The critics are experts in the field of LD, especially reading, but they do not demonstrate expertise in the area of intellectual assessment and they are not current in their knowledge of contemporary instruments. They cite research to buttress their points, especially that IQ is unrelated to reading ability and to successful remediation, but they ignore other pertinent research that supports differences between individuals with LD versus low achievement (LA). These and related points are addressed here in summary fashion.
Selective Examination of Pertinent ReSEARCH. The critics do present an impressive compilation of research to support their claims that IQ does not discriminate between individuals with LD and LA in reading. However, they ignore other lines of research that support the opposite position. Kavale and Forness (2000), for example, cite several studies that show "that students with LD and LA could be reliably distinguished with the population with LD being the lowest of the low on the achievement distribution but equivalent to the LA population on the ability (i.e., IQ) distribution" (p. 248). Nicholson takes Stanovich and other antidiscrepancy theorists to task for (1) switching the focus from research on the causes of dyslexia to research on the causes of poor reading; and (2) treating the much acclaimed notion of phonological deficit as the cause, rather than as a symptom, of dyslexia. Nicholson (1996) provides a well-reasoned logical attack on Stanovich's line of thinking, strongly disputing his conclusions about the irrelevance of IQ for diagnosing reading disability as well as his dismissal of the term dyslexia. Furthermore, genetic research conducted with identical and same-gender fraternal twins, when at least one member of each pair was diagnosed with a reading disability (RD), found different heritabilities for individuals based on their WISC-R or WAIS-R Full Scale IQs (Wadsworth, Olson, Pennington, & DeFries, 2000). The heritability was .43 for those with IQs below 100 and .72 for those with IQs 100 and greater, a significant difference. Wadsworth et al. concluded: "The results of the current study support the hypothesis of a differential genetic etiology of RD as a function of IQ, suggesting that genetic influences may be more important as a cause of RD among children with higher IQ scores [They] suggest that knowing a child's IQ may tell us something about the causes of the reading deficit, which could possibly help focus intervention and remediation efforts" (p. 198).
A more complete picture of relevant research, one that extends beyond the litany of studies cited by Siegel, Stanovich, and others, provides a more balanced view of the issues and does not quickly condemn the IQ to irrelevance in the LD or RD assessment process. For example, in a recent neuropsychological investigation, poor readers did, indeed, differ from those with dyslexia (i.e., those with an IQ-achievement discrepancy) on phonological and related tasks—as Stanovich, Vellutino, and Siegel have claimed— but the dyslexics and poor readers differed significantly and notably on tests of static cerebellar function (Fawcett, Nicolson, & Maclagan, 2001). These investigators concluded from their findings, "that there are indeed theoretically valid reasons for distinguishing between poor readers with IQ discrepancy and those without" (p. 132).
IQ Is Not a Synonym for Wechsler. The critics unabashedly equate "IQ test" with "Wechsler test." Siegel's (1999) criticisms about bonus points penalizing individuals with slow, deliberate styles, and about the failure of IQ tests to measure problem-solving ability are all targeted at Wechsler's tests. Stanovich (1999) dismisses as unimportant the professional disagreements concerning whether verbal or nonverbal scales provide the best IQ criterion for a child assessed for possible LD as if Wechsler's nontheoretical dichotomy is the only available subdivision of intel ligence. These LD experts show no awareness of the many theoretically derived intelligence tests that offer more meaningful divisions of global IQ than Wechsler's armchair dichotomy, created more than 60 years ago.
Yes, Wechsler's tests are the most commonly used instruments for LD assessment in the United States, but that is not an acceptable reason for these LD professionals to fail to consider well-validated options in the face of their uncompromising dismissal of the IQ construct— broadly defined—from LD assessment. These leading spokespersons should consult the cross-battery approach for supplementing Wechsler's scales with tests that measure abilities not covered by the WISC-III or WAIS-III (Flanagan et al., 2000) or for focusing on a variety of alternatives to Wechsler's tests for learning disabilities assessment (Flanagan et al., in press). They should familiarize themselves with the growing number of alternative instruments, such as the CHC-based Woodcock-Johnson III Tests of Cognitive Ability (Woodcock, McGrew, & Mather, 2000), the Horn-Cattell-based KAIT (Kaufman & Kaufman, 1993), the Luria-inspired Cognitive Ability Scales (CAS; Naglieri & Das, 1997), the Differential Ability Scales (DAS; Elliott, 1990), to name a few of the more recent Wechsler alternatives.
Kaufman and Kaufman (2001a), citing the fluid and planning abilities measured by the aforementioned tests, as well as by neuropsycho-logical tests, state: "Consider Siegel's (1999) criticism that IQ tests fail to measure reasoning or problem-solving skills. If one departs from the Wechsler system and examines the available well-constructed, well-designed, theory-driven test batteries (both cognitive and neuropsycho-logical), one finds an abundance of scales or subtests that measure the kinds of abilities that Horn would classify as fluid and Piagetians would consider dependent on formal operational thought" (p. 442). They add, on a related topic:
In addition to the Wechsler tests' shortage of high-level reasoning tasks, the channels of communication mea sured by the various Wechsler subtests fall into one of only two categories: auditory-vocal (Verbal subtests) and visual—motor (Performance subtests). These are important channels, but clinicians who evaluate individuals suspected of SLD will often benefit by assessing other channels of communication. For example, the K-ABC includes subtests for school-age children within the auditory-motor channel (Word Order) and the visual-vocal channel (Gestalt Closure, Faces & Places) as well as the two channels measured by Wechsler's scales Similarly, the WJ-R... includes two visual-vocal subtests (Picture Vocabulary, Visual Closure) and the CAS...includes an auditory-motor subtest, Verbal-Spatial Relations. (Kaufman & Kaufman, 2001a, pp. 442-443)
For an in-depth treatment of these and other alternative instruments for use in SLD assessment, with most chapters written directly by each test's authors, consult Kaufman and Kaufman (2001b). For a thorough discussion of the application of the cross-battery approach to enhance greatly the IQ's validity for predicting achievement and for its utility for issues concerning LD assessment in general, consult Flanagan's texts (Flanagan, McGrew, & Ortiz, 2000; Flanagan & Ortiz, 2001; Flanagan et al., in press).
IQ AS A Measure OF g. The array of studies used to criticize the IQ construct for LD assessment treats IQ as if it is Spearman's (1904) century-old "g" factor, supported by some researchers in the present (most notably, Jensen, 1998), but contrary to most modern theories of intelligence such as Horn's (1989) expansion and elaboration of the Horn-Cattell Gf-Gc theory, or the Luria-based PASS model that Naglieri and Das (1997) used as the foundation of the CAS (Naglieri, 1999). The LD experts show no awareness that the Full Scale IQ is the least interesting and least valuable score yielded by IQ tests. Horn's expanded theory, as well as its integration with Carroll's (1993, 1997) theory to form Cat-tell-Horn-Carroll or CHC theory (Chapter 14), has been quite influential for the development of new and revised intelligence tests (Kaufman & Kaufman, 1993; Woodcock & Mather, 1989; Woodcock, McGrew, & Mather, 2000) and has greatly impacted interpretation of Wechsler's scales (Flanagan et al., 2000; Kaufman, 1994a; Kaufman & Lichtenberger, 1999, 2000).
Consider the various studies cited by Siegel
(1999), Stanovich (2000), and Vellutino et al.
(2000) that feature Full Scale IQ's dismal failure at discriminating among reading groups or remediation groups. Might these groups have differed on other IQ-related scales, such as the Attention or Planning Scales on the CAS, the K-ABC Sequential or Simultaneous Processing Scales, the KAIT Fluid Scale, the DAS Nonverbal Reasoning Ability Scale, or any of the seven Horn-based clusters that comprise the WJ III Tests of Cognitive Ability? Might they have differed on "new and forthcoming neuropsycho-logically-based instruments that are designed to go beyond conventional profiles of scores on IQ tests, such as the WISC-III as a Process Instrument (WISC-III—PI) or the Delis-Kaplan Test of Executive Functions"? (Kaufman & Kaufman, 2001a, pp. 445-446).
Stanovich (1999) states that, "Intelligence has played a major role in the conceptual muddle surrounding the notion of reading failure. The confusion arises because it makes no sense to say that low intelligence... causes reading difficulties, given what is currently known about reading disabilities" (p. 352).
Eliminating the IQ-Achievement Discrepancy Does NOT Mean Throwing Out IQ TESTS. The IQ critics join two issues at the hip when they are quite separate: eliminating the IQ-achievement discrepancy from the LD definition and jettisoning IQ tests from the entire LD assessment process. These topics are not cause-effect, even if the critics act as if deleting the discrepancy criterion from the definition leads directly to getting rid of the IQ test from the assessment process.
It is easy to see why the anti-IQ LD experts have no need for the IQ. Apart from their equation of IQ test with Wechsler test, and their failure to appreciate the last two decades' growth, both in the publication of new and revised IQ
alternatives and in theory-based approaches to clinical interpretation and test selection, they no longer view LD as anything but an achievement deficit.
In fact, there are some excellent reasons to eliminate the discrepancy criterion from the LD definition. For example, as Shepherd (2001) indicates from her own clinical experience and that of her colleagues, data are frequently ignored when diagnostic decisions are made, replaced by practical variables such as available resources and the needs of the parents and teachers. MacMillan and Speece (1999), in a review of three studies conducted after PL 94-142 was enacted, discovered that more than 50% of the students in each study identified with SLD failed to meet relevant diagnostic criteria; they concluded that tests were given mostly to conform to legal requirements, but data from the tests were often ignored. Similarly, Kavale and Forness (2000) reviewed several other pertinent studies that found that "large-scale investigations of LD populations show that only about 50% of students actually classified as LD demonstrate a significant aptitude-achievement discrepancy" (p. 249). Kaufman and Kaufman's (2001a) response to the real-life practices:
What a waste! Why bother having trained psychologists administer 90-minute IQ tests and have other professionals administer time-consuming achievement, adaptive behavior, or processing tests, if these measures are just given so the professionals can cover their own backs? One does not need to weigh the carefully-reasoned (though occasionally flawed) arguments of Siegel (1999), Stanovich (1999), or Vellutino et al. (2000) against the use of the IQ-achievement discrepancy for SLD diagnosis. Their attacks on IQ tests, however motivated, are far less impressive evidence for abandoning the IQ-achievement discrepancy than are the apparent everyday realities of differential diagnosis: The discrepancy is often not used when diagnosing SLD, even though the pertinent test data are invariably obtained. Given the realities of clinical practice, at least in schools, why not delete the IQ-achievement discrepancy from the definition of SLD? (p. 437)
But giving up the discrepancy criterion does not mean abandoning IQ tests from the LD diag nostic process. Nor does it mean giving up the concept of discrepancy, a notion that is intricately woven into the LD fabric at many levels of the definition (Kavale & Forness, 2000; Mather & Woodcock, 2001; Nicholson, 1996), even if it means giving up a formulaic, rigid approach to discrepancies. Kaufman and Kaufman (2001a) state:
The acceptance of error as a necessary prerequisite for measuring IQ, embraced by David Wechsler,... has persisted from one century to the next and continues into the new millennium. Error is a fact of assessment life, a fundamental tenet of a psychologist's clinical training, and antagonistic to the use of any discrepancy formulas or cutoff points One does not have to be a special educator or learning disabilities specialist to criticize the psychologist's tools. We have historical reasons to acknowledge—even embrace—their limitations. One does not need to read Siegel's (1999) or Stanovich's (1999) criticisms of IQ tests to discover that these measures are imperfect. .. . Verbal tasks overlap with the content of achievement tests; process deficits are just as likely to impair performance on IQ tests as on tests of academic skills; neither verbal nor nonverbal measures of IQ are necessarily better or more valid (or valid at all) of the intelligence of an individual with SLD; IQ does not effectively provide a measure of a person's potential; and so forth... . The real problem resides in the federal and state guidelines that mandate the use of these formulas (even if their use is illusory in many real-life situations). (p. 439)
One needs to be creative, and function as a blend of a psychometrician and clinician, to identify appropriate discrepancies between ability and achievement, discrepancies that highlight cognitive strengths as well as deficient basic cognitive processes that contribute to academic failure. Consult the innovative and clever approaches to computing ability-achievement discrepancies described by Mather and Woodcock (2001), as well as the in-depth treatment of the topic by Flanagan and her colleagues (Flanagan et al., 2000, Appendix H; Flanagan et al., in press).
Perhaps the best conceptual argument for keeping the notion of discrepancy in the definition of LD and SLD comes from M. J. Shepherd (personal communication, October 14, 1999), who was Nadeen Kaufman's mentor in the Learning Disabilities program at Teachers Col lege, Columbia University, and who was directly responsible for the intra-individual approach to Wechsler interpretation that Alan Kaufman adopted for his intelligent testing approach:
Siegel and Stanovich's claim that phonological reading disability occurs at all IQ levels is "déjà vu all over again"—Cruikshank, Kephart et al. claimed that specific learning disability (meaning visual-perceptual deficit) occurred at all IQ levels If we accept the hypothesis that mental activity is specific (unique) to the task being performed, it makes sense that all children having difficulty with a particular task (word recognition and spelling) will have similar cognitive deficits. This means (to me) that we will not achieve a full understanding ofspe-cific learning disabilities by looking at deficits alone. In neuropsychological terms we have to document the cognitive traits that have been "spared. " This is the point that Stanovich and Siegel aren't making because (a) they insist on working with a limited conception of reading (word recognition) and/or (b) they have a political agenda— protect the poor against the rich.
The task for professionals involved with the assessment of LD or SLD, whether they are in psychology or special education, is to embrace new research programs (see, for example, Nicholson, 1996) and to make use of novel and better approaches to assessment as part of the diagnostic process. The definition and diagnosis of LD, SLD, dyslexia, and so forth, are complex; the disorders are real, and qualitatively—not quantitatively—different from low achievement; the answers to the pressing practical and theoretical issues facing the field require careful study, not simple, knee-jerk solutions.
Was this article helpful?
How Would You Like To Amaze People With Your Intelligence? Increase your IQ and get prepared to receive accolades in every sphere of life. Do you feel dejected every time your boss praises a colleague for an intelligent professional move? Do you want to become a crucial resource to your company?