Are IQ scores valid for children who are poor readers?

Abstract

Remedial services for children with reading problems are often allocated according to discrepancies between reading and IQ scores. Results of some recent research suggest, however, that IQ scores of poor readers do not covary with their levels of functioning in other cognitive domains. This study evaluated whether the external validity of IQ scores (from the Wechsler Intelligence Scale for Children—Revised) was moderated by reading levels within 2 separate samples of referred children. We found that IQ scores had expected correlations with external measures of verbal, visual-spatial, short-term memory, and arithmetic ability, and that these relations were invariant across levels of reading skill. Many school-age children have reading difficulties, but only some may be classified as reading disabled, which indicates a discrepancy between specific reading proficiency and general cognitive ability. Children so classified may be eligible for general learning disability services or for more specialized programs for the reading disabled. Poor readers who do not meet ability-achievement discrepancy definitions of reading disability may receive other remedial services but, in some instances, may receive none at all. For example, poor readers with IQ scores in the borderline range (e.g., 70-79 on Wechsler scales) may be ineligible for special education services because they are considered to be slow learners. Poor readers with even lower IQ scores (e.g., 60-69), however, may be eligible for placement in classrooms for the educable mentally impaired, where they may at least receive more individual attention than slow learners who remain in regular classrooms. Considering the implications for children's educational careers, it is crucial to demonstrate the validity of allocating special education resources on the basis of discrepancies between reading and general ability levels. There have been two types of criticisms about this method of allocation, however, the second of which provides the main focus of this study. First, as with the broader notion of learning disability, there are numerous conceptual and statistical problems with the operational definition of significant ability-reading discrepancies (e.g., see Kamphaus, Frick, & Lahey, 1991; Reynolds, 1984-1985). Second, there has been much debate about whether IQ tests provide valid estimates of general cognitive capacity for children with reading problems. For example, some researchers have argued that poor readers may obtain artificially low IQ scores because