INFORMING STYLISTIC LEARNING-BEHAVIOR, DISPOSITION, AND ACHIEVEMENT THROUGH ABILITY SUBTESTS - OR, MORE ILLUSIONS OF MEANING

This article reports on a series of empirical studies that assessed co ntinuing claims for the utility of subtest analysis. Large and represe ntative cross-samples were formed from the national standardization sa mples of the Wechsler Intelligence Scale for Children-Third Edition (W echsler, 1991), the Wechsler Individual Achievement Test (The Psycholo gical Corporation, 1992), and the Guide to the Assessment of Test Sess ion Behavior (Glutting & Oakland, 1993) (N = 640) and for the Differen tial Ability Scales (Elliott, 1990) and the Learning Behaviors Scale ( McDermott, Green, Francis, & Stott, 1996) (N = 1,250). Hierarchical re gression and discriminant models were used to determine the maximum po tential of ability subtests to explain variation in academic achieveme nt, stylistic classroom learning, and test-session behavior and dispos ition, and to distinguish among groups of exceptional children. Analys es demonstrated that ipsative subtest scores provide no information be yond that already available through conventional normative subtests. M oreover, even normative subtests add little information to what is kno wn through global ability measures. Implications are drawn for psychol ogical practice.