Professional Documents
Culture Documents
net/publication/235995920
CITATIONS READS
0 17,481
Some of the authors of this publication are also working on these related projects:
Patterns of Adapting to Health, Self-Reported Health Status, and Risk Odds of Type 2 Diabetes View project
All content following this page was uploaded by Frederick H. Navarro on 20 May 2014.
Frederick H. Navarro
Walden University
Running head: WOODCOCK – JOHNSON III 2
Abstract
This paper reviews the Woodcock-Johnson Tests of Cognitive Ability, Third Edition (WJ-III
COG). The paper summarizes its historical development, reviews its psychometric properties,
and describes the materials and administration associated with its use. The reliability and
validity of the WJ-III COG is discussed, as well as its application in variable centered and
person-centered assessment of cognitive function. The paper concludes with a discussion of the
WJ-III COG strengths and weaknesses and implications for further development.
Running head: WOODCOCK – JOHNSON III 3
The Woodcock-Johnson Tests of Cognitive Ability, Third Edition (WJ III COG) is a set
of tests administered individually which assesses an extensive range of intellectual and cognitive
abilities for children up to two years old and adults up to and over the age of 90 (Blackwell,
2001). The WJ III COG is paired with the Woodcock-Johnson III Tests of Achievement (WJ III
ACH) which are both co-normed on the same sample (Cizek, 2003). This allows the exploration
of the link between cognitive abilities and academic achievement (Cizek, 2003). The WJ III
(WJTCA) of 1977 (Blackwell, 2001). The WJ III COG revision is designed to measure
cognitive ability constructs that are a part of the Cattell-Horn-Carroll (CHC) three stratum theory
of cognitive abilities including many narrow abilities (Stratum I), a limited set of broad cognitive
abilities (Stratum II), and a single general intelligence factor (Stratum III; Sanders, McIntosh,
Dunham, Rothlisberg & Finch, 2007). The WJ III COG also has a diagnostic supplement (DS;
Sares, 2003). Many reviewers note that the technical manuals of the WJ III COG clearly follow
many guidelines outlined for reliability and validity in the Standards for Educational and
Historical Development
The historical development of the WJ III COG closely parallels theoretical conceptions of
stages: First, the conception of intelligence as a single ability, second, as a pair of abilities, third,
as a small set of multiple abilities, fourth, is a larger complete set of abilities, and fifth, as a set of
stratum where a few abilities are behind a broader set of abilities. The conception of intelligence
as a single factor or ability was first proposed by Spearman back in 1904 and is generally
Running head: WOODCOCK – JOHNSON III 4
referred to as Spearman's g (Kane, Oakland, & Brand, 2006). This conception of intelligence
was supported by the earliest version of the Stanford-Binet test of intelligence defined by the
distinguished by verbal intelligence (VIQ) and nonverbal or performance intelligence (PIQ) was
The first predecessor of the WJ III appeared in 1979, and at the third stage of intelligence
conceptualization which postulated a set of broad abilities (Woodcock, 2003). The WJTCA was
not based on any theory of intelligence, but was unique in that it offered co-normed tests of
The first revision of the WJTCA occurred in 1989 with the Woodcock-Johnson Psycho-
Educational Battery-Revised (WJ-R; Sanders, et al., 2003). The WJ-R was designed in response
intelligence (Gf) and crystallized intelligence (Gc), and was expanded to measure seven
(Sanders, et al., 2003) to eight (Kane & Brand, 2006) broad cognitive abilities, as well as two
other abilities assessed in the achievement portion of the batteries (Sanders, et al., 2003). The
normative data set for the WJ-R consisted of 6,359 subjects with ages ranging from 24 months to
95 years old drawn from 100 different geographic areas in the U.S. (Kane, et al., 2006). The
Margolis, & Barenbaum, 2001) correspond to the Horn-Cattell model of cognitive abilities which
included fluid processing, visual processing, processing speed, long-term memory, crystallized
intelligence, picture vocabulary, oral vocabulary, incomplete words, sound blending, memory for
Running head: WOODCOCK – JOHNSON III 5
sentences, memory for words, calculation, and applied problems (Kane, et al., 2006). The
addition of measures for g gave the WJ-R a patched together three stratum model (e.g., a single
factor of intelligence, broad cognitive abilities, and a set of narrow cognitive abilities (Kane, et
al., 2006).
The WJ III COG was published in 2001 (Taub & McGrew, 2004) and is firmly grounded
in the Cattell-Horn-Carroll (CHC) theory of cognitive abilities (Taub, et al., 2004) validated by
many factor analytic studies (Evans, Floyd, McGrew, & Leforgee, 2002). CHC theory, which is
the blueprint for the MJ III COG, fuses the Cattell-Horn theory of fluid intelligence (Gf) and
crystallized intelligence (Gc) with John Carroll’s three strata model of intelligence (Evans, et al.,
2002). Carroll’s three strata model of intelligence is supported by over 460 different exploratory
factor analyses (Phelps, McGrew, Knopik, & Ford, 2005) described in Human Cognitive
Abilities: A Survey of Factor-Analytic Studies (Sanders, et al.). As such, the CHC theory of
cognitive abilities postulates three strata as discussed above, namely: A single factor of
intelligence (Stratum III) comparable to Spearman’s g, a second stratum (Stratum II) with ten
more broad cognitive abilities, and a third stratum (Stratum I) of 73 distinct narrow cognitive
abilities which underlie the ten broad ones (Taub, et al., 2004; Phelps, et al., 2005). Phelps lists
the ten broad abilities of the CHC model as follows: “short-term memory (Gsm), crystallized
intelligence (Gc), quantitative knowledge (Gq), reading/writing (Grw), visual processing (Gv),
auditory processing (Ga), long-term storage and retrieval (Glr), fluid intelligence (Gf),
Psychometric Properties
Normative Sampling
Running head: WOODCOCK – JOHNSON III 6
The WJ III COG is standardized using a national sample of 8,818 subjects aged 2 through
90 years and older drawn from 100+ geographic regions of the United States and stratified to
match the U.S. population (Blackwell, 2001; Naglieri, Lauder, Goldstein, & Schwebech, 2006).
The total sample consists of “1,143 preschool-aged children, 4,783 students in kindergarten
through 12th grade, and 1,843 adults participants” (Cizek, 2003). In the sampling, communities
were targeted first, followed by the schools both private and public, and then the students
including those homeschooled (Cizek, et al.). Students with disabilities were included, but not
by design, as were students learning English with at least a year's worth of time practicing it
(Cizek, et al.). Key criteria addressed in sample stratification included geographic region, the
size of communities, ethnicity and Hispanic origin, school or college type, and gender (Sares,
2003).
According to Cizek (2003), the WJ III COG is made up of 20 narrow tests that are
separated into a standard version and an extended version. Woodcock (2002) describes the WJ
III COG as consisting of “21 narrow abilities” (p. 13). In addition to these there is a supplement,
the WJ III Diagnostic Supplement to the Tests Of Cognitive Abilities (WJ III COG DS) that
contains eleven more tests that are intended to improve the diagnostic capability of the WJ III
COG (Sares, 2003). Both the standard version of the WJ III COG and the extended version have
ten subtests each (Cizek, et al.). The standard battery addresses verbal comprehension, visual-
auditory learning, concept formation, sound blending, spatial relations, visual mapping,
incomplete words, numbers reversed, auditory working memory, and visual auditory learning-
delayed (Cizek, et al.; Phelps, et al., 2005). The extended battery includes picture recognition,
general information, auditory attention, retrieval fluency, analysis synthesis, memory for words,
rapid picture naming, planning, decision speed, and pair cancellation (Cizek, et al.; Phelps, et al.,
Running head: WOODCOCK – JOHNSON III 7
2005). The WJ III COG DS has 10 more tests which are designed to correspond with the 20
making up the regular WJ III COG (Sares, et al.). The supplementary tests for memory of
names, visual closure, and memory for sentences can be combined with the verbal
comprehension, visual matching, and incomplete words from the regular WJ III COG to create
clusters measuring broad intellectual and cognitive abilities (Sares, et al.). In fact, Phelps, et al,
describes how the WJ III COG approach is based on focusing only a few broad cognitive
abilities at a time. For example, Thinking Ability consists of four of the broad factors including
Gv, Ga, Gf, and Glr, while Cognitive Efficacy is derived from Gs, and Gsm (Blackwell, 2001).
Woodcock, et al, describes some of the key tools used to construct the WJ III COG including
item response theory (IRT), confirmatory factor analysis, and the Rasch single-parameter logistic
model that provided the benefits of “sample-free calibration,” “item free measurement,” and
several others.
Materials
The WJ III COG comes with a technical manual, an examiner's manual, examiner
training workbooks, response booklets for response recording for both examiners and test takers,
overlays for hand scoring (Cizek, 2003). Blackwell (2001) adds “two easel test books” (p 233),
a computer program for scoring, and audio taped presentations. Blackwell also notes that
examiners are expected to supply pencils, an audio tape player, ear phones, and a stop watch for
the timed tests. According to Cizek, and to a lesser extent Sandoval, et al., (2003), the
examiner's manual contains thorough documentation on the WJ III COG’s grounding in theory,
it’s development, instructions for administering the tests, guidelines to aid interpretation, and
Administration
Running head: WOODCOCK – JOHNSON III 8
According to Blackwell (2001) all of the tests in the WJ III COG can be completed in 45
to 50 minutes, and a little over another hour is needed for the extended batteries. With respect to
the WJ III COG DS, Sares (2003) reports some tests given to preschool age children can run five
to 10 minutes per test. Most of the testing items are read to subjects by the examiner, and the
examiner records responses on an easel test book beyond the view of test takers. For some tests
an audio recording is used, and for timed tests the examiner uses a stopwatch. Blackwell
describes administering the WJ III COG as “relatively easy to learn” (p. 233), but then back
peddles by saying that examiner’s manuals need to be studied thoroughly along with the testing
materials, and that training in cognitive ability and educational testing should be at the graduate
level (p. 234). Cizek (2003) describes how the manuals address the qualifications of examiners
and test confidentiality issues. In fact, the testing materials include a “test session observation
checklist” that the examiners are supposed to use to record test taker cooperation, concentration,
attention, and self-confidence (Cizek, et al.). Finally, as many as 13 pages of the examiner's
manual is dedicated to describing special accommodations for people with language issues, or
Abbreviated Screening. An important feature of the WJ III COG which saves time in
administration is the ability to use a selected portion of test items to gauge the cognitive-
intelligence level of individual subjects (Blackwell, 2003) with the key goal of establishing the
base level of a test takers ability (Sandoval, 2003). Two tests offer this: the General Intelligence
Ability (GIA) score and the Brief Intellectual Ability (BIA) score (Cizek, 2003). The WJ III
COG BIA is a shorter version of the full WJ III COG which uses scores for verbal
comprehension, concept formation, and visual mapping (Newton, McIntosh, Dixon, Williams, &
Youman, 2008). The verbal comprehension scores indicate language skills; the concept
Running head: WOODCOCK – JOHNSON III 9
formation scales provide information on categorical reasoning, flexibility, and visual mapping
Scoring. According to Sandoval (2003), the WJ III COG cannot be scored by hand.
Cizek (2003) disagrees and flatly states that the WJ III COG tests can be “hand scored”, but the
Compuscore software eliminates this hassle. Yet, both Blackwell (2001) and Cizek describe that
testing materials come with a scoring template because some calculations are necessary in the
early going to establish upper and lower baseline levels. For the entire battery, the WJ III COG
Output. Scoring tables are included with the technical manual and used to change raw
scores into age based and grade based scores (Woodcock Johnson III) with means of 100 and
standard deviations of 15 (Sanders, et al., 2007). Because scale levels are derived from a
normative sample, all scores can be converted to standard scores, T scores, and percentile ranks
(Gregory, 2007)
Reliability
Blackwell (2001) gives a good accounting of the reliability data for the WJ III COG. He
describes that split half methods were used for most of the tests, while reliability for the timed
tests and the multipoint items were established by Rasch analysis. He reports median reliability
coefficients of .80, with higher median reliability scores for clustered items (.90 or higher). Cizek
(2002) describes two test-retest reliability studies in the technical manual where time intervals of
less than a year and as many as 10 years returned good reliability coefficients (.70 to .90).
Finally, like the full WJ III, the WJ III COG BIA can be applied to subjects two years old to 90
plus with reliability coefficients from .94 to .98, and concurrent validity correlations in the range
of .60 to .69 with other measures of intelligence (Newton, et al., 2008). Reliability data for the
Running head: WOODCOCK – JOHNSON III 10
WJ III COG DS as reported by Sares (2003) is derived from Rasch analysis and split half
methods. Median reliability coefficients ranged from a low of .63 to a high of .91
Validity
According to Blackwell (2001), “Evidence for the validity of the WJ III is provided for
three categories: content, construct, and concurrent.” And, “The authors present an extensive list
of studies that have provided a broad variety of content and construct validity evidence
correlations “in the moderate to high range” (p 234) with other intelligence or cognitive
assessments. Cizek (2003) points out that much more information in technical manual addresses
the validity of scores from the WJ III COG versus the WJ III ACH, and that a key focus is the
description of attempts to link the WJ III COG to the CHC theoretical model. Phelps, et al.,
(2005) describes this collection of joint validation studies as CHC Broad Confirmatory Factor
Analysis because the focus is on the broad cognitive factors in the Stratum II. Phelps points out
that the validation of the narrow abilities in Stratum I primarily rest on content validity, and the
lack of factor analytic studies that focus on the narrow factors, and the broad and narrow factors
together, is a weakness of the WJ III COG. The article by Sanders, et al., (2007) is one example
of factor analysis applied to both narrow and broad cognitive that represents a concurrent validity
study actually intended to be part WJ III technical manual. In this study different factor analytic
models are explored using both the WJ III COG and the Differential Ability Scales (DAS) to
explore narrow ability loadings from Stratum I on overall g (Stratum III), narrow ability loadings
on the broad cognitive abilities in Stratum II, and finally, the narrow abilities loading on the
broad abilities, and the broad abilities loading on overall g to match the CHC model. Another
study contributing to this relationship is described by Taub, et al., (2004). Using data from the
Running head: WOODCOCK – JOHNSON III 11
original WJ III COG standardization sample comprising nearly 7,500 individuals age 6 through
90 and older, and 14 tests from the WJ III COG applied to calculate the GIA and CHC factor
scores, Taub, et al. obtained results supporting the factor structure of the WJ III COG over five
age groups. Because many broad cognitive ability scales in the WJ III COG are carried over
from previous versions, studies which supported the validity of these earlier versions contribute
to the validity of its present form. An example of this is the canonical correlation analysis study
conducted by Estabrook (1984) which looked at the relationship between the WJTCA and the
Wechsler Intelligence Scale for Children-Revised (WISC-R). Using a sample of children from 7
to 12 years old, the squared canonical correlation obtained (Canonical R2 = .53) was sound
Applications
Based on this review, I could employ the WJ III COG and its supplement to investigate a
wide range of cognitive abilities as potential sources of poor achievement. I could also explore
Exploring Comprehension
Floyd, Bergeron, and Alfonso (2006) used the WJ III COG to develop a profile of
subjects with poor comprehension across the cognitive domains specified by the CHC model.
Using a sample of children in grades 2 through 12 from the original WJ III normative sample, the
children were selected based on poor scores in reading skills, comprehension, and numerical
calculating skills based on the WJ III ACH. The analysis sample consisted of 20 poor
comprehenders, 30 low achievers, and 50 children with average scores. The findings of Floyd, et
Running head: WOODCOCK – JOHNSON III 12
al. (2006) showed key cognitive differences shared by poor comprehenders and low achievers
compared to average achievers. Among poor comprehenders and poor achievers WJ III COG
reasoning (Gf), processing speed (Gs), and working memory (Gsm) were lower than those
obtained by average achievers (Floyd, et al.). Interestingly, poor comprehenders tended to score
better on long-term retrieval (Glr) as well as visual spatial thinking (Gv) and phonemic
awareness (Floyd, et al.). The use of the WJ III provided ample evidence that poor
comprehenders share a different cognitive ability profile when compared to average achievers.
This is a good example of how the WJ III COG can successfully identify cognitive weaknesses
Predicting Giftedness
In a test of the comparative predictive power of the WJ III COG BIA and two other brief
tests of intelligence, Newton, et al., (2008) found that the WJ III COG BIA did the best job of
Crockett, Moilanen, Raffaelli, and Randall (2006) make the distinction between variable
centered analysis and person-centered analysis. In their definition, the focus on traits as
individual items divorced from their intrapersonal organization within an individual represents
variable level analysis. Person-centered analysis, on the other hand, examines the nonlinear
interactive structure of traits as they exist within individuals. Based on this review, the WJ III
COG and its supplement take the variable centered perspective. An individual assessed with the
WJ III COG is evaluated at the variable level based on scores obtained for the narrow and
brought cognitive abilities. Personally, I have a great interest in person centered analysis
Running head: WOODCOCK – JOHNSON III 13
techniques and the application of statistical procedures like q type factor analysis (Asendorpf,
Borkenau, Ostendorf, & Van Aken, 2001), cluster analysis (Crockett, et al., 2006), and profile
analysis using multidimensional scaling (Kim, Frisby, & Davison, 2004) applied to exploring
intrapersonal structure. Given this interest, I was pleased to find examples of this type of person
tool, even if it wasn't the WJ III. Konold, Glutting, and McDermott (1997) applied repeated
cluster analyses to random samples of 262 children drawn from the WJ-R standardization sample
of 2,620 children. The cluster analyses conducted in several steps yielded eight “types” with
homogeneous interactive patterns across four scales of scholastic aptitude and four scales of
broad achievement. The importance of this kind of person centered analysis is that it
consistently reveals how single variables have interactive relationships with other variables
within a person, and that the form of that interaction can be different for different people. In my
view, this expands cognitive assessment dimensions beyond a focus on trait scale levels to
the population. In examining the profiles of the eight types validated by Konold, et al., there
were two types with above-average aptitude and achievement, one homogenous type of low
achievement and low aptitude, and five homogenous types existing around the middle which
showed quite a lot of overlap depending on the traits examined. It is interesting to note that an
eight type solution is common when large samples are used (see Brennan, Breitenbach, &
Dieterich, 2008; Matthews, Yousfi, Schmidt-Rathjens, & Amelang, 2003, & Lesser & Hughes,
1986).
Strengths
Running head: WOODCOCK – JOHNSON III 14
The new standardization sample used to norm the WJ III COG is large and developed
with great care (Cizek, 2003). The inclusion of the WJ III ACH set of tests in the standardization
sample makes the WJ III COG unique in that many areas of achievement can be related to
diverse areas of cognitive function (Blackwell, 2001). All reviewers cited see the CHC theory
foundation of the WJ III COG as positive and a feature that adds to its uniqueness. The fact that
the WJ III COG retains many features and scales from previous versions allows it to share the
many validity studies conducted with these predecessors. The technical manual’s attention to
AERA, APA, and NCME standards is a clear strength, as are the administration procedures
which uses simple aids, a computer program to simplify and speed up scoring, and procedures
Weaknesses
Even though the WJ III COG is modeled on CHC theory, Cizek (2003) noted that the WJ
III COG manual does not describe how this was done. According to Cizek, the manual also
lacks examples of computer-score output, and any description of recommended interventions for
different test outcomes. Sandoval (2003) noted several weaknesses in the WJ III COG, as well.
On the topic of accommodation, Sandoval noted that in tests where non-English (e.g., Spanish)
responses are allowed, understanding those responses must rely on the examiner’s knowledge of
the non-English language, which may or may not be present. In addition, Sandoval, noted that
extreme scores related to either age or ability are estimated rather than derived from actual
subject data. Sares (2003) also expresses frustration with the little information provided about
how the Rasch model contributed to test item selection, and the lack of any information
describing the factor analysis which confirmed model fit. As noted above, there is also the issue
Running head: WOODCOCK – JOHNSON III 15
addressed by Phelps, et al., (2005) related to lack of factor analytic studies supporting the
In an interesting study of the WJ-R by Kane, Oakland, and Brand (2006), differing levels
of cognitive function discriminated between those with high and low cognitive ability. It was
found that the impact of the g factor is less important among those with high versus low
cognitive ability. The authors speculated that non-cognitive factors like motivation and
personality might be the cause, and noted that the 1993 WJ-R incorporated a cognitive
performance model (CPM) made up of non cognitive factors. Based on the various reviews
(Cizek, 2003; Sandoval, 2003; Sares, 2003; Thompson, 2003), there is no evidence that
response data to functional magnetic resonance (fMRI) results. For example, the Grw scale
addresses reading and writing, and such tasks should engage areas such as Broca’s areas, the
angular gyrus, and the lateral temporal lobe (Diaz, & McCarthy, 2007). This may also open the
door to the formal inclusion of non-cognitive factors (e.g., motivation) which also have their
cognitive neuroscience counterparts (Daw & Shohamy, 2008). There is also a growing
Mellott, and Farnham, 2002; Nosek, 2007). Depending on how this research evolves, there may
come a day when implicit cognition, along with fMRI study data, expands the CHC model and
References
Running head: WOODCOCK – JOHNSON III 16
Asendorpf, J., Borkenau, P., Ostendorf, F., & Van Aken, M. (2001, May). Carving personality
description at its joints: Confirmation of three replicable personality prototypes for both
children and adults. European Journal of Personality, 15(3), 169-198. Retrieved August
8, 2008, doi:10.1002/per.408
Blackwell, T. (2001, Summer). Test Review. Rehabilitation Counseling Bulletin, 44(4), 232.
Brennan, T., Breitenbach, M., & Dieterich, W. (2008, June). Towards an Explanatory Taxonomy
doi:10.1007/s10940-008-9045-7
Crockett, L. J., K. L. Moilanen, M. Raffaelli, & B.A. Randall.(2006). Psychological profiles and
Daw, N., & Shohamy, D. (2008, October). The cognitive neuroscience of motivation and
learning. Social Cognition, 26(5), 593-620. Retrieved August 7, 2009, from Academic
Diaz, M., & McCarthy, G. (2007, November). Unconscious Word Processing Engages a
Evans, J., Floyd, R., McGrew, K., & Leforgee, M. (2002, June). The Relations Between
During Childhood and Adolescence. School Psychology Review, 31(2), 246. Retrieved
Floyd, R., Bergeron, R., & Alfonso, V. (2006, July). Cattell–Horn–Carroll cognitive ability
profiles of poor comprehenders. Reading & Writing, 19(5), 427-456. Retrieved July 19,
2009, doi:10.1007/s11145-006-9002-5
Greenwald, A., Banaji, M., Rudman, L., Nosek, B., Mellott, D., & Farnham, S. (2002, January).
295X.109.1.3
Gregory, R. J. (2007). Psychological testing: History, principles, and applications (5th ed.).
Kane, H., & Brand, C. (2006, December). The Variable Importance of General Intelligence (g) in
Kane, H., Oakland, T., & Brand, C. (2006, September). Differentiation at Higher Levels of
Cognitive Ability: Evidence From the United States. Journal of Genetic Psychology,
167(3), 327-341. Retrieved July 18, 2009, from Academic Search Premier database
Running head: WOODCOCK – JOHNSON III 18
Kim, S., Frisby, C., & Davison, M. (2004, October). Estimating Cognitive Profiles Using Profile
Konold, T., Glutting, J., & McDermott, P. (1997, June). The development and applied utility of a
Matthews, G., Yousfi, S., Schmidt-Rathjens, C., & Amelang, M. (2003, March). Personality
McCabe, P., Margolis, H., & Barenbaum, E. (2001, October). A comparison of Woodcock-
instructional reading levels. Reading & Writing Quarterly, 17(4), 279-289. Retrieved
Naglieri, J., Lauder, B., Goldstein, S., & Schwebech, A. (2006, Spring). WISC-III and CAS:
Which Correlates Higher with Achievement for a Clinical Sample?. School Psychology
Quarterly, 21(1), 62-76. Retrieved June 27, 2009, from Academic Search Premier
database.
Newton, J., McIntosh, D., Dixon, F., Williams, T., & Youman, E. (2008, July). Assessing
Schools, 45(6), 523-536. Retrieved July 18, 2009, from Academic Search Premier
database.
Nosek, B. (2007, June). Understanding the individual implicitly and explicitly. International
doi:10.1080/00207590601068159
Phelps, L., McGrew, K., Knopik, S., & Ford, L. (2005, Spring). The General (g), Broad, and
Narrow CHC Stratum Characteristics of the WJ III and WISC-III Tests: A Confirmatory
Sanders, S., McIntosh, D., Dunham, M., Rothlisberg, B., & Finch, H. (2007, February). Joint
confirmatory factor analysis of the differential ability scales and the Woodcock-Johnson
Sares, T. (2003). [Review of the Woodcock-Johson® III Diagnostic Supplement to the Tests of
Cognitive Ability.] In Woodcock, R., McGrew, K., Mather, N., & Schrank, F. (1977,
Abilities. Retrieved July 23, 2009, from Mental Measurements Yearbook database.
Taub, G., & McGrew, K. (2004, Spring). A Confirmatory Factor Analysis of Cattell–Horn–
Abilities III. School Psychology Quarterly, 19(1), 72-87. Retrieved June 27, 2009, from
Tests of Cognitive Ability.] In Woodcock, R., McGrew, K., Mather, N., & Schrank, F.
Cognitive Abilities. Retrieved July 23, 2009, from Mental Measurements Yearbook
database.
Woodcock Johnson III – Tests of Cognitive Skills retrieved June 13, 2009 from
http://www.cps.nova.edu/~cpphelp/WJIII-COG.html
Woodcock, R. (2002, April). New Looks in the Assessment of Cognitive Ability. PJE. Peabody
Journal of Education, 77(2), 6-22. Retrieved July 2, 2009, from Academic Search
Premier database.