You are on page 1of 10

Peabody Individual Achievement Test -

Revised/Normative Update
-------------
Don R. Bartels, Wanaque Borough Board of Education, New Jersey

Background Information
Author
The Peabody Individual Achievement Test-Revised was designed
by Frederick C. Markwardt, Jr. He received his Ph.D in educational psy-
chology from the University of Minnesota. While at St. Cloud State
University in Saint Cloud, Minnesota, Dr. Markwardt spent eight years
teaching counseling, statistics, and measurement, while also providing
individual assessments of elementary aged students for educational plan-
ning and placements. Dr. Markwardt went on to serve on the Executive
Council of the Minnesota Psychological Association and the Minnesota
State Board of Examiners of Psychologists. He is currently associated
with a Saint Paul consulting firm specializing in assessment, while pursu-
ing his consulting services on educational research and test development.

Publisher
American Guidance Service, 4201 Woodland Road, PO Box 99,
Circle Pines, MN 55014-1796, www.agsnet.com, 1-800-328-2560

Price
PIAT-R/NU Complete Kit $289.95
Includes 4 easels, 50 Combined Test
Record and Written Expression
Response Booklets, NU Manual,
Carry bag
Above with nylon briefcase 319.95
PIAT-R/NU Manual 84.95
PIAT-R Combined Test Record
and Written Expression Response
Booklets (50) 69.95
Computer ASSIST Macintosh 199.95
Computer ASSIST Windows 199.95
Pronunciation Guide Cassette 15.95
Nylon Briefcase 49.95
DIAGNOSTIQUE 24(1-4),211-220 (1998-1999)

211
212 DIAGNOSTIQUE VOL. 24, NO. 1-4,1998-1999

Copyright Dates and Revisions


The Peabody Individual Achievement Tesr-Revised-Normative
Update (PIAT,R/NU) (Markwardt, 1998) is a revision of the 1970
Peabody Individual Achievement Test (PlAT) (Dunn and Markwardt,
1970), with updated norms, more items, and more contemporary item
content. The Written Expression subtest addition is a major feature of
the PIAT,R (Markwardt, 1989). This 1998 Normative Update has "NU"
norms based on a national sample of over 3,000, allowing for accurate
comparison (including score equating) among three other achievement
batteries (Kevlvlath-R, WRMT,R, and KTEA).

Groups for Whom the Instrument is Intended


The PIAT,R/NU is designed as an individually administered school
achievement measure for students from kindergarten through grade
twelve (ages 5-0 to 22,11). According to the manual, due to its
nature/design, the actual instrument appeals to those students who are
disadvantaged, poorly motivated, and/or physically handicapped.

Forms of the Instrument


The PIAT,R/NU is available solely in one form. The optional
Written Expression subtest Level II has two forms.

Purpose and Recommended Uses


The PIAT-R/NU is an individually administered instrument that
provides a wide-range assessment in six content areas: General
Information, Reading Recognition, Reading Comprehension,
Mathematics, Spelling, and Written Expression. According to the
author, although useful diagnostic hypotheses can be generated, the
PlAT,R/NU was not developed as a diagnostic tool in any of the six con,
tent areas measured. The PlAT-R/NU was developed to provide a stu'
dent's general level of achievement and not that student's total level of
attainment. Additionally, the actual items presented to the student rep'
resent a cross section of various subject areas and not any individually
selected school's curriculum (Taylor, 1997). Consequently, a student's
subject matter background/exposure needs to be taken into account
when interpreting the results.
The author's suggested uses include individual evaluation, program
planning, student grouping, follow up evaluation, program evaluation.
PEABODY INDIVIDUAL ACHIEVEMENTTEST-REVISED/NORMATIVE UPDATE 213

The research uses include demographic studies, validation studies, longi-


tudinal studies, and basic research studies.

Dimensions Measured

The PIAT,R/NU measures a student's current academic perfor-


mance levels in the areas of Reading, Mathematic, Knowledge, and
Written Expression. Six different subtests assess the following dimen-
sions:
Knowledge - student's general knowledge from past exposures and
experiences in the fields of science, social studies, geography, history,
sports, fine arts, social awareness, and cultures;
Reading - student's ability to match alphabet letters, phonetic
sounds to pictures and letters, orally read words of progressing difficulty,
and provide literal comprehension to silently read sentences;
Mathematics - student's ability to match and recognize numbers,
apply mathematical facts, and compute mathematical concepts, equa-
[ions, and processes;
Spelling - student's ability to recognize letters presented, to iden-
tify a letter or word from dictation, and to disriminate a correctly spelled
word from four choices given;
Written Expression - student's ability to write name, copy letters
and words, write from dictation, interpret a picture through a written
story like response with appropriate logic, grammar, and composition
techniques.

Administration
The PIAT,R/NU is an individually administered untimed test {the
only timed administration is the 20 minute limit requirement on the
higher level written expression (writing sample) subtest). Total adminis-
tration time is approximately one hour. Each subtests' items are arranged
in increasing difficulty levels. The subtests in general information, read,
ing recognition, mathematics, and spelling all include 100 verbal or mul-
tiple choice items. The reading comprehension subtest includes 82 items,
whereas the written expression subtest involves, at the younger age lev,
els, copying and dictation items. At the higher ages, students are expect,
ed to write a passage relating to one of two pictures shown. Based upon
the standardization, the manual requests that the order of the subtests
follow the structure of the test booklet with general information, reading
recognition, reading comprehension, mathematics, spelling, and written
214 DJAGNOSTIQUE VOL. 24, NO. 1-4,1998-1999

expression subtests, given in sequence. This sequential procedure allows


each previous subtest to determine the actual starting point for the fol-
lowing subtest.
Critical ranges, as determined by basal and ceiling points, can be
easily identified thus enabling the examiner to assess only the pertinent
test items and not to have to administer the entire subtest, Basal and ceil-
ing levels are identified in the same pattern for all the subtests. This uni-
formity is provided by the basal item being five consecutively numbered
correct answers and the ceiling item being five incorrect responses out of
seven consecutive items. Any incorrect item below the basal is cited as
correct; any incorrect item above the ceiling is cited as incorrect. Case
study examples are given in the manual to show various critical range
development outcomes.
Training items are given at the beginning of each subtest.
Directions given to the student are provided for the examiner on the
back of each page of the testing easels. Expected answers are also provid-
ed the examiner on the test easel and in the actual test record. Students
are not allowed to use additional materials (pencil/paper), except on the
Written Expression subtest.

Summation of Data
Procedures
The PIAT-R/NU examiner must have a strong background in psy-
chological testing. statistics, and education. This test user has completed
graduate training in individual assessment, guidance, mesurement along
with knowledge of special appraisal methods appropriate for the
PIAT-R/NU instrument.
PIAT-R/NU subtests are scored objectively except for the Written
Expression subtest, which is scored based upon criteria as described in the
manual. Raw scores (items correct) along with SEM (standard errors of
measurement) can be recorded in three confidence levels (68%, 90%,
95%). Grade equivalents, age equivalents, standard scores, and per-
centiles are obtainable from the manual's norm tables or from the
ASSIST computer program.

Types of Scores Available


A student's raw scores (items correct) can be converted to derived
scores (means by which that student's performance can be compared to
PEABODY INDIVIDUAL ACHIEVEMENT TEST-REVISED/NORMATIVE UPDATE 215

the performance of the standardization sample). For the five basic sub-
tests (General Information, Reading Recognition, Reading
Comprehension, Mathematics, Spelling) and the two composites (Total
Reading, Total Test), the student's derived scores can be converted to:
age equivalents, grade equivalents, standard scores (for both age and
grade - age by three months intervals, grade by fall/winter/spring, based
upon time of administration), percentiles, normal curve equivalents, and
stanines. Total Reading composites are obtained by adding the raw scores
of the Reading Recognition and Reading Comprehension subtests. The
Total Test composites are calculated by adding all the raw scores of the
subtests, except for those scores of the Total Reading composite and the
Written Expression subtest. For the Written Expression subtest (Level I
& 11), derived scores can be converted to grade based stanines, while for
Level II the derived scores can be also converted to developmental scaled
scores.
Since a student's "true score" is impacted by a measurement error
(that is seen in every assessment), the PIAT~R/NU scores are considered
an estimate of a true score. Statistically, one can determine the estimate
of the measurement error through the utilization of standard error of
measurement (SEM). From the standard error of measurement, one can
determine a range of scores {confidence interval} wherein the "true
score" will likely lie. These confidence intervals are provided for grade
and age equivalents, standard scores, and percentiles at the 68%, 90%
and 95% levels.
Graphic profiles can be developed for a student's subtest scores
(except for Written Expression) in two formats: a Standard Score Profile
for showing standard scores and a Developmental Score Profile for graph-
ically showing age equivalents or grade equivalents. A student's confi-
dence intervals are plotted on the graph provided on the student's test
booklet. A reference line can then be drawn with actual grade place-
ment, chronological age, mental age, and IQ score, in order to compare
the student's PIAT~R/NU performance with expected performance.

Interpretation of Scores
The PIAT~R/NU, through the manual, allows the interpretation of
the student's derived scores in relation to the standardization sample. A
detailed case study is given to show the derived scores' relationship in
terms of comparisons to peers and in terms of the student's actual indi-
vidual performance. A student's confidence intervals allow one to inter-
216 DIAGNOSTIQUE VOl. 24, NO. 1-4, 1998-1999

pret a functioning range and thus allow highlighting a student's strengths


and weaknesses for educational planning. Caution must, however, be
given since the nature of the PIAT~R/NU's multiple choice format could
allow for unreliable reporting of strengths and weaknesses between the
academic subtests.
Looking at the student's plotted confidence intervals, one can
determine whether a difference between subtest scores is a real achieve-
ment difference: overlaps by more than half their length are unlikely to
reflect a real difference; overlaps by half or less their length indicate a
possible real difference; and overlaps that do not exist indicate a proba-
ble real difference.
Evaluating differences can also be completed by viewing the
Minimum Differences tables (H.3 for grade, H.4 for age, in the manual's
appendix) showing statistically significant subtest standard score differ-
ences. Unusual differences can be evaluated by viewing the Unusual
Scatter tables (H.5 for grade, H.6 for age, in the manual's appendix).
As with any assessment scores evaluation, the examiner must be
cautioned to utilize the PIAT~R/NU as part of a full battery of tools in
order to provide comparison between instruments and the results they
generate. Furthermore, any minimum numeral difference, by its face
value, can obscure small but true individual differences in the student's
academic/educational profile.

Standardization
Sampling Procedures
The 1998 PIAT~R/NU was renormed based on data collected
between 1995 and 1996. This renorming was coordinated with the
renorming of the Key-Math-Revised, the Woodcock Reading Mastery
Tests-Revised, and the Kaufman Test of Educational Achievement. No
changes were made to the batteries' contents.
Between October, 1995 and June, 1996, along with between
September 1996 and November 1996, a forty state national standardiza-
tion of 3,184 kindergarten through grade 12 students and an additional
245 age eighteen to twenty-two young adults were tested. Students were
selected by a multistage sampling based on the March 1994 Bureau of the
Census Current Population Survey (U. S. Bureau of the Census, 1994).
Site coordinators, psychologists, or educators with prior experience in
the test batteries, were recruited for school-based sampling. The fall and
spring norm samples were derived by using a stratified random process by
PEABODY INDIVIDUAL ACHIEVEMENTTEST-REVISED/NORMATIVE UPDATE 217

computer. A target distribution, within each age or grade group, was


developed for the variables of educational placement, race/ethnicity, sex,
geographic region, and parental education. As each student wasselected,
he/she was randomly assigned one of the test batteries available for
administration at that site. Consequently, one-fifth of the students at
each grade took the PIAT~R/NU battery.

Sample Characteristics
The PIAT~R/NU manual gives tables listing the stratified sample
according to grade, sex, socioecononic status (parental education),
race/ethnicity, region, and educational placement. Upon review, the
total standardization sample appears to closely resemble the country as a
whole. Utilizing the March 1994 Census data of U.S. population (U.S.
Bureau of the Census, 1994) figures of 49.3% female, the sample provides
50.1%j and 50.7% male, the sample provides 49.9%. Distribution by
parental education level and by race/ethnicity shows that the sample
closely alines with the general population. From the U.S. Census classi-
fication, four categories of race/ethnicity were utilized: Hispanic,
African-American, White, and Other (including Asian, Pacific Islander,
Native American, and Alaskan Native). A new component of this
renorming included the designation of educational placement in order to
ensure the appropriate proportions of special education and gifted stu-
dents. The special education category was broken down into the follow-
ing classifications that closely resemble the U.S. percentages (U.S.
Department of Education, 1995): Learning Disabled 5.9%/5.5%US,
Speech/Language Delayed 2.4%/2.3%US, Emotionally/Behaviorally
Disturbed 0.8%/0.9%US, Mentally Retarded 0.8%/1.1%US, Physically
Impaired 0.1%/0.1%US, Other 0.8%/0.3%US. The Gifted classification,
unfortunately, did not resemble the U.S. percentages as closely:
2.3%/14.2%US.

Reliability
Estimating reliability of the PIAT~R/NU was completed by the four
methods of: test-retest, split half, item response theory, and
Kuder-Richardson. Tables in the manual cite the data.

Test-Retest Reliability
Stability of scores from one test session to another was determined
after a two to four weeks delay from the initial testing, with substantial
218 DIAGNOSTIQUE VOL. 24, NO. 1-4, 1998-1999

agreement between first and second testing scores. Based upon reported
data, most of the subtest correlations fell within the mid .80s to high .90s
with lower figures seen in grade four Mathematics (.78), grade six
Reading Comprehension (.78), and in grade six Mathematics (.79). The
composite correlations were reported as being from the low to upper .90s.

Split-Half Reliability
High levels of split-half reliability were documented for all age and
grade levels, due, in all likelihood, to the usage of basal and ceiling oper-
at ions of the test itself. Bygrade, for most of the first five subtests and two
composites, the coefficients are in the .90s for subtests and from .95 to
.99 for the composites. Kindergarten levels for Spelling (.86) and
Mathematics (.84) showed lower reliability. By age, the coefficients
reflect similar patterns: with one exception, the age subtest reliabilities
fall in the .90s, while the composite reliabilities fall within the upper
.90s. At the age five level (lowest level tested), Mathematics provides
the lowest coefficient (.83).

Item Response Theory Reliability


Using the classical test theory, according to the Rasch model, given
the difference between the item's difficulty and the student's ability, the
probability of the student answering a certain item correctly can be
determined. The item response theory reliability coefficients are in the
mid to high .90s with the lowest being the Kindergarten and age five
levels in Spelling (.93) and Mathematics (.91). These very high coeffi-
cients attest to strong reliability.

Kuder-Richardson Reliability
A strong degree of content homogeneity has been shown with the
reliability coefficients falling in the low to mid .90s, except for
Mathematics age five and Mathematics and Spelling grade Kindergarten.
With a comparison between these coefficients and those from the
split-half study, the little difference between the two reliabilities indicate
that the PIAT,R/NU provides unambiguous and clear content domains.

Validity
Content Validity
Based upon the reliability estimates of the split-half and
PEABODY INDIVIDUAL ACHIEVEMENT TEST-REVISED/NORMATIVE UPDATE 219

Kuder-Richardson reported in the previous discussion and the extensive


developmental procedures followed for each of the PIAT-R/NU subtests
(cited in an entire section of the manual), individual test items and the
test as a whole appear to adequately address the content domain of the
PIATR/NU. The revision of the 1970 PlAT, completed in 1989, led to
the increase in the number of test items in each of the first five subtests
and the inclusion of a new sixth subtest (Written Expression). Field test-
ing of the new Written Expression, item analysis, and extensive devel-
opment of the previous five subtests, addressed the content validity of
the instrument as a whole.

Construct Validity
Although there are a few exceptions (not named in the manual),
high correlations between PIAT-R/NU subtests and those same subtests
on the PlAT are evidence as a measure of PIAT-R/NU correlation with
other tests. The two forms of the Peabody Picture Vocabulary
Test-Revised (PPVT-R) (Dunn & Dunn, 1981) were given with the
PIAT-R/NU during the standardization. As reported, the median corre-
lations between the two instruments are moderate (.52 to .70). Studies
of developmental changes (nature of achievement over time) and factor
analysis are presented in tables (in the manual) to support the
PIAT-R/NU validity. These tables list the noted increases in the stu-
dents' scores at each grade and age level showing that the test's content
sequence corresponds with achievement expectations. In the manual,
fifty one validity studies are listed with descriptive information and ref-
erences. It appears, however, that further validity data is needed for the
PIAT-R/NU. In fact, in the manual itself, research on the PIAT-R/NU is
encouraged and solicited.

Summary and Conclusions


The PIAT-R/NU appears to be a well standardized wide range
achievement test for students in Kindergarten through grade twelve. The
PIAT-R/NU consists of five subtests and two composites: General
Information, Reading Recognition, Reading Comprehension, Total
Reading, Mathematics, Spelling, Total Test, along with an optional
Written Expression subtest. The five subtests come from the original
PlAT instrument, but include new items, content updates, and a re-stan-
dardization. The Written Expression is a new subtest for the PIAT-R.
The construction and format of the instrument are user friendly with
220 DIAGNOSTIQUE VOL. 24, NO. 1-4, 1998-1999

scoring guidelines and space for observations allowing the examiner ease
in interpretation and analysis.
The PIAT,R/NU allowsfor quick administration and scoring, use of
easels for testing, and use of training examples and demonstration. The
clearly understood format involves a multiple choice visual stimulus that
requires a motor (pointing) or verbal response for the Reading
Comprehension, Mathematics, and Spelling subtests. These response
modes make the instrument easy to use with students with motor or
speech difficulties.
The Written Expression subtest does allow one to evaluate the
practical aspects of the entire writing process, however, the utility of the
scores has been questioned. Consequently, an informal analysis of a stu,
dent's writing sample should be an additional component of an examin-
er's test battery.
The current new standardization updates the sampling to show the
inclusion of school age special education and gifted students in direct
correlation to the overall public school population. This allows the
PIAT,R/NU to be utilized for information regarding in-class inclusion
and for evaluating IEP goals and objectives.
Throughout the literature (including the actual manual), the need
for further research on the PIAT,R/NU is cited (and welcomed by the
author). In fact, concurrent validity studies are being called for in order
to delineate the correspondence of PIAT,R/NU with other current
achievement instruments.

References
Dunn,L.M., & Dunn,L.M. (1981). Peabody Picture Vocabulary Test,Revised. CirclePines,
MN: American Guidance Service.
Dunn, L.M., and Markwardt, Ee. (1970). Peabody Individual Achievement Test. Circle
Pines, MN: American Guidance Service.
Markwardt, He. (1989). Peabody Individual Achievement Test-Revised. Circle Pines, MN:
American Guidance Service.
Markwardt, HC. (1998). Peabody Individual Achievement Test-Revised NU. Circle Pines.
MN: American Guidance Service.
Taylor, R.L. (1997). Assessment of Exceptianal Students (4th ed.). Needham Heights, MA.
Allyn & Bacon.
UiS, Bureau of the Census. (1994). Current Population Survey. Washington. DC:
U.S.Government Printing Office.
U.S. Department of Education. (1995). Seventh Annual Report to Congress on the
Implementation of the Individuals with DisabiUties Education Act. Washington, DC: U.S.
Government Printing Office.

You might also like