Analyzing WISC-III Profiles: A Comparison of Two Approaches

Abstract

The effects of the use of a constant difference score (Kaufman, 1979, 1994) to identify subtest scores that were significantly different from the mean score on the Verbal, Performance, or Full Scale of the WISC-III were compared to a more complex approach that employed tables from Sattler (1992) and Wechsler (1991). It was hypothesized that the use of a specified number of points as indicative of a significant discrepancy, as suggested by Kaufman (1994), would result in over- or underinterpretation because the minimum deviation required for statistical significance depends on the intercorrelations of the subtests and on the number of subtests in the composite (e.g., Silverstein, 1984; Willson, Stanton, & Olivarez, 1989). Analysis of the WISC-III profiles of 392 participants, who ranged in age from 6 to 16 years, generally supported this hypothesis. The difference scores on which the over- and underinterpretation were based occurred with relative frequency in the standardization sample. Implications were discussed.