Four
How to Interpret the WJ IV ACH

The WJ IV ACH provides a rich variety of interpretive options and scores. You must know what scores are available, when to use the various scores, and how to interpret these scores. The purpose of the assessment dictates what scores are most appropriate to use. In addition, you need to know what each test and cluster measure, what skills and abilities are required to perform the tasks, and what implications may be derived. Interpreting the WJ IV ACH requires a higher level of skill than administering the test. Prior to scoring, you must first decide whether to use grade- or age-based norms.

Grade- or Age-Based Norms

The WJ IV ACH provides the option for using either grade- or age-based norms. When making school-based decisions regarding achievement levels, grade-based norms are generally preferable. Age-based norms are more applicable in clinical settings or with adults, and should be used in cases in which the results will be compared to scores from another test that only provides age norms (e.g., comparing the WISC-V to the WJ IV ACH). Selection of grade or age norms does not affect the obtained age equivalents (AEs) or grade equivalents (GEs), but differences will be noted in the standard scores, percentile ranks, and relative proficiency index scores. The Woodcock-Johnson online scoring and reporting program (Schrank & Dailey, 2014) provides the option to make grade or age comparisons. The grade norms are available for grades K.0 to 17.9. Although it is usually clear which grade norm group to select in the scoring program, there may be some confusion if you wish to compare an examinee's performance to individuals in the first year of graduate school (17.0 to 17.9). There is not a separate entry for graduate school. You must select the 4-year college norms, which include the first year of graduate school. Be sure to clearly indicate in reports which norm group was used for scoring.

Types of Scores

When reporting results to parents, teachers, and examinees, select the scores that are most meaningful and easily explained. Some metrics are easier to interpret than others. Age and grade equivalents or percentile ranks can be useful for discussions with parents and teachers who may more easily understand these types of scores than standard scores. As an examiner, you need to understand the purpose and interpretation of each type of score and be able to describe and discuss each type of score accurately.

The next section presents a detailed description of the WJ IV ACH scores. This overview includes AEs, GEs, relative proficiency indexes, instructional zones, Cognitive Academic Language Proficiency (CALP) levels, percentile ranks, and standard scores. It also presents additional optional standard score scales (i.e., z scores, t scores, normal curve equivalents, and stanines). The scoring program generates all of the derived scores for all tests and clusters.

Raw Score/Number Correct

For most tests, the raw score is the number of correct responses plus the number of items below the basal that were not administered, each receiving one point. Two tests (Test 6: Writing Samples, Test 8: Oral Reading) use multiple-point scoring so the raw score is based on the total number of points earned; Test 9: Sentence Reading Fluency uses both the number correct and the number incorrect on the range of items attempted in obtaining the scores; and Test 12: Reading Recall uses the points earned on administered stories. As shown in Figure 4.1, the number correct is listed in the first column on the left in the Scoring Table that appears for each test in the Test Record. This information is followed by the corresponding AEs and GEs. Chapter 3 presents procedures for calculating the number correct.

img

Figure 4.1 Scoring Table for Test 1: Letter-Word Identification From the Test Record

Copyright © 2014 by The Riverside Publishing Company. Reproduced from the Woodcock-Johnson® IV (WJ IV®), WJ IV® Tests of Achievement. Reprinted with permission of The Riverside Publishing Company. All rights reserved.

Zero Scores

In cases in which an individual receives a zero on any test, you must decide whether that score represents a true assessment of ability or reflects an inability to perform the task. If the individual has not been exposed to the type of task in question, it may be more appropriate not to score the test rather than interpreting a zero raw score. For example, Jamie, a kindergarten student, obtained a zero on the Word Attack test. This score may indicate that she has not yet been exposed to letters and sounds and, therefore, this score should not be interpreted. Max, a fifth-grade student, also had a score of zero on the Word Attack test. In this case, that score may be an accurate reflection of his very limited reading ability and, therefore, should be interpreted. Even when a zero score is considered an accurate reflection of ability and is entered into the scoring program, it produces only age and grade equivalents. Because of the problems associated with interpretation, no other derived scores are available for zero scores at the individual test level. All of the derived scores, however, are reported for clusters that include a test with a zero raw score. If all tests within a cluster have a raw score of zero, then no derived scores are reported.

W Score

The W score (Woodcock, 1978; Woodcock & Dahl, 1971) is a special transformation of the Rasch ability scale (Rasch, 1960; Wright & Stone, 1979). The W scale for each test is centered on a value of 500, which is set to approximate the average performance at age 10-0 and grade 5.0. Cluster scores are the average (arithmetic mean) W score of the tests included in that cluster. For example, the cluster score for Broad Written Language is the average W score of Test 3: Spelling, Test 6: Writing Samples, and Test 11: Sentence Writing Fluency. In the WJ IV, the W score is not shown in the Test Record but appears in the score report generated by the online scoring program. Each raw score has a corresponding W score. The scoring program converts raw scores into W scores. Because W scores are equal-interval scores, they are the preferred metric for use in statistical procedures.

Age and Grade Equivalents

An age equivalent (AE), or age score, reflects performance in terms of the age level in the norm sample at which the median raw score and corresponding W score are the same as the examinee's score. For example, if the median raw score or number correct for 10-year-olds in the norm sample is 43 on a particular test, then any examinee who obtains a raw score of 43 on that test receives an AE of 10-0. AEs may be more useful in some applications than grade equivalents, especially as they relate to the abilities of young children or adults not attending school. The AE is useful when attempting to determine an approximate level of development.

A grade equivalent (GE), or grade score, reflects the examinee's performance in terms of the grade level in the norm sample at which the median raw score and corresponding W score are the same as the examinee's score. In other words, if the median W score (or corresponding raw score) on a test for students in the sixth month of the second grade is 488, then an individual who earns a W score of 488 would receive 2.6 as a grade score. The GEs on tests like the WJ IV ACH represent the midpoint of the individual's instructional zone and can be used for instructional planning. Because the test includes items distributed over a wide range of difficulty levels (rather than a limited range typically found on group-administered tests), the age and grade scores reflect the actual level of task difficulty an individual can perform. The GE is useful when attempting to determine an appropriate, approximate level for instructional materials.

At the lower ends of the age and grade scales, less-than signs (<) are used for levels of performance that fall below the median score of the lowest age or grade group reported. Greater-than signs (>) are used for levels above the median for the age or grade of peak performance. For example, if the age of peak performance is 34 years, an examinee who scored above the median for that age group receives an age score of >34.

You can obtain an estimate (Est.) of age and grade equivalents by using the scoring tables for each test in the Test Record (as shown in Figure 4.1). Precise age or grade equivalents for tests are obtained from the Woodcock-Johnson online scoring and reporting program (Schrank & Dailey, 2014). Both age equivalents and grade equivalents can be displayed in the same score report by selecting the additional score option. For example, if you are using grade norms, the score report automatically displays grade equivalents. To add in age equivalents, you would select a new score selection template and choose to include age equivalents. Age and grade equivalent scores for clusters are available only when using the scoring program. Examples of the types of statements that are used for grade and age equivalent scores are:

  • Jacob's grade equivalent score (GE) on basic reading skills indicates that he is reading at approximately a beginning third-grade level (GE = 3.1).
  • Marco, a 10th-grade student, scored similarly to the average student in early seventh grade (GE = 7.2) on the Applied Problems Test.
  • Diana's science knowledge as measured by Test 18: Science is comparable to average 8-year-olds.
  • The number of items Tom, a seventh grader, answered correctly on the math calculation task is comparable to the average student in early fourth grade.

W Difference Score

The W difference scores are based on the difference between an examinee's test or cluster W score and the median test or cluster W score (REF W) for the reference group in the norm sample (same age or same grade) with which the comparison is being made. W difference scores are used when deriving other scores such as the relative proficiency index and the standard score.

Relative Proficiency Index

The relative proficiency index (RPI) was originally called the relative mastery index (RMI) in the WJ-R (Woodcock & Johnson, 1989). The RPI allows statements to be made about an examinee's proficiency on the task and the predicted quality of performance on tasks similar to the ones tested. The RPI is expressed as a fraction. The denominator is a constant of 90. The numerator ranges from 0 to 100 and reflects the examinee's proficiency on the task and the predicted level of performance on similar tasks.

RPIs are based on the distance along the W scale that an individual's score falls above or below the median score for the reference group (Reference W or REF W). This is the W difference (W Diff) and reflects the individual's absolute distance from the average performance for age or grademates on the task. You would interpret an RPI of 45/90 on the Spelling test for Natalie, a sixth-grade girl, to mean that she is only half as proficient on the task as her average age or grademates. When others at her age or grade show 90% success, Natalie is predicted to show only 45% success on similar tasks. On the other hand, if her RPI were 99/90, you would predict that she would perform spelling tasks with 99% success that average age or grademates perform with 90% success. Rapid Reference 4.1 provides verbal labels for describing performance when using the RPI. Technically, the verbal labels are driven by the W Diff score rather than the RPI itself. This means that on occasion the same RPI may yield two different verbal labels. Note that there is overlap between the RPI ranges in Rapid Reference 4.1, but there is no overlap in the W Diff ranges. For example, the RPI of 100/90 appears in both the “advanced” and “very advanced” categories. Looking at the W Diff ranges, we can see that if one person has a +30 W Diff and another person has a +31 W Diff, they will get the same RPI of 100/90 but have different labels.

Depending on the purpose of the evaluation, one type of label may be more appropriate than another. For example, a developmental label may be more appropriate when testing a younger child, whereas a proficiency label may be more appropriate for describing the performance of a school-age student. Examples of the types of sentences used to describe RPIs are:

  • Akeisha's RPI of 30/90 on the Phoneme/Grapheme cluster indicates that when average sixth-grade students have 90% proficiency, Akeisha has only 30% proficiency. Her knowledge of sound–symbol relationships is limited.
  • Marla's RPI of 99/90 in Broad Mathematics indicates advanced proficiency in basic math skills and mathematical concepts. When average age peers have 90% success on similar tasks, Marla is predicted to have 99% success. Grade-level tasks in mathematics will be easy for Marla.
  • Bennett's RPI of 92/90 on the Academic Knowledge cluster indicates average proficiency compared to peers.
  • Although Judith's standard scores on both Broad Reading and Broad Writing are in the low range (standard scores of 70 and 72), her proficiency in reading (RPI = 4/90) is lower than her proficiency in mathematics (RPI = 35/90).

CALP Levels

Cognitive Academic Language Proficiency (CALP) levels, another application of the RPI, are helpful in determining an individual's language proficiency. CALP is defined as language proficiency in academic situations and includes those aspects of language that emerge with formal schooling. Aspects of language that are acquired naturally, without formal schooling, are referred to as Basic Interpersonal Communication Skills (BICS). Cummins (1984) formalized the distinction between these two types of language proficiency (BICS and CALP). When evaluating English language learners, it can be helpful to consider an individual's CALP levels in determining language proficiency, language dominance, or for planning an appropriate educational program. CALP levels are also helpful when considering the instructional needs of native English speakers with delayed or limited language. In addition, the WJ IV OL provides three parallel English and Spanish tests that can be helpful in evaluating language dominance and proficiency (see Chapter 5 for more information). CALP levels are available for certain WJ IV ACH clusters (see Table 4.1).

Table 4.1 Clusters for Which CALP Levels Are Available

Reading Clusters Written Language Clusters Other ACH Clusters
Reading Written Language Academic Skills
Basic Reading Skills Basic Writing Skills Academic Applications
Reading Comprehension Written Expression Academic Knowledge
Reading Comprehension–Extended Brief Achievement

To display CALP levels for any of these clusters, select the option to include CALP when using the online scoring program. Rapid Reference 4.2 illustrates the six CALP levels available plus two regions that fall between two levels, the relationship to the RPI, and the instructional implications of each.

Examples of the types of sentences used to describe CALP levels are:

  • Kai met the criteria for fluency on all of the WJ IV Oral Language clusters (CALP 4–4.5).
  • Raul's performance on the Basic Reading Skills cluster suggests extremely limited functioning in reading English (CALP = 1). It will be nearly impossible for him to read grade-level English books.
  • Daniel's CALP level of 2 on the Written Expression cluster suggests that he will find 12th-grade-level writing tasks to be extremely difficult.

Percentile Rank

A percentile rank uses a scale from 1 to 99 to describe performance relative to a specific age- or grade-level segment in the norm sample. The examinee's percentile rank indicates the percentage of people in the selected segment of the norm sample who had scores the same as or lower than the examinee's score. Percentile ranks are particularly useful for describing a person's relative standing in the population. For example, Ivan's percentile rank of 6 on the Reading Cluster indicates that his performance was the same as or better than only 6% of the population, whereas Susie's percentile rank of 94 indicates that her score was the same as or better than 94% of the population. In other words, only 6% of her age peers would have a higher score.

Extended percentile ranks (Woodcock, 1987) provide scores that extend down to a percentile rank of one tenth (0.1) and up to a percentile rank of 99 and nine tenths (99.9). If an individual's percentile rank is 0.2, this indicates that only two persons out of 1,000 (0.2%) would have a score as low or lower. If an individual's percentile rank is 99.8, this indicates that the person's performance was as good as or better than that of 998 persons out of 1,000 (99.8%) in the reference group, or that two persons out of 1,000 would have a score as high or higher. Extending the percentile rank scale adds discriminating measurement to the range of a traditional percentile rank scale—about three fourths of a standard deviation at the top and three fourths of a standard deviation at the bottom.

Examples of statements describing percentile ranks include:

  • Martha's percentile rank of 99.5 on the Basic Math Skills cluster indicates that only five out of 1,000 students would have a score as high or higher (or her score was as high or higher than 995 out of 1,000 people).
  • Angela's percentile rank of 1 on the Basic Writing Skills cluster indicates that only one out of 100 third-grade students would obtain a score as low or lower (or only 1% of grademates scored as low or lower than Angela).
  • On the Broad Mathematics Cluster, Dominick's percentile rank of 0.5 indicates that only five out of 1,000 people would have a score as low or lower.

Standard Score

The standard score scale used in the WJ IV is based on a mean of 100 and a standard deviation of 15. This scale is the same as most deviation-IQ scales and may be used to relate standard scores from the WJ IV to other test scores based on the same mean and standard deviation. The WJ IV standard score range (0 to 200+) provides more discrimination at the ends of the scale. Because standard scores are more difficult for parents and other nonprofessionals to understand, many examiners use the more meaningful and equivalent percentile rank to interpret the standard score.

In writing reports or communicating test results to parents and others, you may prefer to use verbal labels rather than numbers to describe test performance. Rapid Reference 4.3 provides suggested verbal labels to use when describing test results. Exercise care when using descriptors of a disability. Use person-first language when describing an individual with a disability (e.g., “an individual with an intellectual disability” rather than “an intellectually disabled individual”). Use caution and professional judgment in the selection and application of verbal labels to describe a range of scores. Although labels may assist in communicating test results, the terminology is at times ambiguous or the meaning of the labels can be misunderstood.

Examples of statements using standard scores would be:

  • Arnold's low to low average performance on the Math Problem Solving cluster reflects his current difficulty with mathematical concepts and reasoning.
  • When Mr. Begley's performance is compared to students entering graduate school, his Academic Knowledge cluster is competitive (Standard Score = 120).
  • Rhia's standard score of 125 (+/– 1 SEM: 120–130) indicates that her performance on the WJ IV Broad Written Language cluster is in the superior range.

Four other types of standard scores—z-scores, T-scores, stanines, and normal curve equivalents (NCEs)—are available when using the online scoring program. To display any of these scores, add a new score selection template that includes the additional score or scores you wish to report. The additional standard score selected appears for all tests and clusters. Rapid Reference 4.4 shows the means and standard deviations for each type of standard score available. Appendix C of this book presents a table for converting z-scores to percentile ranks or standard scores.

Standard Error of Measurement

Every test score, from any test, has error associated with it. To improve accuracy when describing an individual's scores, the standard error of measurement (SEM) is used to determine the range of scores around the obtained score. This provides an indication of the degree of confidence professionals can have in an obtained score. Rather than using an average SEM, every possible score in the WJ IV has a unique SEM reported. These SEMs are primarily used to build confidence bands for standard scores and percentile ranks that are presented on the score report.

Levels of Interpretive Information

The WJ IV provides four hierarchical levels of information in an interpretive framework. Information from one level cannot be used interchangeably with information from another. Each level provides unique information about a person's test performance and builds on information from the previous level. Consider information from all four levels when describing a person's performance. Rapid Reference 4.5 provides a brief summary of the four levels of information.

Level 1 information is useful in interpreting results and planning the appropriate instructional program. Informal and qualitative in nature, this information is obtained through behavioral observations during testing and through error analysis of responses to individual items. Each test in the Standard Battery (Tests 1–11) has a Qualitative Checklist that can be used to help document the examinee's performance on the task. In addition, a framework for recording behavioral observations, the Test Session Observations Checklist, is located on the first page of the WJ IV ACH Test Record. Level 1 information can include comments that the examinee makes (e.g., “I really don't like math”), as well as behavioral observations (e.g., the student was inattentive during timed tests). It also can help with forming hypotheses about the nature of a problem. For example, Phillip, a fourth-grade student, missed several items on both the Letter-Word Identification and Word Attack tests that involved short vowel sounds. One instructional recommendation was that Phillip should be retaught short vowels and practice reading consonant-vowel-consonant words with these sounds.

Watching how the individual performs the task can also provide insights into instructional recommendations. Two ninth-grade students, John and Rebekkah, solved this problem in different ways, although they both arrived at the correct solution:

Form A, Item 31, Applied Problems: The Roberts have four people in their family. For breakfast they each eat three muffins. If the muffins come in packages of six, how many packages do they need each morning?

John multiplied 4 by 3 and then divided 12 by 6 to get the answer of two packages. Rebekkah drew four people and then put three muffins under each of the people. She then drew a package around two groups of six. Although both students have a conceptual understanding of the problem, Rebekkah's approach is more concrete and reveals that she does not understand how to use multiplication and division to solve such a word problem.

Level 2 information indicates an individual's stage of development and is expressed as age or grade equivalents. You can use these scores to help estimate an appropriate developmental or instructional level.

Level 3 information indicates the quality of a person's performance on criterion tasks of a given difficulty level and can be helpful in determining an appropriate instructional level. The RPI compares the examinee's proficiency on a task to the proficiency of average age- or grademates. The RPI can be used to help determine an appropriate instructional level. The instructional zone, based on the RPI, defines the range of tasks that a person would perceive as quite easy (96% successful) to a level that a person would perceive as quite difficult (75% successful). As noted, this is similar to the independent and frustration levels typically found on informal reading inventories. CALP levels, also determined by the RPI, can be helpful in describing an individual's language proficiency.

Level 4 information provides a basis for making peer comparisons. In educational and clinical settings, percentile ranks and standard scores are the most common metrics used to describe an individual's relative group standing as compared to age or grade peers. These scores are often the scores used in making placement decisions, such as a need for special education.

Interpreting Tests

The interpretation of test results is a complex process. Simply reporting the derived scores or printing out a computer report is not an interpretation. An analysis of the individual's performance on a test must consider the stimulus material, task demands, task complexity, language requirements needed to complete the task, developmental nature of the task, and any factors that may have affected performance. Furthermore, understanding the Cattell-Horn-Carroll Theory of Cognitive Abilities (CHC theory) (Carroll, 1993; Cattell, 1963; Horn, 1988, 1991; Horn & Cattell, 1966; McGrew, 2005, 2009; Schneider & McGrew, 2012; Woodcock, 1990, 1998) can provide guidance in interpreting test results. Rapid Reference 4.6 provides definitions of the various factors in CHC theory.

Interpreting the Reading Tests

Many children find learning to read a formidable task. For about 20% of children, learning to read is the most difficult challenge they face in school (Lyon, 1998; Shaywitz, 2003). Because reading is a major foundational skill for all school-based learning, it is critical that examiners evaluate reading performance and plan appropriate instructional programs. Rapid Reference 4.7 provides a list of some of the common characteristics of individuals with low reading performance.

To interpret the reading tests, you must be aware of the skills involved in each task, know what abilities underlie each test, and recognize any additional factors that may facilitate or inhibit performance, such as attention. You will also want to consider the impact of poor oral language skills when evaluating reading performance as well as the relationships between reading and written language performance.

The skills measured in the eight reading tests range from lower-level (less complex) abilities, such as recognizing letters and sounds in isolation, to the higher-level (more complex) abilities, such as comprehending and recalling connected discourse. Figure 4.2 displays an interpretive model of the skills measured by the WJ IV reading tests. Using the framework of CHC theory to interpret the reading tests shows that the tests are primarily measures of reading ability, an aspect of reading/writing ability (Grw).

img

Figure 4.2 Skills Measured by the WJ IV ACH Reading Tests

Other aspects of processing, particularly auditory processing (Ga), long-term retrieval (Glr), comprehension-knowledge (Gc), cognitive processing speed (Gs), and short-term working memory (Gwm), are also measured by the WJ IV reading tests. McGrew and Wendling (2010) report that Gc, Ga, Gs, and Gwm have a consistent, significant relationship with reading achievement. Gc is strongly related to basic reading skills and reading comprehension across all ages. Phonetic coding, an aspect of auditory processing (Ga), is important to basic reading skills in particular. Short-term working memory (Gwm) appears to contribute to reading achievement. This seems to be particularly true as it relates to working memory (Gwm-WM), which is significant at all ages for both basic reading skills and reading comprehension. Memory span (Gwm-MS) is related to reading comprehension for individuals at the high school level. In addition, naming facility and associative memory, aspects of long-term retrieval (Glr), appear to be important predictors of early reading failure. Fluid reasoning (Gf) is important to comprehension (Nation, Clarke, & Snowling, 2002), but research has not linked Gf to decoding. Perceptual speed, a narrow (Gs) ability, adds significantly to the explanation of reading performance. Some developmental differences are noted. For example, the significance of comprehension-knowledge (Gc) and working memory (Gwm-WM) to reading achievement increases with age, whereas the significance of long-term retrieval (Glr) and processing speed (Gs) declines with age. Visual processing abilities (Gv) have little significance in explaining or predicting reading achievement. However, some visual abilities, such as orthographic coding (the ability to recall letters and letter strings) (Berninger, 1990; Hale & Fiorello, 2004), are important to reading success.

Test 1: Letter-Word Identification

This test requires the individual to read isolated letters and words orally. It is a measure of reading decoding (sight recognition), including reading readiness skills. The items are presented in a list rather than in context. The reader does not need to know the meaning of the words. Individuals with good sight-word recognition skills recognize the letters and words rapidly and with little effort. Automatic and fluent sight-word identification facilitates reading performance.

Individuals with inefficient or limited strategies for word identification typically have low performance on this test. These individuals may identify several words accurately but require increased time and greater attention to phonic analysis to determine the correct response. Word identification skills are not automatic. Individuals with poor word recognition skills tend to read slowly and make numerous errors. They may be unwilling to try, are easily frustrated, or are afraid to risk making an error. In many cases, you will want to evaluate the types of errors that the individual makes during testing. Table 4.2 indicates the task demands and categories of the phonic elements or syllable types for items in Forms A, B, and C. Although many of the words could be placed in multiple categories, each word was assigned to only one category. For example, the word must could be placed in both the short vowel and consonant blend categories. The examples in each category are designed to help you identify other phonic elements or syllable types that may appear in each word. Rapid Reference 4.8 defines the various categories and phonics terminology used in Table 4.2. These terms also apply to categories in Table 4.3 for Word Attack and Table 4.6 for Spelling. Analyzing the types of errors an individual makes on a test or across tests can provide valuable information that helps guide instruction.

Table 4.2 Categories of Phonic Elements for Items in Letter-Word Identification, Forms A, B, and C

Phonic Element/SyllableType Item Form A Item Form B Item Form C
Letter Recognition(pointing response) 1 L 1 T 1 D
2 A 2 S 2 E
3 S 3 B 3 S
4 W 4 C 4 G
5 k 5 i 5 d
6 y 6 r 6 a
Letter Identification(oral response) 7 R 7 X 7 A
8 F 8 s 8 c
9 p 9 z 9 e
10 J 10 T 10 D
Word Recognition(pointing response) 11 car 11 cat 11 dog
12 sun 12 rug 12 mat
13 dog 13 tree 13 see
Word Identification(oral response)
Irregular 14 the 17 will 15 to
and high frequency 22 have 22 was 17 one
23 into 29 only 22 two
38 often 30 does
31 been
38 piece
Short vowel closed syllable (CVC or VC) 15 at 14 in 14 is
16 and 15 dog 16 can
18 man 16 it 18 if
20 cup 18 but 20 hot
21 fish 20 big 29 rabbit
25 them 25 then 35 quick
26 must 27 family 40 million
31 animal 32 children 41 studying
54 veteran 34 different
59 stamina 48 toxic
70 minuend 71 nuptial
Open syllable (ends on vowel with long vowel sound) 17 no 19 play 32 because
19 she 31 grow 44 decide
27 going 43 diamond
34 become 62 diatom
36 library
Long vowel –igh 29 light
Long vowel 61 breathes 57 leagues 19 like
25 time
26 cake
43 island
Consonant -le syllable 28 people 76 inveigle 60 whittles
67 chortle
Vowel team(digraph or diphthong) 24 keep 28 new 24 soon
32 could 33 great 27 each
37 point 37 own 28 about
40 however 41 through 34 town
41 brought 45 sausage 39 build
42 jewel 49 statue 55 seized
53 knead 72 poignant
66 heuristic
R-controlled syllable 30 morning 26 more 50 fierce
33 garden 44 liberty 51 whirled
44 natural 54 portrait
52 guarantee
Soft c or soft g 39 special 40 sentence 37 change
45 distance 42 pledge 61 porcelain
48 imagine 47 social 65 gist
56 ancient 74 taciturn
63 prestige
67 deficiencies
Hard c or hard g 47 signal 39 guess 58 domesticated
77 casuistry
Consonant blend 55 sphere 21 from 57 scowled
24 just 59 justifiable
30 hand
Consonant digraph 35 knew 23 when 21 this
43 whose 46 bachelor 23 they
63 thoroughfare 50 nephew 36 father
69 blithe 48 character
49 phrase
66 labyrinth
Affixes 46 overwhelm 35 faster 33 really
49 investigate 36 beautiful 42 famous
50 reverse 38 building 45 lacked
51 doubtful 51 curious 47 announcer
56 accustomed 52 hesitate 56 courageous
57 contrary 53 tremendous 62 controversial
60 ferocious 55 evidence 63 ravenous
64 staunchest 59 particularly 64 nutritive
65 millinery 61 infectious 69 subsidiary
68 municipality 64 trajectory 71 denotative
72 aggrandizement 68 unscrupulous 73 quadruped
74 tertiary 73 impetuosity 75 ubiquitous
75 septuagenarian 76 perspicacity
French origin 58 cologne 66 opaque 52 moustache
62 silhouette 74 pique 78 frisson
73 milieu 78 bourgeois
76 echelon
77 coiffure
78 macaque
Greek origin 69 idiosyncrasy 60 psychology 53 thermostat
71 rhetoric 70 pterodactyl 68 aesthetic
75 rhetorician 70 euphemism
77 chimerical 72 psoriasis
Qu: /kw/ 58 quarreled 54 quite

You may wish to compare the results of this test with Passage Comprehension, Oral Reading, Sentence Reading Fluency, Reading Recall, Word Reading Fluency, and Reading Vocabulary to develop insights into the individual's level of reading skills with and without the context of meaning. Although all of the tests require word identification, these other tests also require knowledge of word meanings, sentence structure, and comprehension. You may also wish to compare the results of Letter-Word Identification and Word Attack to determine whether differences exist between the reader's word identification and his or her ability to apply phonic skills.

Test 4: Passage Comprehension

This test requires the individual to read a short passage silently, comprehend the information, and provide a missing word. It is a measure of reading comprehension and lexical knowledge. This modified cloze task requires the ability to use both syntactic and semantic clues in comprehending text.

Low performance on Passage Comprehension may be a function of limited basic reading skills, comprehension difficulties, or both. Analysis of the types of errors made will help in determining the most appropriate instructional plan. You may consider three different types of errors: (1) syntactically correct but semantically incorrect, (2) semantically correct but syntactically incorrect, or (3) both incorrect.

To help clarify whether the problem is with reading alone or reflects limited language comprehension, you can compare the results from the Passage Comprehension test to Test 2: Oral Comprehension in the WJ IV OL, a similar task that does not require reading (see Chapter 5 for more information). If the individual does well on the oral test, then language comprehension is not likely to be the reason for poor performance on the reading test. If the individual does poorly on Oral Comprehension, however, then limited language comprehension must also be considered as a contributing factor to poor reading performance.

Test 7: Word Attack

This test requires the individual to orally read phonically regular nonsense words. Word Attack measures aspects of both phonological (using speech sounds to read words) and orthographic (using common letter patterns or strings to read nonsense words) coding. Phoneme–grapheme knowledge is necessary to perform well on this test.

Low performance on Word Attack may result from poor phonological processing, limited phoneme–grapheme knowledge, poor decoding skills and strategies, or a lack of fluency. Impaired decoding is frequently thought to be the basis of reading problems. You may wish to evaluate the types of errors the individual makes on this test in order to make the most appropriate instructional recommendations. Table 4.3 shows the tasks and phonic elements for the items on Word Attack, Forms A, B, and C. Each item appears in just one category in Table 4.3, although many items could be assigned to more than one category. For example, gusp appears in the Consonant Blend category but could also be assigned to the Short Vowel category. Review of the examples in Table 4.3 will help you determine additional phonic elements that may apply to each item. This information, along with error analysis, can be helpful in planning instruction.

Table 4.3 Categories of Phonic Elements for Items in Word Attack, Forms A, B, and C

Phonic Element Item Form A Item Form B Item Form C
Sound Recognition(point to picture that starts with sound specified) 1 /k/ 1 /k/ 1 /k/
2 /f/ 2 /f/ 2 /f/
Letter/Sound Recognition(point to letter or letters thatmake the sound specified) 35 /p//sh/ 35 /r//gr/ 35 /f//sh/
Sound Identification(says sound of letter or letters) 4 /k/ 4 /m/ 4 /k/
6 /t/ 6 /n/ 6 /z/
8 /sp/ 8 /gl/ 8 /ck/
12 /tw/ 12 /kl/
Nonsense Word Identification (reads nonsense word aloud)
Short vowel 7 tiff 7 hap 7 lat
10 ven 8 mell 10 fim
11 wugs 9 lish 11 bix
13 mip 10 rox 14 jop
16 jox 13 gugs 15 zent
23 centizen 15 tisp
27 hudned 27 depnonlel
Long vowel 17 bine 18 cade 22 rotion
19 blighten 19 sluke 25 ligtite
26 cythe 23 frime
Vowel digraph or diphthong 9 zoop 21 comfrain 12 oy
14 foy 26 mafreatsun 17 pawk
21 baunted 18 loast
24 saist
R-controlled 18 artible 24 cirdon 16 snirk
20 yerdle
Consonant blend 29 fleighted 14 floxy 19 thrept
16 gusp 21 trond
17 splist
Consonant digraph 15 leck 20 shomble 9 fash
20 wreet 27 phigh 13 vack
23 phintober 27 paraphonity
Multisyllabic 28 intestationing 29 mivocative 28 subdirement
31 sylibemeter 30 apertuate 29 botrationary
32 armophodelictedness 31 psypenoptimeter 30 redigitation
32 pretrationistic 31 sophomibistry
Qu: /kw/ 22 quade 22 quantric 23 quog
Silent g 25 gnib 24 gnobe
Silent k 26 knoink
Soft g 30 coge 32 mefgest

If an individual has a low score on Word Attack, you will want to evaluate the individual's phonological awareness. You may administer Test 3: Segmentation, Test 7: Sound Blending, and Test 9: Sound Awareness from the WJ IV OL, as well as Test 5: Phonological Processing from the WJ IV COG. An individual who has strong auditory processing (Ga) but does poorly on Word Attack has an excellent prognosis for improvement. The underlying abilities to analyze and synthesize sounds are intact, so systematic instruction in phoneme–grapheme relationships should help improve reading skills.

An individual who has poor auditory processing (Ga) in addition to poor performance on Word Attack would benefit from an instructional program that focuses on both developing phonological awareness and phonic skills. Individuals need a prerequisite amount of phonological awareness to develop decoding skills. Specific training in sound blending and phonemic segmentation, in particular, improves decoding skills (Ehri, 2006; National Reading Panel, 2000). Focused instruction can help an individual increase his or her understanding of the relationship between the sounds (phonemes) and the written symbols (graphemes).

Test 8: Oral Reading

This test requires oral reading of a set of sentences that gradually increases in complexity. The test measures decoding skill, automaticity with reading, and prosody (reading with expression). Within the Test Record, you can use the Qualitative Observation Tally to code the type of errors an individual makes while reading the sentences; this may lead to specific instructional implications. For example, a fourth-grade student, Roberto, made numerous hesitations and repetitions while reading the passages. This suggests that Roberto has a problem with fluency and automaticity, so a method to address these issues would be appropriate.

Test 9: Sentence Reading Fluency

This test requires the individual to read simple sentences quickly and indicate whether the statement is true or false by circling Yes or No. It is a measure of reading speed and automaticity. Low performance on Sentence Reading Fluency may be a result of difficulty sustaining attention, limited basic reading skills, slow processing speed, or comprehension difficulties. An individual's processing speed (Gs) (see Cognitive Processing Speed Cluster from WJ IV COG) may facilitate or inhibit performance on this test. The speed and fluency with which an individual performs basic skills can influence performance on higher-level skills. You may wish to compare Sentence Reading Fluency with Test 15: Word Reading Fluency to see if performance is similar on both timed tests. If Sentence Reading Fluency is higher than Word Reading Fluency, this suggests that increased context improves word recognition skill.

Test 12: Reading Recall

This test requires the individual to read a story and then recall the elements of that story. Both reading and expressive language skills are required to perform this story-retelling task. Reading Recall measures reading comprehension, meaningful memory, and language development. Poor attention, poor memory, poor decoding, limited vocabulary, low expressive language, or limited comprehension may negatively impact performance on this test. You may compare performance on the Reading Recall test to Test 6: Story Recall on the WJ IV COG. If performance is higher on Story Recall than Reading Recall, this suggests that the problem is not with memory or attention, but rather with reading skills. Conversely, if Reading Recall is higher than Story Recall, this suggests that reading comprehension is stronger than listening comprehension.

Test 15: Word Reading Fluency

This test is a timed test that requires the examinee to silently and quickly read four words and mark the two that share a semantic relationship. Low performance on this test may result from poor basic reading skills, limited vocabulary, or slow processing speed. An individual's processing speed (Gs) (see Cognitive Processing Speed Cluster from the WJ IV COG) may facilitate or inhibit performance on this test.

As noted previously, you may wish to compare performance on this test to Test 9: Sentence Reading Fluency since both are timed reading tests. If performance is higher on Test 15: Word Reading Fluency, this suggests that the increased context of language in Sentence Reading Fluency does not help the individual. You also may wish to compare the results of Word Reading Fluency to tests that require vocabulary knowledge such as Test 17: Reading Vocabulary, or Test 1: Picture Vocabulary from the WJ IV OL, or Test 1: Oral Vocabulary from the WJ IV COG, all of which are untimed tests. This comparison may help determine whether speed or vocabulary knowledge is influencing performance.

Test 17: Reading Vocabulary

This test has two parts: 17A: Synonyms and 17B: Antonyms. The individual is required to read words and to orally supply synonyms in 17A and antonyms in 17B. Low performance on this test may result from poor basic reading skills, limited vocabulary, or both. If the individual reads the stimulus words correctly but provides an incorrect response, he or she may be better at decoding than comprehending. If the individual misreads the stimulus words but provides the correct response, he or she may be better at comprehending than decoding. For example, a ninth-grade student read the word receive as recover and then said that retrieve was a synonym. Although this response would be scored as incorrect, the examiner noted that it was a decoding, not a vocabulary or comprehension, error.

You may wish to compare Reading Vocabulary directly to the individual's performance on Test 1: Oral Vocabulary in the WJ IV COG, a similar task that does not require reading. If the individual's performance is higher on the oral task than on the reading task, then the focus of instruction should be on developing basic reading skills. If the individual's performance is low on both the oral task and the reading task, then the focus of instruction should be on developing oral vocabulary, as well as basic reading skills.

Reading Clusters

The WJ IV ACH has seven reading clusters: Reading, Broad Reading, Basic Reading Skills, Reading Comprehension, Reading Comprehension-Extended, Reading Fluency, and Reading Rate. You can use each of these clusters when calculating intra-achievement variations or ability/achievement comparisons. Rapid Reference 4.9 illustrates typical instructional implications for individuals with reading difficulties.

Reading Cluster

The Reading cluster consists of two tests: Letter-Word Identification, a measure of word recognition skill, and Passage Comprehension, a measure of reading comprehension. Because it includes a measure of basic skills and comprehension, this cluster may be used when you want an estimate of a person's general reading ability. The cluster does not include a timed test, so it provides a measure of reading accuracy rather than reading speed.

Broad Reading Cluster

Composed of three tests, Letter-Word Identification, Passage Comprehension, and Sentence Reading Fluency, this cluster provides a broad overview of the individual's overall reading level. Because it is a mix of three different aspects of reading (basic skills, comprehension, and fluency), interpretation of this cluster is most meaningful when performance is similar on all three tests.

Basic Reading Skills Cluster

For this cluster, Letter-Word Identification and Word Attack provide a broad view of the individual's basic word reading skills, including sight-word recognition and phonic skills. Comparing the results of the two tests will help you determine if word recognition skills, phonic skills, or both are limited and require remediation. Analysis of errors made on both tests can help you target specific instructional elements.

If an individual has trouble decoding words, few resources are left for comprehension. Slow, labored reading with many errors will have a negative impact on comprehension. Analyzing an individual's performance on the basic reading skill tests is important to the interpretation of the reading comprehension tests. Be sure to consider the effect of decoding problems before identifying a problem in reading comprehension.

Reading Comprehension Cluster

The Reading Comprehension cluster is composed of Passage Comprehension and Reading Recall. This cluster provides a broad view of the individual's reading comprehension skill. Both tests measure comprehension in the context of connected discourse. Reading Recall also has a meaningful memory component that requires reconstruction of the passage that was read. Because of the retelling aspect of Reading Recall, expressive language demands are increased and must be considered when interpreting results.

The Reading Comprehension-Extended cluster also includes Passage Comprehension and Reading Recall, but adds a third test, Reading Vocabulary. Reading Vocabulary measures comprehension in a decontextualized format, that is, the knowledge of word meanings in isolation. Comparing the results of the three tests in this cluster will help determine whether a meaningful context helps or interferes with an individual's comprehension.

You may want to compare the results of the Reading Comprehension cluster to the Basic Reading Skills cluster, the Academic Knowledge cluster, and the Listening Comprehension cluster available in the WJ IV OL. Low performance on reading comprehension tasks may result from low basic reading skills or limited oral language or background knowledge. Considering the impact of these various factors will help you determine the most appropriate instructional program for the individual.

Reading Fluency

The Reading Fluency cluster contains two tests: Oral Reading and Sentence Reading Fluency. This cluster includes a measure of untimed oral reading and a measure of silent reading that is timed. Although the Oral Reading test is not timed, you can evaluate a person's ease of reading, accuracy, and expression (prosody) as he or she reads the set of sentences aloud. In addition, you can document the number and type of errors the person makes while reading. You can record this information in the Qualitative Observation Tally located in the Test Record and then use the results to help inform instructional planning. Sentence Reading Fluency is a timed silent reading task that requires comprehending simple sentences quickly. When comparing the individual's performance on the two tests in this cluster, consider how the differences in task and response demands may have influenced performance.

Reading Rate

The Reading Rate cluster consists of Sentence Reading Fluency and Word Reading Fluency. Both of these tests are timed and read silently. Sentence Reading Fluency requires comprehension of simple sentences, whereas Word Reading Fluency involves the understanding of vocabulary and the semantic relationships among pairs of words. Results on this cluster can be compared to Basic Reading Skills and Reading Comprehension. In addition, you may want to investigate how the person performs on other timed measures (e.g., Math Facts Fluency and Sentence Writing Fluency), as well as on measures of processing and perceptual speed on the WJ IV COG. In some cases, the results of this cluster, when substantiated with additional information, can be used to help document a need for extended time on tests.

Sheila was applying for graduate school. Although she had been diagnosed with dyslexia in third grade, she had never received special education services at school. She was concerned that she would need certain accommodations on tests, such as extended time, if she was going to be successful. In the past, she remembered never being able to complete exams even though she could have answered more questions if she had been provided with more time. The WJ IV ACH assessment results indicated superior Academic Knowledge, but very low scores on Word Attack, Spelling, Sentence Reading Fluency, and Word Reading Fluency. Additional scores on the WJ IV COG and OL helped establish the rationale for Sheila's need for more time (e.g., slow perceptual speed but above average oral language and reasoning abilities).

Interpreting the Mathematics Tests

Interpretation of the mathematics tests requires that you are aware of the skills involved in each task, know what abilities underlie each test, and recognize additional factors that may influence performance. Rapid Reference 4.10 indicates common characteristics of individuals with low math achievement. The four mathematics tests measure skills that range from lower-level abilities, such as writing numbers, to higher-level abilities, such as analyzing and solving problems. Figure 4.3 provides an interpretive model of the various skills measured by the WJ IV mathematics tests.

img

Figure 4.3 Various Skills Measured by the WJ IV ACH Mathematics Tests

Using the CHC theory as the interpretive framework, the mathematics tests primarily measure quantitative knowledge (Gq). The WJ IV math tests also measure aspects of fluid reasoning (Gf), comprehension-knowledge (Gc), cognitive processing speed (Gs), short-term working memory (Gwm), auditory processing (Ga), and visual processing (Gv). McGrew and Wendling (2010) found that Gf, Gc, and Gs abilities, as well as some narrow abilities, were correlated consistently and significantly with basic math skills and math reasoning. Some developmental differences were also noted. For example, the Gc relationship to math reasoning increases with age, whereas Gf was related consistently and significantly across all ages. The narrow abilities of perceptual speed (Gs-P) and working memory (Gwm-WM) were also significantly related to both basic skills and math reasoning across all ages. The narrow Ga ability of phonetic coding (Ga-PC) correlated with mathematics, especially at the younger ages. Gv appears to be related to math tasks that require higher-level skills and thinking, but is not related to basic math skills (Flanagan, Ortiz, & Alfonso, 2013).

Test 2: Applied Problems

This test requires the individual to analyze and solve practical math problems. It is a measure of quantitative reasoning, math achievement, and math knowledge. Because no reading is required, low performance will most likely be related to limits in mathematical knowledge (Gq). Low performance may result from poor attention, limited fluid reasoning (Gf), limited math skills, or oral language comprehension difficulties (Gc).

Compare results on this test to reading and writing clusters to determine whether the individual does better when no reading is required. Oral language comprehension, limited vocabulary (Gc), working memory (Gwm-WM), perceptual speed (Gs-P), and fluid reasoning (Gf) can impact performance on Applied Problems. Analyzing the individual's errors may help provide ideas for instructional planning. Table 4.4 indicates the categories for the problems in Forms A, B, and C.

Table 4.4 Categories of Problems on Test 2: Applied Problems, Forms A, B, and C

Category Form A Items Form B Items Form C Items
Number Concepts 1–10 1–8, 10 1–12
Subtraction 11–14, 21–23, 26, 29–30 9, 11, 13, 15–17, 26, 30, 37 13–14, 16, 20, 22, 24–25, 29
Addition 15, 17–18, 33 12, 14, 19 15, 17, 19, 21
Money 22–23, 28 20, 23, 28, 32 27–28
Division 24, 31, 35 24, 31, 33, 40, 42 23, 31
Multiplication 25, 32 27, 29, 34 26, 32, 34, 38
Fractions 27, 42 35, 38, 41 21
Time 34, 37 18, 25, 36, 39, 44 30, 35–37
Probability 36, 39, 46–47 51, 56 55
Algebra 38, 43–44, 49–50 48, 50, 54–55 33, 43, 48–49, 53–54, 56
Percentages 40–41 49 40, 42, 46
Interest 45 46 44, 52
Geometry 48, 52–56 43, 45, 52–53 41, 47, 50–51
Measurement 33, 42
Averaging 47 39, 45

Test 5: Calculation

This test requires the individual to perform a variety of calculations ranging from simple addition and subtraction to complex calculus. Tasks progress from: (a) basic addition, subtraction, multiplication, and division; (b) advanced calculations of each operation with regrouping; (c) advanced calculation of each operation with negative numbers (except division); (d) fractions; (e) percentages; (f) algebra; (g) trigonometry; (h) logarithms; and (i) calculus. Calculation measures the ability to perform mathematical computations that are fundamental to more complex math reasoning and problem solving. Fluency with basic math skills is fundamental to more complex math. Low performance may result from limited basic math skill, weaknesses in short-term working memory (Gwm-WM), limited fluency or automaticity with math facts (Gs-P), poor or limited instruction, or difficulty with attention.

Be sure to observe the examinee's behaviors; categorize errors that are made; note which concepts are known and unknown; and interview the examinee, if needed. For example, Anna, a sixth-grade girl, made numerous errors in Calculation. To try to understand the reasons for her errors, the examiner presented Anna with a set of similar problems and asked her to talk through what she was doing as she performed the calculations. This helped the examiner pinpoint Anna's confusions regarding math algorithms. Bart, an eighth-grade student, made errors on all problems involving fractions. For example, in adding fractions, he would combine the numerators and denominators (e.g., ½ + ¼ = img). Observations of these types of errors lead to specific instructional recommendations. Bart needs explicit instruction in how to add, subtract, multiply, and divide with numbers that have fractions. Table 4.5 indicates the number of items in each category for Forms A, B, and C.

Table 4.5 Categories of Problems on Test 5: Calculation, Forms A, B, and C

Category Form A (Number of items) Form B (Number of items) Form C (Number of items)
Basic addition 12 10 10
Basic subtraction 6 8 8
Basic multiplication 4 4 4
Basic division 2 3 2
Advanced addition 3 3 4
Advanced subtraction 1 1 2
Advanced multiplication 3 3 4
Advanced division 2 2 2
Algebra 5 4 4
Derivatives 2 2 2
Factorials 1 1 1
Fractions 4 4 4
Geometry 1 1 1
Integrals 2 3 3
Logarithms 1 1 1
Matrices and determinants 1 1 1
Percentages 2 1 1
Powers and roots 2 2 1
Trigonometry 2 2 1

Test 10: Math Facts Fluency

This test requires the individual to quickly solve simple addition, subtraction, and multiplication facts. Low performance on this test may result from poor attention, limited basic math skills, lack of automaticity, or slow processing speed (Gs-P). Some individuals work slowly and make no errors, whereas other individuals work very quickly and make several errors. If numerous errors are made, attempt to determine the reasons for the mistakes. As examples, a student may be confused or not pay careful attention to the signs, may not know multiplication facts, or may be confused about the properties of zero (e.g., 6 + 0 = 0).

Test 13: Number Matrices

The task in Number Matrices requires the individual to look at a matrix of numbers, figure out the pattern, and then provide the missing number. It is a measure of quantitative reasoning (RQ), an aspect of fluid reasoning (Gf), and also requires working memory (Gwm-WM) and perceptual speed (Gs-P).

Low performance on this test will most likely result from limited fluid reasoning (Gf), especially the narrow ability of quantitative reasoning (RQ). In addition, limited attention, working memory, and perceptual speed may impact performance.

Math Clusters

The WJ IV ACH has four math clusters: Mathematics, Broad Mathematics, Math Calculation Skills, and Math Problem Solving. Each of these clusters may be used in the intra-achievement variation procedure or the ability/achievement comparisons. Rapid Reference 4.11 lists possible instructional implications for individuals with low math achievement.

Mathematics

The Mathematics cluster consists of two tests: Applied Problems and Calculation. Because it includes a measure of basic math skills and a measure of math reasoning, this cluster may be used to estimate a person's general mathematics ability. The cluster does not include a timed test, so it provides a measure of math knowledge, rather than automaticity with math facts.

Broad Mathematics

This cluster is composed of three tests: Applied Problems, Calculation, and Math Facts Fluency. It provides a broad, comprehensive view of the individual's math achievement level. The Broad Mathematics cluster measures computational skill, automaticity with math facts, and problem solving and reasoning. Because this cluster measures three different aspects of math ability, interpretation of the cluster is most accurate when performance is similar on all three of the tests.

Math Calculation Skills

This cluster is composed of two tests: Calculation and Math Facts Fluency. It provides a measure of basic math skills including computational skills and automaticity with basic math facts. Relative ease with computations is an important factor in predicting math performance. You may wish to compare the results of the two tests in this cluster to help determine whether processing speed (Gs) or mastery of basic math facts is impacting performance.

Math Problem Solving

This cluster is composed of two tests: Applied Problems and Number Matrices. It provides a measure of mathematical knowledge and reasoning. You can compare results of this cluster to the Math Calculation Skills cluster to help determine whether basic math skills are impacting performance. In addition, you can compare results of this cluster to oral language if using the WJ IV OL and to comprehension-knowledge (Gc) and fluid reasoning (Gf) clusters if using the WJ IV COG. Individuals with low oral language may have difficulty with quantitative terminology or math vocabulary. Individuals with low comprehension-knowledge may lack prerequisite knowledge for acquiring and identifying mathematical concepts. Individuals with low fluid reasoning may have difficulty identifying and thinking through the various steps of a mathematical problem.

Interpreting the Written Language Tests

Interpretation of the written language tests requires awareness of the skills involved in each task, what abilities underlie each test, and the additional factors that may be affecting writing performance. Rapid Reference 4.12 lists common characteristics of individuals with low written language performance.

The skills measured in the four written language tests range from lower-level abilities, such as copying shapes, to the higher-level abilities, such as expressing ideas in writing. The WJ IV ACH measures several aspects of writing skill: spelling, usage and proofreading, writing fluency, and the quality of written expression. Figure 4.4 provides an interpretive model of the various skills measured by the WJ IV written language tests.

img

Figure 4.4 Various Skills Measured by the WJ IV ACH Written Language Tests

Using CHC theory as the interpretive framework, the written language tests primarily measure writing ability, an aspect of reading/writing ability (Grw). The WJ IV written language tests also measure aspects of comprehension-knowledge (Gc), cognitive processing speed (Gs), short-term working memory (Gwm), fluid reasoning (Gf), long-term retrieval (Glr), and auditory processing (Ga) (Flanagan et al., 2013; Floyd, McGrew, & Evans, 2008).

Test 3: Spelling

This test requires the individual to produce, in writing, single letters or words in response to oral prompts. Several factors that may influence performance include: fine-motor skill, handwriting, phonological coding, and orthographic coding. This test measures prewriting skills and spelling. A careful analysis of errors can often result in specific instructional recommendations. For example, an examiner noted that Caleb, a third-grade student, was having trouble spelling medial short vowel sounds and was also reversing certain letters (e.g., b and d). The examiner provided specific recommendations for spelling instruction in consonant-vowel-consonant words, as well as a strategy to use to help eliminate b–d reversals.

Table 4.6 indicates phonic elements in the items for Forms A, B, and C. Many words could appear in more than one category. For example, the word comb could appear in the long vowel and silent b categories. However, each word was assigned to only one category. Review of the examples in Table 4.6 will help you determine additional phonic elements that may be in each word. This information, along with error analysis, can help you make recommendations for planning instruction. (See Rapid Reference 4.8 for the definitions of the categories and phonics terminology in Table 4.6.)

Table 4.6 Categories of Phonic Elements for Items in Spelling, Forms A, B, and C

Skill Item Form A Item Form B Item Form C
Prewriting(copy after demonstration) 1 horizontal line 1 box 1 horizontal line
2 triangle 2 vertical line 1 circle
(copy from model) 3 H 3 E 3 S
Print letters 4 A or a 4 M or m 4 O or o
5 T or t 5 B or b 5 C or c
6 P or p 6 X or x 6 U or u
7 W or w 7 R or r 7 E or e
8 L 8 F 8 B
9 i 9 e 9 i
Spelling(write dictated words)
Short vowel 10 is 10 hat 13 had
11 fun 16 under 14 ten
12 got 27 gallon 19 funny
13 am 30 second
14 with
Consonant blend 15 from 14 jump
16 camp 36 crystal
Consonant digraph 26 laugh 21 who 17 mother
31 elephant 23 shoe
38 whisper 31 laughing
59 pharaoh 50 zephyr
Vowel digraph or diphthong 17 saw 11 book 15 bee
25 because 19 yellow 18 house
26 already 24 friend 25 early
17 juice 25 young 32 mountain
36 calorie 29 unfair 40 tomorrow
44 leisure 46 crouton 42 coax
52 nuisance 52 maneuver
Long vowels 19 nice 12 she 10 he
23 fight 13 green 12 my
34 clothes 18 cake 21 place
42 cocoa 30 remember 26 before
38 concrete
R-controlled vowels 31 important 26 tore 22 first
43 calendar 49 separate 44 squirrel
55 questionnaire
Irregular 18 water 15 are 11 the
21 once 17 they 16 was
23 any 20 said
42 yacht 24 were
37 beautiful
Affixes 20 cooked 39 actually 34 international
24 walked 41 profession 49 anonymous
30 vacation 51 repetitious 54 conscientious
37 subscription 43 carriage 53 exacerbate
32 manager 47 sufficient 41 advertisement
33 electric 56 vacillate 45 necessary
41 enthusiastic 47 arrogance
35 reliable 52 acquaintance
39 skiing
51 treacherous
Silent b 29 comb 36 doubt
Silent w 35 sword
Silent g 34 design
45 foreign 58 impugn
Qu: /kw/ 57 soliloquy 29 question
48 exquisite
French origin 48 bouquet 48 bureau 56 camouflage
50 lacquer 54 rendezvous
57 crevasse 55 parquet
59 bouillon 58 liaison
Greek origin 40 gymnasium 60 hemorrhage 51 dilemma
55 chimerical
Words ending in consonant -le 47 tentacle 20 table 27 handle
22 people 39 bicycle
Soft c or g 45 cyst 53 vengeance 35 general
49 exaggerate 43 congenial
Hard c or g 38 league 44 dialogue 33 garage
46 negotiate
-ough pattern 28 bought
33 cough
-eigh pattern 38 neighbor
-sc pattern 40 scenery 57 iridescent
59 scintillant
Multisyllabic 53 vituperative 32 adventure 30 popular
54 omniscient 50 anthropomorphic 60 lachrymose
56 sergeant
58 lackadaisical
60 camaraderie

Test 6: Writing Samples

This test requires the individual to produce meaningful written sentences in response to a variety of tasks. Low performance may result from limits in: (a) oral language, (b) vocabulary, (c) organizational ability, (d) knowledge, or (e) spelling skill. The individual's attitude toward writing may also influence performance. Although errors in spelling and usage are not penalized on this test (unless the response is illegible), you can gain additional qualitative information through a careful analysis of the individual's responses. For example, you may note that the individual does not: (a) start sentences with capital letters, (b) use ending punctuation, (c) spell words accurately, or (d) write using complete sentences. Or you may observe that the individual makes errors in spelling and usage, but writes descriptive, interesting sentences that address the task demands. You may also wish to compare Writing Samples to measures of oral language to see if writing ability is commensurate with receptive and expressive language.

Test 11: Sentence Writing Fluency

This test requires the individual to produce, in writing, legible, simple sentences with acceptable English syntax. Low performance on this test may result from poor attention, poor motor control and handwriting, limited spelling or reading skills, slow cognitive processing speed (Gs), or a response style that interferes with performance (slow and accurate, fast and inaccurate, slow and inaccurate, etc.). You may wish to compare the results on this test to other timed measures to see whether the person generally works slowly or has a particular difficulty with writing speed.

Test 14: Editing

This test requires the ability to identify and correct errors in punctuation, capitalization, spelling, and usage in short written passages read by the examinee. Low performance may result from limited instruction, lack of knowledge of writing conventions, failure to self-monitor or self-correct errors, or poor reading skill.

Test 16: Spelling of Sounds

This test requires the individual to spell nonsense words that conform to conventional phonics and spelling rules. Both phonological coding (Ga) and orthographic coding are measured by this test. Low performance may result from poor attention, poor phonological processing, poor orthographic awareness, or low phoneme–grapheme knowledge. You can compare results from this test to the Word Attack test or to tests from the WJ IV OL (Segmentation, Sound Blending, and Sound Awareness) to help determine whether difficulties result from phonological problems or limited knowledge of phoneme–grapheme relationships.

An analysis of errors can be particularly valuable on this test. Brady, a sixth-grade student, spelled the nonsense words exactly like they sound (e.g., quib as kwib and scritch as skrich). Although this indicates good skill with phonemic segmentation, it indicates a lack of sensitivity to common English spelling patterns. Brady does not need practice sequencing sounds, but rather needs to be taught common English spelling patterns, such as -tch.

Written Language Clusters

The WJ IV ACH has four written language clusters: Written Language, Broad Written Language, Basic Writing Skills, and Written Expression. All of these clusters may be used when calculating intra-achievement variations and ability/achievement discrepancies. Rapid Reference 4.13 lists several possible instructional recommendations for individuals with low written language performance.

Written Language

This cluster contains two tests: Spelling and Writing Samples. Because it includes a measure of basic writing skills and a measure of written expression, this cluster may be used to estimate a person's general writing ability. This cluster does not include a timed test, so it provides a measure of written language rather than automaticity and ease of writing.

Broad Written Language

This cluster includes three tests: Spelling, Writing Samples, and Sentence Writing Fluency. It provides a broad, comprehensive view of the individual's written language achievement. Task demands include spelling single-word responses, expressing ideas to various tasks, and writing simple sentences quickly. You may want to compare the results from this cluster to the results on the Broad Reading cluster and to the Oral Language cluster from the WJ IV OL to help determine the impact of oral language and/or reading skills on written language performance.

Basic Writing Skills

This cluster comprises two tests: Spelling and Editing. It provides a measure of basic writing skills in isolated and context-based formats. Task demands include writing letters, spelling single words, and identifying and correcting errors in spelling, usage, capitalization, and punctuation. Because Editing requires the examinee to read the items and then identify and correct the errors, you will want to consider the impact of the examinee's reading ability on performance. You may also wish to compare the results of this cluster with the Written Expression cluster to determine whether basic writing skills are affecting performance. Mastery of basic skills is a fundamental part of complex meaningful written expression. Error analysis of the two tests in this cluster may help identify the individual's developmental level in spelling.

Written Expression

This cluster includes two tests: Writing Samples and Sentence Writing Fluency. It provides a measure of meaningful written expression and automaticity with writing. You may compare the results of this cluster to the results on the Reading Comprehension and Basic Writing Skills clusters, or the Oral Expression cluster from the WJ IV OL. Low performance may result from low oral language or limited basic writing skills.

Interpreting the Cross-Domain Clusters

Cross-domain clusters represent a mix of reading, writing, and math tasks grouped in the areas of academic skills, academic fluency, or academic applications. Additional clusters include Brief Achievement, Broad Achievement, Academic Knowledge, and Phoneme–Grapheme Knowledge.

Academic Skills Cluster

This cross-academic cluster is composed of the three basic skill tests: Letter-Word Identification, Calculation, and Spelling. It provides a general, basic skills achievement level and can help determine whether the individual's level of basic skills is similar or variable across the three academic areas. Consider whether basic skills facilitate or inhibit the individual's performance. Low performance may suggest particular curricular adaptations, such as use of books on CD during reading time or a calculator during math activities. Students with low performance in one or more areas often require direct, explicit instruction to improve the accuracy of skills.

Academic Fluency Cluster

This cross-domain cluster is composed of the three fluency tests: Sentence Reading Fluency, Math Facts Fluency, and Sentence Writing Fluency. It provides a general academic fluency level. This cluster can help determine whether the individual's level of automaticity with basic skills is facilitating or inhibiting academic performance. Note whether the individual's speed of performance is similar or variable across the three academic areas. For example, a person may work slowly on reading and writing tasks, but at an average rate on measures of mathematics. In many cases, low performance on this cluster when contrasted with measures of (Gc), such as the Academic Knowledge cluster, may suggest a need for extended time or shortened assignments. Additional information can be obtained by comparing the Academic Fluency cluster to the Perceptual Speed (Gs-P) and the Cognitive Processing Speed (Gs) clusters in the WJ IV COG. This may help determine whether speed is generally slow or performance on one or more academic tasks is slowed because of low basic skills.

Academic Applications Cluster

This cross-domain cluster is composed of the three application tests: Passage Comprehension, Applied Problems, and Writing Samples. It provides a general measure of an individual's ability to reason and apply academic knowledge. As with the other cross-domain clusters, note whether the individual's performance on the tests in this cluster is similar or variable. Also, consider the impact of basic skills, fluency, and oral language proficiency when interpreting this cluster. Low performance may suggest a need for adjusting the difficulty level of the instructional materials. Students with low Academic Applications often need adjustment in the difficulty levels of instructional tasks.

Academic Knowledge

Academic Knowledge contains three tests: Science, Social Studies, and Humanities. These tests require the individual to respond orally to questions. Early items require the examinee only to point to a response. This test measures acquired curricular knowledge, an aspect of comprehension-knowledge (Gc). A broad sample of the individual's range of scientific knowledge, general knowledge, cultural knowledge, and geographic knowledge is measured by the Academic Knowledge cluster. Some individuals will have high scores in one area, but not another. Todd, a ninth-grade student, knew a lot about science, but little about humanities. Nicole, a college sophomore and English major, knew a lot about humanities, but recalled little about science.

Low performance may result from limited vocabulary, limited exposure to curricular areas, or both. This cluster can provide valuable insights into the interests of the individual as well as the level of crystallized intelligence. No reading is required. Therefore, the results can help determine whether the individual's knowledge base is impacting performance in the other academic areas. If the WJ IV COG has been administered, results from this test can be compared to the comprehension-knowledge (Gc) cluster. If the WJ IV OL has been administered, results can be compared to the Oral Language and Listening Comprehension clusters. In addition, within the WJ IV ACH, this cluster can be used as the ability score in an ability/achievement comparison procedure to help determine whether reading, writing, and/or mathematics are in line with the person's content knowledge.

Phoneme–Grapheme Knowledge

This cluster is a combination of two tests: Word Attack and Spelling of Sounds. It may be used to evaluate the individual's proficiency using phonics for reading and spelling. The tasks require the individual to decode (read) and encode (spell) pseudo words or nonsense words. Although they have no meaning, these words all use possible English spelling patterns (orthography). Low performance may result from poor phonological skills or limited orthographic skills, or both. A low score may also be attributable to poor or limited instruction. To help understand the reasons for an individual's low performance, compare this cluster to the auditory processing (Ga) cluster in the WJ IV COG or the Phonetic Coding cluster in the WJ IV OL. The Phoneme–Grapheme cluster can be especially useful in documenting specific reading disabilities (dyslexia).

Brief Achievement Cluster

The Brief Achievement cluster is a combination of three tests: Letter-Word Identification, Applied Problems, and Spelling. This cluster provides a quick screening of the person's levels of performance in reading, writing, and math.

Broad Achievement Cluster

This cluster is composed of nine tests in the Standard Battery. These nine tests are used to create the Broad Reading, Broad Mathematics, and Broad Written Language clusters. The purpose of the Broad Achievement cluster is to provide a general level of academic proficiency. The cluster can be used to identify students with very limited or advanced performance levels across curricular areas. It is also helpful when you need a global view of the individual's overall performance across the various achievement domains. The tests required for obtain-ing the Broad Achievement cluster also yield the four additional cross-domain clusters (Skills, Fluency, Applications, and Brief Achievement) previously described.

Interpreting Instructional Zones

The instructional zone, a special application of the RPI, identifies an individual's present level of functioning from easy (the independent level) to difficult (the frustration level). The instructional zone extends from an RPI of 96/90 to an RPI of 75/90. An individual will perceive tasks that fall at an RPI of 96/90 as “easy” (EASY), whereas tasks that fall at an RPI of 75/90 will be “difficult” (DIFF). When using the online scoring program, select the Age/Grade Profile report to get a graphic representation of the examinee's instructional or developmental zone, ranging from the independent level to the frustration level. The grade equivalent (GE) printed below the left end of the shaded band represents the easy or independent level. The GE printed below the right end of the shaded band represents the difficult or frustration level. The obtained GE represents the midpoint of the zone and provides an estimate of instructional level. If you choose age norms, the developmental zone is represented with age equivalents (AEs). The concept of the RPI is similar to that found in informal reading inventories, in which instructional reading levels are identified that range from the independent (easy) level to the frustration (difficult) level.

Vygotsky's (1978) zone of proximal development is a similar concept to the instructional zone. This learning zone represents the distance between an individual's actual developmental level and the level of potential development when assisted by a more knowledgeable other. The concept of developmental ranges is helpful in planning instruction to support growth. Ideally, within a school setting, both homework and seatwork would be at the independent level, whereas work that is supported by others (teachers, peers, parents) would be at the instructional level.

Examples of the types of statements used to describe the instructional zone include:

  • Arnold's instructional zone on the Basic Writing Skills cluster indicates that spelling and editing tasks will be easy at the beginning second-grade level but difficult at the beginning third-grade level.
  • Appropriate instructional materials for June in basic math skills would range from beginning fifth-grade level (easy) to end of sixth-grade level (difficult).
  • Jared's instructional zone on the Math Problem Solving cluster indicates that he will find math reasoning tasks at the 9th-grade level to be easy, whereas tasks at the 11th-grade level will be very difficult.

The instructional zone also depicts the rate of growth in a particular ability at a particular point in time. A narrow range means that growth is rapid during the stage of development, whereas a wide range means that growth is slow. For example, the range on the Word Attack test for a ninth-grade student is wide because the student's ability to pronounce pseudo words (phonically regular nonsense words) is not changing much during this time period. In contrast, the range on a test such as Letter-Word Identification is narrow because the ability to decode increases at a steady rate across the school years.

Table 4.7 illustrates the instructional zones for several achievement clusters for Toby, a student near the end of fifth grade (5.8). First, note the grade equivalents in the column labeled GE. These are the obtained grade equivalents on the task and represent the midpoint of the instructional zone. Next, note the EASY to DIFF columns, which represent the instructional zone ranging from the independent to frustration levels for Toby. A review of this information helps us understand Toby's instructional needs. We can see that reading is problematic whereas math is not. Every instructional zone for a reading cluster is below Toby's grade placement of grade 5.8. His frustration level on the reading clusters is in the third- to fourth-grade range with the exception of reading fluency, which is at the second-grade level. The extreme limits in reading fluency explain why the instructional zone for Broad Reading, which includes fluency, is lower than the instructional zone for Reading, which does not include fluency. Clearly, Toby will have little success on grade-level reading tasks. To ensure success and classroom participation, the level of instruction will need to be adapted for Toby, or he will need the textbooks to be recorded.

Table 4.7 Instructional Zones for Several Achievement Clusters

Cluster/Test W GE EASY to DIFF
Reading 485 3.3 2.6 4.3
Broad Reading 466 2.4 2.0 3.0
Basic Reading Skills 483 3.0 2.3 4.0
Reading Fluency 450 1.8 1.4 2.1
Broad Mathematics 539 13.4 10.5 >17.9
Math Calc Skills 540 13.0 9.1 >17.9

Interpreting Standard Score or Percentile Rank Ranges

Viewing a range of scores is often preferred to interpreting a single score because every score includes some error. To account for this error, confidence bands are used. For example, at the 68% level of confidence, the standard score or percentile rank range portrays the ±1 SEM confidence band. This band, extending from a point 1 SEM below the individual's obtained score to a point 1 SEM above the individual's obtained score, is the area in which the true score falls two out of three times. If a higher degree of confidence is desired, the online scoring program provides an option to select a 90% or 95% level of confidence. A 95% confidence band (±2 SEM) represents the region in which an individual's true ability would fall 19 times out of 20.

Table 4.8 illustrates the standard score confidence bands at the 68% level of confidence for several achievement clusters. Although most of the ranges for the clusters shown are three to four standard scores above and below the obtained score, Reading Fluency has a larger range of scores, meaning there is more error associated with the score. Reporting the range of scores increases confidence that the individual's true score is represented. Select the Standard Score/Percentile Rank Profile report when using the online scoring program to obtain a graphic representation of this information.

Table 4.8 Standard Score Confidence Bands (68% level) for Several Achievement Clusters

Cluster/Test SS (68% Band)
Reading 83 (80–86)
Broad Reading 70 (66–73)
Basic Reading Skills 82 (79–85)
Reading Fluency 64 (58–69)
Broad Mathematics 133 (129–137)
Math Calc Skills 128 (124–132)

You can use the SS confidence bands to evaluate differences in an individual's performance on administered tests. Use caution, however, when interpreting these differences because some degree of variability in performance is to be expected. The significance of observed differences decreases as the number of comparisons increases. In addition, you must determine whether a statistically significant difference has any practical or educational implications. A statistically significant discrepancy may exist between measures but have little educational relevance.

For example, Ralph's standard score of 100 on Broad Mathematics was significantly lower than his advanced performance in other areas of achievement. He is unlikely to need intervention, however, because this ability falls within the average range. As a general rule, do not use single scores on individual tests to attempt precise descriptions of specific academic skills; rather, use scores to generate hypotheses about academic performance. You can derive more reliable estimates of achievement from cluster scores than from individual test scores.

One way to approach test or cluster analysis is to evaluate the scores in reference to the norm group. For the SS ranges, the reference point is a mean of 100 (coupled with a standard deviation of 15). For analysis purposes, (a) standard scores above 115 always indicate a strength (one or more standard deviations above the mean), (b) standard scores of 85 to 115 usually indicate average ability (within one standard deviation from the mean), and (c) standard scores below 85 usually indicate a weakness (one or more standard deviations below the mean).

Sometimes, clinicians discuss “relative” strengths and weaknesses. This concept becomes particularly relevant when analyzing the performance of very low-functioning or high-functioning students. For example, a person may obtain a standard score of 80 in Broad Reading and 60 in Broad Mathematics. Although both scores are below average, reading could be described as a “relative” strength for this individual. A “twice exceptional” student (e.g., a gifted student who also has a reading disability) may have superior scores in Broad Mathematics (e.g., a standard score of 130), but only average scores in Reading (e.g., a standard score of 95). Reading would be viewed as a relative weakness.

Because all of the WJ IV ACH clusters and tests have a mean of 100 and a standard deviation of 15, you can statistically evaluate performance across the clusters and the test scores. To say that one score is meaningfully (i.e., statistically significantly) higher than another score, you first need to determine that the difference between the two scores does not represent chance variation. It is not possible to simply look at two scores and say that one is statistically higher or lower than the other score. A statistically significant difference indicates a high probability that the academic skills measured by the two scores differ. In other words, the difference is greater than that which is expected to occur simply by chance.

The main questions to ask are:

The following list provides three guidelines for determining the likelihood that a true difference in abilities is represented by the difference between the scores on any two tests or any two clusters.

  1. If the 68% confidence bands for any two tests or clusters overlap at all, assume no difference exists between the person's two abilities.
  2. If a separation exists between the ends of the two test confidence bands that is less than the width of the wider of the two confidence bands, assume that a possible difference exists between the individual's two abilities. A difference of this size would occur on average about eight times in 100 if a single comparison is made. A common practice with batteries such as the WJ IV ACH is to compare more than one pair of tests at a time or to look only at the most extreme examples. Under these conditions, the probability increases that at least one difference will be observed even though the two abilities do not differ. For example, if 10 comparisons are being made, a difference of this size would occur by chance about 57 times in 100.
  3. If the separation between the two confidence bands is greater than the width of the wider band, assume that a real difference exists between the individual's two abilities. A difference this size would occur on average about two times in 1,000 if a single comparison is being made. (The actual probability ranges from 0 to five times in 1,000 depending on the amount of separation.) If 10 comparisons were being made, a difference of this size would occur by chance about 20 times in 1,000.

You need to determine the significance of both high and low scores. For example, in testing an adult, Ms. Garrigus, you observed that she had a significant weakness in spelling. Her job as an engineer, however, does not require that her reports and memos have accurate spelling. Thus, in this situation, there are currently no vocational implications because the spelling weakness does not cause any problems. If, however, Ms. Garrigus decides to become an English teacher, her lower spelling abilities may become an issue.

The goal of profile analysis is to generate hypotheses about a person's achievement. The hypotheses generated need to be compared with other information gathered about the examinee, including relevant background knowledge, additional testing, and interviews with teachers and parents. When the hypotheses are reasonable, you can use them to: (a) clarify the specific nature of an individual's academic performance, (b) clarify the severity of any problems in academic development, (c) develop intervention recommendations, (d) select appropriate educational programs, or (e) consider an appropriate academic or vocational placement.

To determine the factors that might account for a certain pattern of scores, consider both the extrinsic (or external) factors, as well as the intrinsic (or internal) factors. Extrinsic factors include the person's socioeconomic status, education, past services and interventions, special training, social and physical environment, family background, and so forth. In some cases, differences among scores may be factors associated with examiner characteristics or factors associated with the assessment situation. Internal factors include the examinee's attention, health, temperament, physical and mental status, and motivation and interest.

Interpreting Variations and Comparisons

Two types of procedures are available on the WJ IV: intra-ability variations (variations among abilities) and ability/achievement comparisons (discrepancies between a predictor score and measured academic performance). The primary purpose of the intra-ability variation procedures is diagnosis, whereas the primary function of the ability/achievement comparison procedures is prediction. The following section describes these options in more detail. Table 4.9 indicates the variations or comparisons that are available when using the WJ IV ACH alone or with the WJ IV OL or the WJ IV COG.

Table 4.9 Variation and Comparison Calculations Available in the WJ IV

Intra-Ability Variation Ability/Achievement Comparisons
Intra-achievement (ACH only) Academic knowledge/achievement (ACH only)
Academic skills/Academic fluency/Academic Applications (ACH only) Oral language/achievement (OL and ACH)
Intra-oral language (OL only) GIA/achievement (COG and ACH)
Intra-cognitive (COG only) Scholastic aptitude/achievement (COG and ACH)
Gf-Gc/other ability (COG and ACH and OL)

Because all norms for the WJ IV ACH, the WJ IV OL, and the WJ IV COG are based on data from the same sample, actual discrepancy norms are available. This eliminates the need to estimate the discrepancy by using a regression equation or a table based on the equation. The co-norming of the WJ IV also provides an accurate means for evaluating the significance of a variation or a discrepancy. These advantages are available to the examiner when using any of the variation or comparison procedures shown in Table 4.9.

Interpreting the Significance of a Variation or Discrepancy

Understanding the “significance” of a variation or discrepancy is based on interpretation of either of two scores: the discrepancy percentile rank (Discrepancy PR) or the discrepancy standard deviation (Discrepancy SD). These scores help interpret the presence and severity of the variations and discrepancies. The Discrepancy PR reflects the percent of the population that possesses a discrepancy (difference score) of that magnitude, such as 1% or 95%. The Discrepancy PR score differs from other percentile ranks in that it compares the person to age or grade mates of the same ability or other score, not to the entire age or grade norm group. The Discrepancy SD is a standardized z-score that changes the obtained percentile rank into standard deviation units, related to a criterion such as –1.5 standard deviations.

When describing certain variations or discrepancies, refer to differences as being relative (compared to the person's other abilities) strengths or weaknesses. For example, on the intra-achievement variation procedure, a student had standard scores in the low 60s on Broad Reading, Broad Mathematics, and Broad Written Language. In contrast, the student's standard score on Academic Knowledge was 80. The score report identifies Academic Knowledge as a significant strength. In the report or description of performance, you would describe this as a relative strength because none of the scores fell within the average range.

Intra-Ability Variation Procedures

The intra-ability variation procedures are based on the practice of examining test performance to determine patterns of strengths and weaknesses. As noted in Table 4.10, four types of intra-ability variations are available on the WJ IV. With these variations, equal interest exists in either strengths or weaknesses in one area relative to the average of all other areas being considered. Only two intra-ability variation procedures are available when using just the WJ IV ACH: Intra-achievement and Academic Skills/Academic Fluency/Academic Applications. The others require using the WJ IV COG or WJ IV OL.

Table 4.10 Four Types of Intra-Ability Variations in the WJ IV

WJ IV ACH WJ IV COG WJ IV OL
Intra-achievement
(Requires a minimum of ACH Tests 1–6)		 Intra-cognitive
(Requires a minimum of COG Tests 1–7)		 Intra-oral language
(Requires a minimum of OL Tests 1–4)
Academic Skills/Academic Fluency/Academic Applications
(Requires a minimum of ACH Tests 1–6 and Tests 9–11)

Intra-Achievement Variation Procedure

The intra-achievement variation allows examiners to analyze an individual's academic test and cluster scores and to explore strengths and weaknesses. For example, a person may have a strength in reading, but a weakness in mathematics. These variations within achievement can help determine a person's present educational needs.

Examiners can calculate intra-achievement variations for three curricular areas (reading, mathematics, and writing) if Tests 1 through 6 have been administered. Additional tests and clusters can be added into this variation procedure (see Table 4.11). Individuals who have specific achievement strengths or weaknesses, such as superior math skills relative to the average of all other achievement areas, exhibit an intra-achievement variation. Table 4.12 illustrates results from an intra-achievement variation for Rebecca, a fifth-grade student. Because this variation includes more than Tests 1 through 6, it is labeled (Extended).

Table 4.11 WJ IV Intra-Achievement Variation Procedure

Intra-Achievement Variations
Required from WJ IV ACH (Tests 1-6) Optional from WJ IV ACH
Test 1: Letter-Word Identification Uses same predictor as Letter-Word Identification
Test 7: Word Attack
Test 8: Oral Reading
Basic Reading Skills
Reading Fluency (also requires Test 9: Sentence Reading Fluency)
Test 2: Applied Problems Uses same predictor as Applied Problems
Test 13: Number Matrices
Math Problem Solving
Test 3: Spelling Uses same predictor as Spelling
Test 14: Editing
Test 16: Spelling of Sounds
Basic Writing Skills
Test 4: Passage Comprehension Uses same predictor as Passage Comprehension
Test 9: Sentence Reading Fluency
Test 12: Reading Recall
Test 15: Word Reading Fluency
Test 17: Reading Vocabulary
Reading Comprehension
Reading Comprehension-Extended
Reading Rate
Test 5: Calculation Uses same predictor as Calculation
Test 10: Math Facts Fluency
Math Calculation Skills
Test 6: Writing Samples Uses same predictor as Writing Samples
Test 11: Sentence Writing Fluency
Written Expression

Table 4.12 Results from the Intra-Achievement Variation Procedure

Variations Standard Scores Discrepancy Interpretation
Intra-achievement (Extended) Actual Predicted Difference PR SD +/−1.50(SEE)
Basic Reading Skills 82 110 −28 0.1 −2.97 Weakness
Reading Comprehension 81 108 −27 0.3 −2.76 Weakness
Reading Fluency 64 109 −45 <0.1 −4.26 Weakness
Reading Rate 71 107 −36 <0.1 −3.12 Weakness
Math Calculation Skills 128 97 31 99.9 +2.98 Strength
Math Problem Solving 141 100 41 >99.9 +3.62 Strength
Basic Writing Skills 98 108 −10 13 −1.11 —
Written Expression 99 106 −7 27 −0.62 —

Interpreting the intra-achievement variation procedure. The Actual Standard Score (SS) column in Table 4.12 shows the standard score the individual actually obtained on the test. The Predicted SS column shows the individual's predicted or expected SS. The Difference column shows the difference between the actual SS and the predicted SS. A negative difference indicates the actual score was lower than predicted. A positive difference indicates the actual score was higher than predicted. The PR column shows the discrepancy percentile rank (Discrepancy PR). This score indicates the percentage of the individual's peer group (same age or grade and same predictor score) that achieved variations or differences of the same size as the individual achieved. The Discrepancy SD column shows the difference, in units of the standard error of estimate, between the individual's actual and predicted scores. A negative value in the Discrepancy SD column indicates the individual's actual achievement was lower than predicted. A positive value indicates the individual's actual achievement was higher than predicted. This statement of significance (Discrepancy SD) can be used instead of the percentile rank in programs with selection criteria based on a difference equal to or greater than, for example, one and one-half times the standard deviation. Any individual who had a Discrepancy SD of –1.5 or lower would meet this selection criterion. The final column indicates whether the variation is a significant strength or weakness. If performance is at or above +1.5 SD, a relative strength is documented. If performance is at or below −1.5 SD, a relative weakness is documented. Dashes in the final column indicate the variation was not significantly different than expected.

Using Basic Reading Skills as an example (see Table 4.12), the actual SS is 82 and the predicted score is 110. Because the actual SS (82) is lower than predicted (110), there is a negative difference (−28). The discrepancy PR of 0.1 indicates that only one out of 1,000 grademates with the same predictor score had a negative difference as large or larger between his or her actual and predicted scores on Basic Reading Skills. In other words, less than 1% of peers (same age or grade and same predictor) scored this low or lower on Basic Reading Skills. The actual score is nearly three standard deviations (−2.97) below prediction, indicating a significant weakness for this individual. The weakness in Basic Reading Skills is relative to performance on the other areas, but it is also a normative weakness (SS < 85).

Other results shown in Table 4.12 indicate that this individual had significant weaknesses in Reading Comprehension, Reading Fluency, and Reading Rate. In addition, this person had significant strengths in Math Calculation Skills and Math Problem Solving, but did not have significant variations in Basic Writing Skills or Written Expression. The following list presents a few examples of how to discuss these scores in a report.

  • In Math Calculation Skills, only one out of 1,000 grade peers with the same predicted achievement score as Rebecca would obtain a standard score of 128 or higher (or her score exceeded 999 out of 1,000 grade peers with the same predicted score) (Discrepancy PR = 99.9).
  • On the intra-achievement variations, less than one in 1,000 grademates with the same predicted score would obtain a score as high or higher on the Math Problem Solving cluster (Discrepancy PR = >99.9).
  • Math Problem Solving and Math Calculation skills are significant strengths relative to performance in the other achievement areas and are also normative strengths (SS > 115).
  • Rebecca has significant weaknesses in reading and significant strengths in math, suggesting the presence of a specific reading disability.

Although professionals must evaluate the examinee's entire performance, clinical history, and background information to arrive at the most reasonable hypothesis that accounts for significant differences among abilities, significant intra-achievement variations may indicate one or more of the following:

  • Varied interests
  • Strengths or weaknesses in processing information
  • Strengths or weaknesses in academic performance
  • Strengths or weaknesses in speed of processing (e.g., as demonstrated on the timed tests versus the skills and/or application tests)
  • Strengths or weaknesses in the curricula
  • Sensory impairments
  • Specific learning disabilities
  • Limited educational opportunities

Academic Skills/Academic Fluency/Academic Applications Variation Procedure

Nine achievement tests are required for this variation procedure: three in reading (Letter-Word Identification, Passage Comprehension, Sentence Reading Fluency), three in written language (Spelling, Writing Samples, Sentence Writing Fluency), and three in mathematics (Applied Problems, Calculation, Math Facts Fluency). The person's performance in skills, fluency, and applications is compared across the academic areas of reading, written language, and mathematics.

Three additional clusters can be added to this variation procedure: Reading Rate from the WJ IV ACH and Cognitive Processing Speed and Perceptual Speed from the WJ IV COG. Table 4.13 illustrates an example of the Academic Skills/Academic Fluency/Academic Applications variation procedure.

Table 4.13 Example of Academic Skills/Academic Fluency/Academic Applications Variation Procedure

Variations
Academic Skills/Academic
Fluency/Academic Applications		 Standard Scores Discrepancy Interpretation±1.50 (SEE)
Actual Predicted Difference PR SD
Academic Skills 103 95 8 85 +1.05 —
Academic Fluency 84 95 −11 14 −1.10 —
Academic Applications 106 103 3 59 +0.22 —

Other Intra-Ability Variation Procedures

The two remaining procedures, intra-cognitive and intra-oral language, require use of the WJ IV COG and the WJ IV OL. When using the WJ IV COG, the intra-cognitive variation procedure helps document specific cognitive strengths and weaknesses often required in determining the presence of a specific learning disability. The intra-oral language procedure requires the WJ IV OL and allows examiners to analyze an individual's strengths and weaknesses across several domains of oral language. See Chapter 5 for more information about the intra-oral language variation procedure.

Ability/Achievement Comparisons

Ability/achievement models use certain abilities to predict academic performance. The five ability/achievement comparison procedures available in the WJ IV provide a means for comparing an individual's current academic performance with the average academic performance for all others in the population with similar ability. Only one ability/achievement comparison option (academic knowledge/achievement) is available when using the WJ IV ACH. The other four procedures require the use of the WJ IV COG or the WJ IV OL in conjunction with the WJ IV ACH.

Academic Knowledge/Achievement Comparison Procedure

Since Academic Knowledge is a strong measure of acquired knowledge, or Gc, and does not require reading, writing, or math, it serves as a good predictor of academic ability. This procedure allows the examiner to determine if current achievement levels are commensurate with the individual's store of acquired content knowledge.

The Academic Knowledge cluster comprises Test 18: Science, Test 19: Social Studies, and Test 20: Humanities. The standard score for this cluster is used as the predictor of expected achievement. The individual's expected achievement is then compared to his or her actual achievement. Examinees with expected scores significantly higher than their actual achievement scores exhibit a relative strength in academic knowledge with weaknesses in one or more areas of achievement. If expected scores are significantly lower than actual achievement, the individual exhibits a relative weakness in academic knowledge with strengths in one or more areas of achievement. In addition, you can include two clusters from the WJ IV OL in this comparison procedure. Table 4.14 lists the various clusters than can be included in this ability/achievement comparison procedure. Table 4.15 illustrates an example of the Academic Knowledge/Achievement comparison for Heather, a high school sophomore. The Academic Knowledge cluster score in this example is a standard score of 127. It is used as the ability score, or predictor, to predict achievement.

Table 4.14 Academic Knowledge/Achievement Comparisons

Predictor Achievement Clusters That May Be Compared Oral Language Clusters That May Be Compared
Academic Knowledge Brief Achievement Phonetic Coding
Broad Achievement Speed of Lexical Access
Reading
Broad Reading
Basic Reading Skills
Reading Comprehension
Reading Comprehension–Extended
Reading Fluency
Reading Rate
Mathematics
Broad Mathematics
Math Calculation Skills
Math Problem Solving
Written Language
Broad Written Language
Basic Writing Skills
Written Expression
Academic Skills
Academic Fluency
Academic Applications

Table 4.15 Example of Academic Knowledge/Achievement Comparisons

Comparisons Academic Knowledge/Achievement Standard Scores Discrepancy Significant at ±1.5 SD(SEE)
Actual Predicted Difference PR SD
Broad Achievement 95 114 −19 7 −1.51 Yes (−)
Broad Reading 70 112 −42 <0.1 −3.27 Yes (−)
Basic Reading Skills 82 113 −31 1 −2.39 Yes (−)
Reading Comprehension 81 111 −30 1 −2.18 Yes (−)
Broad Mathematics 133 113 20 94 +1.55 Yes (+)
Math Calculation Skills 128 111 17 90 +1.26 No
Math Problem Solving 141 116 25 98 +2.08 Yes (+)
Broad Written Language 93 113 −20 8 −1.42 No
Basic Writing Skills 98 114 −16 10 −1.27 No
Written Expression 99 110 −11 23 −0.73 No

The correlations between Academic Knowledge (the ability score) and each achievement area are used to create the predicted score shown in the predicted score column. The final column indicates whether the achievement area was significantly higher (Yes +) or lower (Yes −) than predicted based on the Academic Knowledge cluster score. In the example shown in Table 4.15, achievement was significantly lower than predicted in Broad Achievement and all the reading clusters reported. Achievement was significantly higher than predicted in Broad Mathematics and Math Problem Solving.

Examples of statements that may be used to describe the results of this procedure would be:

  • Using Heather's Academic Knowledge score (SS = 127) to predict her achievement, less than one out of 1,000 of her peers (same grade, same predictor) would obtain a Broad Reading score as low or lower (Discrepancy PR = <0.1).
  • In Reading Comprehension, only one out of 100 grademates with the same predictor would have a negative difference of this magnitude or greater between their actual and predicted scores (Discrepancy PR = 1). Heather's actual Reading Comprehension score (SS = 81) is significantly lower than predicted (SS = 111).
  • Heather's Math Problem Solving score (SS = 141) is significantly higher than predicted. Only two out of 100 peers (same grade, same predictor) would have a score as high or higher on this cluster (or her score exceeded 98 out of 100 grade peers with the same predicted score) (Discrepancy PR = 98).
  • Although Heather's Math Calculation Skills score (SS = 128) falls in the superior range and is a normative strength, it is not significantly higher than predicted based on her Academic Knowledge score (SS = 127).
  • There is not a significant difference between Heather's actual Written Expression score (SS = 99) and her predicted score (SS = 110).

Other Ability/Achievement Comparison and Discrepancy Procedures

The oral language/achievement comparison procedure is available when using the WJ IV OL in conjunction with the WJ IV ACH. This procedure is particularly helpful in exploring the role oral language plays in the individual's academic performance. See Chapter 5 for more information about the oral language/achievement comparison procedure.

The remaining procedures require the use of the WJ IV COG. Examiners may use the General Intellectual Ability (GIA) cluster, the Scholastic Aptitudes, or the Gf-Gc Composite as the ability score. The general intellectual ability/achievement procedure can be used to determine the presence and severity of a discrepancy between general intellectual ability (g) and any particular area of achievement. When the Gf-Gc composite is the predictor, it can be used to determine the presence and severity of a discrepancy between the high g, complex cognitive abilities, and any area of achievement, as well as oral language and other cognitive abilities. In each academic area, the scholastic aptitude/achievement comparison procedure can be used to determine whether an examinee is achieving commensurate with his or her current levels of associated cognitive abilities. More information about these procedures can be found in the WJ IV COG Examiner's Manual (Mather & Wendling, 2014b).

Step-By-Step Interpretation of the WJ IV ACH

A standardized test, such as the WJ IV ACH, is designed to be a tool used during the evaluation process, and is not meant to be the only source of information about an examinee. A test does not interpret or diagnose—that is the role of the examiner. The emphasis on standard scores and determining eligibility has led some examiners to ignore important qualitative information about the examinee's instructional needs. The following provides a step-by-step plan for analyzing performance on the WJ IV ACH and goes beyond consideration of standard scores alone.

Step 1. Interpret the Clusters

Cluster scores are the most valid and reliable scores to interpret. They indicate a more representative array of abilities as well as a larger number of items than are found in individual tests. Twenty-two clusters are available.

  1. Reading: Reading, Broad Reading, Basic Reading Skills, Reading Comprehension, Reading Comprehension-Extended, Reading Fluency, Reading Rate
  2. Mathematics: Mathematics, Broad Mathematics, Math Calculation Skills, Math Problem Solving
  3. Written Language: Written Language, Broad Written Language, Basic Writing Skills, Written Expression
  4. Cross-Domain: Academic Skills, Academic Fluency, Academic Applications, Academic Knowledge, Phoneme/Grapheme Knowledge, and Brief and Broad Achievement
  • Compare results of clusters across academic areas. For example, compare reading clusters to written language, academic knowledge, or mathematics clusters. Examine the Broad Achievement clusters to determine which achievement areas are intact and which are of concern. For example, is achievement low in all academic areas or primarily in one area?
  • Use the Brief or Broad Achievement score as an estimate of academic abilities if a general indicator of overall academic performance is needed.
  • Compare results of clusters within an academic area. For example, when looking at a particular academic area such as reading, compare the scores from Basic Reading Skills to Reading Comprehension. Is performance low on all clusters (generalized) in that area? Is performance lower on Basic Reading Skills than on Reading Comprehension or lower on the Reading Comprehension cluster than on the Basic Reading Skills cluster?
  • Use the cross-domain clusters to determine if the individual has a generalized problem in basic skills (accuracy), fluency (speed), or application of the skills (reasoning).

Step 2. Interpret the Variations and Discrepancies

  • Determine whether significant intra-ability variations exist by using the intra-achievement or Academic Skills/Academic Fluency/Academic Applications variation procedures.
  • Determine whether significant ability/achievement discrepancies exist by using the academic knowledge/achievement comparison procedure.
  • Consider the implications of any significant variations or discrepancies.

Step 3. Interpret Instructional Ranges

  • Examine the grade ranges (EASY/DIFF) to determine instructional zones. For each cluster or test, you can identify an independent (easy), instructional, and frustration (difficult) level.

Step 4. Interpret SS or PR Confidence Bands

  • Examine the SS or PR ranges to determine relative standing of the individual in comparison to the peer group. Interpretation of the confidence bands can help you identify statistically significant discrepancies in the individual's performance. Determine whether the discrepancy has both practical and instructional implications.
  • Analyze the differences among cluster scores to develop hypotheses about strengths and weaknesses.
  • Determine whether significant differences exist among cluster scores.
  • Determine whether significant differences exist among the tests that constitute a specific cluster.
  • Determine whether significant differences exist among the individual test scores.

Step 5. Interpret Individual Tests (20 possible tests)

  • Consider the results of each test in each cluster obtained. Are the results similar or different? It is often informative to analyze performance among the tests measuring a broad achievement area. For example, make comparisons among the three tests of the Broad Reading Cluster: Letter-Word Identification, Passage Comprehension, and Sentence Reading Fluency. Although these comparisons are open to the errors associated with multiple comparisons, they are valuable for generating hypotheses about the examinee's reading abilities. Evaluating results at the individual test level can help determine instructional plans.
  • Analyze the differences among all of the test scores to see the patterns that go across different domains. For example, you may note that a person responds slowly on all timed tests or has difficulty on all tasks involving basic skills, but not on tasks involving reasoning and problem solving.

Step 6. Complete an Error Analysis

  • Analyze the pattern of errors within each test. Evaluating the quality of responses will help determine the individual's level of understanding. Some responses are very close to being correct, whereas others reveal minimal understanding or limited knowledge. Within a test, the items are arranged in order of difficulty, which will help you to evaluate the responses. For example, an examinee may complete addition problems correctly but make errors on subtraction problems involving regrouping.
  • Examine a person's correct and incorrect responses to individual items to help identify error patterns and determine instructional objectives. For example, you may find that on the reading and spelling tests the individual has not learned specific phoneme–grapheme correspondences. This type of observation can lead to specific recommendations and additional assessments to investigate certain findings in more depth.

Step 7. Review and Summarize Behavioral Observations

  • Review the information compiled when completing the Test Session Observations Checklist in the Test Record.
  • Use the Qualitative Observation Checklists that are available for Tests 1 through 11 to help document the individual's performance on the various tasks. Consider the individual's reaction to different tasks as well as his or her response to increased difficulty of tasks. How does the individual handle increasingly difficult tasks? Does the individual respond carefully or impulsively, quickly or slowly, accurately or inaccurately? Is there a noticeable difference in attitude or behavior between academic areas?
  • Consider the strategies used by the individual. Are they appropriate for the task? Is the individual flexible or rigid in applying strategies?
  • Note whether the individual used self-monitoring or made self-corrections on certain tasks.
  • Note any comments that were made by the examinee during the testing session. For example, Sheila, a second-grade student, told the examiner, “I like to read, but I hate math and so does my mother.”

Step 8. Integrate Information

  • Integrate the information obtained from the current evaluation with information from other sources (e.g., classroom work, observations, interviews, other testing data, RTI data, and school records).