• Test Review: Wide Range Achievement Test (WRAT4)


  •   
  • FileName: joped-11-no-1-2009-test-review-2-wrat4.pdf [read-online]
    • Abstract: Test Review: Wide Range Achievement Test (WRAT4)Wide Range Achievement Test 4 (WRAT4)Published by: Pearson Assessment UK, 2006Reviewed by Colleen Makray and Giselle HopeJobcentre Plus Work Psychology Service, North West Regional Office, Ontario House,

Download the ebook

Test Review: Wide Range Achievement Test (WRAT4)
Wide Range Achievement Test 4 (WRAT4)
Published by: Pearson Assessment UK, 2006
Reviewed by Colleen Makray and Giselle Hope
Jobcentre Plus Work Psychology Service, North West Regional Office, Ontario House,
2 Furness Quay, Salford M50 3XZ
Background to the WRAT Whilst the WRAT4 can be used as part of more
comprehensive batteries to assess cognitive
abilities, used alone it can still provide a
The Wide Range Achievement Test was first
straightforward and accurate screen measuring
published in 1946. It was originally developed
the skills of learning. Importantly, the addition
by the psychologist Joseph Jastak to provide
of a sentence comprehension measure in this
information on aspects of word recognition,
version gives a clearer measure of
spelling and arithmetic which would
understanding beyond the ability to decode
complement measures of cognitive ability being
words.
developed by David Wechsler at that time.
There have been several revisions, including
the WRAT-R, the Wide Range Achievement Materials
Test-Revised (1984) and the WRAT3, or Wide
Range Achievement Test: Third Edition (1993). • Professional Manual
Its measures, combined with the Wechsler • Administrator Test Forms - the WRAT4 has
scales, aimed to provide a broader picture of been designed so the user can make a
cognitive abilities. Jastak felt the basic choice in the content of the test. There are
academic codes of word recognition, spelling two forms (blue and green for each subtest).
and arithmetic were carriers of thoughts, rather These were developed and equated during
than thoughts themselves, and would use the standardisation and may be used separately
analogy of a truck carrying cargo to explain their or interchangeably with comparable results
differences, e.g., he saw codes as the vehicle of • Word Reading List/Spelling List Card
cognition, with meanings as their cargo. Jastak
• Sentence Comprehension Sample Card
felt that no complete assessment of academic
performance could be carried out without • Sentence Comprehension Card
measuring both functions. • Participant Response Forms
Additional requirements: pencil, eraser and
The most recent version, the WRAT 4 (2006) stopwatch.
includes:
The cost of a WRAT4 starter set is currently
• Updated norms – age based norms extended £208.69 (inc VAT).
from 5.0 – 74.11 years to 5.0 – 94.11 years
• Inclusion of a Sentence Comprehension Description
subtest using the cloze technique
• In the Reading subtest the number of words The WRAT4 is a norm-referenced test that can
has increased from 42 – 55 be administered to individuals from 5 to 94
• In the Spelling subtest the number of words years of age. It is made up of Word Reading,
has increased from 40 – 42 Sentence Comprehension, Spelling and
• Two forms (Blue and Green) Arithmetic subtests.
Journal of Occupational Psychology, Employment and Disability, Vol 11, No 1, Spring 2009
49
Test Review: Wide Range Achievement Test (WRAT4)
(i) The Word Reading subtest measures letter In this latest version there is also a Reading
and word decoding. It is made up of two Composite Standard Score, based on the sum
components: Letter Reading, which is made up of standard scores on the Word Reading and
of 15 letters, and Word Reading, made up of 55 Sentence Comprehension subtests. This
words, are presented without a context (ie, the Reading Composite score provides a more
words are read in isolation rather than in a comprehensive measure of reading
sentence), which the participant is required to achievement than a single measure.
recognise and pronounce aloud. If an individual
does not read a word after 10 seconds they are As noted above, the four subtests are available
encouraged to try the next word. in two forms (Blue or Green) that can be used
either separately or together (known as the
(ii) The Sentence Comprehension subtest Combined Form) in order to provide the test-
measures the ability to comprehend ideas and user with a choice in the content. If separate
information contained in a sentence through the forms are used, it also allows the user to retest
use of a cloze technique (eg, John had a within a short period of time, with reduced risk
headache so he took an _______.) It comprises of practice effects.
50 questions, of increasing difficulty, with
different starting points (A- F).The participant is The test-user may forego the preliminary items
asked to read through a sentence to him or in Word Reading, Spelling and Arithmetic if the
herself silently, then to verbally provide a word participant is able to answer the first five items
or words (eg, phrasal verbs such as ‘get out’ in any of these subtests.
convey the same meaning as ‘exit’) to complete
the meaning. A participant’s score on the Word Administration Time
Reading subtest establishes the starting point
on the Sentence Comprehension subtest. This will depend on the age and skill level of the
participant, and number of subtests undertaken.
The Sentence Comprehension subtest does not For all four subtests, average administration
include preliminary items. However, five time is estimated at 15-25 minutes for
consecutive items must be answered correctly participants from 5-7 years of age, and 30-45
at any starting point in this subtest, or the minutes for older participants.
participant will be tested ‘backwards’ (in reverse
order) until five consecutive correct answers are
achieved.
Technical Information
(iii) The Spelling subtest measures the ability to Norms
encode dictated words into a written format.
This has two components. First, the participant The WRAT4 norms have been designed to
is asked to spell his or her name, along with 13 produce age- and grade-based standard scores
preliminary letters. This is then followed by the and percentile ranks, as well as optional
dictation of up to 42 words of increasing stanines, normal curve equivalents (NCEs),
difficulty. Participants are allowed 5 seconds to grade equivalents, and Rasch ability scaled
write each letter and 15 seconds to spell each scores for the Blue, Green and Combined
word. forms.
(iv) The Arithmetic subtest involves counting, In the standardisation of the test, a sample of
number and symbol recognition, solving oral approximately 3,000 people was used, with
problems, and performing written computations. between 100 and 150 people assigned to 19
The participant is given 15 minutes to complete age groups (also known as intervals or
the Arithmetic subtest. sampling frames) corresponding to the 2001 US
census population with respect to gender,
race/ethnicity, educational attainment and
Journal of Occupational Psychology, Employment and Disability, Vol 11, No 1, Spring 2009
50
Test Review: Wide Range Achievement Test (WRAT4)
geography. Forty three norm groups were then Research Association,1999), building on
established for age-based standard scores, previous WRAT research and the more recent
starting with twelve 3-month intervals at the process of assessing reliability and validity of
youngest ages (ages 5:0-5:2 to 7:9-7:11) where the WRAT4. The Professional Manual not only
development and growth is considered greatest, summarises reliability measures undertaken,
to seven 10-year intervals (25-34 to 85-94) but outlines the rationale for each method used.
where growth curves decline. Reliability measures include:
Grade-based standard scores (US K-12, Fall • Internal Consistency Reliability
and Spring Terms) were developed with similar • Alternate-Form Reliability, including
methods to those described above. However, Immediate and Delayed Retest
as grade-based scores would not necessarily • Standard Error of Measurement
apply to UK–based grades, and would not be
relevant to adults, norm development will not be • Standard Score Confidence Intervals
summarised. • Rasch IRT (Item Response Theory) Statistics
Percentile ranks, stanines and normal curve Internal Consistency Reliability
equivalents (NCEs) conform to the statistical
properties of the normal curve and therefore link By age, the median coefficient alpha subtest
directly to the age-based standard scores. reliability coefficients range from .87 to .93 for
Further optional scores are available eg, overall the four subtests, with similar reliabilities for the
grade equivalents (median ability scores for the Blue and Green forms. For the Reading
fall and spring testing seasons at each grade), Composite, the median alpha reliabilities range
and Rasch ability scaled scores, but these, like from .95 to .96 for the Blue and Green forms.
the grade-based standard scores, are linked to The Combined Form fares even better with a
US school grades (Kindergarden to Grade 12), Reading Composite median alpha coefficient of
seasons (Fall and Spring) and precise testing .98 for the age group.
periods.
Alternate-Form Reliability, Immediate
As a result, there are a number of appendices
included in the Professional Manual in which to and Delayed Retest
interpret scores on the Blue, Green and
Combined forms. As well as the standard score The average alternate-form subtest reliability
conversions, as outlined above, there are three coefficients range from .82 (ages 19-34) to .90
confidence levels provided for each of the (ages 10-11), based on eight age groupings.
subtests and Reading Composite scores, as Overall, the average subtest reliability
well as information on significance levels for coefficients is .86 for all participants. With
differences between pairs of subtest standard respect to delayed retesting, the average
scores (at four levels of significance) and alternate form-delayed retest reliability
prevalence of differences between pairs by coefficients are .86 (ages 7-18), .84 (ages 19-
percentage of the standardisation sample. This 94) and .84 for all ages combined.
provides a comprehensive range of
measurements in which to assess individual Standard Error of Measurement and
performance. Standard Score Confidence Intervals
Reliability Used in constructing a band or range of scores
or confidence intervals, the median standard
The reliability of the WRAT4 follows guidelines error of measurement (SEM) for the WRAT4
outlined in US Standards for Educational and subtests by age group ranges from 3.0
Psychological Testing (American Educational (Reading Composite) – 5.0 (Arithmetic) on the
Journal of Occupational Psychology, Employment and Disability, Vol 11, No 1, Spring 2009
51
Test Review: Wide Range Achievement Test (WRAT4)
Blue Form, 3.0 (Reading Composite) – 5.4 Arithmetic subtests. Overall, 77% of WRAT4
(Arithmetic) on the Blue Form, and 2.3 (Reading items overall were taken from the WRAT3. With
Composite) – 3.7 (Arithmetic) on the Combined respect to the subtests, Arithmetic contained
Form. These measurements, along with those 30% new items, Word Reading 26.4%, and
obtained from grade level data, were used to Spelling 11.9%.
develop the standard score confidence
intervals. As noted above, three confidence However, the new Sentence Comprehension
levels are provided in the norms tables, and the subtest required item development and review
level chosen (eg, 85%, 90%, or 95%) is at the in establishing suitability of the content for the
discretion of the test-user. WRAT4. The steps used to establish suitability
included content development, item tryout,
Rasch IRT Statistics expert item review, and selection of the final
items based on standardisation data.
Person-separation reliability coefficients range
from .96 to .98 (Blue Form) and .95 to .98 Increase in Mean Subtest Scores
(Green Form), indicating that the subtests
differentiate well amongst individuals in the In a test like the WRAT4 it is proposed that
sample. Moreover, item-separation reliability acquisition of skills tested will increase with an
coefficients on both the Blue and Green forms individual’s age. The data provided support this
are 1.00, which also provide evidence of assertion, with means of raw scores of both the
construct validity. Blue and Green forms increasing incrementally
until the ages 45 – 54, followed by a gradual
Validity decline in the remaining age groups, from 55 –
94.
In the Manual the authors stress that validity of
any test is a cumulative and ongoing process, Intercorrelations among Subtests
and that in particular the WRAT4 builds on
processes developed in the WRAT3 and its Two broad age groups were measured in
predecessors. The exception to this is the subtest intercorrelations: 5 – 18 years, and 19 –
Sentence Comprehension subtest, which has 94 years. As predicted, there was moderate
been newly subjected to validation processes. subtest overlap for these groups, with median
The validity section summarises both internal intercorrelations of .63 for both groups.
and external validity evidence, and includes:
Item Bias or DIF (Differential Item
• Content and development of the subtests Functioning)
• Increase in mean subtest scores in
successive age levels Independent calibrations of items were carried
• Intercorrelations amongst subtests out on both the Blue and Green form subtests
• Item Bias or DIF (Differential Item with data from the following contrast groups:
Functioning) male/female, African American/Caucasian, and
Hispanic/Caucasian. The resulting data indicate
• Correlation of the WRAT4 with other
high agreement with few outliers, suggesting
measures of cognitive ability and
that minimal DIF emerged in the subtests.
achievement
Correlations with other cognitive
Subtest Content and Development ability and achievement measures
Development of the WRAT4 was based on a Whilst the above measures provide internal
review of the WRAT3 content. Minor revisions evidence of validity of the WRAT4, it was also
were applied to the Reading, Spelling and compared with other achievement tests,
Journal of Occupational Psychology, Employment and Disability, Vol 11, No 1, Spring 2009
52
Test Review: Wide Range Achievement Test (WRAT4)
including the Wechsler Individual Achievement requirement in the administration of the WRAT4
Test, Second Edition (WIAT – II); Woodcock– required in the UK.
Johnson III Tests of Achievement (W-J III), and
the Kaufman Tests of Educational Achievement: Relevance of WRAT4 in
Second Edition-Comprehensive Form and
Second Edition-Brief Form (KTEA-II- Employment Assessments
Comprehensive and KTEA-II-Brief). All the
WRAT4 subtests are moderately or moderately- The WRAT4 has a number of positive features:
highly correlated with like subtests on
achievement tests compared, but less so with • It can be administered in less than an hour
subtests of unlike subtest, providing evidence of (approximately 30-45 minutes administration
concurrent validity. time with adults).
• Standard scores can be used for comparison
The WRAT4 has been compared with tests of with other tests, eg, WAIS-III, WASI, to
cognitive ability, including the Wide Range consider discrepancies in reading/spelling.
Intelligence Test (WRIT), Wechsler Intelligence • Standard scores can be used to highlight
Scale for Children, Fourth Edition (WISC IV), discrepancies between each of the subtests
the Stanford-Binet Intelligence Scales, Fifth contained in the WRAT4.
Edition (SB-5) the Wechsler Adult Intelligence
• It has good internal reliability measures: .87 -
Scale, Third Edition (WAIS-III), the Wechsler
.93 for the subtests; .95 - .98 for the Reading
Abbreviated Scale of Intelligence (WASI: The
Composite.
Psychological Corporation, 1999), the Kaufman
Brief Intelligence Test: Second Edition (KBIT-II), • The user can make choices in the content of
and the Reynolds Intellectual Assessment the test (Blue or Green) and in administration
Scales (RIAS), all of which provide of the form (Separate or Combined).
demographic characteristics of the samples, • The inclusion of the Sentence
including age, gender, race/ethnicity, and parent Comprehension subtest allows the
education. Correlations between full scale IQs measurement of an extra dimension to
of these tests and WRAT4 subtest scores reading whilst the Word Reading section can
indicate a moderate relationship between the directly assess the skills required in reading,
subtests and cognitive ability tests, with median (i.e., decoding sounds, making sound/symbol
correlations ranging from .57 for Spelling to .72 relationships and sight recognition separate
for the Reading Composite subtest. It was to comprehension).
predicted that the verbal indices would correlate • The test form includes a full page Score
more highly with measures of academic Summary Table (outlining raw and standard
achievement skills on the WRAT4 than the scores in relation to chosen norms, ie, age or
nonverbal measures, and this was confirmed by grade) a Standard Score Profile in a graph
the data. form and, as noted above, a Standard Score
Comparison Table in which subtests can be
Restrictions on Usage compared (at different significance levels and
in relation to prevalence in the
In its User Qualifications guidelines, the standardisation sample).
Professional Manual states that the WRAT4 has
been designed to be administered and Within an employment assessment this can
interpreted by ‘individuals who have the assist in:
necessary training and experience’ outlined in
the US Standards for Educational and • The provision of information to distinguish
Psychological Testing (American Educational between a general learning difficulty and a
Research Association, 1999). As such, the specific learning difficulty, (when used in
author does not specify a Level A certificate conjunction with other test materials).
Journal of Occupational Psychology, Employment and Disability, Vol 11, No 1, Spring 2009
53
Test Review: Wide Range Achievement Test (WRAT4)
• The provision of guidance and advice to • Standardisation has been done exclusively in
customers with literacy or numeracy the US and norms are derived entirely from
difficulties who may benefit from further the US population, so norms should be used
support or strategies in helping them to with caution with the British population.
achieve their intended job goal, for example, • Whilst the Professional Manual suggests that
determining the type or level of support the WRAT4 can be completed in 30 - 45
which might help an individual in the minutes by adults, as there is no overall time
workplace. limit on 3 of the 4 subtests, administration
could take much longer.
Practical points to consider when deciding
whether and how to use the WRAT4 include:
Summary
• In the Spelling subtest some words may be In summary, the WRAT4 can provide an
more common to the US population (eg, indicative screening measure of fundamental
physician), or reflect US spelling rules (eg, reading, comprehension, spelling and arithmetic
recognize). skills in adults with different learning
• The phonetic system provided for the backgrounds without the need to select four
spelling component is US based. separate tests. It can also be used in
conjunction with other tests to provide a more
• The Sentence Comprehension subtest
comprehensive picture of learning ability or
reflects US spelling rules and uses
needs
vocabulary and situations that may be more
common to that population (eg, references to
A substantial improvement on the WRAT3 is the
apartments, Abraham Lincoln, Americans
Sentence Comprehension subtest, which
abroad, North American wildlife, etc.)
measures the ability to comprehend ideas and
• The newest version uses three test forms. information, and the new Reading Composite
This can prove unwieldy, particularly if standard score, which provides a more holistic
following the suggested administration of measure of the ability to both recognise and
tests (ie, Word Reading, Sentence make sense of words. Moreover, the WRAT4
Comprehension, Spelling and Arithmetic), as can now measure discrepancies between
it involves one form for the Word Reading, a individual subtests of Word Reading, Sentence
second for Sentence Comprehension, then Comprehension, Spelling and Arithmetic, which
returning to the Word Reading form for was not available on the WRAT3.
Spelling Administration instructions, whilst
the test taker completes the subtest on a In an employment assessment, the WRAT4 can
separate (third) form. Instructions for the provide a useful starting point in assessing
Arithmetic subtest are also found in the Word abilities or potential support needs which can be
Reading form, whilst the test taker completes explored in more detail with the individual
this subtest on a separate (third) form. thereafter.
• The 5/7/10 rules, which refer either to the
number of items which must be answered References
correctly to waive preliminary items, or to the
number of incorrect items a participant may
American Educational Research Association,
answer incorrectly before discontinuing a
American Psychological Association, &
subtest, vary between tests and are not
National Council on Measurement in
always easy to monitor. This is particularly
Education (1999). Standards for educational
true of the Spelling subtest, where the
and psychological testing. Washington, DC:
participant is in possession of the response
American Educational Research Association.
form so incorrectly spelt items may not
always be visible to the administrator.
Journal of Occupational Psychology, Employment and Disability, Vol 11, No 1, Spring 2009
54
Test Review: Wide Range Achievement Test (WRAT4)
Publishers/Suppliers of Tests Wechsler Adult Intelligence Scale (WAIS-III).
(2005). San Antonio, TX: Psychological
Kaufman Brief Intelligence Test, Second Corporation.
Edition. (KBIT-II). (2nd ed). Los Angeles:
Western Psychological Services. (2003). Wechsler Individual Achievement Test (WIAT-II).
Lutz, FL: Psychological Assessment (2001) (2nd UK Edition) (WIAT-II UK). San
Resources. Antonio, TX: Psychological Corporation.
Kaufman Test of Educational Achievement, Wechsler Intelligence Scale for Children
Second Edition (KTEA-II). (2005). San (WISC). (2003). San Antonio, TX:
Antonio, TX: Psychological Corporation. Psychological Corporation.
Reynolds Intellectual Assessment Scales Wide Range Intelligence Test (WRIT). (2000).
(RIAS). (2003). Odessa, FL: PAR, Inc.
San Antonio, TX: Psychological Corporation.
Stanford Binet Intelligence Scales (SB-5).
(2003). (5th ed). Itaska, Il: Riverside Woodcock-Johnson III Tests of Achievement
Publishing. (W-J III). (2001). Itasca, Il: Riverside
Publishing
Wechsler Abbreviated Scale of Intelligence
(WASI). (1999). San Antonio, TX:
Psychological Corporation.
Colleen Makray is a Higher Work Psychologist with Jobcentre Plus. Prior to joining the
department, she worked in survey research and training roles. Giselle Hope was also a Higher
Work Psychologist in the North West team until she left in late 2008 to start work as an
independant practitioner.
Journal of Occupational Psychology, Employment and Disability, Vol 11, No 1, Spring 2009
55


Use: 0.0353