We compared three approaches to defining disabilities on predictor developmental measures related to reading and writing and on criterion reading and writing measures in an unreferred sample of 300 first, second, and third graders. For each measure, the children who fell in the lowest 5% of the normal distribution (based on the mean and standard deviation for grade) were assigned to the low-functioning group; those children who, based on the Mahalanobis statistic (.05 level, one-tail test), were significantly different from and below their VIQ were assigned to the underachieving group; finally, children who met the criterion for both underachieving and low functioning were assigned to the learning disabled group. Of the low-functioning group, 44% were learning disabled (also discrepant from their VIQ). Of the underachievers, 36% were learning disabled (also very low functioning). Thus, elimination of IQ in the definition of learning disabilities (as Siegel [1989] and Stanovich [1991] have recommended) would dramatically affect who is identified as learning disabled.
We argue for a two-stage assessment model (cf. Leong, 1987) in which (a) absolute criteria without IQ are first used by classroom teachers to identify children who are low functioning in orthographic coding, phonological coding, word finding, fine motor, and visual-motor integration and who would benefit from early intervention to prevent reading and writing disabilities; and (b) relative criteria with IQ are subsequently used in more comprehensive psychoeducational assessment for learning disabled children whose problems persist despite early intervention. Based on our findings, we recommend that, in future research on reading and writing disabilities, investigators differentiate among low-functioning, underachieving, and learning disabled (low-functioning and underachieving) children and among component reading and writing skills.
Restricted accessResearch articleFirst published May, 1992pp. 74-94
