Please provide a short summary of your impressions of the MCMI-III based on your reading (including reliability and validity information). In reviewing the MCMI-III, what are at least 2 advantages and 2 disadvantages of this assessment versus the MMPI-2/MMPI-2R (based on what you have read)? Do not forget to support your postings and replies with APA style and provide details to support your claims.

Also, please share your initial reactions of the Interpretive Report provided for the MCMI-III and how the report could be helpful to you as a human service provider working with this client. Do not forget to support your postings and replies with APA style and provide details to support your claims.

Material

Interpretive Reports (OPTIONAL RESOURCE)

Chapter 5 Copyright 2009. Wiley. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable copyright law. WECHSLER INTELLIGENCE SCALES The Wechsler intelligence scales are individually administered, composite intelligence tests in a battery format. They assess different areas of intellectual abilities and create a situation in which aspects of personality can be observed. The most recent versions (WAIS-IV and WISC-IV) provide an overall, or “Full Scale IQ,” as well as specific index scores that can be calculated using various combinations of subtests. The Wechsler intelligence scales are considered to be among the best of all psychological tests because they have sound psychometric properties and produce information relevant to practitioners. As a result, they have become the most frequently used tests in clinical practice (Archer, Buffington-Vollum, Stredny, & Handel, 2006; Camara, Nathan, & Puente, 2000; Watkins, Campbell, Nieberding, & Hallmark, 1995). TESTING OF INTELLIGENCE: PRO AND CON The testing of intelligence has had a consistent history of misunderstanding, controversy, and occasional misuse (Bartholomew, 2006: D. Flanagan & Harrison, 2005; Weinberg, 1989). Criticisms have ranged from moral indictments against labeling individuals, to cultural bias, and even to accusations of flagrant abuse of test scores. Although valid criticisms can be made against testing intelligence, such procedures also have a number of advantages. One of the main assets of intelligence tests is their accuracy in predicting future behavior. Initially, Alfred Binet was able to achieve a certain degree of predictive success with his scales, and, since that time, test procedures have become progressively more refined and accurate. More recent studies provide ample support that intelligence tests can predict an extremely wide number of variables. In particular, IQ tests are excellent predictors of academic achievement (A. Kaufman & Lichtenberger, 2006; Neisser et al., 1996) and occupational performance (J. Hunter & Schmidt, 1996; F. Schmidt & Hunter, 1998, 2004; Wagner, 1997), and are sensitive to the presence of neuropsychological deficit (Groth-Marnat, Gallagher, Hale, & E. Kaplan, 2000; Lezak, Howieson, & Loring, 2004). However, certain liabilities are also associated with these successes. First, intelligence tests can be used to classify children into stereotyped categories, which limit their freedom to choose fields of study. Furthermore, IQ tests are quite limited in predicting nontest or nonacademic activity, yet they are sometimes incorrectly used to make these inferences (Snyderman & Rothman, 1987; Sternberg, 2003). It should also be stressed that intelligence tests are measures of a person’s current level of functioning and, as such, are best used for making short-term predictions. Long-term predictions, although attempted frequently, are less accurate because there can be many uncontrolled, influencing variables. Similarly, even short-term academic placements made solely on the basis of an IQ score have a high chance of failure because all the variables that may be crucial for success are not and cannot be measured by an intelligence test. It can sometimes be tempting for test users to extend the meaning of test scores beyond their intended scope, especially in relation to the predictions they can realistically be expected to make. 119 EBSCO Publishing : eBook Collection (EBSCOhost) – printed on 1/18/2023 7:01 PM via POST UNIVERSITY AN: 354708 ; Gary Groth-Marnat.; Handbook of Psychological Assessment Account: ns017336.main.eds c05.indd Sec1:119 9/23/10 11:07:00 AM 120 Wechsler Intelligence Scales In addition to predicting academic achievement, IQ scores have also been correlated with occupation, ranging from highly trained professionals with mean IQs of 125, to unskilled workers with mean IQs of 87 (Reynolds, Chastain, Kaufman, & McLean, 1987). Correlations between job proficiency and general intelligence have been highest in predicting relatively more complex jobs (.58) than less demanding occupations (.23; F. Schmidt & Hunter, 2004). They have also reported moderately high correlations between general intelligence and success for managers (.53), salespersons (.61), and clerks (.54). For intellectually demanding tasks, nearly half the variance related to performance criteria can be accounted for by general intelligence (F. Schmidt & Hunter, 2004; F. Schmidt, Ones, & Hunter, 1992). The use of intelligence tests for personnel selection has demonstrated financial efficacy for organizations (F. Schmidt & Hunter, 1998). In addition, the accuracy of using IQ tests can be incrementally increased by combining the results with integrity tests, work samples, and structured interviews (F. Schmidt & Hunter, 1998, 2004). Another important asset of intelligence tests, particularly the WAIS-IV and WISCIV, is that they provide valuable information about a person’s cognitive strengths and weaknesses. They are standardized procedures whereby a person’s performance in various areas can be compared with that of age-related peers. In addition, useful comparisons can be made regarding a person’s pattern of strengths and weaknesses. The WAIS-IV, WISC-IV, and other individually administered tests provide the examiner with a structured context in which a variety of tasks can be used to observe the unique and personal ways the examinee approaches cognitive tasks. Through a client’s interactions with both the examiner and the test materials, an initial impression can be made of the individual’s self-esteem, behavioral idiosyncrasies, anxiety, social skills, and motivation, while also obtaining a specific picture of intellectual functioning. Intelligence tests often provide clinicians, educators, and researchers with baseline measures for use in determining either the degree of change that has occurred in an individual over time or how an individual compares with other persons in a particular area or ability. These distinctions may have important implications for evaluating the effectiveness of an educational program or for assessing the changing abilities of a specific student. In cases involving recovery from a head injury or readjustment following neurosurgery, it may be extremely helpful for clinicians to measure and follow the cognitive changes that occur in a patient. Furthermore, IQ assessments may be important in researching and understanding more adequately the effect on cognitive functioning of environmental variables, such as educational programs, family background, and nutrition. Thus, these assessments can provide useful information about cultural, biological, maturational, or treatment-related differences among individuals. A criticism leveled at intelligence tests is that almost all have an inherent bias toward emphasizing convergent, analytical, and scientific modes of thought. Thus, a person who emphasizes divergent, artistic, and imaginative modes of thought may be at a distinct disadvantage. Some critics have even stressed that the current approach to intelligence testing has become a social mechanism used by people with similar values to pass on educational advantages to children who resemble themselves. Not only might IQ tests tend to place creative individuals at a disadvantage, but also they are limited in assessing nonacademically oriented intellectual abilities (Gardner, 2006; Snyderman & Rothman, 1987; Sternberg, 2003). Thus, social acumen, success in dealing with people, the ability to handle the concrete realities of the individual’s daily world, social fluency, and specific tasks, such as purchasing merchandise, are not measured by standard intelligence tests (Sternberg, 2003). More succinctly, people are capable of many more cognitive abilities than can possibly be measured on an intelligence test. EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd Sec1:120 9/23/10 11:07:01 AM Testing of Intelligence: Pro and Con 121 Misunderstanding and potential misuse of intelligence tests frequently occur when scores are treated as measures of innate capacity. The IQ is not a measure of an innate fixed ability, nor is it representative of all problem-solving situations. It is a specific and limited sample, made at a certain point in time, of abilities that are susceptible to change because of a variety of circumstances. It reflects, to a large extent, the richness of an individual’s past experiences. Although interpretation guidelines are quite clear in pointing out the limited nature of a test score, there is a tendency to look at test results as absolute facts reflecting permanent characteristics in an individual. People often want a quick, easy, and reductionist method to quantify, understand, and assess innate cognitive abilities, and the IQ score has become the most widely misused test score to fill this need. An important limitation of intelligence tests is that, for the most part, they are not concerned with the underlying processes involved in problem solving. They focus on the final product or outcome rather than on the steps involved in reaching the outcome. They look at the “what” rather than the “how” (Embretson, Schneider, & Roth, 1986; E. Kaplan et al., 1999; Milberg, Hebbgen, & Kaplan, 1996). Thus, a low score on Arithmetic might result from poor attention, difficulty understanding the examiner because of disturbances in comprehension, or low educational attainment. The extreme example of this “end product” emphasis is the global IQ score. When the examiner looks at the myriad assortment of intellectual abilities as a global ability, the complexity of cognitive functioning may be simplified to the point of being almost useless. The practitioner can apply labels quickly and easily, without attempting to examine the specific strengths and weaknesses that might make precise therapeutic interventions or knowledgeable recommendations possible. Such thinking detracts significantly from the search for a wider, more precise, and more process-oriented understanding of mental abilities. A further concern about intelligence tests involves their limited usefulness in assessing minority groups with divergent cultural backgrounds. It has been stated that intelligencetest content is biased in favor of European American, middle-class values. Critics stress that minorities tend to be at a disadvantage when taking the tests because of deficiencies in motivation, lack of practice, lack of familiarity with culturally loaded items, and difficulties in establishing rapport. Numerous arguments against using intelligence tests for the assessment and placement of minorities have culminated in legal restrictions on the use of IQ scores. However, traditional defenses of IQ scores suggest that they are less biased than has been accused (see the “Use with Diverse Groups” section later in this chapter). The issue certainly has not been resolved, but clinicians should continue to be aware of this dilemma, pay attention to subgroup norms, and interpret minority group IQ scores cautiously. Finally, many people feel that their IQs are deeply personal pieces of information. They would prefer that others, even a psychologist who is expected to observe confidentiality, not be allowed access to this information. This problem is further compounded when IQ scores might be given to several different persons, such as during legal proceedings or personnel selection. Intelligence tests provide a number of useful and well-respected functions. They can adequately predict short-term scholastic performance; assess an individual’s relative strengths and weaknesses; predict occupational achievement; reveal important personality variables; and permit the researcher, educator, or clinician to trace possible changes in an individual or population. However, these assets are helpful only if the limitations of intelligence tests are adequately understood and appropriately taken into consideration. They are limited in predicting certain aspects of occupational success and nonacademic skills, such as creativity, motivational level, social acumen, and success in dealing with people. Furthermore, IQ scores are not measures of an innate, fixed ability, and their use in EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 121 9/23/10 11:07:02 AM 122 Wechsler Intelligence Scales classifying minority groups has been questioned. Finally, there has been an overemphasis on understanding the end product of cognitive functioning and a relative neglect in appreciating underlying cognitive processes. HISTORY AND DEVELOPMENT During the 1930s, Wechsler began studying a number of standardized tests and selected 11 different subtests to form his initial battery. His search for subtests was in part guided by his conception that intelligence is global in nature and represents a part of the greater whole of personality. Several of his subtests were derived from portions of the 1937 revision of the Stanford-Binet (Comprehension, Arithmetic, Digit Span, Similarities, and Vocabulary). The remaining subtests came from the Army Group Examinations (Picture Arrangement), Koh’s Block Design (Block Design), Army Alpha (Information, Comprehension), Army Beta (Digit Symbol-Coding), Healy Picture Completion (Picture Completion), and Pinther-Paterson Test (Object Assembly). These subtests were combined and published in 1939 as the Wechsler-Bellevue Intelligence Scale. The WechslerBellevue had a number of technical deficiencies primarily related to both the reliability of the subtests and the size and representativeness of the normative sample. Thus, it was revised to form the Wechsler Adult Intelligence Scale (WAIS) in 1955; another revised edition (WAIS-R) was published in 1981. The 1981 revision was based on 1,880 individuals who were generally representative of the 1970 census and categorized into nine different age groups. The Wechsler Adult Intelligence Scale-III (WAIS-III) replaced the earlier (1981) WAIS-R. The primary reason for the revision was to update the norms. Additional reasons included extending the age range, modifying items, developing a higher IQ “ceiling” and “floor,” decreased reliance on timed performance, developing index/factor scores, creating linkages to other measures of cognitive functioning/achievement, and extensive testing of reliability and validity. Despite these changes, many of the traditional features of the WAIS-R were maintained, including the six Verbal subtests and the five Performance subtests. Maintaining these clusters of subtests still enabled practitioners to calculate the Full Scale, Verbal, and Performance IQs. An added feature of the WAIS-III was the inclusion of three new subtests, which enabled the calculation of four index scores. Thus, the WAIS-III was not merely a renormed “face-lift”; it also enabled the clinician to do more with the different test scores, such as being able to assess persons with either greater age or IQ ranges as well as linking scores with the Wechsler Memory Scales or calculating both IQ and index/factor scores. The Wechsler Adult Intelligence Scale—Fourth Edition (WAIS-IV) is the most recent revision of the evolving Wechsler intelligence scales (Wechsler, 2008a,b). The general purpose of the revision was to update norms, improve floors and ceilings, improve psychometric properties, reduce testing time, and conorm it with the Wechsler Memory Scale-Fourth Edition (WMS-IV; see Table 5.1) and the Wechsler Individual Achievement Test-II (WIAT). One of the most obvious changes has been the elimination of the timehonored verbal versus performance IQ. Instead, the WAIS-IV uses the traditional Full Scale IQ along with four index scores (Verbal Comprehension, Working Memory, Perceptual Reasoning, and Processing Speed). The major rationale for the elimination of the Verbal-Performance IQs is that they are not pure measures but typically combine a number of different abilities. For example, the Verbal IQ included measures of verbal abilities as well as working memory. Thus, it was not a unitary measure of an ability. In contrast, relying on the four index scores ensures that relatively pure, theoretically sound measures EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 122 9/23/10 11:07:02 AM History and Development Table 5.1. 123 Major changes on the WAIS-IV Elimination of Verbal and Performance IQs Updated norms 15 subtests (versus 14 on the WAIS-III) Computation of Full Scale IQ and Indexes based on 10 core subtests General Ability Index (optional index that combines Verbal Comprehension and Perceptual Reasoning) 3 newly developed subtests (Visual Puzzles, Figure Weights, Cancellation) Deletion of 2 subtests (Object Assembly, Picture Arrangement) Organization of subtests into core and supplemental Renaming of the Perceptual Organization Index to the Perceptual Reasoning Index Inclusion of process scoring options for Block Design, Digit Span, Letter-Number Sequencing Potential for shortened administration using only the 10 core subtests (for FSIQ + Indexes) Greater attention to floor and ceilings Normed linkages with the Wechsler Memory Scale-IV Upgrade kit for specialist neuropsychologists and geropsychologists of abilities have been made. This reliance on a Full Scale IQ plus the four indexes parallels a similar development for the WISC-IV (Wechsler, 2003). In addition, the WAIS-IV includes an optional General Ability Index that combines the Verbal Comprehension and Perceptual Reasoning index scores. An upgrade of the WAIS-IV for neuropsychologists and geropsychologists became available in 2009 (WAIS-IV/WMS-IV Advanced Clinical Solutions, Pearson, 2009c). A further feature of the WAIS-IV has been the deletion, addition, and revision of subtests. In addition, subtests have been organized according to core and supplemental subtests. The core subtests are used to develop the index scores (see Table 5.2). However, if a core subtest is “spoiled” (i.e., made invalid) or if practitioners are unable to administer a core subtest, they can use one of the supplemental subtests instead. Supplemental subtests can also be administered to find additional information regarding a client’s level of functioning. For example, the new Cancellation subtest might be added to Symbol Search and Coding to add further information related to a client’s ability to process information rapidly. Many of the subtests have undergone revisions to enhance clarity of instructions, refine scoring rules, change stimuli, and include different items. New norms have been developed for the WAIS-IV derived from 2,200 persons between the ages of 16 and 90 stratified according to sex, education, ethnicity, and geographical region. These norms closely correspond to the 2005 U.S. census data. Whereas 200 examinees were included for the age bands between 16 and 60, only 100 examinees have been included for the age bands between 70 and 90. The WAIS-IV has been conormed with the WMS-IV and Wechsler Individual Achievement Test-II. Norms and patterns of responses have been developed for special groups, including mild cognitive impairment, borderline intellectual functioning, traumatic brain injury, Alzheimer’s disease, attention deficit hyperactivity disorder (ADHD), reading disorder, math disorder, autism, Asperger’s syndrome, and depression. The original Wechsler-Bellevue Scale was developed for adults, but in 1949, Wechsler developed the Wechsler Intelligence Scale for Children (WISC) so that children from the age of 5 years 0 months could be assessed in a similar manner. Easier items, designed EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 123 9/23/10 11:07:02 AM 124 Wechsler Intelligence Scales Table 5.2. Organization of WAIS-IV subtests Index Verbal Comprehension Core subtests Supplemental subtests Similarities Comprehension Vocabulary Information Perceptual Reasoning Block Design Figure Weights Matrix Reasoning Picture Completion Visual Puzzles Working Memory Digit Span Letter-Number Sequencing Arithmetic Processing Speed Symbol Search Cancellation Coding for children, were added to the original scales and standardized on 2,200 European American boys and girls selected to be representative of the 1940 census. However, some evidence shows that Wechsler’s sample may have been overrepresentative of children in the middle and upper socioeconomic levels. Thus, ethnic minorities and children from lower socioeconomic levels may have been penalized when compared with the normative group. The WISC was revised in 1974 and standardized on a new sample that was more accurately representative of children in the United States. The WISC-III (Wechsler, 1991) was released in 1991; its major change was the inclusion of four factor/index scores (Verbal Comprehension, Perceptual Organization, Freedom from Distractibility, and Processing Speed). The new Processing Speed factor involved the inclusion of a new Symbol Search subtest along with the older Coding subtest. As with the earlier WISC-R, the standardization and reliability were excellent. The scales were standardized on 2,200 children between the ages of 6 and 16 who closely matched the 1988 census. The sample consisted of 100 boys and 100 girls for each of the different age groups. The WISC-IV (2003) was noteworthy in that it contained more changes than any other previous edition (see Table 5.3). The most obvious change was the elimination of the time-honored Verbal and Performance IQ. There was instead a greater reliance on interpretation using a combination of the four index scores along with the global Full Scale IQ (see Table 5.4). The indexes have also been refined by the inclusion of a five new subtests (and the deletion of Picture Arrangement, Object Assembly, and Mazes). Completely new norms for the WISC-IV were developed that closely represented the U.S. census. A further potentially useful feature was the publication of the WISC-IV Integrated (Wechsler et al., 2004), which allows for 12 additional procedures that enable specialty practitioners the option of analyzing the underlying processes clients go through when making their responses (see McCloskey & Maerlander, 2005). One of the motivations for the WAIS-IV and WISC-IV revisions was to update the instruments’ theoretical foundations. To a certain extent, this has been done. The importance of fluid intelligence has been reflected in the introduction of a new subtest that assesses this area of intellectual functioning (Matrix Reasoning, Picture Concepts, Word Reasoning, Visual Puzzles, and Figure Weights). In addition, the concepts of working memory and processing speed have been incorporated and refined. The result has been changes and refinements in the subtests and psychometric properties included in the Working Memory and Processing Speed indexes. It should also be noted that the factor EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 124 9/23/10 11:07:03 AM History and Development Table 5.3. 125 Major changes on the WISC-IV Elimination of Verbal and Performance IQs Updated norms Fifteen subtests (versus 13 on the WISC-III) Computation of IQ and Indexes based on 10 core subtests Introduction of five new WISC-IV subtests: Three newly developed subtests (Word Reasoning, Cancellation, and Picture Concepts) plus downward extensions of two WAIS-III subtests (Matrix Reasoning, and Letter-Number Sequencing) Three deleted subtests (Picture Arrangement, Object Assembly, and Mazes) New organization of subtests into core and supplemental Introduction of seven Process scores Longer administration time with all 15 subtests (65–80 minutes versus 50–70 minutes for the WISC-III) Potential for shortened administration using only the 10 core subtests (for FSIQ + Indexes) Greater attention to floor and ceilings Normed linkages with the Wechsler Individual Intelligence Test-II IQs below 70 now called “Extremely Low” versus the “Intellectually Deficient” title on the WISC-III Table 5.4. Organization of WISC-IV indexes and subtests Index Core subtests Supplemental subtests Verbal Comprehension Similarities Information Word Reasoning Vocabulary Comprehension Perceptual Reasoning Block Design Picture Completion Picture Concepts Matrix Reasoning Working Memory Digit Span Arithmetic Letter-Number Sequencing Processing Speed Coding Cancellation Symbol Search structure of the more recent revisions of the Wechsler intelligence scales has resulted in a de facto theory of intelligence defined by the Full Scale IQ in combination with the functions measured by the four indexes. However, neither the WAIS-IV nor the WISC-IV has been organized around a specific theory of intelligence. In contrast, revisions of the KABC, Stanford Binet-5, and the Woodcock Johnson-III, each of which took place in the first few years of the 21st century, were closely aligned to the Cattell-Horn-Carrol (CHC) theory of intelligence. Some have stated that the Wechsler intelligence scales have been overburdened by their traditions, resulting in a failure to make major adaptations to evolving knowledge related to intelligence (Flanagan & Kaufman, 2004). In 1967, the Wechsler Preschool and Primary Scale of Intelligence (WPPSI) was first published for the assessment of children between the ages of 4 and 6 years 6 months. Just as the WISC is a downward extension of the WAIS, so the WPPSI is generally a downward EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 125 9/23/10 11:07:03 AM 126 Wechsler Intelligence Scales extension of the WISC in which easier but similar items are used. Although most of the scales are similar in form and content to the WISC, a number are unique to the WPPSI. The WPPSI was revised in 1989 (WPPSI-R; Wechsler, 1989) and again in 2002 (WPPSI-III; Wechsler, 2002c). RELIABILITY AND VALIDITY WAIS-IV Reliability and Validity The reliabilities for the WAIS-IV are generally quite high (Wechsler, 2008b). Areas of note are that average split-half reliability for the Full Scale IQ (FSIQ) is an extremely high .98 (Wechsler, 2008b). Average split-half reliabilities for the other combined or “composite” scores are only slightly lower and range from a high of .96 for the Verbal Comprehension index to a low of .90 for Processing Speed. Average split-half reliability for the subtests ranged from excellent (i.e., Vocabulary r = .94, Digit Span r = .93) to an acceptable .78 for Cancellation. All subtests except Cancellation were above .81. These good to excellent reliabilities were found not only for the standardization sample but also among various clinical populations (i.e., brain injured, ADHD, Alzheimer’s disease). The average standard error of measurement for various WAIS-IV scores indicates a small band of error (i.e., Full Scale IQ = 2.16, Verbal Comprehension Index = 2.12, Processing Speed Index = 4.24). Calculating the Standard Error of Measure for each of the composite scores is a standard procedure on the WAIS-IV Record Form so that it is reported for all examinees. The Average test-retest reliabilities (8–82 days, M = 22 days) for the Full Scale IQ was quite high (r = .96) and the composite scores were similarly high, ranging from .96 for Verbal Comprehension to a low of .87 for Processing Speed. These reliabilities are among the best for any test available and, in almost all cases, represent a slight improvement over the WAIS-III. While these test-retest reliabilities indicate a high degree of temporal stability, there is still some degree of improvement on retesting because of practice effects. Improved performance due to retesting is important to understand; clinicians need to know when to attribute an increase in scores to practice effects and when this might indicate actual clinical improvement. Over the retesting interval (8–82 days, M = 22 days), the Full Scale IQ was found to increase by 4.7 points. The lowest increase was for Verbal Comprehension (2.5 points) followed by Working Memory (3.1), Perceptual Reasoning (3.9), and Processing Speed (4.4). These increases are not only statistically significant but may have clinical significance when making inferences about the extent to which real improvement/deterioration has occurred for a particular client. Thus, a client who has a Perceptual Reasoning increase of 4 points on retesting may not really be improving in everyday functions but merely demonstrating practice effects. A difference of 15 points on the earlier WAIS-III Full Scale IQ (for ages 16 to 54) was found to be necessary to infer that there had been an actual improvement in abilities (A. Kaufman & Lichtenberger, 2002). Research with the WAIS-R indicated that these practice effects can occur up to 9 months later even among head-injured patients. However, retest gains have also been found to diminish with advancing age (J. Ryan, Paolo, & Brungardt, 1990; Wechsler, 2008a, 2008b). Because extensive validity studies exist for the WAIS-III, one of the most important steps in WAIS-IV validation was to determine the comparability between the two tests. As expected, correlations were found to be quite high. The WAIS-IV and WAIS-III Full Scale IQ correlation was .94 (Wechsler, 2008b). The four indexes were similarly high, ranging from .91 for Verbal Comprehension to .84 for Perceptual Reasoning/Perceptual Organization. This finding suggests that the WAIS-IV measures essentially the same constructs as the WAIS-III. Noteworthy high correlations between WAIS-IV and WAIS-III EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 126 9/23/10 11:07:04 AM Reliability and Validity 127 subtests were .90 for Information, .87 for Vocabulary, and .85 for Coding. In contrast, a relatively low correlation was found for Picture Completion (.65). Correlations between the WAIS-IV and WISC-IV for a group of 16-year-olds were quite high (Full Scale IQ = .91, Verbal Comprehension = .88, Perceptual Reasoning = .77, Working Memory = .78, Processing Speed = .77). The correlations between the WAIS-IV and Wechsler Memory Scale-III were .61 for the Working Memory Index. A similar .59 correlation was found between the Full Scale IQ and the WMS-III General Memory Index. These moderate correlations are expected, given that the WAIS-IV and WMS-III measure somewhat different but still overlapping constructs. Correlations between achievement were, as expected, in the moderate to high range (WAIS-IV Full Scale IQ and Wechsler Individual Achievement Test-II Total Achievement = .88). The WAIS-IV has also been found to produce expected patterns of correlation with a number of additional standard ability measures (Wechsler, 2008b). The Delis-Kaplan Executive Functioning System (Delis, Kaplan, & Kramer, 2001) is a series of subtests that measure various aspects of a client’s ability to initiate, plan, and flexibly monitor their behavior. Representative correlations were a .22 between Perceptual Reasoning and the ability to fluidly produce the names of classes of objects (Category Fluency) and a correlation of .77 with Full Scale IQ and the ability to flexibly and rapidly connect combinations of letters and numbers (Trail Making). The California Verbal Learning Test-II (Delis, Kramer, Kaplan, & Ober, 2000) measures how well a person can recall lists of words that are read to them. Correlations between the WAIS-IV Full Scale IQ and a series of trials on learning word lists ranged from .48 to .32 (Wechsler, 2008b). A final representative test is the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS; Randolph, 1998), which measures various domains of cognitive functioning (see Chapter 12). Correlations between the WAIS-IV Full Scale IQ and the RBANS Total Score was .75. The WAIS-IV Index scores and RBANS Total Score correlations were slightly lower, ranging from a high of .72 for Perceptual Reasoning to a low of .54 for Processing Speed. This overview of WAIS-IV correlations with various standardized measures provides strong support for WAIS-IV validity. Factor analysis of the WAIS-IV has supported the presence of g in that most subtests correlate with each other as well as with the FSIQ at least to a moderate extent (Wechsler, 2008b). Dividing subtests into four indexes is supported by current theories of intelligence as well as factor-analytic procedures (Wechsler, 2008b). However, Arithmetic was found to load on both the Verbal Comprehension as well as the Working Memory factors. This is consistent with conceptualizations of Arithmetic in that it involves both verbal abilities as well as working memory. In addition, the new Figure Weights subtest was found to load highly on both Perceptual Reasoning and Working Memory. Again, these high loadings are expected, given that Figure Weights involve visual reasoning, but this reasoning is related to quantitative manipulations. Future research likely will further refine these findings and present alternative factor alternatives. Various clinical populations have patterns of deficits in learning, cognition, and memory (see Wechsler, 2008b). It thus is to be expected that the WAIS-IV would be sensitive to these patterns. This finding was somewhat supported in that the mean WAIS-IV Full Scale IQ (M = 81.2) and index scores for Alzheimer’s disease patients were low compared with their age-related peers. Comparisons among the index scores indicated differential cognitive abilities in that the mean Verbal Comprehension Index was relatively higher (86.2) than Processing Speed (76.6). However, it would have been expected that the Working Memory Index would have been somewhat lower than the mean of 84.3, given the considerable memory complaints among this population. Traumatic braininjured patients had a somewhat similar pattern in that verbal abilities were relatively spared EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 127 9/23/10 11:07:04 AM 128 Wechsler Intelligence Scales (relatively higher Verbal Comprehension mean of 92.1) when compared with a relatively lower Processing Speed (80.5). This finding indicates that the WAIS-III is sensitive to the difficulties these patient populations have with rapidly processing and consolidating information. The mean WAIS-IV Full Scale IQ (96.9) for clients diagnosed with attention deficit hyperactivity disorder (ADHD) did not differ substantially from the standardization sample. In addition, their mean Working Memory Index score (94.7) was only slightly lower than their Verbal Comprehension mean scores (100.9). This finding suggests that the WAIS-IV is not particularly sensitive to the attentional problems of this group, perhaps because the WAIS-IV is administered in a structured testing situation with few distractions. In contrast, real-world environments are likely to have multiple concurrent attentional demands that would be much more difficult for these clients to ignore. Subjects diagnosed with reading-related learning disabilities were found to have mean Working Memory scores (88.9) that were significantly below the WAIS-IV standardization sample (101.1). Learning-disabled persons with mathematical difficulties similarly had the greatest difficulty with Working Memory (84.1) when compared with matched controls derived from the standardization sample (98.7). This finding reflects the common difficulties related to tasks requiring short-term memory and attention. These examples of research described in the WAIS-IV Technical and Interpretive Manual suggest that the WAIS-IV is sensitive to the types of cognitive difficulties found among various patient groups. WISC-IV Reliability and Validity Reliability on the WISC-IV is generally excellent. Internal consistency reported in the Technical Manual for the Full Scale IQ ranges from .96 to .97 (M = .97). The mean internal consistencies for the individual index scores range from .91 to .92. The mean internal consistencies for 12 of the 15 subtests ranged from a low of .79 for Symbol Search and Cancellation to a high of .90 for Letter-Number Sequencing. Test-retest reliability (average 32-day interval) for the Full Scale IQ was .89. The four index test-retest reliabilities range from a high of .79 for Processing Speed to a relative low of .89 for Verbal Comprehension. Average test-retest stability for the subtests ranged from a high of .85 for Vocabulary to a low of .68 for Symbol Search. It should be noted, however, that this data represents reliabilities that were averaged across the 11 age groups. Reliabilities for certain age groups can vary substantially. For example, 12 of the 15 subtest coefficients for the 6- to 7-year age group were found to be below .80. An important issue related to test-retest reliability is how to interpret changes in scores. The Technical Manual indicates that, over a 9-week interval, there was an average increase of 5.6 points on the Full Scale IQ. The index scores had increases of 2.1 points for Verbal Comprehension, 5.2 points for Perceptual Reasoning, 2.6 points for Working Memory, and 7.1 points for Processing Speed. This finding means that increases over a time interval of 9 weeks or less will need to be substantially greater than the listed scores to indicate actual increases in ability. If the increases are within the magnitude indicated here, they are more likely to be due to practice effects. It should be particularly noted that Processing Speed and Perceptual Reasoning have quite large practice effects (7.1 and 5.2, respectively). More research is needed to determine the impact of practice effects over a longer interval and with different groups (i.e., brain injured, learning disabled). The WISC-III had well-substantiated validity, and it can be cautiously assumed that much of this research can be transferred to the WISC-IV. This idea is partially supported in that 56% of the items are shared. In addition, there are moderate to high correlations between the WISC-III and WISC-IV Full Scale IQs, index scores, and subtests EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 128 9/23/10 11:07:04 AM Assets and Limitations 129 (i.e., Full Scale IQ = .87, Verbal Comprehension = .85, Perceptual Reasoning/Perceptual Organization Index = .70, Working Memory/Freedom from Distractibility = .74, Processing Speed = .81). Criterion validity has been performed on several Pearson Assessment (previously Psychological Corporation) tests with generally favorable results. For example, the WISC-IV Full Scale IQ and General Memory and Attention/ Concentration scores on the Children’s Memory Scale correlated .61 and .72 respectively. Correlations with the BarOn Emotional Quotients were low to nonsignificant, which is what would be expected, given that they are theoretically assessing different variables. Finally, correlations with the WISC-IV Full Scale IQ and the Adaptive Behavior Assessment Scale-Teacher were also supportive (i.e., General Adaptive = .58 and Conceptual = .63). The Technical Manual also provides studies of 16 different groups. For example, a closed head injury group had a mean Processing Speed score of 85 compared to a mean Full Scale IQ of 90. In contrast, children who were considered to be intellectually gifted had Full Scale IQ scores of 123.5 with Verbal Comprehension scores of 124.7. Factor analyses generally has supported the four-factor model that was originally reported in the Technical Manual and is reflected in the four index scores (Sattler & Dumont, 2004; Watkins, Wilson, Kotz, Carbone, & Babula, 2006). However, even though Sattler and Dumont (2004) found support for the factor structure, it can also vary according to the age of the sample. For example, Information was found to have a factor loading of only .25 with the Verbal Comprehension Index at age 8. A similar low factor loading was a .22 for Symbol Search on Processing Speed. In addition, nearly all the subtests are good to fair measures of g (Sattler & Dumont, 2004) and have good to fair specificity. The exception is Cancellation, which is a poor measure of g and has poor subtest specificity. A contrast to these factor-analytic findings is that the factor structure of the WISC-IV does not optimally reflect the four indexes but instead is best reflected in various dimensions of the Cattell-Horn-Carrol (CHC) theory of intelligence (Flanagan & Kaufman, 2004; Keith, Fine, Taub, Reynolds, & Kanzler, 2006). Specifically, Keith et al. (2006) found that their factor structure best fit measures of crystallized ability (Gc), visual processing (Gv), fluid reasoning (Gf), short-term memory (Gsm), and processing speed (Gs). The practical implications of this finding are that interpretations are likely to be most accurate and most consistent with theory when clusters of subtests are arranged according to these constructs (Flanagan & Kaufman, 2004; Keith, Fine, Taub, Reynolds, & Kranzler, 2006). Thus the Verbal Processing Speed Index still can be organized according to the traditional index cluster (Digit Symbol-Coding and Symbol Search). Verbal Processing Speed similarly measures the speed by which the client processes information (or general speed/Gs, in CHC terminology). In contrast, Block Design and Picture Completion measure the CHC dimension of visual processing (Gv), which represents a different cluster than the indexes presented in the WISC-IV Technical Manual. ASSETS AND LIMITATIONS Since their initial publication, the Wechsler intelligence scales have been used in numerous research studies and have become widely used throughout the world. Thus, they are familiar to both researchers and practitioners and also have a long and extensive history of continued evaluation. This enormous research base allows practitioners to make relatively accurate predictions regarding clients. Inconsistencies between an individual’s performance and relevant research can also be noted, alerting the practitioner that he or she needs to develop and pursue further hypotheses. Furthermore, the subtests are relatively easy to administer, and the accompanying manuals provide clear instructions, concise tables, and excellent norms. EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 129 9/23/10 11:07:05 AM 130 Wechsler Intelligence Scales Norms for both the WAIS-IV and WISC-IV represent a further clear strength. The size is adequate and, for the most part, has corresponded to the demographics of the U.S. census. Cross-national use has been developed through research on how residents in other countries perform. Sampling on the WAIS-IV and WISC-IV for African American and Hispanics closely approximated U.S. Census data. A further important feature is that the WAIS-IV was conormed with the Wechsler Memory Scale-IV (WMS-IV) and the Wechsler Individual Achievement Test. Thus, a high degree of confidence can be placed in comparing scores between these tests. Finally, the WAIS-IV has extended its age range to include the performance for persons in the 70- to 90-year range. This is an important feature, given the increases in knowledge related to this age group along with the expanding number of persons over 65. Of perhaps even more practical importance to the clinician is the clear, precise data obtained regarding the person’s cognitive functioning from the IQ, index, and subtest scores. For example, high scores on the Verbal Comprehension Index indicate good verbal abilities and that the person has benefited from formal education. In contrast, a low score on Processing Speed suggests the person would have difficulty processing information quickly. Clinicians can become extremely sensitive to the different nuances and implications of various patterns of scores. Thus, many of these interpretive guidelines, particularly for the IQ and index scores, have substantial theoretical and empirical support. A final, but extremely important, asset of the Wechsler scales is their ability to aid in assessing personality variables. This assessment can be done by directly observing the individual as he or she interacts with the examiner, studying the content of test item responses, or evaluating information inferred from the individual’s pattern of subtest scores. For example, a person who scored low on Digit Span, Arithmetic, and Coding is likely to be experiencing anxiety, to have an attentional deficit, or a combination of both. On the other hand, it might be hypothesized that a person who scored high on Comprehension is likely to have good social judgment. Despite attempts to establish patterns of subtest scores for different psychiatric groups, few clear findings have emerged (Piedmont, Sokolov, & Fleming, 1989a, 1989b). Thus, the Wechsler scales should not be seen as similar to “personality scales” or “clinical scales.” Rather, the subject’s subtest patterns, behavior surrounding the test, and qualitative responses to the items should be considered as a means of generating hypotheses related to personality. In this context, the Wechsler intelligence scales are noteworthy in the degree to which they can provide personality variables and clinical information. One significant criticism leveled at the Wechsler scales has been their lack of data supporting their ecological (or everyday) validity (Groth-Marnat & Teal, 2000; Reinecke, Beebe, & Stein, 1999; Sbordone & Long, 1996). Knowing a test’s ecological validity is particularly important as referral questions are increasingly related to a client’s everyday levels of functioning (i.e., extent of disability, ability to function independently, everyday aspects of memory). Although the Wechsler scales have been correlated with other measures, including the Stanford-Binet and academic achievement, for the most part, there has been a notable lack of comparisons with behavior external to the scales themselves, despite the belief that many significant areas of a person, such as adaptive behavior, personal competence, and need for achievement, are separate (but related) constructs (Greenspan & Driscoll, 1997; Sternberg, 2003). In particular, the meanings associated with subtest scores should be investigated in more depth. For example, Picture Completion has traditionally been considered a measure of a person’s ability to distinguish relevant from irrelevant details in his or her environment, yet this assumption has not been adequately tested. Likewise, no studies have been made to determine if high or low Digit Span scores relate to actual day-by-day behaviors, such as recalling telephone numbers, facility with computer programming sequences, or following directions. EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 130 9/23/10 11:07:05 AM Assets and Limitations 131 An extension of this concern is that a number of authors have criticized what they believe is an overinterpretation of subtest and index scores (Glutting, Watkins, Konold, & McDermott, 2006; Konold, Glutting, McDermott, Kush, & Watkins, 1999). Specifically, they believe that individual subtest reliabilities are too low and not sufficiently specific for interpreting individual profiles. For example, they note that, compared with g (as represented by the Full Scale IQ), WISC-IV index scores do not account for a sufficient proportion of the variance in predicting achievement (Glutting et al., 2006). As a result, index interpretation does not demonstrate sufficient incremental increases in prediction. In addition, the ipsative patterns of subtest strengths and weaknesses are not sufficiently stable over time (Macmann & Barnett, 1997). Clinicians might therefore be advised to rely on the Full Scale IQ rather than index scores when making academic (and possibly other) predictions. Various authors counter this belief by emphasizing the importance of hypothesis testing, combining interpretations with external criteria, and noting the conceptual importance of the complexity of intelligence (A. Kaufman, 1994, 1999; A. Kaufman & Lichtenberger, 2000, 2002, 2006; Lezak, 1988, 2004; Milberg et al., 1996). There are several additional limitations to the Wechsler scales. Some critics believe that norms may not be applicable for ethnic minorities or persons from lower socioeconomic backgrounds (see next section, “Use with Diverse Groups”). In addition, the complexity of scoring, particularly the numerous calculations required for the Wechsler intelligence scales, is likely to increase the probability of clerical errors by examiners (Linger, Ray, Zachar, Underhill, & LoBello, 2007; Loe, Kadlubek, & Marks, 2007; Slate & Hunnicutt, 1988; Slate, Jones, & Murray, 1991). A further potential difficulty is that when supplementary subtests are substituted for core subtests, it is unclear how these supplementary subtests will affect the Full Scale IQ or index scores. As a result, supplementary subtests should be given only under unusual circumstances, such as when one of the core subtests has been “spoiled.” A further issue is that there is a certain degree of subjectivity when scoring many of the items on Comprehension, Similarities, and Vocabulary. Thus, a “hard” scorer may develop a somewhat lower score than an “easy” scorer. This is particularly true for Similarities, Comprehension, and Vocabulary, where scoring criteria are less clear than for other subtests. The Wechsler scales, like other tests of intelligence, are also limited in the scope of what they can measure. They do not assess some important factors, such as need for achievement, motivation, creativity, or success in dealing with people (Gardner, 2006; Sternberg, 2003). It should finally be noted that the WAIS-IV and WISC-IV have continued the traditional measurement of intelligence as represented by the Stanford-Binet scales and the earlier versions of the Wechsler scales. Although their revisions have provided features such as updated norms and index scores (especially the inclusion of Working Memory and Processing Speed), the underlying theories and essential construction of these scales have remained relatively unchanged for well over 50 years, despite numerous developments in both theory and measurement. These developments include the Cattell-Horn-Carroll (CHC) theory (see Flanagan & Kaufman, 2004), Luria’s PASS (Planning-AttentionSuccessive-Sequencing: Luria, 1980) model, Gardner’s independent competencies (Gardner, 2006), various theories on emotional intelligence (Bar-On, 1998; Ciarochi, Chan, & Caputi, 2000), and commonsense problem solving (Sternberg et al., 1995). Thus, one criticism of the Wechsler intelligence scales is that they have not responded to more current views on intelligence (A. Kaufman & Lichtenberger, 2002, 2006; Sternberg, 2003; Sternberg & Kaufman, 1998). It remains to be seen whether newer models and assessment tools will have much of an impact on assessing either intelligence or, especially, the frequency to which the Wechsler scales will be used in this process. EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 131 9/23/10 11:07:06 AM 132 Wechsler Intelligence Scales USE WITH DIVERSE GROUPS Each of the considerations discussed in Chapter 2 (see “Assessing Diverse Groups”) should be taken into account when evaluating a person from a diverse background. Doing this involves evaluating the client’s level of acculturation and language proficiency as well as the cultural competency of the examinee. In addition, a high degree of flexibility should occur both when conducting the assessment and when making interpretations. Different accommodations and strategies need to be made based on the outcome of these considerations. One of the key issues when assessing diverse clients is determining the extent that the instruments used might be biased. Evaluating for test bias can be partially informed by the considerable research devoted to evaluating the extent that intelligence tests such as the Wechsler intelligence scales are biased when used to assess various minority groups in the United States. The majority of these studies have been done with groups that are reasonably well acculturated and with moderate to good English skills. Reviews of this research have generally concluded (i.e., Kaplan & Sacuzzo, 2005; Sattler, 2008) that intelligence tests are not as biased as has been assumed. For example, deletion of items that appear to have biased content seems to make little difference in overall scores (i.e., Sandoval, 1979). Numerous validity studies have also found that intelligence tests make academic predictions as accurately for minority groups as for majority groups (i.e., Weiss, Prifitera, & Roid, 1993; Sattler, 2008). Factor-analytic research indicates that the same construct is being measured across various minority groups (i.e., Gutkin & Reynolds, 1981). Finally, Japanese populations, who come from a quite different culture than that of the United States, had mean scores that were actually higher than the U.S. standardization group (Lynn, 1977). Thus, research supports the use of the Wechsler intelligence scales for use with minority groups in the United States. The central problem does not seem to be the tests themselves. Although these tests are far from perfect, they provide the sort of information they were intended to provide. The main problem seems to be unequal opportunities that are accurately reflected in how various disadvantaged groups perform on intelligence tests. Despite the conclusion that cognitive tests generally measure what they intend to measure, clinicians still need to take extra care to insure that accurate data and conclusions are developed. These general and specific guidelines seem appropriate: • Make extra efforts to insure that clients feel comfortable and welcomed. • Make extra efforts to increase motivation; encourage clients to do their best. • Make sure that communication is as clear as possible, especially if there are differences in accents between the clinician and the client. • Resources beyond merely tests should have a greater significance with diverse clients (teacher reports, discussions with parents, history, behavioral observations) than for majority clients. • If language and culture appear to have been a factor in lowering the client’s performance, subtests that seem to be less influenced by culture and language should be the focus of interpretation. (Deemphasize language-based subtests such as Vocabulary and instead emphasize nonverbal tests such as Block Design, Matrix Reasoning, and Visual Puzzles.) • When assessing persons from cultures that deemphasize performing tasks rapidly (i.e., South Pacific islands), deemphasize speeded subtests (Processing Speed Index, Coding, Symbol Search, Cancellation). EBSCOhost – printed on 1/18/2023 7:01 PM via POST UNIVERSITY. All use subject to https://www.ebsco.com/terms-of-use c05.indd 132 9/23/10 11:07:06 AM Use with Diverse Groups 133 • Be cautious interpreting PRI