SLD and Unexpected School Failure
The RTI approach to SLD identification essentially eliminates the notion of SLD as unexpected learning failure in the presence of adequate general cognitive ability. Without an assessment of general cognitive ability, the "unexpected" notion would be impossible to document. Although underachievement defined by a discrepancy appears integral to the SLD concept, the field appears to be moving away from the idea of unexpected learning failure with the mantra that IQ is not necessary in the identification of SLD. This suggestion is reinforced by the idea that there is no need to distinguish between high ability - low achieving versus low ability - low achieving students because of small differences found between high and low ability groups on measures related to reading and behavior (Vellutino, Scanlon, & Lyon, 2000). But it would be illogical to assume that students with low ability (meaning they tend to score low on nearly every assessment) would in some way outperform students with SLD (i.e., high ability) on measures of phonological processing, rapid-automatic naming, verbal memory, vocabulary, or any reading ability measure. Similarly, students with problems learning to read will by definition score low on reading related measures.
A low ability - low achievement group represents an inappropriate comparison for research investigating the nature of SLD. An SLD group, by definition, does not function in the low average to borderline IQ range. For example, if an IQ cut-off score is set too low, then there are difficulties in distinguishing IQ-discrepant from non-discrepant students. Stuebing et al. (2002) showed a modest difference (ES = .30) in general cognitive ability between IQ discrepant and low ability groups primarily because they were not really identifying different groups. The reluctance to use full scale IQ scores has also led to the inappropriate use of partial IQ test batteries. For example, using only visual-perceptual measures to measure IQ (known to be unrelated to reading) leads to the self-fulfilling prophecy that no relationship exists between IQ and reading. In another inappropriate practice, a limited number of subtests from intelligence assessments are given and full-scale IQ scores are pro-rated. In this case, no IQ effects are found primarily because IQ has not actually been measured (Hale & Fiorello, 2001).
A more useful comparison would be to evaluate low ability students who receive a diagnosis of SLD versus those who do not receive such a designation. Such a comparison is likely to demonstrate that important ability differences account for the adequate versus inadequate academic development between these groups. Another potentially useful comparison would be between achievement discrepant students (i.e., LA) diagnosed with RD versus students with an IQ-achievement discrepancy who are not diagnosed. Ultimately, the goal of these comparisons is to uncover factors that lead to a valid diagnosis; only then will we better understand the complex interactions that shape the nature of SLD.
The demonization of IQ has led to the suggestion that IQ is unrelated to academic achievement. This is patently not the case (Nagleri, 2001) and it has been clear for a long time that intelligence tests are more highly correlated with scores on achievement tests than they are with grades given by teachers (Donahue, Coombs, & Travers, 1949). For example, Siegel (2003), believing that IQ was not related to reading, chose a specific measure of intelligence that was previously found to be not related to reading. Not surprisingly, Siegel's investigation found no relationship between IQ and reading. Siegel had earlier rejected a measure like verbal intelligence because of its overlapping variance with reading even though Konold (1999) showed that the best predictors of reading achievement were the Verbal Comprehension and Freedom from Distractibility factor scores from the WISC-III. In a subsequent article, Siegel reported that IQ was, in fact, associated with reading even though the influence of IQ was reduced because of its use as a grouping rather than continuous variable (see Jimenez, Siegel, & Lopez, 2003). In yet another study, to prove no relationship between IQ and reading, D'Angiulli and Siegel (2003) used an outdated version of an IQ test (i.e., WISC-R) suggesting that the desire to prove IQ irrelevant for SLD identification far outweighed the tenets of sound experimental design.
When placed in proper perspective, it is possible to conclude that discrepancy is really not problematic and can be included as a variable in making eligibility decisions (Kavale, 2002). When additional factors are also included in the eligibility decision, the number of false negative and false positive diagnoses would be greatly reduced. What is also clear is that eliminating IQ-achievement discrepancy would result in a significant number of students with SLD not being identified when using only a relative discrepancy or low achievement criterion for determining eligibility. Thus, discrepancy remains useful as a fundamental element in SLD identification and discussion about its demise are simply unwarranted (see Aaron, 1997).
Previous Page | Next Page
(SLD & Intelligence) | (SLD Diagnosis)
