You are on page 1of 12

A Critical Review: Informal Reading Inventories

Author(s): John Pikulski


Source: The Reading Teacher, Vol. 28, No. 2 (Nov., 1974), pp. 141-151
Published by: Wiley on behalf of the International Reading Association
Stable URL: http://www.jstor.org/stable/20193701
Accessed: 14-09-2016 14:03 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

International Reading Association, Wiley are collaborating with JSTOR to digitize, preserve and extend
access to The Reading Teacher

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
A critical review:
Supervisor of Diagnostic Services
informal at the Reading Study Center,

reading inventories University of Delaware,


Pikulski's special interests lie in
the prevention of reading
JOHNPIKULSKI problems by early identification.

mal diagnostic procedures until the


PROBLEMS in evaluating
ing performance and achieve read mid-1930s. In his 1936 book, Betts
ment continue to plague reading spe began to give some general consid
cialists and researchers. The United eration to evaluation based on sam
States has always been a very "test ples taken from instructional materi
conscious" nation and this emphasis als. By 1941, he was using the term
has not diminished. "subjective reading inventory" and
Evaluation of reading in this coun was using the informal approach on
try surged ahead in the 1920s and a regular basis at the Penn State
1930s with the advent of group Clinic. In 1942, Killgallon completed
standardized reading tests. However, a doctoral dissertation which is usu
limitations in these standardized ally cited as the first attempt to
measures were rapidly recognized, empirically validate some of the
and the search for more complete concepts involved in informal
and more diagnostic measures of evaluation. Beldin (1970) also sum
reading was begun. Beldin (1970) marizes the major studies that have
concisely traces the rise of informal dealt with attempts to validate infor
diagnostic procedures. For example, mal diagnostic procedures.
he notes that in a 1923 book by The concept of informal diagnosis
Wheat, the general concept of infor continues to receive widespread at
mal diagnosis is outlined rather well. tention as a means of evaluating the
Wheat recommended spending the reading performance of children.
reading periods of the first week of Nevertheless, in spite of widespread
each school year in "testing" the oral recognition, decades of discussion
and silent reading skills of the chil and a few research attempts to deter
dren by having them try to read from mine the most appropriate proce
several books which vary in level of dures and criteria to use with infor
difficulty?the core of informal diag mal reading inventories, there are
nosis. still several very vexing problems re
There was then a lull in the devel garding the use of Informal Reading
opment of strategies for use of infor Inventories. The purpose of this pa

141

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
per is to review and explore some of The second major strategy in
these areas and to invite greater ac volves the classroom use of informal
tivity aimed at approaching a solu reading inventories, based on poten
tion to these problems. tial instructional materials. A pri
William Powell (1970) summa mary reason for use of this strategy
rized what may be a very valuable has been to appropriately "group"
distinction in this area: "The children, and also to gain some in
strength of the IRI is not as a test in sight into specific strengths and
strument, but as a strategy for study weaknesses, both of the child and the
ing the behavior of the learner in a program. They have also been used
reading situation and as a basis for to assess growth, and to set indepen
instant diagnosis in the teaching en dent, instructional and frustration
vironment." levels. With this approach, it is rea
Far too many teachers and reading sonable to use a teacher-constructed
specialists continue to think of the informal inventory that samples the
IRI as a "test" that appears in material that is under consideration
printed and mimeographed form and for instructional use or to use an IRI
that is based on materials different that a publisher includes with his in
from those that are being considered structional materials. In both cases,
for instruction. They attempt to the IRI is simply a sampling of the
avoid what should be the ongoing ac materials that will later be used for
tivity of continuously constructing teaching.
and revising informal reading in Many of the problems to be dis
ventories. There have been several cussed in this paper are much easier
well-received attempts to "standard to work out when informal proce
ize" or package informal reading in dures are used in a classroom con
ventories. Using a published in text. They are far more challenging
ventory such as that by McCracken when they are applied to informal
(1966), Silvaroli (1965) or Smith reading inventories that are used as
(1959) certainly simplifies the task of clinical instruments on a recurring
testing a child. However, it may give basis. In the latter situation, ques
rise to some very difficult questions tions of reliability and validity de
that will be pointed out below. serve attention. However, as this
There appear to be two major happens, do these inventories remain
ways in which IRIs have been used. "informal" or do they become an
First, they have been used in reading other standardized diagnostic read
clinics and by school reading special ing test? Perhaps the primary
ists to diagnose individual children; strength of an IRI, that of close cor
set independent, instructional, and respondence between the test mate
frustration levels for them; and to de rial and the teaching material, is lost.
sign a plan for the remediation of Some reading specialists would
any difficulties found. In this frame maintain that unless there is this cor
work, since the examiner sees chil respondence, the diagnostic in
dren who are receiving instruction strument is not an IRI.
with materials that are different from Lowell (1970) in a very critical ar
case to case, it has not been feasible ticle about informal reading in
to construct or have available IRIs ventories states that "despite the
that fit each child's instructional pro value placed on the concept of infor
gram. In this situation, one or two mal analysis the practice of using in
mimeographed or printed informal formal reading inventories by class
reading inventories are used on a re room teachers is not widespread" (p.
curring basis. 120). In the pages that follow, he

142 The Reading Teacher November 1974

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
makes interpretations based on evi evaluation instrument are validity
dence presented in studies by Emans and reliability. Reliability is a meas
(1965), Ladd (1961) and Milsap ure of the consistency of an eval
(1962) which generally have sug uation instrument, while validity is a
gested that teachers are inaccurate reflection of the extent to which the
and incapable of adequately adminis test measures what it purports to
tering and interpreting the results of measure. Full consideration of the
informal reading inventories. Ladd, several forms of reliability and valid
for example, found that even after ity is impossible in the context of this
thirty hours of training, teachers still report. In addition, validity is diffi
failed to record 33 to 37 percent of cult to assess without reference to the
the errors. Some counterbalancing purpose for which and the population
positive evidence is supplied by Kelly with which the measure is being em
(1970) who reports very satisfactory, ployed. Classroom use of informal
practical application of information reading inventories that are based on
from informal reading inventories, if materials that are being considered
the teachers received an inservice for instruction probably does not
training program early in the school need to be subjected to most tradi
year. He also reported that 95 per tional evaluation procedures by
cent of the teachers trained under which reliability and validity are
these conditions used IRIs to place judged. There is a strong case for
children in reading groups. "face validity." The adequacy of the
Some teachers are reluctant to try sampling of the materials and skills
reading inventories because they are would probably be the main validity
often represented as a highly formal question. Testing is generally a sam
ized and very precise technique. pling of behavior. In classroom use
However, if the IRI is viewed as a of informal inventories, the sample
sampling procedure designed to give used in the test comes directly from
a teacher some direction with regard the material that will be used in
to instructional strategies, it may be teaching. Testing procedures closely
that a very high degree of precision is approximate, but do not duplicate,
unnecessary. This would be the case the procedures that are most fre
if the teacher continued to evaluate quently used for teaching reading.
the child's performance during in On the other hand, it should be
structional periods and altered her possible to expose informal reading
diagnostic impressions. It is difficult inventories that are being used on a
to argue against the idea that diag recurring basis to many formal eval
nosis should be a continuous process, uations of reliability and validity.
and that extant measures of reading Some questions that might be pur
provide data that are meaningful sued include: Do two examiners ob
only within the context of an instruc serving and recording the same read
tional program. ing performance obtain similar
More information needs to be col results? If alternate IRI forms are
lected regarding the practicality of used, is the performance of one child
using informal procedures. How similar on the two instruments? If it
widely are they used and how accu is assumed that the several levels of
rate are or need teachers be in an IRI become progressively more
recording and interpreting results? difficult, do a group of children per
form increasingly more poorly as
Reliability and validity the assumed difficulty increases? The
The central concepts that are gen final question is probably the most
erally considered with regard to any important: Do the results of an infor

PIK U LS KI :. . . reading inventories 143

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
mal reading inventory predict which the material becomes so diffi
whether or not a child will be suc cult that the child can no longer cope
cessful with particular classroom with it. The word recognition de
reading materials? Will the instruc mands or the level of understanding
tional level established by an IRI required by the comprehension ques
predict the level of book in which the tions exceed the skills that the reader
child can profit from instruction? possesses. Quantitatively the frustra
There have been some research at tion level is defined as the lowest
tempts to deal with almost all of the level of reading material at which the
above with some tests resembling child makes more than 10 percent of
IRIs, but there is a serious need forerrors in word recognition (less than
additional evidence in this area. As 90 percent accuracy) or where he can
noted earlier, some would maintain answer less than half of the compre
that if reliability and validity meashension questions he's asked.
ures are carried out the material so Finally, there is the instructional
studied is no longer an "informal" level, which is probably the level of
inventory. However, it seems unjusti greatest interest to the teacher. De
fiable to recurringly use an IRI not scriptively, this is the level of mate
taken from the materials planned for rial where the child encounters some
instruction without raising reliability difficulty, but where the difficulty is
and validity questions. not so great that the child will not be
able to successfully deal with it if
Establishing "levels" given help from a teacher. Quan
One of the main purposes of an titatively the instructional level is the
IRI is to set "functional reading lev highest level at which the child is
els" for a child. Betts is usually cred able to correctly deal with 95 to 98
ited with suggesting that a child doespercent of the word recognition
not have a reading level, but instead demands and can answer 75 to 90
that he has three reading levels?an percent of the questions he is asked.
independent level, an instructional A number of questions and objec
level and a frustration level. There tions have been raised with regard to
are two ways to describe each of the concepts involved in functional
these levels?descriptively and quan reading levels. Most of these deal
titatively. with the quantitative aspects, but at
Descriptively we talk about an in least one objection has been raised
dependent level as the highest level of regarding the qualitative descrip
difficulty of a book that a child can tions.
cope with on his own. He can sustain George Spache (1963), for ex
an almost perfect performance. He ample, takes issue with the concept
has almost no difficulty with word of independent level as described
recognition, understands the passage, above. He writes, "A pupil's inde
can remember most of its contents pendent reading level need not be
and answer questions about it. Quan limited to material graded at or be
titatively the independent level is de low his instructional level. Rather,
fined as the highest level of material pupils should be encouraged to read
with which a child can sustain ap from books available at all levels up
proximately 99 percent word recog to the limits of their independent
nition and can answer 90 percent of level. It is not essential that pupils
the comprehension questions that an read only very easy materials for un
examiner asks. derstanding or enjoyment" (p. 20).
At the other end of the spectrum is Thus, according to Spache, a com
the frustration level, the point at prehension score of 85 percent is gen

144 The Reading Teacher November 1974

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
erally recommended for an instruc from the third reader level book that
tional level; but a score of only 60 is being considered for instructional
percent is required for an indepen use. But it seems quite unsafe to as
dent level. Spache maintains that his sume that a child designated as read
concept of an independent level em ing at the third grade level as deter
phasizes the role that interest plays mined by an IRI based on one series
in making a book readable. of books will be able to read a third
While most reading teachers grade level book from another series,
would support the position that a the one that he might meet in class
reader's independent reading level is room instruction. One has only to
lower, consisting of easier materials, compare the same reader level books
there would be almost universal from more than one series or even
agreement that readers who are various editions of basal readers
highly motivated to read something done by the same company, to be
can enjoy that material, even if the come impressed with the nature of
word recognition and comprehension the problem. For example, those pro
challenges are very significant. How duced in the 1970s are generally
ever, we need evidence to show that more challenging than those pro
children who are guided to read at duced in the mid-1950s and 1960s. If,
the more traditional independent as is sometimes recommended, the
level read more, enjoy it more, de recurringly used IRI is based upon
velop long-term reading habits to a outdated material in order to be sure
greater extent, and in general have a the child is not familiar with it, and if
more positive attitude toward read the child is going to be receiving in
ing. struction with a different basal series
As noted, most of the questions, or even a non-basal approach, the
however, have focused upon the problem becomes increasingly com
quantitative aspects of IRI reading pounded, and the need to establish
levels. validity becomes a matter of critical
importance. The problem becomes
Evaluating validity more acute when trade books, maga
The IRI shares a problem with zine and newspaper articles, books in
most other evaluation instruments in content areas and instructional mate
reading. The child's independent, in rials other than basais are being cho
structional and frustration levels are sen on the basis of IRI results.
usually expressed in grade or reader The child who has been taught us
level scores. For example, a child's ing a non-basal approach is fre
independent level may be the third quently particularly penalized when
reader level, his instructional level evaluated with an IRI which gener
may be fourth and his frustration ally follows basal reader procedures.
level, sixth. He is not "test-wise" with respect to
The above means, among other the procedures. Some of these chil
things, that he should be able to work dren show improvement in perform
effectively, with a teacher's help, us ance as the testing proceeds even
ing a third reader level book. How though the material is supposedly be
ever, even a very inexperienced coming more difficult. They seem to
teacher knows that the skill demands quickly "catch on" to what is ex
made even by basal readers, which pected.
are highly controlled, vary consid Many of the above problems can
erably from one series to the next. be avoided by simply emphasizing
Again, there is no problem when the areas of strength and weakness dem
evaluation has been made sampling onstrated by the reader in an IRI

PIK U LS KI :. . . reading inventories 145

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
performance and de-emphasizing the Should all errors be counted equally
setting of levels. Many reading clini or should they be weighed in some
cians do an outstanding job in this re way? For example, look at the fol
spect. They use the IRI as a criterion lowing two readings of the sentence:
referenced test. However, the grade The boy sits on the chair and waits
level designations for independent, for his mother.
instructional and frustration levels Child 1: The boy is sitting on a
are widely used and appear to have chair waiting for his mother.
some utility. It will help considerably Child 2: The (hesitation) boy s-s-s
1) if teachers are carefully alerted to (examiner must give the word sits) on
the difficulties and limitations in the (hesitation) champ and (hesita
volved in the grade level designa tion) water for his mother.
tions, 2) if researchers move toward In the first, the child read fluently
attempting to provide a clearer and with meaning. The substitutions
method of designating level of func and insertions did very little to
tioning, and 3) if attempts are used to change the meaning of the sentence.
locally validate recurringly used He made four scoreable errors. The
IRIs. second child read in a word-by-word
fashion, needed examiner help with
Using quantitative criteria one word and substituted champ for
Two criteria are used in setting chair and water for waits. A very
functional reading levels: the word substantial amount of difficulty with
recognition in context score and the word recognition is suggested and
comprehension score. The first num the child received very little meaning
ber, the word recognition score, is de from the sentence. Yet, the first child
rived by having the child read a para made four scoreable errors and the
graph or paragraphs of material. The second made only three. Quan
examiner makes an effort to record titatively, the second child did better
all "errors" that the child makes in according to conventional IRI scor
reading. A percentage score is then ing systems.
derived by dividing the number of It should be strongly emphasized
"errors" or "miscues" by the total that users of informal reading in
number of words in the passage and ventories do make interpretations
subtracting from 100. Even the term beyond the quantitative information
"errors" has been challenged. Ken and take symptoms and qualitative
Goodman, based on his linguistic information into consideration. If er
analysis of children's errors in oral rors similar to those above were
reading, suggests that the term "mis made by the two children in com
cue" is much better than error. pleting the reading of a passage so
The first question, of course, is: that both achieved a score of 88 per
What constitutes an error? In the cent, it is conceivable that a reading
early work by Betts (1936) and Kill diagnostician might conclude that
gallon (1942), repetitions were the first child was "independent"
counted as errors; most current with respect to that material while
writers do not suggest that repeti the second was "frustrated." In most
tions be counted. Some include self courses which prepare teachers to
corrections while others do not. use informal reading inventories, stu
There is, however, more agreement dents are cautioned to go beyond the
than disagreement about what numerical data. This points out why
should and what should not be the following statement is heard so
counted as an error. frequently: "An informal reading in
A much more difficult problem is: ventory is no better than the person

146 The Reading Teacher November 1974

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
using it." error, it was grammatically accept
Research such as that conducted able but not semantically acceptable
by Weber (1970) and Goodman and there was a meaning change. By
(1965) might be translated into a set analyzing each of the child's miscues
of specified guidelines for inter in this fashion a detailed analysis of
preting the extent to which different his word recognition strategies can
types of errors should be considered be made and his strengths and weak
diagnostically significant. A recently nesses evaluated. From this type of
published manual by Yetta Good evaluation, the authors maintain that
man and Carolyn Burke (1972) pro lessons which use the reader's
vides some valuable clues as to how strengths to help overcome his weak
this can be done. They suggest, for nesses can be devised and used.
example, that nine questions be Throughout this manual are provoc
asked by the diagnostician regarding ative but helpful suggestions; for ex
oral reading miscues: ample, "often substitutions of words
Dialect: Is a dialect variation in like a for the, by for at, in for into, do
volved in the miscue? not cause a change in meaning. The
Intonation: Is a shift in intonation story context is the determining fac
involved in the miscue? tor in how extensive the change has
Graphic similarity: How much been. Substitutions like daddy for fa
does the miscue look like what was ther, James for Jimmy cause no
expected? change in meaning. These types of
Sound similarity: How much does habitual association miscues are gen
the miscue sound like what was ex erally produced by proficient readers
pected? and are not reading problems" (pp.
Grammatical function: Is the 101-102). The important point is that
grammatical function of the miscue if numerical criteria are going to be
the same as the grammatical func used, efforts should be made to refine
tion of the word in the text? existing procedures so that the inter
Correction: Is the miscue cor pretation of results would be some
rected? what less reliant on the judgment of
Grammatical acceptability: Does the examiner.
the miscue occur in a structure which There are striking similarities as
is grammatically acceptable? well as differences between the IRI
Semantic acceptability: Does the and Reading Miscue Inventory. The
miscue occur in a structure which is latter may, however, have some valu
semantically acceptable? able suggestions to make with regard
Meaning change: Does the miscue to the weighting of oral reading er
result in a change of meaning? rors or miscues. The oral Miscue In
Each of the reader's miscues is cat ventory is based on extensive re
egorized according to this scheme. search and a well-defined linguistic
For example, if the sentence: "I theory of reading. It may be that
looked up and had my first view of a most teachers would find it difficult
lion" were read: "I looked up and to use in its present form and that
heard my first view of a lion," the there are too many and overlapping
substitution would not be considered categories, so that it probably needs
the result of dialect, there was no simplification in order to come into
shift in intonation, there was a strong widespread use. However, it certainly
graphic similarity, a moderate deserves attention from reading spe
amount of sound similarity, the two cialists and researchers.
words serve the same grammatical There are also considerations that
function, the child did not correct the need to be made regarding the quan

PIKULSKI:... reading inventories 147

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
titative aspects of comprehension. extent so that less reliance is placed
Generally, in IRI procedure, a child on examiner judgment?
is asked to read two selections at
each grade level. One is read orally The question of questions
and one read silently. Comprehen There is also the question of what
sion questions are asked after each types of questions should be asked of
selection is read. A percentage score the child, since this is an important
is then calculated for each, and the contributor to scores. Johnson and
two are then averaged. This seems Kress (1965) in an unusually compre
acceptable when the two scores are hensive and lucid discussion of IRI
similar, but many questions arise procedures suggest the use of four
when the two are dissimilar. Con types of questions: factual recall, in
sider the following scores actually at ferential, vocabulary and back
tained by a sixth grade student. The ground of experience. Valmont
word recognition was relatively per (1972), on the other hand, in an ar
fect throughout. ticle in The Reading Teacher sug
gests the categories of main ideas,
Oral Silent seeing cause and effect relationships,
Grade Compre- Compre and several other specific compre
Level hension hension Average
hension skill areas. It has also been
Fourth 60 80 70
Fifth 80 90 85 argued that taxonomies such as that
Sixth 50 85 68 offered by Bloom (1956) and modi
Seventh 50 80 65 fied by Sanders (1966) would work
better.
If the criteria for IRI levels were Twoappoints seem important here.
plied strictly, and once again it First,
is notthe problem is not unique to in
suggested that diagnosticians are formal
en reading evaluation; it is
shared by other approaches to eval
couraged to do this, this youngster
would have no independent level uationesand by the area of reading in
tablished, her instructional struction.
level Simons (1971) does an ex
would be fifth and her frustration
cellent job of describing some of the
level would remain unestablished.
difficulties that remain in conceptual
However, even the novice examiner
izing and evaluating reading compre
would almost certainly concludehension.
that Also, this seems a far less
perplexing
so literal an application of the cri question for those who
teria would be inappropriate use IRIs as a classroom technique.
since
comprehension of silently readThey
mateshould employ in the IRI ques
tions
rial approximates much more closely which are like those that they
employ
the type of reading demands that are in their classroom instruc
placed on a sixth grade child. Is tion.
thereThe question is more serious for
the clinic evaluator or diagnostician
a point then at which the averaging
who evaluates children from a vari
of comprehension scores is no longer
ety of
appropriate? Is there a point, for exclassrooms.
ample, at first grade level where com is yet another problem with
There
respect to questions which revolve
prehension of orally read material
around the concept of reading-de
should be considered more impor
pendent questions. Tuinman (1971)
tant? Unquestionably, the adequate
examiner takes such questions
hasinto
made a strong plea for using only
consideration when formulating
those questions that can be answered
diagnostic decisions; however,only
once
on the basis of reference to in
again, might it not be possibleformation
to re contained in the selections
read. This would largely eliminate
fine IRI methodology to a greater

148 The Reading Teacher November 1974

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
vocabulary questions like: What is a gested that different criteria a
beach? What is a ticket? What is a called for at various grade levels. H
lunch basket? What is a mongrel? It suggests that at grades one and tw
would also eliminate background of word recognition scores between 8
experience questions like: Where do and 98 percent signify an instru
birds who do not stay in the north go tional level; whereas, at third throug
in the wintertime? What do insects fifth grade, 91 to 98 percent is the ap
do in the winter? Although vocabu propriate range and that at sixt
lary and background of experience grade, 95 to 98 percent is acceptable
are factors which unquestionably After noting that there is a deart
contribute to a good reading per of experimental evidence to suppo
formance, should they not be kept the Betts criteria, he conducted th
separate from questions which can be following study. He began by adop
answered only with reference to in ing the rationale that if a child
formation contained in the selection comprehension score remaine
read? above 70 to 75 percent, the word r
Content area considerations. One ognition scores achieved at those le
often finds quite inconsistent scores els were "tolerable." He selecte
in selections, especially beyond sixth groups of children in grades on
reader level. Comprehension scores through six who were average in ab
such as the following do occur: ity and achievement. The compr
hension scores of each child were su
Oral Silent veyed, and the highest level at whi
Grade Compre- Compre the child was able to attain 70 p
Level hension hension Average
Sixth 80 100 90 cent comprehension was identified
Seventh 60 50 55 Scores in word recognition were th
Eighth 70 90 80 surveyed at and below that level t
determine the lowest word recogn
Repeatedly-used selectionstionvaryscore
as achieved. This score w
to subject area. Usually, science, so
recorded for each child, and th
cial studies and fiction are included.
mean was calculated. Mean score
When this is the case, the child who is
were calculated for each grade leve
generally interested in geography,
with the following results:
Grade Percent
and who enjoys this area, frequently
First 83
does well in answering comprehen
Second 87
sion questions based uponThird
a selection
91
which taps geographicalFourth
concepts
92
and vocabulary. He may Fifthdo
92 very
Sixth 94
poorly with a science selection car
rying the same grade level designa
Although the results of the study
tion. Talking about grade
have beenlevel
available since 1968, there
scores may be particularly
are fewuseless
published attempts to verify
beyond sixth. or refute them. Therefore, two ex
Criteria for selecting levels. Wil were undertaken at
ploratory studies
liam Powell has been particularly
the University of Delaware. Both fo
prominent in challenging cused
one on ofsecond
the graders, since the
Betts andan
numerical criteria for setting Powell
incriteria were most
structional level: the 95 percent
disparate atscore
that level. The first, con
for word recognition whichducted by Bassett and Hutchison
is usually
attributed to Betts. Where(1972)
did used
he "average"
get second grad
ers.numbers
it? Critics talk about pulling Using an informal reading in
from the air. Powell (1971) has similar
ventory sug to Powell's, they

PIK U LS KI :. . . reading inventories 149

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
administered it to twenty-eight sub ity of informal reading procedures,
jects, the exact number of subjects but seeks only to refine or verify the
used by Powell at that level. Using utility of our present practices. Pre
the same calculation procedure em liminary work at the University of
ployed by Powell they obtained the Delaware has suggested that the
following results: method used for teaching the child to
School A 94.9 percent read, and whether the child is an
School B 96.7 percent achieving or non-achieving reader
Total Population 95.8 percent may influence the balance of word
In spite of every effort to follow Pow recognition and comprehension
ell's procedure, the results were sup scores. Perhaps the weighting of
portive of the original Betts criterion scores in terms of type of word rec
and in fact exceeded it. ognition errors will need to be devel
There was one very important oped before the numerical criteria
population characteristic that is dif can be adequately specified.
ferent in the two studies. The chil The problems raised in this paper
dren in Powell's study had been can be solved. Pointing to difficulties
taught to read using a popular basal in no way suggests that informal
reader series, while those in the Bas reading inventory procedures are not
sett-Hutchison study had been in useful. As a matter of fact, it is prob
structed using programed reading. ably because they have proven so
The investigators concluded, and useful and because they have been
Powell concurred (personal commu placed into the hands of extremely
nication, 1970), that the strong em competent professionals who com
phasis on phonics probably created pensated for many of the limitations
the high word recognition scores rel noted, that research has been so lim
ative to the comprehension. Thus, it ited. The purpose here was simply to
appears that the method used for point out some pitfalls that should be
teaching reading will influence the guarded against and to suggest the
configuration of scores. need for more study to make the IRI
A second phase was undertaken by an even more useful instrument.
the author using a sample of second Classroom use of informal reading
grade children who had been referred inventories based on instructional
to a clinic for evaluation. When the materials seems strongly indicated,
same procedure was applied to this since this procedure provides the
group, the resultant word recognition closest match between testing and
score was 83.8, a sharp contrast to teaching. This alone compensates for
the score found by Bassett and Hut some of the imprecision and uncer
chison, in spite of the fact that the in tainty that exist with regard to in
formal testing materials were the formal procedures. On the other
same. The results of this second in hand, there are very important meth
vestigation are almost identical to odological questions that deserve
those found by Powell. As with so closer scrutiny by examiners who use
many studies, these two suggest the informal reading inventories on a re
need for more exploration. They are curring basis.
helpful in trying to identify some of
the factors that will influence the References
Bassett, D. and M. L Hutchison. Criteria for In
scores.
structional Levels for Second Graders. Unpub
Powell has offered a lished manuscript. University of Delaware,
challenge to 1972.
the frequently employed criteriaBeldin,
for H. L. "Informal Reading Testing: Histori
cal Review and Review of the Research."
establishing a child's instructional
Reading Difficulties: Diagnosis, Correction and
level. He does not challenge the Remediation. W. Durr. Ed. Newark, Delaware:
util

150 The Reading Teacher November 1974

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms
International Reading Association. 1970. Levels with Informal Reading Inventories."
Betts, E. A. The Prevention and Correction of Reading Difficulties: Diagnosis, Correction and
Reading Difficulties. Evanston. Illinois. Row Remediation. W. Durr. Ed. Newark, Delaware:
Peterson, 1936. International Reading Association, 1970.
Betts, E. A. "Reading Problems at the Inter McCracken, R. A. Standard Reading Inventory.
mediate Grade Level." Elementary School Klamath Falls, Oregon: Klamath Printing.
Journal, vol. 40 (1941), pp. 737-46. 1966.
Bloom, B. (Ed.) Taxonomy of Educational Objec Milsap, L. N. A Study of Teachers Awareness of
tives. Handbook I: Cognitive Domain. New Frustration Reading Level among Their Pupils
York: David McKay, 1956. in Basal Readers. Unpublished doctoral dis
Cooper, J. L. The Effect of Adjustment of Basal sertation. University of Oregon, 1962.
Reading Materials in Reading Achievement. Pikulski, J. Criteria for Instructional Levels for
Unpublished doctoral dissertation, Boston Disabled Readers. Unpublished manuscript,
University, 1952. University of Delaware, 1972.
Emans. R. "Teacher Evaluation of Reading Skills Powell, W. "Validity of the I.R.I. Reading Levels."
and Individualized Reading." Elementary Eng Elementary English, vol. 48 (1971). pp. 637
lish. vol.42 (1965), p. 258. 42.
Goodman, K. S. "A Linguistic Study of Cues and Sanders, N. M. Classroom Questions: What
Miscues in Reading." Elementary English, vol. Kinds? New York: Harper and Row. 1966.
42 (1965), pp. 639-43. Silvaroli, N. Classroom Reading Inventory. Du
Goodman, Y. and C Burke. Reading Miscue In buque, Iowa: William C Brown, 1965.
ventory. New York: The Macmillan Co., 1972. Simons, H. "Reading Comprehension: The Need
Johnson, M. S. and R. A. Kress. Informal Reading for a New Perspective." Reading Research
Inventories. Newark: Delaware: International Quarterly, vol. 6 (1971), pp. 338-63.
Reading Association, 1965. Smith. N. B. Graded Selections for Informal
Kelly. D. "Using an Informal Reading Inventory Reading Diagnosis. New York: New York Uni
to Place Children in Instructional Materials." versity Press. 1959.
Reading Difficulties: Diagnosis, Correction and Spache, G. D. Diagnostic Reading Scales. Mon
Remediation. W. Durr, Ed. Newark. Delaware: terey, California: California Test Bureau, 1963.
International Reading Association, 1970. Tuinman, J. J. "Asking Reading-Dependent
Killgallon, P. A. A Study of Relationships among Questions." Journal of Reading, vol. 14 (Feb
Certain Pupil Adjustments in Language Situ ruary 1971), pp. 289-92.336.
ations. Unpublished doctoral dissertation, Val mont. W. J. "Creating Questions for Informal
Pennsylvania State University, 1942. Reading Inventories." The Reading Teacher.
Ladd, E. M. A Comparison of Two Types of Train vol. 25 (March 1972). pp. 509-12.
Weber. R. M. "A Linguistic Analysis of First
ing with Reference to Developing Skill in Diag
nostic Oral Reading Testing. Unpublished doc Grade Reading Errors." Reading Research
toral dissertation, Florida State University, Quarterly, vol. 5 (1970). pp. 427-55.
1961. Wheat, H. G. The Teaching of Reading. Boston:
Lowell, R. E. "Problems in Identifying Reading Ginn, 1923.

2nd INTERNATIONAL READING INSTITUTE


The Institute will emphasize Comparative and Early Childhood
Education while visiting in Scotland, England and the Netherlands.
Following will be 4 weeks on campus ?n Leysin, Switzerland where a
variety of courses related to reading will be offered. Classes in Leysin
will meet four days a week with optional weekend trips. Up to seven
graduate units possible, accredited by La Verne College.
Option 1 ? July 10?26, two credits Comparative Education (plus 4
days in Paris for a 21-day tour)
Option 2 ? July 27 ? August 23, up to five credits in Leysin
Option 3 ? July 10 ? August 23, up to seven credits (45-day tour,
omitting Paris)
CHRISTMAS READING INSTITUTE
Spend six days in study and relaxation on the beach at Sarasota.
Classes three hours each morning, earning one semester graduate credit.
Date Dec. 21-26 or 26-31, 1974.
Spache Educational Consultants, Inc.
4163 Shell Road, Sarasota, Florida, 33581

Send brochure and registration forms for the 2nd International Reading
Institute, 1975. or the Christmas Institute
Name

Address

City_State _ Zip

PIK U LS KI :. . . reading inventories 151

This content downloaded from 129.100.58.76 on Wed, 14 Sep 2016 14:03:10 UTC
All use subject to http://about.jstor.org/terms

You might also like