WISC-R subtest variability in a clinic-referred sample of Canadian children.

Abstract
The degree of WISC-R [Wechsler Intelligence Scale for Children Revised] subtest variability was investigated in a clinic-referred group of 2000 Canadian children between the ages of 6-0 and 15-11 yr. Comparisons were made with normal Canadian children and the American WISC-R standardization sample. A 1-sample Hotelling T2 test was employed to examine subtest variability with reference to the scale grand means, and a series of MANOVA [mean analysis of variance] examined the effects of Full Scale IQ, age, and sex and their interactions. The discrepancy index, range index and deviation index of subtests were also calculated. Subtest variability comparisons with the American normative population and samples of normal Canadian children revealed a few distinct differences in degree and pattern of scatter on selected indices. The relationship of these findings to the literature on WISC-R subtest variability is discussed.