70 50.22 53.19
35-39 50.50 49.20 51.81
40-44 49.30 48.13 50.48
48.10 46.99 49.22 Standard chmatioo for all age groups
4lJ-49
is 8.95.
IJO....S4 46.90 45.78 48.03
SlJ-59 45.70 44.49 46.92
60-64 44.51 43.14 45.87
~ 43.31 41.74 44.87

Predkted tat KOrW = 59.50402 - 0.2399832 • age
SiguJflcance tests for regression with the test scores

R2 0.747
F<dfJ• p F< t,s) = 62.09, p < 0.000
Age - 0.2399832 0.030 -7.88 0.000 -0.310 to-0.170

Constant 59.50402 1.597 37.25 0.000 55.82 to 63.19
Prediction
SEe 0.68
95%CI 46.77-49.44
(continued)
948 APPENDIX 22M
eo
0
55 0
50
~ ~-----.------.-----.------.-----.------,-
20 30 50 eo eo
•
Fipre A22m.1. A scatterplot illustrating the dispersion of the data points around the regressioo line for the
Dynamometer, Male, Dominant Hand. The size of the bubbles reflects the weight of the data point. with
Tests for assumpticms and moclel8t
Tests for heterogeneity In the 8aal data set

Q(dl).p Q(l4) =177.86, p < 0.000
Tests for model flt_.Liftfon of a quadratic term
Model BIC BIC'
Unear 0.747 0.728 29.699 -17.924

Quadratic 0.752 0.711 32.125 -15.498
Tests for panmeter speeifleatioas
Sbapiro-Wdk W test W = 0.953, p = 0.574
Homoscedasticity
Signiflcance tests for regressicm with the SDs
A regression of SDs on age yielded an B_2 of 0.076 (Fu.s> = 1.36, p = 0.278). Therefore, the SD for the
aggregate sample is suggested for use with all age groups.
EtTects of demographic variables
EdueatiOD
Education did not contribute to grip strength beyond its inverse association with age. There was a negligible
lmpRM!mentof0.0008intheB_iwithadditicmoftheeducationtermtotbeabowregressicmmodel(forlOdata
points that report education). Similarly, significance tests for educaticm yielded t =0.30, p =0.77. Therefore,
further analyses were not performed.
APPENDIX 22M 949
Hand Dynamomter Test: Males, Nondominant Hand
Only those studies reporting data for males and females separately were included in the analyses
Number of studies included in the aaalysis 7
Yean of publieatioa 1974-1993
Number of data points used in the aaalysis 13
Total number of participaats 641
s-p~.e.a.
Mean 13 35.59 42.45 11-178

Age
Mean 13 40.91 15.33 22.6-70
SD 12 3.38 3.78 1.3-17.2
Educadon
Mean 8 13.66 2.48 10.4-16.7
SD 7 2.52 0.91 1.0-3.6
TM ICGre tneCitll
Combined mean 13 47.16 4.25 38.5-53.6
Combined SD 13 8.39 2.17 5.1-12.2
Predicted grip strength in kilograms averaged over trials and SDs per age group•
(Dynamometer, males, nondominant band)
95%CI
A&• Predleted Lower Upper

JS-19 50.1! 47.86 52.37

30-34 49.0! 46.93 51.11
35-39 47.91 45.94 49.89 Standard deviation for all age groups is 8.39.
40-44 46.81 44.90 48.72
45-49 45.71 43.81 47.61
IJO...S4 44.61 42.65 46.56
IJS.-.69 43.50 41.43 45.57
60-64 42.40 40.17 44.63
65-69 41.30 38.87 43.72

Predieted tat 1eore =56.18508- 0.2205636 • age
(continued)
950 APPENDIX 22M
55
0 0
50
40
0
35
20 30 40 50 eo 70 eo
• \
· rsion of the data points around the regression line for the
Fip'e A22m.2. A scatterplot illustrating thr:··
Dyoamometer, Male, Nondominant Hand. size of the bubbles reflects the weight of the data point, with
larger bubbles indicating larger standard e , smaller weight.
I
I
Sigaiflcance tests for regressicm with!the test scores
I
I
Ordmary least-squares regression o£
Number of observations
tft
meaDS OD age (linear)
I 13
Number of clusters , 7
R2 lI 0.630
F<dO· p Fo.s> = f.l5, p < 0.0007
Age -0.2205636 0.034 i -6.34 0.001 - 0.306 to - 0.135

Constant 56J8508 1.852 . 30.33 0.000 51.65 to 60.72
Precliction
Predicted age range
Mean predicted score 25-69ye*
45.71 (3. )
SE, 1.07
95%CI 43.62-47.$o
I
I
Tests for assumptioos and model 8t !
Tests for heterogeneity in the 8aal ~ aet
Pooled estimates for 6xed effect ! 45.173
Pooled estimates for random effect I 46.825
Q(dOo p Q(l2) = 211.33, p < 0.000
Tests for model8t-ddjtion of a + t e r m
Model Adjusted :a 2 BIC BIC'
Linear 0.630 0.5971 32.373 -10.370

Quadratic 0.642 0.571\ 34.505 -8.237
BIC' difference of 2.132 provides positive supfort for the linear model.
I
!
APPENDIX 22M 951
Tests for parameter speei8eations

Shapiro-Wilk W test W=0.962, p=0.787
Homoscedasticity
A regression of SDs on age yielded an R2 of 0.180 (F(l.&J = 7.18, p =0.037). Therefore,

Efl'eets of demographic variables
Education
Education did not contribute to grip strength beyond its inverse association with age.
There was a small improvement of0.038 in the R2 with addition of the education term
to the above regression model (for eight data points that report information on
education). Similarly, significance tests for education yielded t = 1.32, p = 0.24.
Therefore, further analyses were not performed.
Hand Dynamomter Test: Females, Dominant Hand
Only those studies reporting data for males and females separately were included in the
analyses
Yean oE publieatioa 1979-1988
Total number oE partieipaots 454
Satrtpk.U.
Mean 11 28.22 33.14 13-187
Age
Mean 11 40.16 15.96 22.83-70
SD 10 2.85 2.83 1.3-16.1
EtlucGtion
Mean 6 14.48 1.95 10.4-16.2
SD 5 2.33 0.31 2.0-3.1
rataeoremetlftl
Combined mean 11 31.47 2.45 26.9-34.0
Combined SD 11 6.74 2.03 3.6-10.'1
(continued)
952 APPENDIX 22M
Predicted grip strength in ldlograms averaged over bials and SDs per age group•
(I>yuamometer, females, dominant band)
95%CI

25-!9 3lUl 31.76 33.86

30-34 3lU7 32.15 32.99
35-39 32.22 31.01 33.44
40-44 31.76 29.99 33.53 Stanclanl cJ.matloa for all age groups is 6.74.
4lJ-49 31.19 29.18 33.19
50-S4 30.50 28.60 32.40
55-SIJ 29.70 28.23 31.17
~ 28.79 28.03 29.54
65-69 27.76 27.09 28.43

Predicted teat aeore = 32.0932 + 0.087928 • age - 0.0022535 • age2
Sigoificance tests for regression with the test scores
Onliaary least-squares regression of test means on age (quadratic)

R2 0.673
F<dl)• p F<2.4l = 151.82, p < 0.0002
Age 0.087928 0.336 0.26° 0.807" - 0.845 to 1.021

Age squared -0.0022535 0.003 -0.65 0.552 - 0.012 to 0.007
Constant 32.0932 7.128 4.50 0.011 12.30 to 51.88
"Significance test for age centered (sample means- aggregate mean): t = -1.61, p =0.182.
Predietioo
SEe 0.64
95%CI 29.56-32.06
Tests for assumptions and model flt
Tests for heterogeneity in the 8nal dataset

Q<dl)• P Qoo> = 92.11, p < 0.000
APPENDIX 22M 953
34
0
co
32
30
21
21 ~----~----~-----r-----r-----r----~
80 80
20 30
•
FIIPft A22m.3.A sc:atterplot illustrating the dispersion of the data points around the regression line for the
DyDamometer, Female, DomiDant Hand. 'l1le size of the bubbles reflects the weight of the data point, with
larger bubbles indicating larger standard error, smaller weight
Model BIC BIC'

Unear 0.639 0.599 17.097 -8.820
Quadratic 0.673 0.591 18.413 -7.504
BIC' difference of 1.317 provides weak support for the linear modeL
Testarar,...........~
Sbapiro-WIIk W test W=0.888, p=0.131
Homoscedastidty
White's general test 6.1915, p=0.185
Slpiflcance teats for regreuion with the SDs
A regression of SDs on age yielded an It'- of 0.166 (Fu.41 = 4.30, p = 0.107). Therefore, the SD for the aggre-
gate sample Is suggested for use with all age groups.
Edaeatloa
Education did not contribute to grip strength beyond its inverse asiiOCiltkm with
age. '11lere was a negligible impnwement of 0.00071n the It'- with addition of the
education term to the above regression model (for six data-points that report
education). SUnilarly, sign!&cance tests for educalloa yielded t=0.06, p=0.96.
'11lerefore, fUrther analyses were not performed.
954 APPENDIX 22M
Table A22m.4. Results of the Meta-.AnJ).ysis and Predicted Scores for the
Hand Dynamomter Test: Females, Nondominant Hand
(Relevant values are weighted on the s!fDdard error for the test mean)
Only those studies reporting data for males ~d females separately were included in the analyses.
Number of studies included in the aaalysis 4

Number of data points used in the all!dysis 10
Variable n• ~ sot Range
SGtnplelhe
Mean 10 2&86 36.58 13-187
Age
Mean 10 4t97 16.77 22.8-70
SD 10 2;93 3.04 1.3-16.1
EductJtion
Mean 5 14;41 2.12 10.4-16.2
SD 5 9;37 0.36 2.0-3.1
r•acore-
Combined mean 10 2~92 2.96 23.1-31.1
Combined SD 10 ~44 1.00 3.8-6.6
•Number of data points differs for different knalyses due to missing data.
Predicted grip strength in ldlosr:m1

averaged over trials and SDs
per age group• (Dyllamometer, fedaies, nondominant band)
I
95%CI
Age Predieted l..a6rer Upper

Bange Seore B~d Band
J5-J9 !9.69 28.35 31.02

30-34 !9.65 29.!0 30.11
35-39 !9.41 28lls 30.13
40-44 28.95 27:14 30.16
Standard deviation for all age groups is 5.44.
4S-49 28.28 26.80 29.75
50-S4 27.39 25,00 28.88
55-59 26.!9 25,()3 27.54
60-64 !4.97 24,17 25.77
65-69 !3.44 22$)7 23.92

PretJict.d ,_, acore = 26.0349 + 0.2504831• age - 0.00428 • age2
APPENDIX 22M 955
0
30 0 0 0
25
~ ~-----.----~-----.r-----r-----~----,-
30 40 80 80
•
fi&ure A22m.4. A scatterplot iJlustrating the dispersion of the data points around the regression for the DyDamometer,
Female, Nondomlnant Hand. '11le size of the bubbles reflects the weight of the data point. with larger bubbles indicating
larger standard error, smaDer weight
Sipfflcanee tests for regrellioa. with the test scores
Onliaary leMt ........ repellioa ol telt - Oil . . (qudlldic)

R2 0.833
F<df)• p F<u> =157.44, p < 0.0009
Term Coef&cient SE t 95,_, Cl
Agel 0.2504831 0.270 0.93• 0.422• - 0.609 to 1.109

Age -0.00428 0.003 -1.59 0.209 - 0.013 to 0.004
Constant 26.0349 6.046 4.31 0.023 6.79 to 45.28
•stgn•ficanc:e test for age centered (sample means- aggregate mean): t = - 2.38, p =0.098.
Predletioa
Predicted age range 25-69yean
SEe 0.52
95,_, Cl 26.54-18.59
Tests for usumptioal and moclel&t
Tells far heteropaeltr fa the flaal data eet

Pooled estimates for 8xed eft'ect 27.064
Pooled estimates for random eft'ect 27.544
Q(df). p Q(9) = 87.71, p < 0.000
(contmued)
956 APPENDIX 22M
Tests for model&t~ of a q~ term
Model BIC BIC'
Linear 0.739 0.706 17.141 -11.132

Quadratic 0.833 0.785 14.993 -13.281
BIC' difference of 2.148 provides positive su~rt for the quadratic model.
Tests for parameter speei8catioas
Normality of the residuals '
Shapiro-Wille W test w=0.~. p =0.835
Homoscedasticity
White's general test s.197,r=o.268
Sigoifleance tests for regression with jthe SDs

I
A regression of SDs on age yielded an R2 of4229 (Fu.a)=4.23, p=0.132). Therefore, the

SO for the aggregate sample is suggested for ~e with all age groups.
!'
Education
Education did not contribute to grip ~ beyond its inverse association with
age. There was a small improvement of 0.48& in the R2 with addition of the
education term to the above regressio:f.ml (for five data points that report
information on education). Similarly, si · ce tests for education yielded
t = 1.40, p =0.30. Therefore, further were not performed.
Appendix 23: locator and Data Tables for
the Grooved Pegboard Test (GPT)
Table A23.1. Locator Table for the Grooved Pegboard Test (GPT)
Study Age· n Sample Composition IQ/ Education• I..ncation
GPI'.l Rounsaville 24.9 29 CETA workers, SK male 11.2 Massachusetts

et al.• 1982
page 462
TableA23.2
GPI'.S Bornstein, 20-39 365 178 M, 187 F; paid 12.3 Western
1985 40-59 volunteers free of neurological (2.7) Canada
page 462 60-69 or psycbiatric illness; <HS
Tables A23.3, A23.4 data are presented by age x ~HS
education x gender
GPI'.3 Heaton et al., 32.7 100 79 M, 21 F; controls with no 14.15 Denver, CO
1985 (13.5) neurological illness, head trauma, (2.84)
page 463 or substance abuse
TableA23.5
GPI'.4 Heaton et al., 15-81 553 356M, 197 F; normal subjects with no 0-20 Colorado,
1986 39.3 history of neurological illness, 13.3 California,
page 463 (17.5) head trauma, or substance abuse; (3.4) Wisconsin
TableA23.6 <40 7.2% left-handed; data are <12
40-59 presented in 3 age and 3 12-15
~60 education groups ~16
(continued)
957
958 APPENDIX 23

GPr.5 Bomstein, 18-39 365 178 M, 187 F; paid volunteers free 12.3 Western
1986a 40-59 of neurological or psychiatric (2.7) Canada
page 463 60-69 illn~; 91.5% right-handed; data are <HS
Tables A2:3.7, A2:3.8 stratiled by age x education x gender; ;:::HS
proportion of participants classified
as ~ is presented
GPr.6 Polubinski &: 18-24 120 Underg,aduate students (60 M, 60 F); College Ohio
Melamed, 1986 data tre partitioned by firm and students
page464 mixe4-handedness groups
Table A2:3.9
GPr.7 Ryan et al., 21-30 55 Blue-<dlar workers with no history 12.3 (1.4) Eastern
1987 31-40 45 of ~c exposure; all males; 11.9 (2.2) Pennsylvania
page 464 41-50 44 data are partitioned by 4 age 11.3 (1.8)
Table A2:3.10 51-59 38 groups 11.0 (1.8)
GPr.8 Bomstein 32.3 23 Volunteers from a university VIQ: 105.8
et al., 1987a (10.3) comiJunity (9 M, 14 F); (10.8)
page465 test-~st data over 3- PIQ: 105.0
Table M:l.ll week. period are presented (10.5)
GPr.9 Thompson 40.59 426 279M, 147 F; normal subjects; 13.15
et al., 1987 (18.27) percent falling in (3.49)
page 465 late~ dysfunction
Table A2:3.12 rang~ is presented
GPr.IO Bomstein &: ~70 134 49 M, !J5 F; paid volunteers screened 11.7 Western
Suga, 1988 62.7 for a fllstory of neurological or (2.9) Canada
page 466 (4.3) psycltiatric disorders;
Table A2:3.13 clivi~ into 3 education groups:
17M.~F Range ~10, M8.5
16M, S8 F Range 11-12, M11.7
16M, i8 F Range> 12, M15.0
GPr.ll Miller et al., 21-72 Homosexual/bisexual men MACS centers at
1990 Baltimore,
page 466 37.20 769 HIV-1-seronegative 16.36 Chicago,
Tables A2:3.14, A2:3.15 (7.52) (2.34) Los Angeles,
35.66 727 HIV-1-ieropositive, asymptomatic 15.70 &: Pittsburgh
(6.47) (2.44)
36.90 84 HIV-1-seropositive, symptomatic 16.06
(7.04) (2.50)
GPr.l2 Heaton et al., 42.1 486 Urban tnd rural volunteers; data 13.6 California,
1991,2004 (16.8) colleCted over 15 years through (3.5) Washington,
page 467 multicenter collaborative efforts; FSIQ: Texas,
Data are not groups: strict exclusion criteria; 65% M; 113.8 Oklahoma,
reproduced 20-34 data are presented in T-score (12.3) Wisconsin,
in this book ~9 equNalents for M and F separately Groups: Illinois,
40-44 in 10 age groupings by 6 6-8 Michigan,
45-49 edu<:Jtion groupings; 9-11 New York.
50-S4 in 2004 edition, age range is 12 V'uginia.
55-S9 expaaded to 85 years and data 13-15 Massachusetts,
60-64 are presented for African-American 16-17 Canada
~ and Caucasian participants ;::::18
70-74 sepaStely
7~
APPENDIX 23 959

Study Age· n Sample Composition IQ/ Education• Location
GPT.l3 Seines et al., 25-34 733 HOIIlCISt!lalall men, HIV-1- <CoDege MACScenters
1991 35-H seronegative. stratified by age. College at Baltimore,
page 467 45-54 education, and age x education >College Chicago,
Tables A23.16-A23.18 Los Angeles, ~
Pittsburgh
GPT.l4 Ruff~ 16-70 360 Normal volunteers screened for 7-22 California,
Parker, 1993 psychiatric hospitalization, Michigan,
page 468 16-39 chronic poly-drug abuse, or :S12 eastern seaboard
Tables A23.19, A23.20 40-54 neurological disotders; ~13
~70 data are stratified by age x
education X gender; data for
left hand-dominant sample are
also presented
GPT.IS Russelllk 45.5 113 Norms are collected from standardization 12.8 Cincinnati,
Starkey, 1993 (14.1) sample for the HRNES (2.9) Miami
page 468 manual; controls are V.A. patients
Table A23.21 without CNS pathology (95 M, 18 F)
GPT.l8 Dikmen 43.6 121 Healthy adults; 68,. M; data on 12.0 Washington,
et al., 1999 (19.6) test-retest reliabilities and (3.3) Colorado,
page469 practice effect are provided California
Table A23.22
GPT.l7 Strenge 19-30 49 Medical students (23 M, 26 F) Germany
et al., 2002
page 469
Table A23.23
•Age column and IQ/educatlon column contain information regarding range and/or mean and standard deviation for the
Table A23.2. [GFI'.l] Rounsaville et al., 1982: Data for a Sample of

Comprehensive Employment Training Act (CETA) Workers
,. Right- Dominant Nondominant
n ,. Male Education Handed Age Hand Hand
29 59 11.2 90 24.9 70.52 75.59
960 APPENDIX 23
Table A23.3. [GPT.2a] Bornstein, 1985: Data for a Sample of Healthy Canadian Adults
Number Number of Preferred Nonpreferred

Age Education" of Males Females Hand Hand
S,age
20-39 13.0 107 64 60.9 66.2
(2.3) (16.2) (17.1)
40-59 11.9 31 66 68.6 74.2
(2.8) (15.0) (15.7)
60-69 11.8 40 57 75.5 83.1
(2.9) (14.6) (15.5)
By education
48.5 <HS 51 57 72.4 78.1
(16.6) (17.7) (19.1)
41.1 2:HS 127 130 64.2 70.2
(16.8) (15.5) (16.6)
By gend.r
39.2 12.4 178 Males 68.7 74.5
(17.2) (2.9) (20.8) (21.3)
47.3 12.2 187 Females 64.6 70.8
(16.1) (2.5) (10.8) (13.3)
"HS, high school.
Table A23.4. [GPT.2b] Bornstein, 1985: Data Presented in Age x Education (<High School, ~High
School) x Gender Groupings
Male Female
<HS 2:HS <HS ~s
Age Groups n M (SD) n M (SD) n M (SD) n M (SD)
Preferred lumd
20-39 21 65.3 85 62.1 13 60.4 49 57.2
(8.5) (20.8) (6.4) (9.6)
40-59 13 86.8 17 69.5 22 66.5 43 63.9
(30.1) (11.0) (7.0) (7.1)
60-69 16 84.8 23 75.7 22 75.4 34 70.9
(22.3) (14.4) (12.2) (9.2)
Nonptefo•• ed lttmtl
20-39 71.3 67.6 64.1 62.2
(12.2) (20.9) (9.2) (11.8)
40-59 91.2 74.2 71.1 70.6
(30.0) (12.6) (8.5) (9.5)
60-69 93.6 81.9 82.0 79.6
(21.7) (13.3) (14.0) (12.6)
APPENDIX 23 961
Table A23.5. [GPI'.3] Heaton et al., 1985: Data

for a Control Sample
11me for
MIF Both
n Age Education Ratio Hands
100 32.7 14.15 7M1 131.40
(13.5) (2.84) (18.76)
Table A23.6. [GPI'.4] Heaton et al., 1986: Data for a Sample of Healthy Adults
Stratified by Age and Education•
DomiDant Nondominant
n Hand Hand
Age groupe
15-40 319 61.1 65.7
40-59 134 68.1 74.7
60-81 100 85.1 90.0
Etlucadola groupe
~12 132 74.6 79.3
12-15 249 66.0 71.3
U~20 172 62.3 67.6
•Mean education 13.3 (3.4).
Table A23.7. [GPI'.5a] Bornstein, 1986a: Proportion of Participants Classified as

Impaired (Exceeding a Criterion of 66 Seconds) in a Sample of Healthy Canadian
Adults (n =365)
% CJassifled
Mean Median Mode as Impaired
Preferred 66.6 65 65 40.8

band (seconds)
Nonpreferred 72.6 70 75 61.5
band (seconds)
Table A23.8. [GPI'.Sb] Bornstein, 1986a: Proportion of Participants Classi6ed as Impaired in a Sample
Preferred Hand Nonpreferred Hand
Males Females Males Females
Age <HS• ?;HS <HS ?;HS <HS ?;HS <HS ?;HS
18-39 25 19.5 23.1 13 60 47.1 38.5 26.1

(5120) (17187) (3113) (6146) (W20) (41187) (5113) (1~46)
~ 76.9 75 55.5 27.3 84.6 75 59.1 65.9

(10113) (W16) (12122) (1~44) (11113) (1~16) (13122) (29144)
60-69 81.2 82.6 77.3 57.6 100 87 95.5 84.8
(13116) (19123) (17122) (19.133) (16116) (20.'23) (21.122) (28133)
•as. high school.

962 APPENDIX 23
Table A23.9. [GPT.6] Polubinski. and ~elamed, Table A23.11. [GPT.8] Bornstein et al., 1987a:
1986: Data for Undergraduate Students ijlrtitiooed Test-Retest Data for 23 Healthy Participants (9
into Firm and Mixed Right-handedness Croup Men, 14 Women) Between 17 and 52 Years of Age
Rip)t Left Dominant Nondominant
Handedness n Age Education Haad Hand Hand Hand
Men Test 1 56.6 59.3

(5.9) (6.6)
Firm 30 19.7 13.3 60.1 67.1
(1.4) (0.7) (7.6) (9.8) Test 2 (3 weeks later) 58.8 58.8
(8.9) (6.6)
Mixed 30 20.1 13.6 55.3 60.7
(1.6) (0.9) (5.9) (5.9)
w-
Firm 38 19.8 13.4 54.~ 60.2
(1.2) (0.7) (6.2) (6.2)
Mixed 22 19.4 13.3 54.9 60.1
(0.8) (0.6) (6.6) (7.3)
Table A23.10. [GPT.7] Ryan et al., 1987: Data for

Blue-Collar Workers (All Males) without t History
of Exposure to Industrial Toxins
Age Dominant Noncbninant
Group Age Education n Hand Hind
21-30 26.1 12.3 55 69.7 7..5

(2.3) (1.4) (11.5) (10.9)
31-40 36.8 11.9 45 67.2 72.8
(2.7) (2.2) (10.2) (119)
41-50 45.7 11.3 44 76.1 71.1
(2.9) (1.8) (11.9) (Q.7)
51...59 54.8 11.0 38 78.7 8L2
(2.8) (1.8) (13.0) (12.9)
Table A23.12. [GPT.9] Thompson et al., ~987: Percent of Healthy Participants

Scoring in the LateraJized Dysfunction ~ge
lntermanual
Dominant Nondominant Percent
Hemisphere Hemisphere Difference
Groups n Dysfunction Dysfunction Scores
All right 167 20.96 14.97 -8.2

(11.3)
Mixed right 226 25.89 18.75 -7.9
(12.8)
Left 33 36.36 6.06 -4.4
(10.8)
Total 426 24.76 16.27 -7.8
(12.1)
APPENDIX 23 963
Table A23.13. [GPT.lO] Bomstein and Suga, 1988: Data for Healthy
Canadian Volunteers Between 55 and 70 Years of Age

Education Age of Males of Females Hand Hand
~10 62.3 17 29 78.5 83.1

(19.9) (17.2)
11-12 62.9 16 28 74.2 84.2
(16.0) (20.8)
<12 63.0 16 28 71.9 77.7
(14.2) (13.7)
Table A23.14. [GPT.lla] Miller et al., 1990: Demographic Characteristics for the Sample of Homosexual!
bisexual Males Participating in the Multi-Center AIDS Cohort Study
%Handedness Race
CES Depression
n Right Ambidextrous Left White Black Hispanic Other Scale CD4
Seronegative 769 87 0 13 92 2 4 2 9.08 970,42

(9.03) (332.46)
Asymptomatic, 727 86 0 14 91 2 6 2 9.44 561.90
seropositive (9.27) (277.98)
Symptomatic, 84 90 1 8 90 2 5 3 15.21 277.22
seropositive (11.19) (269.45)
Table A23.15. [GPT.llb] Miller et al., 1990: Data for the Multi-
Center AIDS Cohort Study (All Males)
Seronegative 769 37.20 16.36 64.28 69.28

(7.52) (2.34) (9.10) (9.91)
Asymptomatic, 727 35.66 15.70 63.39 68.90
seropositive (6.47) (2.44) (9.17) (12.52)
Symptomatic, 84 36.90 16.06 66.57 73.27
seropositive (7.04) (2.50) (11.42) (16.39)
964 APPENDIX 23
Table A23.16. [GPT.l3a] Seines et al.,!l991: Demographic Characteristics

for the Seronegative Homosexual!BiJext;d Males Participating in the MACS
Smdy :
I
Handedness ; Race(%)
African
Right Ambidextrous i Left Caucasian American
Br1 age
25-34 84.8 0.3 14.9 96.4 3.6
35-44 87.2 1.1 11.7 96.6 3.4
45-54 86.6 2.1 11.3 95.9 4.1
Brl.,._,_
<College 87.8 0.4 11.8 94.8 5.2
College 85.2 0.0 14.8 96.0 4.0
>CoDege 85.8 1.3 12.0 96.7 3.3
I
Table A23.17. [GPT.l3b] Seines et al., 1991: Data Stratified by Age and Education
Dominant Nondomiuant
Hand Hand
Percentile Percentile
Age II Mean Age Educa~ Mean (SD) 5th lOth Mean (SD) 5th lOth
By age
25-34 a 31.0 16.1 62.0 76 72.5 67.0 85 80.5
(2.6) (2.2)" (7.8) (9.3)
35-44 290 39.3 16.4 : 64.4 78 75 69.2 85 82
(2.9) (2.3) 1 (8.1) (9.1)
45-54 9T 48.5 16.7 I 67.9 85 80 73.7 90 86
(2.6) (2.6) i
I
(9.0) (11.1)
Brl educ:adcm
<College 229 36.1 13.7 I 64.1 77 74 69.6 89 84
(7.4) (1.2) : (8.5) (10.3)
College 201 35.6 16.0 1 64.0 79 75 68.4 87 83
(7.2) (0.0) i (8.7) (10.3)
>College 301 38.4 18.6 I 63.4 80 75 69.0 85 81
(7.8) (1.3)! (8.3) (9.1)
APPENDIX 23 965
Table A23.18. [GPT.13c] Seines et al., 1991: Data Stratified by Age x Education (Personal
Communication)
Percentile Percentile
n Dominant Hand 5% 10% Nondominant Hand 5% 10%
<College
25--34 107 62.4 76 73 69.1 89 83
(7.7) (10.4)
35--44 93 65.3 78 75 68.5 84 82
(8.4) (8.2)
45-60 42 65.7 81 77 73.0 93 90
(10.0) (12.2)
College
25--34 104 62.4 76 75 65.6 83 74
(8.4) (8.1)
35--44 77 65.2 80 77 70.4 92 87
(8.2) (10.5)
45-60 35 67.1 85 80 73.4 93 86
(10.0) (13.5)
>College
25--34 111 61.1 76 71 66.5 84 80
(7.6) (8.7)
35--44 150 63.4 78 74 68.8 84 80
(8.4) (9.1)
45-60 64 66.8 84 79.5 72.3 85 81
(8.7) (8.3)
Table A23.19. [GPT.14a] Ruff and Parker, 1993:

Data for the Left Hand-Dominant Sample of
Healthy Participants
Men Women
n 17 18
Age 37.9 38.7
(18.0) (16.1)
Education 13.7 14.1
(2.8) (2.6)
(13.5) (11.6)
Nondominant hand 70.3 73.0
(15.7) (18.6)
966 APPENDIX 23
Table A23.20. [GPT.14b] Ruff and P!p'ker, 1993: Data for a Sample of Healthy Adults Stratified by
Gender x Education x Age
Males Females Combined
Age/Education n :M (SD) n M (SD) n M (SD)
.DolllinGnt ,.,.
16-39
~12 29 67.8 30 62.8 59 65.3
(9.2) (8.9) (9.3)
~13 60 ti4.7 60 57.8 120 61.2
'(10.9) (6.2) (9.5)
All 89 65.7 90 59.5 179 62.5
<10.4) (7.5) (9.6)
40-S4
~12 15 ~1.9 14 63.1 29 67.7
p5.1) (4.4) (12.0)
~13 30 J0.4 30 63.3 60 66.8
~10.9) (7.4) (9.9)
All 45 ~0.9 44 63.2 89 67.1
~12.3) (6.5) (10.6)
55-10
~12 15 ~.7 15 78.6 30 81.1
k10.2) (11.7) (11.1)
~13 30 ~4.1 29 75.3 59 74.7
,13.0) (11.3) (12.1)
All 45 '77.3 44 76.5 89 76.9
i12.8) (11.4) (12.1)
All age,..,.,.
~12 59 72.9 59 66.9 118 69.9
(12.8) (11.2) (12.3)
~13 120 ~.5 119 63.4 239 66.0
{12.0) (10.7) (11.6)
All 179 69.9 178 64.6 357 67.3
(12.5) (10.9) (12.0)
Nondomitlcane ,_,
16-39
~12 29 74.5 29 66.8 58 70.7
(10.9) (10.7) (11.4)
~13 59 67.8 60 65.2 119 66.5
(10.8) (10.3) (10.6)
All 88 '70.0 89 65.7 177 67.9
(11.2) (10.4) (11.0)
40-S4
~12 15 ~9.1 14 69.6 29 74.5
J14.9) (6.5) (12.4)
~13 30 73.7 30 70.8 60 72.3
l9.9) (8.9) (9.4)
All 45 '15.1 44 70.4 89 73.0
J11.9) (6.5) (10.5)
APPENDIX 23 967

Males Females Combined
Age/Education n M (SD) n M (SD) n M (SD)
ISIS-10
~12 15 91.0 13 84.3 28 87.9
(12.7) (15.3) (14.1)
~13 28 83.5 29 82.0 59 82.8
(13.4) (12.5) (12.9)
All 43 86.1 42 82.8 85 84.5
(13.5) (13.3) (13.4)
All age lewJ.
~12 59 79.9 56 71.6 115 75.8
(14.0) (13.1) (14.1)
~13 117 73.1 119 70.7 238 71.9
(12.9) (12.5) (12.7)
All 176 75.4 175 71.0 351 73.2
(13.6) (12.7) (13.3)
Table A23.21., [GPI'.15] Russell and Starkey, 1993: Data for a Sample of Veterans
Administration Patients without Central Nervous System Pathology
n Age Education Race" Gender Hand Hand
113 45.5 12.8 106W 95M 74.4 78.4

(14.1) (2.9) 7B 18 F (24.4) (25.9)
•w, white: B, black.

Material from the manual for the Halsteod-IWssell Neuropsychological Eooluation System-Revised
(HRNES-R) copyright © 1993, 2001 by Western Psychological Services. Reprinted by permission of the
publisher, Western Psychological Services, 12031 Wdsbire Boulevard, Los Angeles, California, 90025,
U.S.A. Not to be reprinted in whole or in part for any additional purpose without the expressed, written
permission of the publisher. All rights reserved.
Table A23.22., [GPI'.16] Dikmen et al., 1999: Test-Retest Data for Healthy Adults•
Dominant Hand Nondominant Hand
WAIS Test-Retest
n Age Education MIF Ratio FSIQt Interval Time1 Trme 2 Time 1 Time2
121 43.6 12.0 68132% 108.8 5.4 69.66 68.68 75.80 73.70
(19.6) (3.3) (12.3) (2.5) (19.27) (21.04) (21.56) (19.69)
"Demographic information is provided for a larger sample of 125 participants.

tThe mean Wechsler Adult Intelligence Scale full-scale IQ (Wechsler, 1955) is reported for three groups used in this study
combined.
968 APPENDIX 23
Table A23.23. [GPT.l7] Strenge et al., 2002: Data for a Sample

of Medical Students•
%Right- Dominant Nondominant
Age n MIF Ratio Handed Hand Hand
24.5 49 23126 100 54.2 57.9

(2.75) (6.9) (5.6)
"Study was conducted in Germany.

Appendix 23m: Meta-Analysis Tables for
the Grooved Pegboard Test (GPT)
GPT, DomiDant Hand
Number of studies ineluded in the aaalysis 6

Variable n xt sot Range
Satnpk-
Mean 15 111.69 116.30 22-727
Age
Mean 15 38.62 16.55 19.4-65
SD 15 4.81 5.56 0.8-19.6
Education
Mean 15 13.54 1.49 11.8-16.7
SD 15 2.33 1.20 0.6-3.8
Percent melle 15 57.28 34.64 0-100
Tatacore-
Combined mean 15 64.94 7.42 54.8-76.9
Combined SD 15 11.36 4.50 5.9-19.3
"weighted means and SDs.

(continued)
969
970 APPENDIX 23M
Predicted number of seeonds to completion and SDs per age group•

(GPT, dominant hand)
959& CI 959& CI

.Rtmge Seore Band Band SD Band Band
20-U 57.95 57.32 58.58 8.3J 6.78 9.85
J5-J9 80.11 59.45 60.79 10.31 7.50 13.12
30-34 6J.J9 61.54 63.04 11.91 7.97 15.85
35-39 64.46 63.59 65.32 13.13 8.48 17.78
40-44 66.63 65.63 67.62 13.95 9.05 18.86
45-49 68.79 67.65 69.94 14.39 9.68 19.10
50-S4 70.96 69.66 72.26 14.44 10.36 18.52
55-89 73.13 71.67 74.59 14.10 10.99 17.21
60-64 75.30 73.67 76.93 13.38 11.22 15.54

Predicl.d ,_, ec:orw = 48.18889 + 0.4337963 • age
Predicl.d SD = - 5.442114 + 0. 7862791• age - 0.0077628 • ag~
Ordiaary least-squares regression o£ test meaDS OD age <li-ar)

R2 0.936
F<dl)• p Fu.s> =544.34, p < 0.000
Term Coefficient SE t p 959& CI
Age 0.4337963 0.019 .23.33 0.000 - 0.386 to 0.482

Constant 48.18889 0.509 94.58 0.000 46.88 to 49.50
Predietioa
SEe 0.54
959& CI 65.58-67.68
APPENDIX 23M 971
80
70
80 0
50
20 30 40 50 80 70 80
age
Grooved Pegboard Test, Dominant Hand. The size of the bubbles reHects the weight of the data point, with
Tests for assumptions and model&t

Q(dl)• p Q(l4)=423.67, p <0.000
Tests for model &t~n of a quadratic term
Linear 0.936 0.931 25.327 -38.426

Quadratic 0.963 0.925 27.968 -35.785
Tests for parameter speci&cations
Homoscedasticity
Ordinary least-squares regression of SDs on age (quadratic)

R2 0.489
F<df>· p F<2.s> = 22.61, p < 0.003
(continued)
972 APPENDIX 23M
Term Coefticient SE i p 95% CI
Age
Age2
0.7862791
-0.0077628
0.423
0.005 -
l86·
Jr56
0.122"
0.179
-0.300 to 1.873
-0.021 to 0.005
Constant -5.442114 6.641 -q.82 0.450 -22.51 to 11.6
•significance test for age centered (sample r+ans -aggregate mean): t =4.33, p =0.007.
Prediction I
Mean predicted SD 12.66 (2.11f
SEe
95% CI
1.81
9.11-16.20
I
!I
Etrects of demographic variables
Education !
(Analysis of the effect of education on t h t scores was performed on a separate
data set, which contained data broken by education groups.)
Regression of test means on education an age. •
Number of clusters 8 1
R2 0.8251
Education -0.685 0.252 -12.71 0.000 - 1.28 to- 0.09
"Regression with education was ran on a ~ set comprising data stratified by

education rather than by age, when available.·
Geuder
t-test by gender
n X male X female Mi-F Difference t p

4M,4F 63.500 59.725 3.775 0.841 0.216
Table A23m.2. Results of the Meta-Analysis and Predicted Scores for

the GFf, Nondominant Hand
(Relevant values are weighted on the ,dard
wrror for the test mean)
Description of the aggregate sample :
Number of studies iaeluded in the ....alym 6

Yean of pab1ieatfoo : 1985-1999
Number of data points used in the ~ 15
(a data point denotes a study or a cell r
APPE N DI X 23 M 973
Variable n xt sot Range
Sample size
Mean 15 115.13 125.04 22-727
Age
Mean 15 38.43 16.48 19.4-65
SD 15 4.83 5.59 0.8--19.6
Education
Mean 15 13.58 1.52 11.8--16.7
SD 15 2.32 1.20 0.6-3.8
Percent male 15 58.78 34.86 0-100
Test score means
Combined mean 15 70.98 7.98 60.1--84.5
Combined SD 15 12.50 4.78 5.9-21.6
Predicted number of seconds to completion and SDs per age group•

(GPT, nondominant hand)
95% CI 95% CI

Range Score Band Band SD Band Band
20-24 63.64 62.99 64.30 9 .40 7.96 10.84

25-29 65.95 65.20 66.70 11.53 8.59 14.47
30-34 68.25 67.38 69.12 13 .23 9.00 17.45
35-39 70.56 69.55 71.56 14.49 9.47 19.52
40-44 72.86 71.71 74.01 15.33 10.02 20.64
45-49 75.16 73.86 76.47 15.74 10.67 20.81
50-54 77.47 76.01 78.93 15.72 11.39 20.04
55-59 79.77 78. 15 81.39 15.26 12.11 18.42
60-64 82.08 80.30 83.86 14.38 12.40 16.36

Predicted test score = 53.27121 + 0.460912 • age
Predicted SD = - 5.48594 + 0.8551187 • age - 0.0085961 • age2

1umber of observations 15
umber of clusters 6
R2 0.907
F(dfl· p F(l.sl = 680.98, p < 0.000
(continued )
974 APPENDIX 23M
Term Coe&icient SE t p 95% CI
Age 0.460912 0.018 26.10 0.000 - 0.416 to 0.506

Constant 53.Z7121 0.384 13f.59 0.000 52.28to54.26
Prediction
Predicted age range 20-65 yellS
SEe 0.60
95% CI 71.68-74.84
90
80
70
80r§
20 30 40 50 eo 70 80
age
fiaure A23m.2. A scatterplot illustrating the cJtspersion of the data points around the regression line for the
Grooved Pegboard Test, Nondominant Hand. 11te size of the bubbles reflects the weight of the data point,
with larger bubbles indicating larger standard b r and smaller weight.
Tests for 8SIUIIlptions and model 8t
Tem £or heterogeneity iD. the flaal ~

Q(clf). p Q(l4) = 435.41, p < 0.000
Tests for moclel &t-clclittma of a quam.tic term
Model BIC BIC'

Unear 0.907 0.900 • 32.957 -32.960
Quadratic 0.910 0.896 35.141 -30.775
Tests for parameter speetfleatioas
Shapiro-Wilk W test =
W 0.813. p =0.005
Homosoedasticity
White's general test 2.136, pt=0.344
APPENDIX 23M 975
Sigalflcance tests for regression with the SDs
Onlmary least-squares rep11i1111 of SDs oa age (quadratic)

R2 0.468
Fecit)• p Fc2.5>=25.20, p <0.002
Term Coefficient SE t p 95'll CI
Age 0.8551187 0.475 1.8° 0.13a" - 0.366 to 2.076

Age2 -0.0085961 0.006 -1.54 0.185 - 0.023 to 0.006
Constant -5.48594 7.506 -0.73 0.498 -24.78 to 13.8
"Significance test for Age centered (sample means- aggregate mean): t=3.99,
p=0.010
Prediction
MeanpredictedSD 13.90 (2.16)
SEe 1.90
95%CI 10.18-17.62
EfFects of demographic \'IU'iables
Ed-tiOD
Regreslion of test means on education and age•
a2 0.836
Term Coeflicient SE t p 95'll CI
Education -0.628 0.159 -3.96 0.005 - 1.00 to- 0.25
"Regression with education was ran on a dataset compriling data stratified by

education rather than by age, when available.
Geader
t-test by gender
n Xmale Xfemale M-F Difference t p

4M,4F 69.425 65.525 3.900 0.840 0.217
Tables for the Category Test (CT)
Study numbers and page numbers p~vided in Locator table also provides a reference for
these tables refer to study numbeq and de- each study to a corresponding data table in
scriptions of studies in the text of Cbapter 24. this appendix.
Table A24.1. Locator Table for the Ca~gory Test

Study Age• n Sample Composition IQ Education• Location
cr.I Halstead, 1947 15-50 28 14 subjects without psychiatric diag- Education: Chicago
page 483 nosis, or history of brain 7-18
TableA24.2 injury; 14 with psychiatric IQ:
diagnosis 70-140
Cf.2 Reitan, 1955b, 32.36 50 35 M, 15 F volunteers Education: Indiana
1959 (10.78) hospitalized with paraplegia 11.58 (2.85)
page484 and neurosis were included FSIQ:
Table A24.3 112.6 (14.3)
Cf.3 Reitan &: No information is provided USA
Wolfson, 1985 regarding the normative
page 484 sample; cutoffs for "severity
Data are not ranges" (perfectly normal,
reproduced normal, mildly normal, im-
in this book paired, and seriously impaired)
are presented
cr.4 Klove &: 31.6 22 American and Norwegian controls Education: Wisconsin,
Lochen (cited in Klove, 32.1 22 11.1 Norway
1974) 12.2
page 485 FSIQ:
TableA24.4 109.3
111.9
976
APPENDIX 24 977

Study Age· n Sample Composition IQ Education• location
Cf.5Wien& 23.6 48 All males, neurologically normal; Education: Portland,

Matarazzo, 1977 24.8 divided into 2 groups; random 13.7 OR
page485 sample of 29 retested 14-24 14.0
Table A24.5 weeks later FSIQ:
117.5
118.3
Cf.6Mack& 60-80 41 Older subjects: 3M, 38 F; no history of Education (Older):
Carlson, 1978 69.76 neurological impairment; younger 14.05 (3.39)
page 486 (4.87) subjects: 9 M, 31 F; no screening for FSIQ:
Table A24.6 20-37 40 neurological impairment was 119.90 (15.14)
25.03 conducted computerized Education (Younger):
(3.70) administration was used 15.43 (2.65)
FSIQ:
113.76 (4.89)
cr.7 Anthony 38.88 100 Normal volunteers, no history of Education: Colorado
et al., 1980 (15.80) medical or psychiatric problems, 13.33 (2.56)
page486 head injury, brain disease, or FSIQ:
Table A24.7 substance abuse 113.5 (10.8)
cr.s Harley, et al., 55-79 193 V.A-hospitalized patients; Education: Wisconsin
1980 55-59 56 T-score equivalents are reported 8.8
page487 6()....64 45 IQ>80
Table A24.8 65-69 35
70-74 37
75-79 20
Cf.9 Pauker, 1980 19-71 363 Volunteers Huent in English; 152 WAIS IQ: Toronto
page 487 19-34 M, 211 F; subjects had no physical 89-102
Table A24.9 35-52 disability, sensory deficit, 103-112
53-71 current medical illness, brain 113-122
disorder, or alcoholism; data 123-143
are presented in age x IQ cells
Cf.IO Fromm-Auch 15-64 193 111 M, 82 F; participants described Education: Canada
& Yeudall, 1983 25.4 (8.2) as nonpsychiat:ric and non- 8-26
page 488 15-17 32 neurological; 83% are right-handed; 14.8 (3.0)
Table A24.10 18-23 75 5 age groupings FSIQ:
24-32 57 119.1 (8.8)
33-40 18
41-64 10
cr.u Heaton, et al., 15-81 553 356 M, 197 F; exclusion criteria Education: Colorado,
1986 39.3 included history of neurological 0-20 California,
page 488 (17.5) illness, significant head trauma, 13.3 (3.4) W'JSCOnsin
Table A24.11 and substance abuse; sample was <12 (132)
<40 319 divided into 3 age groups and 3 12-15 (249)
40-59 134 education groups; % classification ~16(172)
~60 100 as normal is provided
Cf.l2 Dodrill, 1987 27.73 120 60 M, 60 F volunteers; data for Education: Washington
page 489 (11.04) various intelligence levels are 12.28 (2.18)
Table A24.12 presented FSIQ:
100 (14.35)
Cf.l3 Yeudall et al., 15-40 225 Volunteers: 127 M, 98 F; classified Education: Canada
1987 15-20 62 in 4 age groupings; 88% 14.55 (2.78)
page490 21-25 73 right-handed FSIQ:
Table A24.13 26-30 48 112.25 (10.25)
31-40 42
(conHnued)
978 APPENDIX 24

Study Age· n Sample Composition IQ Education• Location
cr.I4 Ernst, 1987 65-75 110 51,M, 59 F volunteers Education: Brisbane,

page490 69.6 10.3 Australia
Table A24.14 (2.7)
cr.ts Alekoumbides 1~ 112 Medical and psychiatric V.A. Education: S. California
et al., 1987 46.85 r;::nts without cerebral lesions or 1-20
page 491 (17.17) ries of alcoholism or cerebral 11.43 (3.20)
Table A24.15 Oontusions; all subjects except for one FSIQ:
~remale 105.9 (13.5)
Cf.16 Bornstein 17-52 23 Volunteers: 9 M, 14 F; no history VIQ:
et al., 1987a 32.3 tf neurological or psychiatric 88--128
page 491 (10.3) ilness; test-retest data are 105.8 (10.8)
Table A24.16 'rovided PIQ:
~121
105.0 (10.5)
Cf.17 El-Sheikh 17-24 32
•
Urtlergraduate and graduate Cairo,
et al., 1987 20.6 students; no history of brain damage; Egypt
page492 (1.4) test-retest data are provided
Table A24.17
Cf.18 Russell, 1987 46.19 155 Patents in V.A. hospitals; 148 M, 7 F; Education: Cincinnati,
page 492 (12.86) Ppected of neuro-logical disorders 12.29 (3.00) Miami
Table A24.18 but negative findings FSIQ:
111.9
Cf.19 Elias et al., 138 Healthy participants; data are Education: Maine
1990 20-31 partitioned into 3 age 15.4
page 493 37-49 f<mps 15.7
Table A24.19 55-67 14.9
Cf.IO Heaton et al., 42.1 486 Volmteers: urban and rural; data Education: California,
1991,2004 (16.8) ciollected over 15 years through 13.6 (3.5) Washington,
page 493 Groups: ~ulticenter collaborative efforts; Groups: Texas,
Data are not 20-34 lirict exclusion criteria; 65% M; 6-8 Oldahoma,
reproduced 35--39 data are presented in T-score 9-11 WISCODSin,
in this book 40-44 ttquivalents for M and F separately 12 Iillnois,
45-49 in10 age groups by 6 education 1~15 Michigan,
50--54 fOups; in 2004 edition, age range 16-17 New York,
55-59 ii expanded to 85 years and the ~18 V'uginia,
60-64 data are presented for African- FSIQ: Massachusetts,
65-69 American and Caucasian participants 113.8 (12.3) Canada
70-74 ~tely
75-80
Cf.Jl Elias et al., 427 Healthy participants; data are Education: Maine
1993 15-24 partitioned into 6 age groups 12-19
page 494 25-34 ~gender
Table A24.20 35-44
45-54
55-64
~65
cr.u Barrett 43.9 1,052 Air Force veteran controls; High school
et al., 2001 (7.6) J?resumably all male; and college
page 494 SDs not provided
Table A24.21
• Age column and IQ/education column cont4l information regarding range and/or mean and standard deviation for the
whole sample and/or separate groups, whichdfer information is provided by the authors.
APPENDIX 24 979
Table A24.2. [Cf.1] Halstead, 1947: Data for the

Control Group (Included Patients with Psychiabic
Diagnoses): Mean Number of Errors for the Total
Group and for Three Subgroups
n Number of Errors
Total 28 36.72
Civilian 14 26.8
(10-46)
MiJibuy 8 50.8
(29-75)
Miscellaneous 6 34.8
(16-93)
Table A24.3. [Cf.2] Reitan, 1955b,1959: Data for Individuals Referred for Neuropsychological Evaluation
with Negative Neurologi.cal Findings
n Age Education VIQ PIQ FSIQ Number of Errors
50 32.36 11.58 110.82 112.18 112.64 32.38

(10.78) (2.85) (14.46) (14.23) (14.28) (12.62)
Table A24.4. [Cf.4] Klove and Lochen (cited in

Klove, 1974): Data for American and Norwegian
Controls
Category
n Age Education IQ Errors
Americans 22 31.6 11.1 109.3 34.6

NoiWegians 22 32.1 12.2 111.9 45.5
Table A24.S. [Cf.5] Wiens and Matarazzo, 1977: Data for Male Applicants to Patrolman Program: Mean
Number of Errors and SDs for Two Equal Subject Groups•
WAIS Category Errors
n Age Education FSIQ VIQ PIQ Test Retest
24 23.6 13.7 117.5 117.4 115.4 23.5

(21-27) (12-16) (8.3) (8.4) (10.5) (21.3)
24 24.8 14.0 118.3 116.4 118.2 22.8
(21-28) (12-16) (6.8) (6.9) (8.6) (11.8)
29 24 14 118 116 ll8 22.83 11.21
(21-28) (12-16) (19.15) (9.32)
-Test-retest data are also provided for 29 subjects who were assessed twice.
980 APPENDIX 24
Table A24.6. [CT.6] Mack and Carlson, 1978: Data for Two Age Groups of Healthy Participants: Mean
Number of Errors and SDs for the Whole Test and Subtests III- VII
Subtest
n Age Education IQ III IV v VI Vll Total
40 25.03 15.43 113.76 14.95 10.00 12.55 7.00 3.72 4 .82

(3.70) (2.65) (14.89) (11.93) (9.54) (6.55) (5.39) (2.92) (27.93)
41 69.76 14.05 119.90 25.07 23.46 19.22 15.32 7.93 91.73
(4.87) (3.39) (15.14) (10.74) (8.57) (6.34) (7.51) (2.62) (26.26)
Table A24.7. [CT.7] Anthony et a!. , 1980: Data for Normal
WAlS
n Age Education FSIQ VIQ PIQ Category Errors
100 13.33 113.54 113.24 112.26 32.59

(2.56) (10.83) (11.59) (10. ) (21. 0)
Table A24.8. [CT.8] Harley et a!., 1980: Data for eterans Administration- Ho pitalized Patients: Means,
SD , and Ranges fo r the umber of Errors per Five Age Intervals for the Who! amp! and for th
Alcohol-Equated Sample
WAlS
n Age FS IQ VIQ PIQ Education Errors
To tal sample
56 55-59 98.57 99.39 97.00 10.1 64.13
(11.43) (12.92) (10.65) (2 .47)
80-129 77-131 72-129 19- 115
45 98.58 101.27 95.00 9. 59.7
(9.93) (11.42) (9.82) (19.6 )
80-121 78-123 7 116 30-110
35 ~9 97.51 100.37 93.66 .7 72.65
(11.18) (12.51) (10.20) (2 .96)
80-130 80-135 68-120 22-141
37 70-74 100.41 102.95 97.24 8. 85.60
(9.92) (11.81) (10.08) (36.27)
82-125 80-133 75-114 21- 162
20 75-79 101.75 101.40 102.15 6.5 69.60
(10.18) (11.40) (9.95) (26. 9)
81-119 77-117 83-119 19- 110
Alcohol-equated sample
47 55-59 99.00 100.00 9 .00 10.1 65.43
(11.73) (13.02) (11.13) (2 .51)
80-129 77-131 72-129 20-115
33 60-64 96.00 99.00 93.00 9.3 63.42
(9.43) (11.33) (9.30) (19.24)
80-117 78-123 78-112 34-110
APPENDIX 24 981

WAIS
n Age FSIQ VIQ PIQ Education Errors

23 65-69 99.00 102.00 95.00 8.8 71.68
(12.06) (13.06) (11.52) (31.14)
80-130 80-135 68-120 22.-141
37 7~74 10.00 103.00 97.00 8.8 85.60
(9.92) (11.81) (10.08) (36.27)
82.-125 80-133 ~114 21-162
20 ~79 102.00 101.00 102.00 6.5 69.60
(10.18) (11.40) (9.95) (26.89)
81-119 77-117 83-119 19-110
Table A24.9. [CT.9] Pauker, 1980: Data for Canadian Volunteers: Means and SDs for Total
Errors for the Whole Sample and for Three Age Groups by Four WAIS IQ Levels
WAIS IQ
Age 89-102 1~112 113-122 123-143 89-143
19-34 n=21 n=53 n=60 n=28 n=162

61.67 40.08 29.82 23.64 36.24
(18.95) (17.47) (14.26) (11.93) (19.33)
35-52 n=20 n=34 n=56 n=25 n=135
75.80 59.06 42.77 37.52 50.79
(24.12) (17.03) (15.25) (18.79) (21.98)
53-71 n=4 n=15 n=27 n=20 n=66
90.00 63.80 58.85 47.60 58.45
(15.25) (14.99) (18.17) (20.96) (20.59)
19-71 n=45 n=102 n=143 n=73 n=363
70.47 49.89 40.37 34.96 45.69
(22.69) (19.76) (18.69) (19.58) (22.37)
Table A24.10. [CT.10] Fromm-Auch and Yeudall,

1983: Data for Canadian Volunteers: Mean Number
of Errors, SDs, and Ranges for Each Age Grouping
n Age Category Errors
32 15-17 35.8
(16.2)
16-68
74 18-23 35.9
(21.2)
9-106
56 30.5
(13.6)
1~
18 36.3
(14.3)
11-67
10 41-64 53.0
(21.0)
29-96
982 APPENDIX 24
Table A24.11. [CT.ll] Heaton et al., 1986: Data for a Sample of Normal
Controls•
WAIS Category % Classified

n Age Education Mean sst Errors Normal
319 <40 11.9 29.3 89.0

134 40-59 11.2 42.6 70.2
100 ~60 9.7 66.4 31.0
132 <12 9.5 53.8 49.2
249 12-15 11.2 38.6 76.7
172 ~16 12.9 28.9 89.0
"Mean number of errors for the six subgroups as well as percent of subjects classi6ed as
normal using Russell et al.'s (1970) criteria.
tMean scaled scores for the Wechsler Adult Intelligence Scale subtests are reported.
Table A24.12. [Cf.12] Dodrill, 1987: Data for a Sample of Volunteers: Mean Number
of Errors and SDs for the Whole Sample and for Various Levels of Intelligence
Category
n Age Education FSIQ VIQ PIQ Errors
120 27.73 12.28 100.00 100.92 98.25 35.74

(11.04) (2.18) (14.35) (14.73) (13.39) (22.76)
Category Category
n FSIQ Errors n FSIQ Errors
7 130 29 60 95 36
18 125 23 48 90 41
34 120 21 33 85 47
64 115 22 19 80 54
93 110 26 10 75 70
101 105 30 70 77
75 100 33
Table A24.13. [Cf.13] Yeudall et al., 1987: Data for Canadian Volunteers: Mean
Number of Errors and SDs for the Whole Sample and for Each Age Group
%Right- Category
n Age Education Handed FSIQ VIQ PIQ Errors
62 15-20 12.16 79.09 111.75 111.18 108.30 33.88

(1.75) (10.16) (10.92) (10.47) (18.25)
73 21-25 14.82 86.03 109.79 110.48 105.88 35.10
(1.88) (9.97) (10.43) (11.20) (19.82)
48 26-30 15.50 89.58 113.95 114.40 110.28 30.52
(2.65) (10.61) (11.45) (8.72) (14.00)
42 31-40 16.50 90.48 116.09 117.76 109.72 36.28
(3.11) (9.51) (9.32) (11.45) (13.66)
225 15-40 14.55 85.78 112.25 112.60 108.13 33.97
(2.78) (10.25) (10.86) (10.63) (17.20)
APPENDIX 24 983
Table A24.14. [CT.14] Ernst, 1987: Data for Australian Volunteers: Mean Number of Errors and SDs for
Each of the Seven Subtests as Well as Total Errors for Each Gender Separately
Category Test
n Gender n III IV v VI VII Total
51 Male 0.1 0.5 19.9 13.7 16.5 10.6 5.8 66.7
(0.6) (0.7) (10.0) (11.1) (7.5) (7.1) (2.6) (27.3)
59 Female 0.2 0.5 25.4 20.5 17.8 12.4 6.7 83.3
(0.6) (0.8) (7.3) (10.6) (6.5) (5.7) (2.3) (21.6)
Table 24.15. [CT.15] Alekoumbides et al., 1987: Data for Veterans

~tionlnpatienb
WAIS
Category
n Age Education FSIQ VIQ PIQ Errors
112 46.85 11.43 105.89 107.03 103.31 62.04
(17.17) (3.20) (13.47) (14.38) (13.02) (28.16)
Table A24.16. [CT.16] Bornstein et al., 1987a: Data for Healthy Volunteers: Means, SDs, and Ranges for
Total Number of Errors for Both Testing Sessions•
n Age VIQ PIQ Test Retest
23 32.3 105.8 105.0 46.7 23.8

(10.3) (10.8) (10.5) (25.3) (19.0)
16-112 4-56
Raw Score Change Medlan Raw Score Change Mean % of Change
23.5 22 46
•Raw score change, median raw score change, and mean percent of change from the test to the retest are also reported.
Table A24.18. [CT.18] Russell, 1987: Data for

Veterans Administration Patienb Referred for
Table A24.17. [CT.17] El-Sheikh et al., 1987: Neuropsychological Evaluation with Negative Neu-
Data for Egyptian Studenb: Mean Number of rological Findings
Errors and SDs for the Total Sample for the Two
Testing Probes WAIS
- - - - - Category
n Age Test Retest n Age Education FSIQ VIQ PIQ Errors
32 20.6 29.5 9.84 155 46.19 12.29 111.9 112.3 109.9 52.11
(1.4) (18.78) (6.37) (12.86) (3.00) (26.31)
984 APPENDIX 24
Table A24.19. [CT.19] Elias et al., 1990: Data for 183 Healthy Volunteers
Partitioned into Three Age Groups
n WAIS
Age Category
Group Male Female Education VIQ PIQ Errors
20-31 41 47 15.7 119 116 26.60

(19.10)
37-49 23 38 15.4 122 122 33.90
(22.60)
55-07 12 22 14.9 124 121 56.80
(32.60)
Table A24.20. [CT.21] Elias et al., 1993: Data for 427 Healthy Volunteers Partitioned into Six
Age Groups by Gender
n Category Errors
Age
Group Male Female Male Female
1~24 37 24 27.51 25.96

(3.43) (3.05)
~ 40 56 37.70 30.11
(4.23) (2.40)
35-44 36 56 37.58 41.86
(3.59) (3.64)
45-54 25 46 39.88 51.91
(5.37) (3.62)
~ 25 35 57.72 58.14
(5.68) (4.90)
~65 24 23 63.04 74.57
(4.79) (5.60)
•Education range for the sample is 12-19 yeab.
Table A24.21. [CT.22] Barrett et al., 21l41: Data for Air Force
Veteran Controls•
n Age Category Errors
1,052 43.9 37.29

(7.6)
•sDs are not provided for the test scores.

for the Category Test (CT)
Category Test
Deserlption of the aggregate sample
Number of stucliea IDeluded Ia the aaalysla 11

Number ol data poUdl used Ia the aaalysla 25
s-p~..-
Mean 25 50.98 39.41 1~162
Age
Mean 25 44.65 19.66 16.~77.0
SD 25 4.35 4.75 1.~17.2
&luc:cdion
Mean 16 11.11 2.92 6.5-16.5
SD 11 2.17 0.86 Uh'3.2
IQ
Mean 16 106.06 7.39 97.5-118.3
SD 15 10.53 1.80 6.8-14.4
Perwa~ -'e 1 70.00 70.00 70.0
r..,_,.._
Combined mean 25 49.84 18.43 22.8-85.6
Combined SO 25 22.57 6.44 11.8-36.3
•Number of data points differs for different malyses due to missing data.
'weighted means and SDs.
(continued)
985
986 APPENDIX 24M
Predicted number of errors and SDs per age group• (Category Test)
95%CI 95%CI

Bt.mge Score Band Band SD Band Band
l~l9 26.96 21.81 32.11 16.59 14.48 18.71

JS-!9 30.83 26.24 35.42 17.60 15.61 19.58
30-34 35.12 31.12 39.12 18.72 16.87 20.57
35-39 39.41 35.96 42.86 19.84 18.07 21.61
40-44 43.70 40.73 46.68 20.96 19.24 22.69
45-49 47.99 45.38 50.61 U.09 20.35 23.82
S0-54 52.29 49.88 54.71 !3.21 21.41 25.00
55-59 56.58 54.16 59.00 24.33 22.44 26.22
~ 60.88 58.25 63.50 25.45 23.42 27.48
6S-69 65.17 62.17 68.17 26.57 24.38 28.77
10-14 69.46 65.99 72.94 27.69 25.31 30.08
15-19 73.75 69.73 77.78 28.82 26.22 31.41

Predicted tat acore = 11.50841 + 0.8585716 •age
Predicted SD = 12.55292 + 0.2243168 • age
Onlmary least-squares regression of test meaDS OD age (linear)

R2 0.839
F0 .1o1= 138.24, p < 0.000
Age 0.8585716 0.073 11.76 0.000 0.659 to 1.021

Constant 11.50841 3.841 3.00 0.013 2.949 to 20.067
Prediction
SEe 1.73
95%CI 46.79-53.57
APPENDIX 24M 987
Table A24m.1. (Contd)
100
0
80
80
40
~ ~~----~-----,-----,----~r-----r-----~
30 40 50 80 70 80
age
fipre A24m.1. A scatterplot illustrating the dispersion of the data points around the regression line for the
Category Test. The size of the bubbles reflects the weight of the data point, with larger bubbles indicating
Tests for heterogeaeity Ia the 8aal data aet

Q(df). p Q(!4) =580.80, p < 0.000
Tests for modelftt...adcJtion of a quadratic tenn
Model BIC BIC'
Linear 0.839 0.832 95.944 -4.2.431

Quadratic 0.843 0.829 98.474 -39.901
Tests for parameter speeifJeatloas
Homoscedasticity
988 APPENDIX 24M
Ordinary least-tCfuares regression ol SDs on age (linear)

~ 0.469
Fu.1o> =57.54, p < 0.000
Tenn Coefficient SE p 959& Cl
Age 0.2243168 0.030 7.59 0.000 0.158 to 0.290

Constant 12.55292 1.453 8.64 0.000 9.315 to 15.791
Prediction
SEe 1.02
959& CI 20.65-24.66
EiJeds oE demographic variables
Education
~ 0.874
Tenn Coefficient SE t p 959& CI
Education -0.079 2.079 -0.04 0.970 - 4.99 to 4.84
IQ
Est. tau2 without IQ 354.80
Est. tau2 with IQ 106.90
Regression of test means and IQ on age
~ 0.903
Tenn Coefficient SE p 959& CI
IQ -0.8418495 0.299 -2.81 0.026 - 1.549 to- 0.134

Appendix 25: locator and Data Tables for
the Wisconsin Card Sorting Test
Table A25.1. Locator Table for the Wisconsin Card Sorting Test (WCST)
WCST.l Heaton, 6-19 899 551 M, 348 F; nonneurological or $8 Southwestern USA,

et al., 1993 20-29 psychiatric 9-11 Colorado, Texas,
page 513 30-39 12 Detroit, Washington
Data are not -ID-49 13-15 DC
reproduced in 50-59 16-17
this book 60-64 ~18
65-69
70-74
75-79
80-84
WCST.! Beatty, 18-34 65 31 M, 34 F; nonneurological or 15.0 California
1993 35-59 psychiatric {2.2}
page 514 60+ 15.8
Table A25.2 {2.5}
15.5
{4.2}
WCST.3 Boone 45-49 91 35 M, 56 F; fluent English speakers; 14.5 Los Angeles,
et al., 1993 60-69 nonneurological or psychiatric; {2.5} California
page 514 70-83 71 were white, 10 FSIQ:
TableA25.3 were African American, 115.89
5 were Asian, and 5 were Hispanic {12.97}
{continued}
989
990 APPENDIX 25

WCST.4 Stratta 31.93 61 Is; nonneurological or 0-8 L'Aquila,

et al., 1993 (5.95) ~c 9-13 Italy
page 515 ~ ~14
I
Table A25.4 12.65
(4.30)
WCST.S Kramer 18-28 32 26 ~· 36 F; healthy volunteers 16.4
et al., 1994 60-74 30 (1.1)
page 516 16.3
Table A25.5 (1.8)
WCST.6 Spencer & 18-35 32 23 ¥· 41 F; undergraduate students; 13.5
Raz, 1994 65-80 32 nctnneurological or psychiatric; 15.3
page 516 n; color blindness
Table A25.6
WCST. 7 Paolo 69.74 187 69 •• 118 F; nonneurological or 14.91 Kansas City, KS
et al., 1995 (6.96) !Ehiatric; 97% were White, (2.57)
page 517 % were Black, and 1.1% were
TableA25.7 panic
WCST.8 Artiola i 27.32 119 51 M, 68 F; Spanish-speaking; non- 14.35 Madrid,
Fortuny& (9.11) neurological or psychiatric (2.25) Spain
Heaton, 1996
page 517
Table A25.8
WCST.9 Hoff et al., 32.1 54 All-male sample; nonneurological 15.4
1996 (9.7) o. psychiatric; 48 (2.4)
page 517 -re White, 4 were African VIQ:
Table A25.9 APterican, and 2 were Hispanic 115.1
(12.6)
WCST.lO Paolo 68.79 87 25M, 62 F; nonneurological or 14.80
et al., 1996 (6.21) pfycluatric; 95% were White, (2.42)
page 518 29& were African American,
Table A25.10 aad 2% were Hispanic;
~tial testing and retesting
irl1year
,
WCST.ll Rosselli 25.61 63 All-Jilale sample; no exclusion criteria 10.51 Bogota,
& Ardila, 1996 (7.54) ate provided (4.58) Colombia
page 518
Table A25.11
WCST.l2 Salthouse 54.1 259 27~ M, 63% F; healthy participants 15.0
et al., 1996 (18.4) r;cruited; no exclusion criteria
page 519 life provided
Table A25.12
WCST.l3 Compton 47.74 52 30 J,l, 22 F; participants were non- 18.44
et al., 1997 (11.77) ~ology faculty members of the (1.69)
page 519 Georgia College and State University;
Table A25.13 no exclusion criteria are provided
WCST.l4 Fristoe 18-35 48 P~cipants were divided into younger and 13.3
et al., 1997 65-80 49 c{der groups; younger group consisted (1.3)
page520 cf 25% male, older group of 35% 13.9
Table A25.14 Itaaie; no exclusion criteria are (2.0)
pded
APPENDIX 25 991

Study Age" n Sample Composition IQ/ Education• Location
WCST.l5 39.25 390 138 M, 252 F; 205 &om Madrid, 11.15 Madrid, Spain,
Artiola i Fortuny (14.8) Spain, 185 from US-Mexico (5.25) Mexico; Arizona
et al .• 1998 border; nonneurological or psychiatric
page520
Table A25.15
WCST.I6 Boone, <65 155 53 M, 102 F; nonneurological or 14.57 California
1998 >65 psychiatric (2.55)
page 521 63.07 FSIQ:
Table A25.16 (9.29) 115.41
(14.11)
WCST.l7 Mejia 55-70 60 21 M, 39 F; Spanish-speaking; 2-5 Medellin,
et al .• 1998 71-a5 nonneurological or psychiatric 6-11 Colombia
page 521
Table A25.17
WCST.I8 Basso 32.50 82 82 M participants were recruited; 14.98
et al .• 1999 (9.27) 50 were retested in 1 year; (1.93)
page522 48 were Caucasian, 1 African
Table A25.18 American, and 1 Hispanic; non-
neurological or psychiatric
WCST.19 Gooding 18.72 104 43 M, 61 F; undergraduate students; CoUege Madlson, WI
et al., 1999 (0.86) nonneurological or psychiabic
page522 IQ:
Table A25.19 116.26
(12.56)
WCST.SO Merriam 26.08 61 Healthy volunteers; nonneurological 14.66
et al .• 1999 (7.67) or psychiatric (2.39)
page 522 IQ:
Table A25.20 103.90
(9.22)
WCST.Jl Rey et al., 33.45 75 19 M, 56 F; primarily Spanish-speaking; 12-15 Dade County,
1999 (19.75) 53 from Cuba, 3 &om Peru, >15 FL
page 523 1 &om Venezuela, 6 &om
Table A25.21 Puerto Rico, 1 &om Panama, 6 14.53
from Colombia, 1 &om Honduras, (3.25)
8 from Nicaragua. and 19 &om
"other" nationalities
WCST.B Snitz 36.0 54 19 M, 35 F; nonneurological 15.0 Minneapolis,
et al .• 1999 (13.4) or psychiatric (1.7) MN
page523 IQ:
Table A25.22 109.7
(13.4)
WCST.J3 Tallent 19.11 63 22M, 41 F; undergraduate students; CoUege WJSCODsin
&: Gooding. 1999 (1.03) primaly language was English; students
page5J4 nonneurologlcal or psychiatric IQ:
Table A25.23 114.63
(11.93)
WCST.J4 Compton 30-39 102 53 M, 49 F; healthy subjects; English was Atlanta,
et al.• 2000 40-49 primaly language GA
page5J4 50-59
Table A25.J4 60+
WCST.JS Ismail 35.9 75 59 M, 16 F; nonneurological or 13.2
et al.• 2000 psychiatric (3.8)
page525
Table A25.25
(continued)
992 APPENDIX 25

WCST.26 Laiacona 1~29 205 100M, 105 F; nonneurological or ~7 Costa Masapa.

et al., 2000 30-30 psychiatric 8-12 Italy
page 525 40-49 13-16
Table A25.26 50-59 17-24
60-69
70-85
WCST.27 Rossi 26.4 64 30 M, 34 F; employees and 14.69 L'Agul]a.
et al., 2000 (5.44) relatives/acquaintances of (2.99) Italy
page526 hospital staff; nonneurological
Table A25.27 or psychiatric
WCST.28 Razani 60.36 104 33M, 71 F; nonneurological or 14.82
et al., 2001 (9.64) psychiatric (3.31)
page526 FSIQ:
Table A25.28 116.81
(14.06)
WCST.29 Salthouse 18-39 261 35% M; no specific exclusion 15.9
et al., 2003 40-59 criteria are provided; data (2.8)
page 527 60-84 are stratified into 3 age
Table A25.29 48.2 groupings
(17.2)
WCST.30 Kongs, 6-19 897 435 M, 462 F; nonneurological or ~8 Southwestern USA.
et al., 2000 ~29 psychiatric 9-11 • Colorado, Teas,
page 527 30-39 12 Detroit, Washmgtm
Data are not 40-49 13-15 DC
reproduced in 50-59 16-17
this book 60-64 2;18
~
70-74
7~79
80-84
WCST.31 Axelrod ~29 20 55 M, 85 F; undergraduate students 15.6 Detroit, MI
et al., 1993 30-39 20 from Wayne State University or (1.2)
page528 40-49 20 newspaper ads; no exclusion 15.4
Table A25.30 50-59 20 criteria are provided for participants (1.6)
60-69 20 younger than 50; nonneurological 15.2
70-79 20 or psychiatric conditions (2.2)
80-89 20 in older participants 15.4
(2.5)
14.4
(3.0)
14.5
(4.2)
14.5
(4.1)
WCST.32 Paolo 71.34 35 22 M, 13 F; nonneurological or 13.11
et al., 1996b (5.73) psychiatric conditions (2.03)
page 529
Table A25.31
WCST.33 Lopez- 18-29 115 All male; Spanish speakers; participants 0-6 Los Angeles.
Carlos et al., 2003 30-49 were tested in Spanish; non- 7-10 CA;
page 529 28.89 neurological or psychiatric 5.82 Jalisco,
Tables A25.32, A25.33 (8.37) conditions (2.49) Melico
APPENDIX 25 993

WCST.34 Bondi 71.1 75 27 M, 48 F; nonneurological 13.7 San Diego,

et al., 1993 (7.6) or psychiatric conditions (2.6) CA
page 530
Table A25.34
WCST.35van 35.2 77 19 M, 58 F; Caucasian participants; IQ: United Kingdom
den Broek (12.8) nonneuroological or psychiatric 109.5
et al., 1993 conditions (13.2)
page530
Table A25.35
WCST.36 Isingrini 18-35 35 52 M, 55 F; control healthy participants; 12.05 France
& Vazou, 1997 65-80 72 no other exclusion criteria provided (3.03)
page 530 8.54
Table A25.36 (1.18)
WCST.37 45-49 29 97 M, 132 F; normal older subjects; 78% 1-6 San Diego,
Lineweaver 60-69 84 white, 21% Mexican American 7-12 CA
et al., 1999 7~79 89 or European Spanish American, 13-16
page 531 80-91 27 1% African American, and 17-20
Table A25.37 1% Cuban American; non-
neurological or psychiatric conditions
whole sample and/or separate groups, whichever Information is provided by the authors.
Table A25.2. [WCST.2] Beatty, 1993 (128-Card Administration Version): Data for a
Control Sample Stratified into 'I1lree Age Groups
Age Groups
n MIF Ratio Variables• 18-34 35-59 >60
65 31134 Age 25.5 40.6 70.9

(5.7) (6.1) (6.5)
Education 15.0 15.8 15.5
(2.2) (2.5) (4.2)
CAT 5.8 5.8 4.6
(0.7) (0.8) (1.9)
PR 11.3 11.0 24.2
(7.5) (9.8) (27.6)
PE 9.9 9.9 20.5
(5.8) (8.5) (21.5)
NPR 10.1 7.7 13.0
(7.6) (8.1) (9.4)
Trials to First Category 13.3 13.5 18.8
(5.2) (6.6) (25.6)
Set Failure 0.6 0.5 1.1
(1.1) (0.7) (1.4)
"CAT, Categories Completed; PR, Perseverative Responses; PE, Perseverative Errors; NPR, Non-
perseverative Responses.
Table A25.3. [WCST.3] Boone et al., 1993 (128-Card Administration Version): Data for a Control Sample Stratified into Three Age Groups, Gender, and Three i
Educational Levels
Variables Age Groups Gender Education (Years)
45-49 60-69 70-83 Male Female $12 13-16 >16
Age - - - 62.29 61.45 62.15 61.83 61.00

(9.67) (9.49) (8.75) (9.58) (10.90)
n 38 31 22 35 56 26 48 17
MIF ratio 13125 13/18 9/13 11115 14134 1017
Education 14.55 14.29 14.41 14.94 14.11
(2.69) (2.12) (2.86) (2.89) (2.24)
FSIQ- 114.53 114.48 117.8& .l1Q.U 11i.Q3. - 110.35 . -.. ----..ll5.2a 123JIO
(14.44) (12.76) (12.53) (13.41) (12.89) (12.26) (12.85) (13.49)
WCST
CAT 4.61 5.13 4.14 4.03 5.07 4.42 4.63 5.18
(1.90) (1.43) (1.96) (1.89) (1.62) (1.68) (1.96) (1.38)
PR 19.81 18.23 27.23 27.34 17.09 26.81 20.55 13.77
(16.79) (14.57) (20.54) (19.82) (14.23) (21.22) (15.36) (12.80)
Errors 31.24 29.48 44.68 41.00 27.42 39.62 34.49 23.65
(20.30) (19.21) (20.45) (21.82) (18.78) (21.57) (19.59) (19.64)
%PR 15.15 14.27 19.54 19.43 13.76 19.34 15.62 11.75
(9.68) (8.37) (11.06) (11.28) (8.02) (12.05) (8.67) (7.11)
% Conceptual 62.87 64.82 51.16 54.41 64.67 55.28 60.11 70.52
(18.67) (17.81) (18.91) (20.06) (17.39) (18.77) (18.90) (16.97)
Trials to First Category 13.60 13.81 19.68 17.52 13.66 18.54 14.45 11.94
(5.45) (5.74) (23.64) (18.78) (6.20) (21.89) (6.34) (2.05)
Set Failure 0.92 0.87 1.05 1.20 0.76 0.92 0.94 0.94
(1.23) (1.09) (1.13) (1.30) (1.02) (1.09) (1.09) (1.44)
>
""0
Other Responses 1.35 1.94 2.36 1.67 1.89 1.85 2.17 0.71 ""0
(2.06) (3.35) (3.33) (2.98) (2.83) (3.03) (3.07) (1.72) m
z
•For the majority of subjects, the Satz-Mogel format (Adams et al., 1984) was used to obtain aglHlOrrected FSIQ, but for nine subjects over the age o£74 the Ryan et al. (1990) tables 0
X
were used. CAT, Categories Completed; PR, Perseverative Responses.
N
c.n
APPENDIX 25 995
Table A25.4. [WCST.4] Stratta et al., 1993 (128-Card Administration Version): Data for a
Control Sample Stratified into Three Education Groups
Education Group
0--8 9-13 2::14

n Age Education WCST Measure• (n=18} (n=23} (n =20}
61 31.93 12.65 CAT 3.00 5.08 4.15

(5.95} (4.30} (2.02) (1.41} (2.23}
PE 26.05 20.34 15.95
(12.33} (10.30} (14.73}
Total Errors 46.55 30.39 27.35
(17.65) (12.54} (17.96)
Other Responses 7.38 2.69 2.70
(6.40} (3.93} (4.49}
•cAT, Categories Completed; PE, Perseverative Errors.
Table A25.5. [WCST.5] Kramer et al., 1994 Table A25.6. [WCST.6] Spencer and Raz, 1994
(128-Card Administration Version): Data for a (128-Card Administration Version): Data for a
Control Sample Stratified into Two Age Groups Control Sample Stratified into Two Age Groups
Age Group Age Group
Variables 18-28 60--74 Variables 18-35 65-80
n 32 30 n 32 32
Age• 20.6 67.8 MIF ratio 12/20 ll/21
MIF ratio 12/20 14/16 Age• 23.8 69.5
Education 16.4 16.3 Education• 13.5 15.3
(1.1} (1.8) CAT 5.06 3.14
IQt 117.8 117.6 (1.70} (2.13)
(8.5} (8.4} Total Errors 29.47 46.53
CAT 5.9 4.20 (16.30} (20.77)
(0.5) (2.1} PE 25.75 41.63
Total Errors 14.8 36.7 (16.22} (19.98)
(11.2) (26.4)
•sos were not provided. CAT, Categories Completed;
PR 11.1 26.5
(7.9}
PE, Perseverative errors.
(24.1}
PE 8.3 20.0
(6.1) (18.9)
Trials to First Category 12.2 14.6
(3.9} (8.1}
% Conceptual Level Responses 63.4 57.2
(7.0} (21.9}
•sos were not provided.

t1Q was based on the Kaufman Brief Intelligence Test.
CAT, Categories Completed; PR, Perseverative Re-
sponses; PE, Perseverative Errors.
996 APPENDIX 25
Table A25.7. [WCST.7] Paolo et al., Table A25.8. [WCST.8] Artiola i Fortuny and
1995 (128-Card Administration Ver- Heaton, 1996 (128-Card Administration Version):
sion): Data for a Control Sample Data for a Healthy Sample from Madrid, Spain, for
Standard and Computerized Administrations
Variables Values
Administration
n 187
MIF ratio 69/118 Variables Standard Computerized
Age 69.74 n 60 59
(6.96)
Age 27.32 27.32
Education 14.91 (10.82) (7.06)
(2.57)
Education 14.13 14.58
WCST (2.33) (2.16)
CAT 4.84 WCST
(1.72) Total Trials 96.65 99.68
Total Errors 30.97 (22.74) (21.15)
(20.73) Total Correct 71.60 75.54
PE 17.06 (11.32) (9.58)
(12.86) Total Errors 24.05 24.14
PR 19.55 (19.52) (13.28)
(15.92) PR 15.23 12.47
NPE 13.87 (14.58) (6.95)
(9.93) PE 13.45 11.47
Set Failure 0.91 (11.86) (6.10)
(1.12) NPE 10.60 12.66
Trials to First Category 13.32 (9.77) (8.29)
(7.61) %Conceptual Level Responses 71.50 72.01
%Conceptual Level 63.93 (18.79) (12.04)
Responses CAT 5.33 5.39
(21.48) (1.39) (1.21)
Learning to Learn -2.43 Trials to First Category 14.80 20.91
(4.94) (11.69) (19.55)
Set Failure 0.75 1.20
(1.03) (1.74)
Learning to Learn -1.38 -1.13
(3.23) (5.58)
PR, Perseverative Responses; PE, Perseverative Errors;

NPE, Nonperseverative Errors; CAT, Categories Com-
pleted.
APPENDIX 25 997
TableA25.9. [WCST.9] Hoffetal., Table A25.10. [WCST.10] Paolo et al., 1996 (128-
1996 (128-Card Administration Card Administration Version): Data for a Control
Version): Data for an All-Male Sample
Control Sample
Retest
Variables Values Initial (1.1-Year
Variables Testing Follow-Up)
n 54
Age 32.1
n 87
(9.7) MIF ratio 25162
Education 15.4 Age 68.79
(2.4) (6.21)
Verbal IQ (prorated) 115.1 Education 14.80
(12.6) (2.42)
RawWCST
WCST
CAT 4.84 4.86
CAT 5.4 (1.76) (1.89)
(0.9)
Trials to First 17.76 18.44
Total Errors 17.9 Categol}' (22.79) (23.32)
(14.1)
Set Failure 1.16 0.69
PR 9.0 (1.39) (1.20)
(10.0)
Learning to Learn -2.66 -1.71
CAT, Categories Completed; PR, Per- (5.26) (4.61)
severative Responses. NormGllutl wcsr•
Total Errors 105.76 111.41
(17.69) (19.58)
PR 104.86 111.94
Table A25.11. [WCST.ll] Rosselli and (15.92) (17.90)
Ardila, 1996 (128-Card Administration PE 105.57 112.17
Version): Data for an All-Male Control (16.91) (18.03)
Sample
NPE 105.17 108.63
Variables Values (19.42) (22.02)
%Conceptual 106.21 111.17
n 63 Level Respouses (18.08) (19.74)
Age 25.61
(7.54) •Normalized age- and education~rrected standard
scores (mean=100, SD=15). CAT, Categories Com-
Education 10.51
pleted; PR, Perseverative Respouses; PE, Perseverative
(4.58)
Errors; NPE, Nonperseverative Errors.
WCST
Total Correct 69.70
(7.60)
CAT 5.80
(0.50)
Total Errors 21.00
(12.40)
PE 9.90
(6.80)
PR 11.70
(8.00)
NPE 11.20
(7.70)
CAT, Categories Completed; PE, Persevera-

tive Errors; PR, Perseverative Responses;
NPE, Nonperseverative Errors.
998 APPENDIX 25
Table A25.12. [WCST.12] Salthouse et al., 1996 Table A25.13. [WCST.13] Compton
(128-Card Administration Version): Data for a et al., 1997 (128-Card Administration
Healthy Sample Version): Data for a Sample of University
Faculty
Variables Values
Variables Values
n 259
%Male 27 Age 47.74
Age 54.1 (11.77)
{18.4) Education 18.44
Education• 15.0 (1.69)
WCST 'WCST
Number of Trials 100.7 CAT 4.02
{23.4) (2.23)
CAT 4.96 Trials to First Category 19.73
(1.59)
(17.18)
%Errors 27.6 Total Trials 111.14
{15.2)
(21.11)
%PR 17.7
(14.8) CAT, Categories Completed.
%PE 15.6
{11.4)
%NPE 11.9
(6.8)
% Conceptual Level Responses 65.7
(20.5)
Table A25.14. [WCST.14] Fristoe et al., 1997
Trials to First Categoryt 15.1 (128-Card Administration Version): Data for a
(11.5) Healthy Sample Using the Computerized Version
Set Failure 0.69 Stratified into Two Age Groups
(1.04)
Learning to Learn: -3.64 Age Croup
{6.90) Variables 18-35 65-80
"Educational level is approDmated from the available n 48 49
data, and the SD is not available.
%Male 25 35
tn=256.
Age 26.7 70.1
*n=254. (5.7) (7.!)
CAT, Categories Completed; PR, Perseverative Re- Education 13.3 13.9
sponses; PE, Perseverative Errors; NPE, Nonpersevera- (1.3) (lLO)
tive Errors.
CAT 4.8 3.1
(1.9) (2.0)
Conceptual Level Responses 65.5 55.2
(17.5) (20.3)
{18.7) (20.7)
PE 16.4 25.2
(10.0) (llL1)
%PE 14.1 20.1
(7.1) (8.8)
CAT, Categories Completed; PE, Perseverative Errors.

APPENDIX 25 999
Table A25.15. [WCST.15] Artiola i Fortuny et al.,

1998 (WCST 128-Card Version): Data for Healthy
Spanish-Speaking US-Mexico and Madrid, Spain,
Groups
Age Group
Variables US-Mexico Madrid, Spain
n 185 205
MIF ratio 471138 91/114
Age 42.2 36.3
(13.5) (16.1)
Education 9.6 12.7
(6.1) (4.4)
CAT 3.4 4.6
(1.6) (1.6)
PR 33.6 21.8
(19.2) (19.1)
CAT, Categories Completed; PR, Perseverative Re-

sponses.
Table A25.16. [WCST.16] Boone, 1998 (128-Card Administration Version): Data for a Healthy Sample Stratified into Two Vascular Status Groups by Two Age
Groups and Three IQ Groups ....
Age Group IQ Group §
Vascular <65 years ~65 years High
Demographic Values WCST Measure Status Average Average Superior
n 155 CAT Healthy 4.95 5.10 4.71 4.88 5.27

M/F ratio 53/102 (1.60)" (1.30) (1.70) (1.45) (1.45)
Age 63.07 n=34 n=25 n=44
(9.29)
Education 14.57 Vascular 4.58 3.68 3.53 3.77 4.76
(2.55) (2.00) (2.30) (2.53) (2.01) (1.89)
FSIQ• 115.41 n=17 n=13 n=17
(14.11)
PE Healthy 17.24 20.70 21.94 19.00 15.07
(14.~) . - - ___ (~.70) .. (16.41) (12.63) (14.19)
.... ·;=-~·- -- . --· ·,;-;,25 n=44
Vascular 25.11 32.52 39.31 27.46 21.71
(27.50) (32.40) (42.68) (25.65) (15.00)
n=16 n=13 n=17
%Conceptual Level Healthy 64.95 62.84 59.87 61.10 69.51
Responses (18.60) (18.60) (18.01) (17.17) (18.79)
n=33 n=25 n=44
Vascular 58.02 47.40 45.32 47.19 61.52
(23.40) (24.60) (29.84) (20.79) (18.49)
n=17 n=13 n=17
Total Errors Healthy 29.51 33.33 36.21 33.72 24.71
(20.60) (21.50) (20.55) (19.27) (20.82)
n=33 n=25 n=44
Vascular 35.89 45.41 46.81 46.77 32.41
(24.50) (24.80) (30.51) (21.02) (19.73)
n=16 n=13 n=17
Trials to First Category Healthy 14.12
(6.00)
13.30
(3.50)
14.53
(5.22)
14.16
(6.40)
13.23
(4.94) ..,..,>
n=34 n=25 n=44 rn
Vascular 19.16 30.46 40.76 24.15 12.53
z
0
(26.80) (40.40) (50.14) (30.61) (2.55)
><
n=16 n=13 n=17 "-J
VI
"The Satz-Mogel format (Adams et al., 1984) was used to obtain &g~HX>rrected FSIQ. CAT, Categories Completed; PE, Perseverative Errors.
APPENDIX 25 1001
Table A2S.17. [WCST.17] Mejia et al., 1998 (128-Card Administration Version): Data for a Healthy
Colombian Sample Stratified into Two Age and Two Education Groups
Af,e Group Education Group
Entire
Variables Sample 55-70 71-&5 2-5 Years 6-11 Years
n 60 28 32 42 18
MIF ratio 21139 6122 15117 16126 5113
Age 69.66
(7.09)
CAT 2.60 2.25 2.65 1.89
(1.16) (1.48) (1.29) (1.32)
PR 42.60 51.84 44.73 53.57
(19.63) (30.85) (22.39) (33.43)
PE 40.14 43.68 38.92 48.73
(16.75) (21.65) (16.10) (24.35)
NPE 26.92 25.25 27.07 23.78
(12.34) (12.33) (11.72) (13.38)
Set Failure 1.10 1.00 1.00 1.15
(1.37) (1.19) (1.16) (1.50)
CAT, Categories Completed; PR, Perseverative Responses; PE, Perseverative Errors; NPE, Nonperseverative Errors.
Table A25.18. [WCST.18] Basso et al., 1999 (128- Table A25.18. (Contd.)
Card Administration Version): Data for a Healthy
Sample Using the Computerized Version Stratified Testing Session
into Two Age Groups Variables Baseline 12 Months
Testing Session PR 16.02 9.34
(12.82) (7.70)
Variables Baseline 12 Months
n 50 (17.94) (18.74)
Af,e 32.50 Leaning to Learn -3.14 -0.72
(9.27) (5.76) (3.90)
Education 14.98 Set Failure 1.16 0.80
(1.93) (1.67) (1.16)
CAT 5.16 5.42
(1.38) (1.515) CAT, Categories Completed; PE, Perseverative Errors;
PR, Perseverative Responses.
Total Trials 101.12 84.74
(22.87) (18.59)
%Correct 76.48 80.99
(12.39) (12.45)
Total Errors 26.12 16.68
(18.04) (11.88)
PE 14.20 8.44
(10.53) (6.16)
%PE 12.79 9.84
(7.52) (7.45)
(continued)
1002 APPENDIX 25
Table A25.19. [WCST.l9] Gooding et al., 1999 Table A25.20. (Contd.)

(128-Card Administration Version): Data for a
Variables Values
Control Sample of College Students
NPE 11.16
Variables Values
(8.78)
n 104 Set Failure 0.75
MIF ratio 43/61 (1.15)
Age 18.72 Trials to First Category 13.61

(0.86) (5.74)
WAIS-R IQ" 116.26 % Conceptual Level Responses 74.31

(12.56) (13.90)
WCST Learning to Learn 0.17

(3.10)
CAT 6.00
(0.00) CAT, Categories Completed; PE, Perseverative Errors;
PE 7.58 PR, Perseverative Responses; NPE, Nonperseverative
(3.20) Errors.
NPE 8.29
(4.20)
Set Failure 0.33
(0.50)
Trials to First Category 13.08 Table A25.21. [WCST.21] Rey et al., 1999 (128-
(4.27) Card Administration Version): Data for a Healthy
Conceptual Level Responses 37.21 Spanish-Speaking Sample
(29.80)
Education Group
"Prorated from the Block Design and Vocabulary subtests
of the WAIS-R. CAT, Categories Completed; PE, Entire 12-15 >15
Perseverative Errors; NPE, Nonperseverative Errors. Variables Sample Years Years
n 75 25 30
Age 33.45
(19.75)
Education 14.53
Table A25.20. [WCST.20] Merriam et al., 1999 (3.25)
(128-Card Administration Version): Data for a MIF ratio 19/56
Control Sample CAT 5.5 5.4 5.8
(1.2) (1.0) (0.8)
Variables Values
PE 10.7 11.6 8.9
n 61 {8.7) (9.4) (6.6)
Age 26.08 PR 11.6 12.7 9.6
(7.67) (9.7) (10.6) (7.3)
Education 14.66 Trials to 12.9 12.4 11.7
(2.39) First Category (8.5) {3.3) (2.0)
WAIS-R IQ 103.90 Set Failure 0.6 0.5 0.6
(9.22) (0.8) (0.6) (1.0)
WCST CAT, Categories Completed; PE, Perseverative Errors;

CAT 5.64 PR, Perseverative Responses.
(1.02)
PE 10.26
(7.05)
PR 11.03
(8.26)
(continued)
APPENDIX 25 1003
Table A25.22. [WCST.22] Snitz et al., 1999 Table A25.23. [WCST.23] Tallent and Gooding,
(128-Card Administration Version): Data for 1999 (128-Card Administration Version): Data for a
a Control Sample Control Sample of College Students
Variables Values Variables Values
n 54 n 63
MIF ratio 19135 MIF ratio 22/41
Age 36.0 Age 19.11
(13.4) (1.03)
Education 15.0 WAlS-R IQ" 114.63
(1.7) (11.93)
WAlS-R IQ 109.7
(13.4) WCST
CAT 6.00
WCST (0.00)
CAT 4.6 PE 7.00
(2.0) (2.30)
PE 17.2 NPE 7.71
(14.9) (5.40)
NPE 21.8 Set Failure 0.27
(9.1) (0.50)
Set Failure 1.5 Trials to First Category 13.49
(1.3) (6.53)
Conceptual Level Responses 65.81
CAT, Categories Completed; PE, Perseverative (5.06)
Errors; NPE, Nonperseverative Errors.
"Prorated from the Block Design and Vocabulary subtests
of the WAlS-R. CAT, Categories Completed; PE, Per-
severative Errors; NPE, Nonperseverative Errors.
Table A25.24. [WCST.24] Compton et al., 2000 (128-Card Administration Version): Data for
a Highly Educated Sample Stratified into Four Age Groups
Age Group
Entire
Variables Sample 30-39 40-49 50-59 60+
n 102 30 27 25 20
MIF ratio 53149 13117 12/15 13112 14/6
Age 34.09 46.19 54.03 65.49
(3.71) (2.80) (3.20) (5.72)
Education• 18.95 19.18 19.92 19.47
CAT 5.00 4.93 4.04 3.25
(1.59) (1.75) (2.44) (2.57)
PE 8.05 8.41 11.22 16.43
(8.71) (12.22) (11.01) (14.06)
% Conceptual Level Responses 77.67 77.46 69.55 60.16
(15.62) (15.48) (21.68) (24.26)
"SDs not reported. CAT, Categories Completed; PE, Perseverative Errors.

1004 APPENDIX 25
Table A25.25. [WCST.25] Ismail et al.,

2000 (128-Card Administration Version):
Data for a Control Sample
Variables Values
n 75
MIF ratio 59116
Age• 35.9
'WCST
CAT 5.72
(1.32)
PE 9.21
(6.63)
•so not reported. CAT, Categories Com-

pleted; PE, Perseverative Errors.
Table A25.26. [WCST.26] Laiacona et al., 2000 (128-Card Administration Version): Data for an Italian
Sample
WCST Measures
Education Global• Set

(Years) Age Gender Scores PR NPR Failure
~7 40-49 Males 13.6

(9.6)
Females 69.3 29.2 20.8 0.3
(19.6) (17.8) (2.3) (0.8)
50-59 Males 47.8 18.4 17.4 1.1
(35.9) (17.3) (3.2) (2.1)
Females 51.0 24.7 12.3 0.9
(27.0) (14.2) (8.2) (1.9)
60-69 Males 77.6 40.2 12.5 0.4
(15.9) (9.9) (14.8) (0.5)
Females 75.4 36.0 20.4 0.8
(22.8) (7.5) (6.8) (0.8)
70-85 Males 4.ot 20.0 14.7 0.5
(48.1) (22.6) (8.3) (0.7)
Females 79.5 47.3 16.3 0.3
(18.9) (9.5) (8.8) (0.5)
8-12 1~29 Males 21.2 7.5 7.9 0.1
(13.4) (5.0) (4.8) (0.3)
Females 28.2 11.8 8.1 0.2
(23.1) (9.3) (5.9) (0.6)
30-39 Males 37.8 16.3 12.1 0.3
(25.3) (14.4) (5.4) (0.5)
Females 41.1 16.0 11.4 0.6
(14.3) (6.0) (6.7) (0.8)
40-49 Males 49.6 19.6 13.8 0.6
(32.3) (13.4) (7.5) (1.6)
APPENDIX 25 1005

WCST Measures

Females 39.4 19.1 11.4 0.1

(27.8) (13.8) (8.4) (0.4)
50-59 Males 44.3 17.8 14.3 0.0
(28.6) (8.5) (11.9) (0.0)
Females 60.3 21.3 15.7 1.0
(34.7) (9.9) (9.6) (1.3)
60-69 Males 58.3 27.2 18.5 0.0
(33.3) (20.2) (9.5) (0.0)
Females 59.1 24.4 17.3 0.4
(37.2) (16.4) (8.7) (0.8)
70-85 Males 38.0 20.0 12.0 0.0
(0.0) (0.0) (0.0) (0.0)
Females 80.0 52.8 12.8 0.0
(25.9) (20.2) (7.0) (0.0)
13-16 15-29 Males 33.5 17.3 7.5 0.5
(26.0) (14.6) (4.1) (0.6)
Females 30.5 11.6 8.8 0.3
(21.6) (7.6) (6.9) (0.5)
30-39 Males 36.8 11.0 10.5 1.0
(21.0) (5.4) (3.3) (1.4)
Females 42.5 17.3 12.5 1.0
(27.4) (13.5) (9.8) (1.2)
40-49 Males 17.0 6.0 8.0 0.0
(0.0) (0.0) (0.0) (0.0)
Females 18.3 9.0 5.3 0.3
(4.1) (2.4) (1.9) (0.5)
50-59 Males 34.0 13.5 11.8 0.2
(27.6) (8.4) (12.0) (0.4)
Females 25.3 14.3 5.0 0.0
(20.5) (15.4) (1.0) (0.0)
60-69 Males 33.8 14.0 12.8 0.5
(15.9) (7.3) (6.8) (0.6)
Females 31.8 15.3 6.8 0.3
(16.4) (10.1) (3.9) (0.5)
70-85 Males 57.5 19.5 21.0 0.0
(57.3) (19.1) (21.2) (0.0)
Females
17-24 15-29 Males 13.8 5.0 4.0 0.4

(3.5) (2.0) (1.9) (0.5)
Females 38.3 17.3 14.0 0.3
(20.6) (9.6) (5.6) (0.6)
30-39 Males 23.4 6.2 8.6 0.8
(18.0) (4.0) (5.9) (1.8)
Females 21.8 6.2 9.5 0.5
(22.8) (3.2) (11.0) (1.0)
(continued)
1006 APPENDIX 25

WCST Measures

40-49 Males 20.8. 6.8 7.5 0.3

(9.4) (2.5) (3.1) (0.5)
Females 30.0 14.0 7.7 0.0
(16.1) (10.1) (1.5) (0.0)
S0-59 Males 24.0 6.5 7.0 0.5
(2.8) (0.7) (2.8) (0.7)
Females 27.0 10.3 11.0 0.0
(16.6) (7.1) (7.4) (0.0)
60-69 Males 38.6 17.8 11.6 0.4
(28.3) (12.4) (9.8) (0.5)
Females 19.7 6.3 6.3 0.3
(9.7) (1.9) (2.6) (0.5)
70-85 Males 23.0 12.0 9.0 0.0
(0.0) (0.0) (0.0) (0.0)
Females 88.0 32.0 27.0 0.0
(0.0) (0.0) (0.0) (0.0)
"Global Score= Btrials odminlstered- (ncategories completed X 10).

tThese are the reported values; however, the means and SDs may have been printed in reverse order in the original article.
Table A25.28. [WCST.28] Razani et al.,

Table A25.27. [WCST.27] Rossi et al., Variables Values
Data for a Control Sample n 104
M/F ratio 331'71
Variables Values
Age 60.36
n 64 (9.64)
M/F ratio 30134 Education 14.82
(3.31)
Age 26.4
(5.44) WAIS-R FSIQ" 116.81
Education 14.69 (14.06)
(2.99) WCST
WCST CAT 4.99
(1.54)
CAT 5.33
(1.49) PE 18.25
PE 8.81 (14.76)
(8.69) Total Errors 30.64
Total Errors 17.81 (20.84)
(14.17) Set Failure 0.81
Unique (Other) Errors 1.34 (1.08)
(2.43) % Conceptual Level Responses 64.33
(18.55)
CAT, Categories Completed; PE, Persevera-
tive Errors. "The Satz-Mogel format (Adams et al., 1984) was
used to obtain age-corrected FSIQ. CAT, Cate-
gories Completed; PE, Perseverative Errors.
APPENDIX 25 1007
Table A25.29. [WCST..29] Salthouse et al., .2003 (1.28-Card Administration Version): Data for a
Highly Educated Sample Stratill.ed into Three Age Groups
Age Group
Variables Entire Sample 18-39 40-59 60-84
n 261 79 112 70
~Male 35 43 29 37
Age 48.2 27.7 49.0 70.3
(17.2) (6.4) (5.0) (6.2)
Education 15.9 15.5 16.0 16.4
(2.8) (3.3) (2.4) (2.9)
WAIS-In
Vocabulary 12.3 12.0 12.2 12.8
(2.9) (3.5) (2.7) (2.4)
'WCST
Correct responses 73.9 71.7 67.9
(12.9) (14.7) (16.6)
CAT 4.7 4.1 3.2
(1.8) (2.2) (2.0)
Total Errors 36.7 41.2 53.7
(22.5) (23.7) (22.6)
PE 17.4 20.8 25.3
(12.2) (16.5) (17.3)
PR 19.7 23.8 37.4
(15.4) (21.7) (50.9)
NPR 19.9 22.9 26.6
(13.8) (24.3) (15.1)
Trials to First Category 22.1 20.6 25.1
(20.5) (14.2) (20.3)
Conceptual Level Responses 61.5 58.9 53.1
(20.8) (21.6) (22.2)
Set Failure 1.2 1.4 1.6
(1.2) (1.5) (1.8)
Learning to Learn -0.8 -2.4 -3.2
(5.4) (6.0) (6.1)
CAT, Categories Completed; PE, Perseverative Errors; PR, Perseverative Responses; NPR, Nonperseverative
Responses.
1008 APPENDIX 25
Table A25.30. [WCST.31] Axelrod et a: , 1993 (64-Card Administration Version): Data for a Healthy
Sample Stratified into Seven Age Groups
Age Group
Variables 20-29 30-39 40-49 50-59 60-69 70-79 80-89
n 20 20 20 20 20 20 20
MIF ratio 4116 6/14 8112 9111 10/10 10/10 8/12
Race (blaclc/white) 9111 uw• 1000 1/19 4/16 4/16 1119
Age 24.4 34.2 45.0 55.3 65.2 74.3 83.4
(3.2) (2.6 (3.0) (2.5) (2.6) (2.9) (3.0)
Education 15.6 15.4 15.2 15.4 14.4 14.5 14.5
(1.2) (1.6 (2.2) (2.5) (3.0) (4.2) (4.1)
WCST
CAT 4.1 3.~ 3.6 3.4 2.6 2.6 2.2
(0.8) (U (1.1) (1.3) (1.7) (1.1) (1.6)
Total Errors 14.5 20.~ 18.4 16.6 21.3 22.8 25.7
(5.1) (9."' (7.8) (7.6) (11.6) (6.0) (10.8)
PR 9.2 1U 12.2 10.2 13.5 14.8 20.4
(4.3) (6. (5.5) (5.3) (8.5) (5.4) (13.2)
PE 8.2 1U 11.7 8.4 12.0 12.7 16.2
(3.6) (5."' (7.0) (4.2) (7.6) (4.3) (10.1)
NPE 6.2 9.C 7.7 7.7 9.4 10.7 9.0
(2.8) (5.~ (3.7) (4.4) (6.3) (5.0) (4.3)
"One Hispanic participant was included in tbJ group. CAT, Categories Completed; PR, Perseverative Responses; PE,
Perseverative Errors; NPE, Nonperseverative ~rrors.
Table A25.31. [WCST.32] Paolo et ,al., Table A25.31. (Contd.)

1996h (64-Card Administration Versickt):
Variables Values
NPE 8.89
Variables V~es
(4.41)
n 35 Conceptual Level Responses 64.26
MIF ratio 22113 (15.27)
Age 71.34 Set Failure 0.46
(5.73) (0.74)
Education 13.11 Trials to First Category 13.91

(2.03) (4.59)
DRS" 13t37 "Dementia Rating Scale (Mattis, 1976). CAT,

(3.36) Categories Completed; PE, Perseverative Errors;
WCST PR, Perseverative Responses; NPE, Nonperse-
verative Errors.
CAT 3.1)3
(1.22)
Total Errors 1&54
(7.50)
PE 9.86
(5..()6)
PR 11.29
(6.84)
(CORti~)
APPENDIX 25 1009
Table A25.32. [WCST.33a] Lopez-Carlos et al., 2003: Monolingual SpandSb-Spea}ing Males with
:510 Years of Education Stratified by Education Group
Education Group n M~ Similarities• Correct PR PE CAT
0-6 38 7.29 10.09 32.05 16.74 14.79 1.53

(3.77) (4.85) (10.09) (10.96) (8.12) (1.22)
7-10 21 7.47 11.20 40.38 13.86 11.71 2.24
(3.77) (4.97) (9.07) (7.25) (5.80) (1.00)
•WAIS-m raw scores (Mexican version). PR, Perseverative Respcmses; PE, Perseverative Errors; CAT, Categories Completed.
Table A25.33. [WCST.33b] Lopez-Carlos et al., 2003: Monolingual Spanish-Spea}ing Males with :510
Years of Education Stratified by Age and Education Group
Age x Ed Group n M~ Similarities• Correct PR PE CAT
18-J9
0-6 18 5.70 9.65 31.61 16.78 14.72 1.44
(2.10) (4.78) (8.00) (10.21) (7.33) (0.98)
7-10 12 9.40 10.67 40.75 12.75 10.83 2.33
(4.48) (5.04) (10.47) (8.67) (6.79) (1.07)
30-49
0-6 20 7.00 10.52 32.45 16.70 14.85 1.60
(3.29) (5.27) (11.85) (11.85) (8.96) (1.43)
7-10 9 8.56 12.78 39.89 15.33 12.89 2.11
(4.75) (4.02) (7.37) (4.87) (4.23) (0.93)
•wAIS-m raw scores (Mexican version). PR, Perseverative Respcmses; PE, Perseveratlve Errors; CAT, Categories Completed.
Table A25.34. [WCST.34] Bondi et al.,

1993 (MCST Administration Version): Data
for a Control Sample
Variables Values
n 75
M/F ratio 27/48
Age 71.1
(7.6)
MMSE• 28.9
(1.2)
JICST
CAT 5.16
(1.33)
PE 1.40
(2.81)
%PE 9.40
(15.29)
NPE 8.20
(6.01)
•MiDi-Mental State Examination (Folstein et al.,

1975). CAT, Categories Completed; PE, Perse-
verative Errors; NPE, Nooperseverative Errors.
1010 APPE N DI X 25
Table A25.35. (WCST.35] van den Broek Table A25.36. [WCST.36] lsingrini and Vazou,
et al., 1993 (MCST Administration Version): 1997 (MCST Administration Ver ion): Data for a
Data for a Control Sample Healthy Sample Stratilled into Two Age Groups
Variables Values Age Group
n 77 Variables 25-46 70-99
MIF ratio 19158 n 35 72

Age 35.2 MIF ratio 15/20 37/35
(12.8)
Age 35.54 0.59
FSIQ 109.5 (7.58) (8.58)
(13.1)
Education 12.05 8.54
MCST (3.03) (1.18)
CAT 4.9 CAT 5.25 3.51
(1.7) (0.85) (1.78)
Total Errors 10.7 Total Errors 9.45 20.47
(10.1) (3.62) (8.95)
PE 2.8
PE 3.57 11.68
(3.7)
(2.37} ( .2 )
%PE 26.5
(12.2) CAT, Categories Completed; PE , Perseverative Errors.
NPE 7.8
(7.0)
Unique• 2.0
(3.2)
Runst 0.3
(0.6)
•Errors that were neither color, fonn, nor number.

tThree or more sequential correct sorts but less
than the six required to achieve a category. CAT,
Categories Completed; PE, Perseverative Errors;
NPE, Nonperseverative Errors.
>
"tJ
"tJ
m
z
0
X
N
VI
Table A25.37. [WCST.37] Lineweaver et al., 1999 (MCST Administration Version): Data for a Healthy Sample Stratified into Four Age Groups, Four Education
Groups, and by Gender
Age Groups Education Groups Gender
-
Variables 45-59 60-69 70-79 80-91 1-6 7-12 1~16 17-20 Male Female
n 29 84 89 27 18 68 86 57 132 97
Age 54.31 64.77 73.93 82.19 64.83 68.75 68.88 71.04 68.04 70.44
(4.20) (3.03) (2.88) (2.60) (7.79) (9.65) (7.95) (8.00) (8.67) (8.30)
Education 12.59 13.04 14.20 14.44 2.39 10.79 14.90 18.52 12.83 14.64
(4.82) (5.00) (4.39) (2.99) (2.30) (1.62) (1.05) (1.20) (4.37) (4.65)
MCST
CAT 5.34 5.29 5.03 4.44 3.83 4.90 5.28 5.46 5.09 5.10
(1.32) (1.31) (1.51) (1.76) (1.89) (1.66) (1.26) (1.10) (1.35) (1.55)
NPE 5.34 6.81 8.42 10.89 9.89 8.54 6.97 7.23 8.45 7.20
(5.22) (6.01) (6.94) (7.52) (8.12) (6.83) (6.10) (6.56) (6.23) (6.88)
PE 2.17 2.73 2.22 3.00 9.78 2.94 1.50 1.16 1.84 2.98
(7.70) (7.38) (4.41) (4.65) (13.96) (5.96) (4.07) (1.95) (3.16) (7.54)
"CAT, Categories Completed; NPE, Nonperseverative Errors; PE, Perseverative Errors.
....
=
....
....
Copyright Acknowledgments
Tables reproduced in this book were adapted Tombaugh, T. N., & Hubley, A. M. (1997).
from the sources discussed in the chapters The 60-item Boston Naming Test: Norms for
where the tables are presented. cognitively intact adults aged 25 to 88 years.
Journal of Clinical and Experimental Neuro-
Adapted with permission from Elaevier Sci- psychology, 19, 922-932.
ence:
Appendix 19, Tables A19.15-A19.17 adapted
Appendix 19, Tables A19.36--A19.39 adapted from:
from:
Geffen, G. M., Moar, K. J., O'Hanlon,
Geffen, G. M., Butterworth, P., & Geffen, L B. A. P., Clark, C. R., & Geffen, L. B. (1990).
(1994). Test-retest reliability of a new form of Performance measures of 16- to 86-year old
the Auditory Verbal Learning Test (AVLT). males and females on the Auditory Verbal
Archives of Clinical Neuropsychology, 9, 303- Learning Test. Clinical Neuropsychologist, 4,
316. 45-63.
Appendix 11, Table A11.20 adapted from: Appendix 19, Tables A19.8-A19.10 adapted
from:
Ruff, R. M., Light, R. H., Parker, S. B., &
Levin, H. S. (1996). Benton Controlled Oral Wiens, A. N., McMinn, M. R., & Crossen,
Word Association Test: Reliability and up- J. R. (1988). Rey Auditory-Verbal Learning
dated norms. Archives of Clinical Neuropsy- Test: Development of norms for healthy
chology, 11, 329-338. young adults. Clinical Neuropsychologist, 2,
67-87.
Appendix 19, Table A19.22 adapted from:
Appendix 19, Table A19.42 adapted from:
Shapiro, D. M., & Harrison, D. W. (1990).
Alternate forms of the AVLT: A procedure Friedman, M. A., Schinka, J. A., Mortimer,
and test of form equivalency. Archives of J. A., & Borenstein Graves, A. (2002). Hopkins
Clinical Neuropsychology, 5, 405-410. Verbal Learning Test-Revised: Norms for el-
derly African Americans. Clinical Neuropsy-
Adapted with permission from Sweta {; Zei- chologist, 16, 356--372.
tlinger:
Adapted with permission from the Educa-
Appendix 10, Tables A10.17 and A10.18 tional Publiahing Foundation (American
adapted from: Paychalogkal Aaociation):
1013
1014 COPYRIGHT ACKNOWLEDGMENTS
Chapter 10, Table 10.2 adapted from : Appendix 11, Tables All.44-All.47 adapted
from:
LaBarge, E., Balota, D., Storandt, M., &
Smith, D. S. (1992). An analysis of confron- Acevedo, A., Loewenstein, D. A., Barker,
tation naming errors in senile dementia of the W. W. , Harwood, D. G., Luis, C., Bravo, M.,
Alzheimer type. Neuropsychology, 6, 77-95. et al. (2000). Category Fluency Test: orma-
tive data for English- and Spanish-speaking
Adapted with permission from Cambridge elderly. Journal of the International Neu-
Univemty Prea: ropsychological Society, 6, 760-769.
Index
Absolute zero point, 34 Rey Auditory-Verbal Learning Test, 361, 383,

Abstract reasoning, 12, 475, 496 385,393
Acculturation, effect on test perfonnance, 29, 31, 70, Stroop, 116, 131
181, 315, 481. See also Marin Marin Trailmaking Test, 64-65
Acculturation Scale Verbal Fluency Tests, 201
Activities of Daily Living (ADLs), 14, 59, 314, 476, Wisconsin Card Sorting Test, 503-505
500,506 American Academy of Neurology, report, 13
Adaptations of neuropsychological tests, linguistic and Ammons Quick Test, 75
etdtwnd,28, 178-179,287 Annett Handedness Questionnaire, 429, 449,
Administration. See also Variability in test 456,470
administration and scoring Appearance, as aspect of mental status, 12
procedures, considerations in selection of nonnative Anny Individual Test Battery, 59, 63
data, 18-19 Attention
standard procedures, 9-10, 15 as an aspect of mental status, 12
of a test/test battery, 3, 9, 11, 15, 17, 38 as assessed with
Affective Auditory Verbal Learning Test, 371 Auditory Consonant Trigrams, 134, 143, 158
Affective state, 12, 16 Category Test, 475
Affective symptoms, 12 Color Trails Test, 100, 106
African American Norms Project, 227 Digit Vigilance Test, 162-163, 170
Age-related decline Paced Auditory Serial Addition Test,
in Benton VISual Retention Test perfonnance, 141-143, 158
395,398 Ruff2&7, 160-161, 170
in finger tapping speed, 442 Trailmaking Test, 60-61, 98
in grip strength, 445, 458 Wechsler Memory Scale, 337
in Grooved Pegboard perfonnance, 460, 471 auditory, 202
in Judgment of Line Orientation perfonnance, 286 as cognitive domain, 9
in memory, verbal and visual, 344 divided, 108, 134, 141-142
in naming ability, 176-177, 199 effect on test perfonnance, 5
in Rey-Osterrieth Complex Figure recall, 250 selective, 160, 170
in verbal 9.uency, 207 sensitivity to brain damage, 9-10
in Wisconsin Card Sorting Test perfonnance, 508, susauned, 141, 160, 162, 170
528,532 Attitude, test-taking, 3, 5. See also Effort, Motivation
in word-list learning, 372 Attribution Identification Test, 506
Aggie Figures Learning Test (AFLT), 372 Auditory Consonant Trigrams (ACT), 108,
Alternate forms. See also Equivalent forms; Multiple 134-140,508
fonns; Versions of the test
Auditory Consonant Trigrams, 137 Baltimore Longitudinal Study of Aging (BLSA), 87,
Benton VISual Retention Test, 394 401,404,405,406,408,409,415
California Verbal Learning Test, 363, 368 Barrow Neurological Institute Screen for Higher
Category Test, 477-480 Cerebral Functions, 441
Color Trails, 101 Base rates, 18, 22, 24, 42-43. See also Incidence of a
Complex Figure Tests, 241-242 disorder; Prevalence rates reporting, 107
Judgment of Line Orientation Test, 284 Bateria Neuropsychologica en Espanol, 29
1015
1016 INDEX
I
Bateria Woodcock-Muiioz-R, Pruebas de Habilldad Categol}' Test, 483
Cognitiva-R, 128, 233, 529 i Design Fluency Tests, 303
Battig and Montague categol}' exemplar collecti.... 368 Finger Tapping Test, 423
Bayesian Information Criterion (BIC), 50 Grip Strength Test, 447
Beck Anxiety Inventory (BAI), 128, 233 Grooved Pegboard Test, 462
Beck Depression Inventol}' (BDI and BDI-U), :128, Hooper Visual Organization Test, 275
194,232,233,268,378 ~ Judgment of Une Orientation Test, 288
Behavioral Dyscontrol Scale, 67 Paced Auditol}' Serial Addition Test, 146
Bender-Gestalt Test, 398 Rey Auditoi}'-Verbal Leaming Test, 375
Benton Visual Retention Test (BVRT), 394-41 Rey-Osterrieth Complex Figure Test, 255
Benton's approach to neuropsychological , Tactual Performance Test, 318
assessment, 18 TrailmakingTest,72
Blessed Dementia Scale (BDS), 213, 228 Verbal Fluency Tests, 209
Boston Naming Test (BNT), 39, 17~199, 202 ' Wisconsin Card Sorting Test, 513
Ponton-Satz BNT, 178, 189 j Children's versions, for
Revised Children's BNT, 175 1 Boston Naming Test, 175
Spanish BNT, 178-179 ; California Verbal Leaming Test, 367~
Boston Process Approach to neuropsychologicali Color Trails Test, 102
assessment, 18
Boston Qualitative Scoring System for the
I Trailinalcing Test, 59
Cbcumlocution, 174, 177
Rey-Osterrieth Complex Figure (BQS$, 247 Clinical comparison data, 9
Brain damage I Cochrane Collaboration, 45
claims in forensic proceedings, 10 Coef&cient of determination, 42
effects on cognitive functioning. 9-10, 69, 72 j Cognitive
Brain dysfunction, assessment of, 3, 5, 10, 17, 3J4, abilities, 13, 14, 15, 17, 33
420,444 I domains,3, 10, 13,16,21
general,273 ( dysfunction, 13, 18, 24
lateralized, 273 j
Brain injury (traumatic) 3, 13, 61, 100, 365-366 445,
flexibility, 106, 202
functioning. baseline level, 14
500, 50i-503. See also Head injury 1 inhibition, 108
outcomes, 59, 161, 476 profile, 15, 18
severity, 365-366 slippage, 101
mild, 61, 98, 101, 142, 161 status, 12, 1~14
moderate, 475 strengths and weaknesses, ~. 15, 33
Brain reserve capacity, 101 Cognitive/information-processing mechanisms
Brief Repeatable Battel}' of Neuropsychologic tests, contributing to test performance
142, 153, 158-159 ' Auditol}' Consonant Trigrams, 134, 141
Brown-Peterson Consonant Trigram Memol}' Benton Visual Retention Test, 394
Task, 134 Boston Naming Test, 174, 176
Cancellation Tests, 160, 170
Calculations ability, as a component of mental Categol}' Test, 475
status, 12 Color Trails Test, 99-100, 106
California Card Sorting Test, 514 I Design Fluency Test, 298-299
California Verbal Leaming Test (CVLT, CVLT-n>, 29, Digit Vigilance Test, 162
341, 362-368, 369, 370
Canadian Study of Health and Aging. 220
Cancellation tests, 1~170
I
I Finger Tapping Test, 420
Grip Strength Test, 444
Grooved Pegboard Test, 460
Cardiovascular Health Study (CHS), 195 I Hooper Visual Organizations Test, 27~274, 277
Category Test, 329-330, 442, 475-495, 479, 500,<506 Judgment of Line Orientation, 284
Cattell's Matrices, 511 ' Stroop Test, 108, 113
Ceiling effect, 340-341, 346, 369 ! Paced Auditol}' Serial Addition Test, 141, 143, 158
CERAD List-Leaming test, 370. See also Conso4ium Rey Auditoi}'-Verbal Leaming Test, 359-361
to Establish a Registry for Alzheimer's 1 Rey-Osterrieth Complex Figure, 241, 249-251
Disease Ruff 2&:7 Selective Attention Test, 161
Charlotte County Healthy Aging Study, 92 Tactual Performance Test, 312
Children/adolescents, ieferences to nonnative Trailmaking. 60, 66, 88, 98
data sources Verbal Fluency, 202-203, 204
Auditol}' Consonant Trigrams, 137 Vuual Form Discrimination Test, 278-279
Benton Visual Retention Test, 402 Wechsler Memol}' Scale (WMS, WMS-R, WMS-III,
Boston Naming Test, 182 WMS-IIIA), 337, 338-342
Cancellation Tests, 164 WISOODSin Card Sorting Test, 496, 499, 506, 507
INDEX 1017
Cognitive set, 496 Greek,65

Cohort effect, 180, 195 Hebrew, 62, 66, 110, 123, 206, 371
Color-Form Sorting Test, 496 Hispanic/Spanish, 102, 104, 105, 110, 127,
Color Trails Test (CIT), 65, 67, 99-107 128,178,205-206,251,282,287,
Color vision, 397 370,371,375,511,523,529
Complex Figure Test (CFT), 241 Holland, 423, 436
Comprehensive Employment Training Act (CETA), Indian,274,371,401,406
428, 448, 462 Italian, 55,110,251,401,416,423
Comprehensive Norms for an Expanded Jamaican, 370
Halstead-Reitan Neuropsychological Battery Japanese, 110
(1991 manual and its 2004 revision), 29 Japanese-AJnerican, 440
and Boston Naming Test, 177, 183 Korean, 179,367,401,407
and Cancellation Tests, 164, 166 Latin AJnerican, 371
and Category Test, 481, 493-494 Native AJnerican, 179
and Finger Tapping Test, 420, 437 Netherlands, 401, 416
and Grip Strength Test, 444, 453 New Zealanders, 65, 179, 367
and Grooved Pegboard Test, 467 Norwegian,206,401,412,485
and List-Learning Tests, 362, 363 s~. 110.179,401,415
normingapp~.23 Swiss, 375
and Tactual Performance Test, 316, 332 Turkish, 135, 140
and Trailmaking Test, 82-83 Venezuelan, 401, 408
and Verbal Fluency Tests, 200 ~etnamese, 110, 126
and WISCOnsin Card Sorting Test, 514 Culture, effect on test performance, 8, 11, 15,
Concentration, 10,98, 106,143,158,475 28-30,31,69-70,159,199,399,511.
Conceptformation,475,496,506 See also Acculturation; Language
Concussion, 161 Cultural bias, 369
Con6.dence interval, 41, 48, 50, 51 Culturally fair
Connecticut Pictorial Learning Test (COPLT), 368 items, 179
Connections test, 66 test, 100-101
Consortium to Establish a Registry for Alzheimer's Cutoff criterion
Disease (CERAD), 29, 177-178, 179 classification accuracy 23-24, 44-45. See also
Content of thought, 12 Diagnostic accuracy; Error, false positive
Continuous Performance Test (CPT), 256, 398 nriSclassification rates
Controlled Oral Word Association (COWA), 19, 162 Category Test, 476-477, 478, 484-485, 488, 489,
Coordination, assessment of, 12 491,495
Cornell Medical Index, 84 Finger Tapping Test, 420-421, 422, 424, 434, 443
Corsi Cube Test, 339 Grip Strength Test, 445
Cost-benefit ratio, 38, 44 Grooved Pegboard Test, 460, 471
Cranial nerves, assessment of functioning. 12 Stroop Test, 130
Criterion measure, external, 17, 24, 44 Tactual Performance Test, 313, 333
Cross-cultural assessment, 28-29, 67 Trailmaking Test, 62, 63, 79, 84, 87, 98
with Color Trails Test, 99-100, 106-107 Verbal Fluency, 216
tests, guidelines for development, 28 for dementia, on MMSE, 42
Cross-cultural Neuropsychological Battery, 221 for impairment, 22, 23, 72, 73, 94, 117, 272, 423,
Crovitz-Zener Test of hand dominance, 432, 464 464,484
Culture- and ethnicity-specific test versions and/or for performance levels, 21, 23, 281-282, 424
normative data sets (those hosed on USA and selection of, 44-45, See also Receiver Operating
Canat:lfan samples are not Included), 28-29 Characteristic curve
A£dcan-AJnerican, 179,369,370,375,389-390 for suboptimal effort, 367
A£dcan-Caribbean, 65, 179 utility of, 23, 63, 64
Arabic, including Egyptian, 65-67, 144, 401, 413,
414,492 Data-driven app~ to neuropsychological
Australian, 179,370,381,423,453,456,490 assessment, 17
Brazilian, 206, 362, 370, 371 Decision-making process, 17
Chinese, 54,102-103,105,110,206,251,362 in activities of daily living. 14, 30
Colombian, 423, 436-437 and clinical judgment, 14, 17, 18, 30, 40, 42, 44
Czechoslovakian, 110 as cognitive domain, 14
Danish, 65, 380 non-psychometric factors in, 15, 16
FlenriSh,65, 179,206,362 in Rohling's Interpretive Method, 25-26
French,367,401,410 and statistical properties of tests, 33, 40, 41,
German,110 43,50
1018 INDEX
Decision theory, outcomes, 42-43, 44 with Rey-Osterrieth Complex Figure, 251-253

Delay inteiVal, length of, 242, 358 with Stroop Test, 112--114, 132--133
Delis-Kaplan Executive Function System (D-KEFS), with Tactual Performance Test, 313-315, 333
66,110-111,201,299 with Trailmaking Test, 63, 67-70, 96-98
Delusions, 12 with Verbal Fluency Tests, 206--208
Dementia with VISual Form Discrimination Test, 280
assessment, 13, 14, 17, 62, 87, 250, 368, 371, 395. with Wechsler Memory Scale (WMS-R, WM5-lli,
See also Diagnosis, differential, of dementia WMS-IIlA), 344-345, 355-356
8uency deficits in, 205 with Wisconsin Card Sorting Test, 508-511, 532
frontotemporal, 108, 499-500 Denman Neuropsychology Memory Scale, 244, 340
1..-y body. 285 Depression. See also Disorders and conditions affect-
vascular, 279, 366, 368 ing cognition, electroconvulsive therapy
Dementia of Alzheimer's Type/Alzheimer's disease, 13, and Benton Visual Retention Test, 396
22, 42. See also Consortium to establish a calculation of selection ratio, 43
registry for Alzheimer's Disease and Category Test, 476
and Benton Visual Retention Test, 395 in the context of mental status examination, 12
and Boston Naming and Verbal Fluency tests, effect on test performance, 62, 70, 161
174, 176 and Grip Strength Test, 445
calculation of base rates, 42-43 and Judgment of Line Orientation, 285, 297
and California Verbal Learning Test, 366, 368 and Stroop Test, 108
and CERAD List-Learning Test, 370 and Verbal Fluency Tests, 208
and Design Fluency Tests, 299 and Wisconsin Card Sorting Test, 501, 522--523
and Finger Tapping Test, 421 Design Fluency Tests, 298-311
and Hopkins Verbal Learning Test, 369 Diagnosis,3, 10,17,24
and Judgment of Line Orientation Test, 285 differential, 17, 33
and Rey Auditory-Verbal Learning Test, 360 of dementia, 13, 18, 27, 42, 98, 206, 361, 395.
and Stroop Test, 109 See also Dementia; Dementia of
and Trailmaking Test, 63, 84, 92 Alzheimer's Type
and Verbal Fluency Tests, 205, 213 of depression, 13. See also Depression
and Visual Form Discrimination Test, 279 of malingering. 22, 27, 476, 502-503
and Wisconsin Card Sorting Test, 499-500, 530 of vascular dementia, 13
Dementia Rating Scale (DRS; Mattis, 1988), 185, 192, Diagnostic accuracy, See also Cutoff criterion,
214,216,277,293,347,349,517 classification accuracy; Discriminant
Demographic factors, effect on test performance 3, function analysis
7-9,16,18-20,22,24,28-29,30-31,32,38, Cancellation Tests, 161
46, 50, 54, 55. See also Culture; Occupation; Color Trails Test, 99, 101
Reporting of test results Wisconsin Card Sorting Test, 499-500, 502-503
age,60, 62, 98,132,235,269,270,354,391,393, Diagnostic use of Neuropsychology, 3, 13-14
430,441-442,457,470,495 Digit Cancellation Test, 160
with Auditory Consonant Trigrams, 135 Digit repetition task, 29
with Benton VJSUal Retention Test, 398-399 Digit Span test, cognitive processes, 28, 134
with Boston Naming Test, 178, 180-182, 197-199 Digit Symbol Test (DST), 108
with Cancellation Tests, 162, 163, 170 Digit Vigilance Test (DVI'), 160, 162--164, 166-170
with Category Test, 476, 480-481, 489, 493 Direct Assessment of Functional Status (OAFS)
with Color Trails Test, 101-102 scale, 28
with Design Fluency Tests, 301, 310-311 Disability, evaluation for, 14
education,60,98, 132,199,236-237,269,270,354, Discriminant function analysis, classification accuracy,
391,470-471,495,501 112, 178, 506, 526
ettuucity, 28-30,31,208,287,511,532 Disorders and conditions (medical and psychiatric)
with Finger Tapping Test, 421-422, 431, 442 affecting cognition, 13. See also Brain injury;
gender, 72,103-104,237,304,392,422,442,445, Dementia; Dementia of Alzheimer's Type;
448,450,458,464,471 Depression; Head injury; Postconcusslon
with Grip Strength Test, 445, 451 syndrome
with Grooved Pegboard Test, 460, 463, 467 apolipoprotein E phenotype, 70, 395
with Hooper Visual Organization Test, 272, 274 alcohol exposure in utero, 108, 134, 300
intellectual level, 236--237, 426, 447, 455, 495, 531 alexia, 279
with Judgment of Line Orientation Test, 286 amnesia, 249-250, 360, 361, 371. See also Disorders
with List-Learning Tests, 372-374 and conditions affecting cognition, Korsakoff
with Paced Auditory Serial Addition Test, syndrome
143-145, 159 transient global, 108, 299
with Rey Auditory-Verbal Learning Test, 362 amyotropic lateral sclerosis, 299-300
INDEX 1019
aneurism, anterior communicating artery, 360 organic memory impairment, 250

anomia, 273 osteoporosis, 70
antisocial personality disorder, 476 chronic pain/cingulatomy, 142, 300
anxiety, 38, 422 Parkinson's diseaselpallidotomy, 13, 134, 177, 279,
aphasia, 109,279 285,299,360,371,395,499,500
batteries, 182 Post-Traumatic Stress Disorder, 396
Broca's, 177 psychopathic traits, 501--502
ftuent vs. nonftuent, schizoaffective disorder, 300
naming deficit in, 177 schizophrenia, 108, 134, 161, 163, 250, 285, 299,
Wernicke's, 177 359,396,475,498,500,502,506,523-524,
attention deficit hyperactivity disorder, 108, 134, 525,526
142,249,299,396 Schizotypal Personality Disorder/traits, 142, 501,
autism, 299 522,524
bipolar affective disorder, 396, 476, 506, 526 sex hormone levels, 70
borderline personality disorder, 300 sleep apnea, 162
brain tumors, 371 sleep disruption, 142
cardiac surgery, 13 somatization tendency, 15
cardiopulmonary bypass, 397 somatoform symptoms, 27-28
cardiovascular disease, 208, 502 spinal cord injury, 361
cerebral atrophy/ventricular enlargement, 68 stroke. See Cerebrovascular accident
cerebrovascular substance use/abuse, 60, 63
accident (stroke), 13, 250 alcoholism,314,396,397,411,476,502
brain pathology, 109 cannabis, 142, 396
risk factors, 135, 208 chronic caffeine use, 109
cholinergic system dysfunction, 285 cocaine, 396
chronic fatigue syndrome, 142, 475 heroin, 396
chronic obstructive pulmonary disease (COPD), smoking. 142, 397, 411
162,163,397,449,502 suicidality, 299
conversion disorder, 28 tardive dyskinesia, 396
~. 70,142,397,476 temporal lobectomy, 134
Down syndrome, 499 toxic exposure, 70, 250
dyslexia, 300 lead, 397
electroconvulsive therapy, 366, 371, 396 mercury, 397
endogenous cholesterol synthesis, 109 sarin, 397
end-stage pulmonary disease, 371 solvents, 142, 397
epilepsy surgery, 13 workplace chemical exposure, 396-397
epilepsy/seizures, 299, 396 Turner's syndrome, 396
frontal lobe, 108 unilateral neglect, 250
temporal lobes, 250, 371, 396 white matter lesions, 135, 502
temporal lobe, right, 396 Distribution
factitious disorder, 15 heterogeneous/homogeneous, 40, 46-47
fragile X syndrome, 396 kurtosis, 179, 199
frontal lobe syndrome, 361 negatively/positively skewed, 20, 23, 25, 38-39
health status, 113, 481 in Boston Naming Test, 179, 199
HIV infection/AIDS, 13, 67, 100, 101, 142, 161, 360, in California Verbal Learning Test, 363
370,396 in Hopkins Verbal Learning Test, 369
hormone replacement therapy, 95, 163, 397 in Rey Auditory-Verbal Learning Test, 392
Huntington's disease, 396, 499-500 in Rey-Osterrieth Complex Figure, 270
hypertension, 70, 208, 397, 476 in Trailmaking Test, 64, 82, 98
hypnosis, 300 in Wechsler Memory Scale, 346
Klinefelter's syndrome, 108, 134 normal,20,36-37
Korsakoff syndrome, 14, 134, 249 not meeting assumption of normality, 38-39, 52,
late-life psychosis, 108 197,270
learning disabilities, 14, 18, 67, 314, 396, 460 sampling, 19
leukemia, acute childhood, 250 standardized, 38
liver transplant, 397 Drug Abuse Treatment Outcome Study (DATOS), 60
lung cancer treatment, 397 Dual-baseline approach to repeated testing. 367
medical condition, 70 Dual-code cognitive neuropsychological model, 250
miotonic dystrophy, 109, 134
multiple sclerosis, 13, 142, 158, 371, 396, 475, 502 Edinburgh Handedness Inventory, 294, 456
Obsessive-Compulsive Disorder, 250, 299, 314, 501 Effect size, 26, 46
1020 INDEX
;r.:·,
Effort In test-taking. 5, 16. See also Attitude;
differential. of malingering; Motlva ·
sequential. 62, 63
sequencing. 62
assessment of, 5, 27-28, 62, 63, 250, 279, . , 367 shifting. 62
below chance pemrmance, 5, 279 1 size, 394
probability theory-based tests, 5 sources of, 15, 53-54
Emoticmal dysfunction, 4. See also Depression; ~ania; systematic, In data reporting. 45, 46
Rapid cycling ! In test construction, 15
Emoticmal factors In test-talcing. 38 ) In test Interpretation, 15
Emotional status, 4, 15 • In test performance, 59, 111, 298-299
Equipercentile equating. 363 vertical vs. horizontal, 285
Epilepsy. See Disorders and medical conditions Ethical concerns In test administration and scoring.
affecting cognition; Seizures 6,10
Equivalent forms, 101, 107, 178, 368, 368. See also Evidence-based clinical decisions, 46
Alternate forms; Multiple forms; Ver+ns of Executive
the test dysfunction,98, 133,140,237,250,298,359,
Error 501,531
altered sequence, 138 ~.3, 14,27,30,60, 106,108,202,389,476,
analysis method, · 496, 506, 507. See also Nei'YOUS system,
Benton V"uual Retention Test, 394, 395, 4de function/dysfunction of, frontal lobes
Boston Naming 'fest, 174-176, 186 system,142
Rey Auditory-Verbal Learning Test, 360 Expectancy tables, 34
Trailmalcing Test, 62-63, 93, 98 Experimental tests, ~11
In calculation of sQtlstics, 38
In clinical decision-malcing. 40, 42, 44 Factor analytic studies/factor structure of tests, 41
color-sequence, 101 Auditory Consonant Trigrarns, 134
commission, 398 Benton Vuual Retention Test, 397
oostof,40 Boston Naming Test, 189
In data recording. 38, 46, 49 California Verbal Learmng Test, 362, 365, 367
distortion, 394, 395, 398 Cancellation Tests, 162, 163
execution, 38 Category Test, 476
false positive misclassi6cation rates Design Fluency Tests, 300
for Boston Naming Test, 199 Finger Tapping Test, 420
for Category Test, 476, 489, 495 , Rey Auditory-Verbal Learning Test, 359
for Finger Tapping Test, 420-421, 433, 44:l Stroop Test, 108
for Grip Strength Test, 445, 450, 458 ! Trailmalcing Test, 60
for Grooved Pegboard Test, 460, 471 Verbal Fluency Test, 202, 219
for Rey Auditory-Verbal Learning Test, 371f Vuual Form Discrimination Test, 397
for Tactual Performance Test, 313 ' WHO-UCLA Auditory Verbal Learning Test, 370
for Trailmalcing Test, 62, 94 i Wisconsin Card Sorting Test, 501, 506-508, 517, 531
for Wisconsin Card Sorting Test, 502-503 1 Federal Aviation Administration/Equal Employment
false positive vs. falle negative, 38. See also ~on Opportunity Commission (FAAIEEOC),
theory, outcomes i 103, 121
Inter- vs. intraquadrant, 285 File-drawer problem, In meta-analysis, 46
Intrusion, 360, 364, 365 Finger Oscillation Test, 419
letter sequence, 136 Finger Tapping Test (FTT), 162, 41~. 460
of measurement, ~1 Five Point Test, 298, 299
mispe~tion, 173, 177 Fixed vs. flexible battery approach, 17-18, 25, 316, 481
misplacement, 394 Flight of ideas, 12
near-miss, 101, 107 Flynn effect, 19
nonperseverative, 497 Forensic setting. use of neuropsychological
number-sequence, 101 ewduations,S-6.~10, 15, 16,17,27,61
omission, 394, 395, 398, 399 Forest plot, 49
peripheral, 279 1 Fragmentatlon,273
perseverative, 62, 63, 298-300, 307, 310, 365, ~. Fuld Object-Memory Evaluation, 201
395. 497, 499-S02 I Functicmal Assessment Scale, 215
publication bias, 46 Fund of information, as a component of mental
rate, 218 status, 12
ratio, 300
repetition,360 Gates-MacGinite Reading Vocabulary Test, 181
rotation, 394, 398, 399 General Processing Tree Approach, 176
scores, 116 Generation tasks, 140
INDEX 1021
Genetics, 12 Design Fluency Tests, 301

Geographic recruitment region for the samples used Finger Tapping Test, 421, 443
in this book (USA and CtJflllda locations 111'W Grip Strength Test, 445, 458
not lndtuled) Grooved Pegboard Test, 460, 464, 471
Australia, 79, 88, 130, 131, 152, 166, 186, 230, 329, Rey-Osterrieth Complex Figure, 253
353, 381,388,453,456,491-492 Tactual Performance Test, 315
China, 105 Head Injury. See also Brain injury
Colombia, 258, 436, 518-519, 521 assessment of, 14, 18, 62, 63, 64, 108, 134, 135, 141,
Denmark, 380 250,279,300,361,369,371,500,506
Egypt. 327, 414, 492 prediction of driving ability, 314
France,410,530-531 prediction of return to work, 142, 161
Germany, 469 recovery rate, 141
Holland, 436 severity, cognitive correlates, 62, 141-142, 161,
Korea, 407 204,299
India, 406 Hillsborough Elder African American Life Study
Israel, 123 (HEALS), 389
Ibdy,416,430,515,525 History, 10, 56
Mexico, 233, 266, 520 educational, occupational, 16
New Zealand, 146 family. 16
Netherlands, 153, 416 of life events, 3
Norway,320,412,485 medical, psychiatric, 4, 5, 16, 19
Spa[n,282,295,309,504,517,520 vocational, avocational, 4
S~en. 70,220,415 Hit rate (in decision theory), 43, 44
Switzerland, 375 Homoscedasticity, test for, 52
Turkey, 140 Hooper Visual Organization Test (HVOT),
United Arab Emirates, 413 27~277
United Kingdom, 72. 150, 153, 213, 255, 256, 293, Multiple Choice HVOT (MC-HVOT), 273
295, 309-310, 379, 530 Hopkins Verbal Learning Test (HVLT, HVLT-R),
Venezuela, 408 368-369, 389-390
Geriatric Depression Scale (CDS), 88, 91, 96, 185, 192, Hypothesis-driven approach to neuropsychological
215,228,235,276 assessment, 18
Global Neuropsychological Deficit Scale, 25
Grade equivalent, 34 Impairment Index. See Halstead Impairment Index
Grip Strength Test, 421, 444-458, 460 Incidence of a disorder, 42, 272
Grooved Pegboard Test (GPT), 5, 20, 421, 459-471 Independence assumption, 53-54
Independent functioning in ADLs, 14, 17, 31,
Hachinski Ischemia Rating Scale, 190 33,38
Hallucinations, 12 Inhibition of responses, 496
Halstead Impairment Index, 25, 419, 424, 476, 485 Initiation
Halstead-Reitan Average Impairment Rating. 25 delicit,298
Halstead-Reitan Battery (HRB) measures, 300, 310
and Category Test, 475-476 Insight, as a component of mental status, 12
in clinical use, 17, 23 Interference
and Finger Tapping Test, 419 factor (in Stroop Test), 120
and Grip Strength Test, 444 index (in Color Trails Test), 100, 107
and Tactual Performance Test, 312, 314 proactive vs. retroactive, 134, 359, 360, 365-366,
and Trailmaking Test, 59 381,386,389
Halstead-Russell Neurops)'Chological Evaluation release from, 366
System (HRNES and HRNES-R) score (in Stroop Test), 123, 126
and Boston Naming Test, 183 Interhemispheric neural transfer, 425
and Category Test, 481 Intermanual differences
and Finger Tapping Test, 420, 438 Asymmetry Index, 440
and Grip Strength Test, 454 in Finger Tapping Test, 421, 433, 435, 442, 443
and Grooved Pegboard Test, 468 in Grip Strength Test, 444-445, 452, 453, 458
norming approach, 22 in Grooved Pegboard Test, 460, 465, 471
and Tactual Performance Test, 316 reversal in lntermanual differences, 425, 433, 443
and Trailmaking Test, 71 in Tactual Performance Test, 313, 319
Hamilton Depression Rating Scale, 229 Interpretation of test performance, 3, 5-6, 7, 10,
Hand Dynamometer. See Grip Strength Test 15,98
Handedness, effect on test performance, 64-65 accuracy, limitations, 10, 199, 237
Category Test, 481 content-referenced, 34
1022 INDEX
Interpretation of test performance (conflnued) Lexi~

criterion-referenced, 34 access, 176
domain-referenced, 34 deficit hypothesis, 176
norm-referenced, 26, 34, 98 network, 176
of raw scores, 11-12 retrieval, 173, 174, 176
lnterquartile range test, 49 storage, 202
Interval scale of measurement, 34 Likelihood ratio, 23, 44-45
Interview, with patient and significant others, 16 in meta-analysis, 46
lntraindividual Measure of Association (UMA), 25 List-learning tests, 29, 357-393
Item Loosening of associations, 12
an~ysis. 175,182,195,273-274,477 Luria-Nebraslca Neuropsycho~ Battery
difficulty, 15, 26, 179, 273 (LNNB),18
dbcrinllnation, 15,26 Luria's approach to neuropsychoiogical assessment,
Item Characteristic Curve (ICC), 26 15, 18
Item Response Theory (IRT), 26-27
Macquarie University Neuropsychologic Normatiw
Johns Hopkins Teaching Nursing Home Study of Study (MUNNS), 353
No~ Aging. 211 M~ring. detection of, 15, 22, '1:1, 286
Judgment Mania, 12
clinical, 4, 10, 17 Marin Marin Acculturation Scale, 128, 189, .219, .233,
limitations of, 17 265, 390, 529. See also Acculturation, effect
as a component of meow status, 12 on test performance
Judgment of Une Orientation Test (JLO), Mathemat~W abilitylsldlls, 143, 144
284-297 Mayo Cognitive Factor Scales. .23
Mayo Older Americans Normat:iw Studies (MOANS),
Kame project on aging and dementia, 440 85, 125, 187, .216, .29.2, 349, 386
Kaplan's approach to neuropsychologi~ assessment, overlapping interval technique, .24
15, 18 scoring system, .23
Kernel Density Estimate, "K-density plot», 52, 392 McCarthy SWes for Children's Abilities, .201
Kinesthetic abilities, 420 Measurement, See also Error, of measurement
Knox Cube Imitation Test, 339 method of, 34
Kurtzke's Disability Status S~e (DSS) scales of, 34
theory of, 34
Lagrange multiplier statistic, 52 units of, 33
Language Mediw-Iegal evaluations, See Forensic setting. use of
as a cognitive domain, 12 neuropsychologi evaluations
bilin~m. 2S-29, 113-114, 159, 199,208,221 Memory and Aging Study (MAS), 91, 195
effect of native language on test performance, 2S-29, Memory, 12. See also Learning
99,107,113-114,199,208,221 control. 365
English as a second language, 2S-29, 99, 107, as information-processing domain, 9, 475
113-114,199,208,221 laterallzation, 343
limited mastery of, 2S-29, 99, 107, 199, 208, 221 metamemory, 361
test translations/adaptations, 2S-29 soun:e/cootelt, 360, 364, 398
Lateral Dominance Examination, 444 Memory compartments
Laterallzation of brain function/dysfunction, 3, 72, 312, immediate, 337,340,359,394
343, 421, 443, 444, 448, 458, 459, 460, 498. long-term, .203
See also Nervous system, interhemispheric short-term, 134, 202
differences working. 60, 108, 134, 140, 141, 143, 341, 501
Learning. 475. See also Memory Memory mechanisms
curve, 359-360 auditory process, 342
incidenw, 361 consolidation, 360, 365
list vs. item, 371 encodinf§'acquisition, 339, 343, 346,
nonverb~. 3, 339 359-360, 363
single trial vs. over trim, 342 forgetting. 9, 345, 347, 35.2, 360
strategy, 360, 365 reWI. 360
verb~. 3, 29, 359-360 cued, 390
Letter and Symbol Cancellation Task, 160 recognition, .279, 339, 360, 364, 366
Letter fluency, 29 recognition discriminatioo index, 390
Level of consciousness, as a component of meoW retrieval, 339, 342, 343, 346, 359-360, 363, 365
status examination, 12 from short-term vs.long-term storage, 371
Levels of data integration, 1~17 storage/retention, 342, 346, 359-360, 365
INDEX 1023
Memory modalities in test-taking. 5, 16, 27-28, 38. See also Attitude;

auditory, 134, 337, 341, 343 Diagnosis, differential, of malingering; Effort
multimodality, 343 in test-taking
spatial, 312, 338 Motor
tactile, 312 ability, 444
verbal, 134, 343, 359 activity. 12
visual, 279, 337, 339, 340, 394, 475 control, 420, 460
Memory Quotient (MQ), 339, 376 functioning. assessment of 5, 20, 162
Mental status, 16 systems, 12
assessment, 12-13 Multicenter Aids Cohort Study (MACS), 83, 95, 122,
components, 12 212,234,268,382,385,389,466,467
Meta-analysis Multilingual Aphasia Examination Battery, 200, 206
advantages and limitations, 46-47, SS-56 Multiple forms of tests, 99. See also Alternate forms;
cluster option, 48, SO, 54 Equivalent forms; Versions of the test
comparative (case-control) vs. noncomparative Multitrait-multimethod matrix, 17
(descriptive), 46 Multicultural neuropsychological assessment, 28-29.
fixed vs. random effects, 48, SO See also Culture, effect on test performance;
heterogeneity in study results, 46-47 Culturally fair test
history, 45
Quality of Reporting of Meta-Analyses (QUO ROM) Naming
statement, 45 ability. 176
weighting data, 48, SO confrontation, 173, 174, 182, 198
Midpoint age interval technique, 24. See also effect on Hooper VOT performance, 273, 277
Overlapping Interval Technique National Multiple Sclerosis Society, 142, 153,
Boston Naming Test, 187 158-159
Judgment of Line Orientation Test, 292 Nervous system, function/dysfunction of
Rey Auditory-Verbal Learning Test, 386 amygdala, 395
Stroop, 126 anterior brain areas, 161
Trailmaking Test, 85 right. 249
Verbal Fluency, 216 anterior vs. posterior brain areas, 395
Wechsler Memory Scale basal ganglia. 204, 395
Miles ABC Test of Ocular Dominance, 435, central nervous system, 3, 4
452,465 cerebellum, 205
Mini-Mental State Exam (MMSE), 42-43 corpus callosum, 460
and Boston Naming Test, 184, 186, 189, 191, diffuse brain dysfunction, 204, 218, 395, 475
194, 196 frontal-extrastriate system, 499
and Benton Visual Retention Test, 410 frontal-limbic-reticular activating system, 135
and California Verbal Learning Test, 366 frontal lobes, 10, 61, 62, 63, 93, 106, 108, 134, 177,
and Finger Tapping Test, 430, 436 299,310,312,359,364,389,498, SOO,S02
and Hooper Visual Organization Test, 276 basal-orbital region, 61, 62
and Judgment of Line Orientation Test, 296 cingulate gyrus. 108, 141, 204, 475
modified MMSE (3MS), 220 anterior, 499
and Rey Auditory-Verbal Learning Test, dorsolateral convexities, 61, 62, 475, 498
378,384 inferior cortex, 204
and Rey-Osterrieth Complex Figure, 261, 266 left, 108, 204, 249
and Stroop Test, 127, 131 left cortex, 204
and Trailmaking Test, 67, 96 medial, 61, 62
use in assessment of mental status, 12 motor strip, 420, 444
and Verbal Fluency, 214, 215, 229, 231, 232 orbitofrontal cortex, 364
and Wechsler Memory Scale, 349 orbitomedial cortex, 475
and Wisconsin Card Sorting Test, 528, 530 prefrontal cortex, 202, 204, 475, 498-499
Minnesota Multiphasic Personality Inventory left, 218
(MMPI, MMPI-2), 36, 125, 320, left dorsolateral, 204, 364, 498
476,487 left inferior, 364
MacAndrew Alcoholism Scale, S02 right, 204, 365, 499
Modified t-test method for small samples, 25 frontostriatal system, 204, 250
Monongahela Valley Independent Elders Survey frontotemporal region, 299, 365
(MoVIES), 92, 230 hippocampal region, 177, 205, 365
Mood, as a component of mental status, 12. See also left, 249
Affective state right, 395
Motivation, 3, 12, 16 interhemispheric differences, 161, 421
1024 INDEX
Neavous system (continued) Normative Aging Study, 231

dominant, 273 Normative data
left, 249, 273, 285, 286, 299 availability of, 7
nondominant, 177,249,273 cultural factors, 28
right,249,273,279,284-285,286,299,3l2,395 multiple sets of, 9
limbic system, 12 in research, 7
multisystem mechaoisms, 249 selection of, 7-9, 18-22, 55
occipital region, left superior, 499 in test results interpretation, 3, 18-27
parietal lobes, 420 Normative Studies Research Project (NSRP), 191
inferior, 499 Norming approaches
supramarginal gyrus, 499 comparative vs. diagnostic norms, 22
nondominant, 273 co-norming. 24, 342, 351, 354, 356
right, 249 continuous norming. 22
superior, 177 regression-based norms, 22-23
peripheral nervous system, 4 single-case approach, 24
posterior brain areas, 312
right anterior quadrant, 273 i Occupation, effect on test performance, 399
subcortical sensory and motor tracts, 420, ~ Odds ratio, 23, 45
temporal cortex, 204 Order of test administration, 421, 506
temporal lobes Organization
anterior, 177, 204 semantic, 363
dominant, 177 in verbal learning. 360
left, 134, 204-205 Organizational
medial, left, 205, 365 abilities, 496
medial, right, 365 spatial249-250, 273-274
posterior inferior, 499 de&cit, 298
right, 134, 204-205, 299 strategy,241
temporoparietal region perceptual clustering. 250
left, 204 Orientation
right, 249, 285 as assessed with Wechsler Memory Scale, 337, 338
thalamus, 395 as a component of mental status, 12
white matter, 140 Overlapping age strata technique, 368
Neural network model, 65, 203 ; Overlapping Interval Technique, 24. See also Midpoint
Neurobehavioral Core Test Battery (NCTB), 39ft age interval technique
Neurobehavioral Cognitive Status Examination, t17
Neuroimaging techniques, 12, 13, 19 1 Paced Auditory Serial Addition Test (PASAT), 141-159
cognitive correlates, 17, 60, 203 i adjusting-PSAT, 143
EEG/evoked potentials, 499 computerized versions, 143
functional imaging. 108, 141, 203-204, 285, Levin version, 142-143
299,499 PASAT-200, PASAT-100, PASAT-50, 143
fMRI, 365, 498 visual version (PVSAT), 157
PET, 364, 395, 498 Paivio, imagery values, 361
rCBF, 498-499 Parallel distributed processing models, 203
SPEer, 299, 498 Patient-centered model, 18
structural imaging Pattern analysis, 21
MRUwhite-matter hyperintensities, 68, 108, :134, Percentile rank, interpretation of 20-21, 34-35
142, 180, 196, 207, 249, 395 Pearson's methods for research synthesis, 45, 46
transcranial Doppler ultrasonography, 499 Perception
Neurological examination, 12 as a component of mental status, 12
NEUROPI,29 tactile,312
Neuropsychiatry, as a clinical discipline, 12 . Perceptual
Neuropsychological evaluation, goals and method., ~entation, 177
3-4, 13, 16, I organization,273
Neuropsychological Screening Battery for ~cs Performance levels. See Cutoff criterion
(NeSBHIS), 104, 107, 189, 219, 265, ~ level vs. pattern, 365-366
Neuropsychology ' Perseverations, 136, 138, 218, 298. See also
apphcations, 13-15 Errors, perseverative
chnical, 3, 10 Personality characteristics, 4, 16, 422
Neurosensory Center Examination for Aphasia, Phonological
200,206 encoding. 177
Nociferous cortex hypothesis, 204 loop, 203
INDEX 1025
Physical Performance Test, modified, 91, 227-228 Rapid cycling. 12. See also Affective state
Pictorial Verbal I.eaming Test (PVLT), 372 Rasch model, 27
Pilots Raven Progressive Matrices, 295, 498
age-related cognitive changes, 14 advanced, 39, 104
normative data for, 423 Raw scores, 33
as subjects, 103, 121 inclusion in reports, 32
PIN test, 20 interpretation-age/grade equivalent, 34
Pittsburgh Occupational Exposures Test Battery number of correct responses, 33
(POET), 464 number of errors, 33
Positive and Negative Syndrome Scale (PANSS), 396 percent of correct responses, 34
Postconcussion syndrome, 14-15, 141, 142, 300 quality of drawings, 33
Posttest probability, 44 time to completion, 33
Practice effect, 5, 38. See also Test-retest stability; Reactivity effect, 40
Dual-baseline approach to repeated testing Reasoning. 475
Auditory Consonant Trigrams, 136 Receiver Operating Characteristic (ROC) curve
California Verbal I.eaming Test, 366-367 in "diagnostic norms" approach, 22
Design Fluency Tests, 300, 305, 306, 311 in differential diagnosis of dementia, 87, 92,
Digit Vigilance Test, 163, 168 499-500, 530
Finger Tapping Test, 443 in meta-analysis. 46
Grip Strength Test, 456, 458 in signal detection theory, ~24
Grooved Pegboard Test, 469 Reflexes, 12
Hopkins Verbal I.eaming Test, 369 Regression approach as used in meta-analysis, 27
Judgment of Line Orientation Test, 286 Regression-based norming techniques, 22-23, 55
Paced Auditory Serial Addition Test, 159 Regression analyses on demographic variables and
Rey Auditory-Verbal I.eaming Test, 361-362, performance indices
383,384 for Auditory Consonant Trlgrams, 139
Rey-Osterrieth Complex Figure, 242 for Benton VJSUal Retention Test, 408, 412
Ruff Uc7 Selective Attention Test, 162 for Boston Naming Test, 192
Stroop,124 for Category Test, 477-478
Tactual Performance Test for Finger Tapping Test, 434
Trailmaking Test, 63, 64, 80, 89, 90 for Grooved Pegboard Test, 460, 464
Verbal Fluency, 202, 223, 226 for List-I.eaming Tests, 373, 374, 378
VJSUal Form Discrimination Test, 279 for Rey-Osterrieth Complex Figure, 252, 257, 263
Wisconsin Card Sorting Test, 505, 508, 510, 518, for Tactual Performance Test, 326-327
522,531 for Trailmaking Test, 69, 78
Pretest probability, 44 for WISCOnsin Card Sorting Test, 502-503
Predictive value, 23 Regression to the mean, 23
in decision theory, 44 Rehabilitation/remediation, 13
Premorbid functional level, 31 cognitive, 14,33,161,360
Prevalence rates, 22, 42 setting. 16
Primacy-recency effect See Serial position effect strategies in, 15, 41
Problem solving. 475, 476 Reitan-Klove Lateral Dominance Exam, 435,
Process of thought, 12 452,465
Professional communication between clinicians, 9 Relative risk analysis, 23
Prognostic predictions, 33 Reliability, 39-41
Project Safety, 86, 351 of meta-analysis, 47
Propriospatial ability, 312 of scoring. 298
Psychiatric evaluation, 12 Reliability, alternate form/interform, 39-40
Psychological assessment vs. testing. 16 Benton VJSUal Retention Test, 398
Psychopbannacology, 12 Boston Naming Test, 178
Psychometric approach to neuropsychological Category Test, 477-480
assessment, 13 Color Trails, 107
Publication bias, in meta-analysis, 46 Hopkins Verbal I.eaming Test, 368
Rey Auditory-Verbal I.eaming Test, 362, 377
Quality of life, 14, 142 Wisconsin Card Sorting Test, 504-505
Qualitative test performance interpretation, 4, 15, 249, Reliability, internal consistency, 40,135, 143, 178,202,
270,273 218,248
and California Verbal I.eaming Test, pictorial
Race, considerations in interpretation of test format, 368
performance, 31. See also Demographic and Category Test, 478
factors, effect on test performance, ethnicity and Hooper Visual Organization Test, 274
1026 INDEX
Reliability (continued) Rey Auditory-Verbal Learning Test (Rey AVLT),

method, 40 7-9, 11, 35, 357~. 372-389,
and Tactual Performance Test, 314 391-393, 341
and Visual Fonn Discrimination Test, 279, AVLTX, 361
280,286 Rey Auditory and Verbal Learoing Test:
Reliability, interrater, 15 A handbook, by Schmidt, 359, 392
Benton VISUal Retention Test, 398 Rey-Osterrieth Complex Figure, 20, 39, 241-271
Design Fluency Tests, 30()...3()1, 309, 310 Fastenau's Extended Complex Figure Test
Hooper VISUal Organization Test, 274 (ECFT), 247-248
Rey-Osterrieth Complex Figure, 248 Handbook of ROCF usage, by Knight and Kaplan,
Tactual Performance Test, 314 241,242,251
Verbal Fluency, 202 Hubley et al.'s simplified versions, 245
Wechsler Memory Scale, 340 Mack Complex Figure Test, 243
WISCOnsin Card Scrting Test, 50S Myers &: Myers' Rey Complex Figure Test and
Reliability, split-half, 40 Recognition Trial, 246
Judgment of Une Orientation Test, 286 Risk factors affecting test perfonnance, 3
Paced Auditory Serial Addition Test, 143 Rohling's Interpretive Method (RIM), 25-26
Reliability, test-retest. 39-40. See also Ruff Figural Fluency Test, 162, 298-306
Practice effect Ruff-light Trail Learoing Test, 162
Benton VISual Retention Test, 398 Ruff 2.&:7 Selective Attention Test, 160-162,
Boston Naming Test, 179-184 163-166, 170
CERAD List-Learoing Test, 370
Design Fluency Tests, 300, 309, 310 Sample
Digit Vigilance Test, 163, 168 biased, 23, 38
Finger Tapping Test, 421, 438, 439 normative, 7-9, 19
Grip Strength Test, 445, 456, 458 random, 42
Grooved Pegboard Test, 460, 465, 468, 469 1 s~. 19,23,24,46-48
Hopkins Verbal Learoing Test, 369 San Diego African American Norms Project, 29
Judgment of Une Orientation Test, 286 San Diego Neuropsychological Test Battery, 160
Paced Auditory Serial Addition Test, 143 Satz-Mogel short fonn of the WAIS, 209, 212, 225,
Rey Auditory-Verbal Learoing Test, 388 263,405
Rey-Osterrieth Complex Figure, 248 Scale of Competence in Independent living SkiUs
Ruff 2.&:7 Selective Attention Test, (SCILS), 476
161-162 Scales of measurement, 34
Trail-Making Test, 90 Schedule for Affective Disorders and Schizophrenia
Verbal Fluency, 201, 206, 218, 223 Interview, 233
VISual Fonn Discrimination Test, 279 Scoring method. See also Midpoint age interval
Wisconsin Card Sorting Test, 50S technique; Test scoring; Variability in test
Reliable change indices, 226 administration and scoring
Rennick Boston Naming Test, 177
Average Impairment Rating. 419 Color Trails Test, 101
index, 435 Paced Auditory Serial Addition Test,
method for HRB administration and scoring. . dyad method, 142
419,484 Rey-Osterrieth Complex Figure, 243-248, 271
procedures for Tactual Perfonnance Boston Qualitative Scoring System
Test, 325 (BQSS), 247
ReDDick-Lafayette Repeatable Battery, 168 Denman Itemized Scoring System, 244-245
Repeatable Battery for the Assessment of . Developmental Scoring System
Neuropsychological Status (RBANS), , . (DSS-ROCF), 245
201,371 1 Myers &: Myers scoring system, 246
Repeatable Cognitive-Perceptual-Motor Battery,!54, other systems, 243-248
162, 163 . Stroop Test, 110, Ill
Reporting of test results, recommendations process-oriented scoring system, Ill
demographically-corrected test results, 30 Verbal Fluency
inclusion of test scores, 31 qualitative scoring system (clustering/switching),
source of normative data, 9 202-203
Research WISCOnsin Card Sorting Test, 497-498
database, 18 Screening tests, cognitive, 12-13
in test development, 11 Secondary gain, considerations in test results
Residual-vs.-fitted (rvf) plot, 53 interpretation, 5
Response stylelbias, 15, 38 Selection ratio, 43
INDEX 1027
Selective Reminding Test (SRT), 370-371 of design production, 298

Free and Cued SRT (FCSRT), 371 of ROCF reproduction, 241, 245, 247, 248, 249
Verbal SRT (VSRT), 371 Stroop
Semantic effect, 108
clustering. 29, 363-364, 369 language-neutral Stroop, 110
clustering index, 364 test, 29, 108-133, 134, 140, 202, 508
cueing. 364, 366 Switching (category vs. semantic), 201
organization, 363 Symptoms, role in data integration, 4, 16
storage, 202 Systematic review, 45, 47
Semantic deficit hypothesis, 176
Sensory systems, 12 Tactual Performance Test (TPI'), 312-334, 491
Sequin-Goddard Formboard, 312 Tanrntiality, 12
Serial position effect (primacy-recency), 360, 361, Tau estimate of residual heterogeneity, 54
366,371,381,384 Taylor Complex Figure, 242
Set shifting. 108 modified, 243
Set Test, 201 Taylor-Russell Tables, 43
Sexual orientation, effect on test Test construction, 15
performance, 286 Testing conditions, optimal vs. standard, 16
Shapiro-Wilk W test for normality of residuals, 52 Test items, 28, 38, 39, 40. See also Item analysis;
Short Blessed Test (Orientation-Memory- Item ditliculty; Item discrimination; Item
Concentration Test), 228 Response Theory
Short forms of tests Test of Everyday Attention, 152, 166
Benton VJSU81 Retention Test, 398 Test scoring. 33, 38. See also Scoring method
Boston Naming Test, 177-178 criteria, 15
California Verbal Learning Test, 363 partial credit, 27
Category Test, 477-479 procedures, 19-20
Judgment of I.Jne Orientation Test, 286, standard, 9-10, 20
288,296 Test selection, 11, 17-18, 39-41
Tactual Performance Test, 313-314, 333 Test usage survey, 18, 59-60, 314, 333, 442
VJSU81 Form Discrimination Test, 279-280 Tbird-party observers, 6. See also Validity of
Wechsler Memory Scale-IliA, 342-344 neuropsychological test performance
Signal detection theory (SDT), 23 Thorndike-Lorge tables, 361
Single-case approach, 24-25 Thought disturbance, 12
Socioeconomic status, effect on test performance, 28, Thought process/content, 12-13
181,481 Trailmalcing Test (TMT), 54, 59-98, 99, 101, 106, 140,
Sources of information, used in neuropsychological 143, 163, 491
evaluation, 4, 16 alphanumeric sequencing. 67
Spanish language test administration, 28-29 comprehensive, 66
Spanish-speaking samples, 28-29 expanded, 61Hi7
Speed mental alternation test, 67
of color naming and color-word reading. 109 midrange expanded, 67
of information processing. 106, 108, 133, 140, 141, oral, 67
143, 159, 364 symbol, 67
motor, 60, 301, 420, 421 Translated versions of neuropsychological tests, 28
psychomoto~98. 162,163,170,460 Treatment
of test performance, 161 effectiveness of, 45
Speeded mental processing. 202 planning. 3, 33, 42, 44
Speed vs. accuracy in test performance, 161 response to, 14
Standard error of measurement, 40-41, 50, 52 strategies, 28, 31
in Item Response Theory, 26-27
Standard scores, 20, 35-37 Validation studies, 7
Standardization, 6 Validity, 178,278,314,334,343
of raw scores, 35-36 of clinical judgments, 17
of test administration, 9, 313 ~t.42,67,279,368,505
Standards for Educational and Psychological ~41, 143,273-274,365,370,475
Testing. American Psychological Association, content, 41
39 convergent,41, 163
Star CanceUation Test, 160 criterion-related, 41-43, 67, 365, 367
Stata, statistical package, 48 discriminant, 41, 369
Strategy ecological, 506
cognitive, 496 of hypothesis testing. 52
1028 INDEX
Validity (continued) Verbal Fluency, 200-201, 205-206

incremental, 43 VISWll Form Discrimination Test, 279
of meta-analysis, 47 Wechsler Memory Scale, 337-344
predictive, 42, 506 Vietnam Experience Study (VES), 149, 222
of symptoms, 279. See also Effort in test-takiag; Vigilance, 160, 162, 170
Motivation VISual. See also VJSUospatial
of test performance, 62. See also Effort in memory, 241, 249
test-taking; Motivation organization,277
of tests, 17,41-45 perception,273,277,278,284,394,397
Variability processing. 397
across studies In meta-analysis, 48-.50, 55 scanning. 60, 396
with Auditory Consonant Trigrams, 135 synthesis, 273
with Boston Naming Test, 180, 182-183 tracking. 60, 162, 170
with Category Test, 476 VISual Form Discrimination Test (VFDT),
in the criterion, 42 278-283, 394
with Finger Tapping Test, 419, 422-423 VISual-motor
with Grip Strength Test, 444, 446 abilities, 420
with Grooved Pegboard Test, 459-460 coordination, 460
intraindividual, 22 Visual Search and Attention Test, 143, 160
with Judgment of Line Orientation Test, 288 VISuospatial. See also VISual
with Paced Auditory Serial Addition Test, abilities/skills, 60, 284, 475
145-146 constructional ability. 3, 5, 241, 249, 274,
of physical parameters, 46 312,394
in recording strate&V of reproduction, 241, 25f deficit, 273
with Rey Auditory-Verbal Learning Test, functioning. 3, 296. See also Organintional abilities
357-359 intelligence, 274
with Rey-Osterrieth Complex Figure, 241-24i, 248, Visuospatial sketchpad, 203
254,270 Vocabulary
with Stroop Test, 110, 115 fund, effect on test performance, 181, 193,
with Tactual Performance Test, 313, 334 198, 199
in test administration and scoring. 9, 15, 38 storage, 202
of test scores, 21-22, 26, 29, 35-36, 3thl9, 40
of traits and abilities, 38 WAlS-R
with Verbal Fluency, 209, 237 norming.22
with Wechsler Memory Scale, 338 relation to neuropsychological test performance, 144
with Wisconsin Card Sorting Test, 496 WAlS-R-NI, 15,341
Verbal fluency, 180, 200-237, 508 WAlS-III, 107
category naming. 201 co-norming with Wechsler Memory Scale, 342, 351,
Controlled Oral Word Association (COWA), 2110 354,356
Controlled Verbal Fluency Task (CVFT), 200 IQ. 20,41
FAS version, 49-54, 108, 134, 200, 299 Mexican version, 128, 529
Fuld Verbal Fluency Test, 206 Warrington Recognition Memory for Faces Test,
relative to nonverbal/design fluency, 298-299, 301 5,340
Thurstone Word Fluency Test, 200 Washington Heights-Inwood-Columbia
Versions of the test. See also Alternate forms; Aging Project, 29, 401, 412, 410,
Culture- and ethnicity-specillc test vei"Sons 413,414
and normative data sets; Equivalent fous; Wechsler-Bellevue Scale, 484
Multiple forms of tests; Short forms oftests Wechsler Memory Scale (WMS), 337-356
Benton VISWll Retention Test, 394 WMS, 337-338
Boston Naming Test. 173 WMS-Revised, 338-339, 355
California Verbal Learning Test, 367-368 WMS-III, 339-342, 344, 356
Cancellation tests, 160 WMS-IIIA, 342-344, 356
Design Fluency, 298-299 White's general test for heteroscedasticity, 52
Hooper Visual Organization Test, 273 WHO-UCLA Auditory Verbal Learning Test,
Paced Auditory Serial Addition Test, 142-143; 369-370, 390-391
145-146 Wide Range Achievement Test-Revised, 208
Rey Auditory-Verbal Learning Test, recognitiop Wide Range Achievement Test-III, 414
list, 358 WISCOnsin Card Sorting Test, 29, 496-532, 108, 134,
Rey-Osterrieth Complex Figure, 241-243 162,398,476,496-532
Stroop, 110-112, 130, 133 MCST, 503, 530-531
Trailmaking Test, 61, 66-67, 88 64-card version, 503-504, 527-529
INDEX 1029
Woodcock-Jolmson-m, Tests of Achievement, 128, World Health Organization (WHO), 67, 99, 101,
233,529 369-370, 396
Word
frequency. 361 Z-scores, 20, 25, 35-37
production set, 202 Zung Depression Scale, 190, 405

Maura Mitrushina, Kyle B. Boone, Jill Razani, Louis F. D'Elia - Handbook of Normative Data For Neuropsychological Assessment (2005, Oxford University Press, USA)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Maura Mitrushina, Kyle B. Boone, Jill Razani, Louis F. D'Elia - Handbook of Normative Data For Neuropsychological Assessment (2005, Oxford University Press, USA)

Uploaded by

Copyright:

Available Formats

When Handbook of Normative Data for

Neuropsychological Assessment was originally

Kyle B. Boone, Ph.D., is Professor-in-

Jill Razani, Ph.D., is an Assistant Professor of

Louis F. D'Elia, Ph.D., is Assistant Clinical

jACKET DESIGN: E\'E SIEGEL

-Kenneth M Adams. PhD. in]oumalofClinical and Experimental Neurops_rcholog.r

-Michael R. Basso, PhD, in Neuropsychiatry, Neuropsychology. and Behavioral Neurology

Oxford University Press, Inc., publishes works that further

Oxford New York

Oxford is a registered trademark of Oxford University Press

All rights reserved. No part of this publication may be reproduced,

Library of Congress Cataloging-in-Publication Data

Printed in the United States of America

2. Use of Methodological Concepts in Neuropsychology Practice, 12

3. Statistical and Psychometric Issues, 33

5. Color Trails Test, 99

6. Stroop Test, 1oa

Relationship Between Stroop Test Perfonnance and Demographic Factors, 112

7. Auditory Consonant Trigrams, 134

8. Paced Auditory Serial Addition Test, 141

9. Cancellation Tests, 160

Modifications and Short Versions bf the BNT, 177

11. Verbal Fluency Test, 200

IV. PERCEPTUAL ORGANIZATIQN: VISUOSPATIAL AND TACTILE

13. Hooper Visual Organization ~-est, 272

14. Visual Form Discrimination Test, 278

15. Judgment of Line Orientation, 284

16. Design Fluency Tests, 298

17. Tactual Performance Test, 312

V. VERBAL AND VISUAL LEARNING AND MEMORY

19. List-Learning Tests, 357

20. Benton Visual Retention Test, 394

VI. MOTOR FUNCTIONS

Results of the Meta-Analyses of the Finger Tapping Test Data, 441

22. Grip Strength Test (Hand Dynamometer), 444

23. Grooved Pegboard Test, 459

VII. CONCEPT FORMATION AND REASONING

25. Wisconsin Card Sorting Test, 496

Copyright Acknowledgments, 1013

lntarpratation --+ REPORT

suggests significant loss in spatial/construc- In neuropsychological practice, one exam-

Table 1.1. Published Rey Auditory-Verbal Learning Test Nonnative Reports

Rey, 1964 70-90 15 No info. 9.5 Switzerland

INTERFACE OF NEUROPSYCHOLOGY commands, calculations, abstract reasoning,

cognitive abilities. If indications of cognitive plementation of rehabilitative efforts to opti-

TBis, infections, metabolic abnormali- possible changes in lifestyle, and prompts

syndrome vs. somatization tendency, with the consistency in performance of

2.17-3.00 99 1 ~133 '5,67 0.63-0.65 74 26

SD, standard deviation.

an Expanded Holstead-Reitan Neuro- Ivnik et al. (2000) demonstrated the

a cutoff involves computation of the area overlapping interval strategies described

impact of heterogeneity, sa~ple size, content of knowledge available to that

The administration, scoring, and interpreta- rehabilitation, evaluating a patient's ability to

In contrast to physical measuring scales that normative or standardization sample is

Rey Auditory-Verbal Learning Test STANDARDIZATION OF RAW SCORES

INTERPRETATION OF INFREQUENT g. Emotional factors and level of endur-

3. The split-half method involves splitting Typical levels of reliability attained by

assessed using a patient's psychiatric di- Decision Theory

a hit rate of 80/125 = 64%. In other words, the Selection ratio

30-24=6 FN Incremental validity

In the past decade, neuropfjychology be insufficient to detect or rule out a

and that variations in study estimates are meta-analysis refers to reproducibility of