The Base Rate Problem and Its Consequences for Interpreting Children's Ability Profiles

Abstract

Base rates refer to the proportion of a population that falls within a diagnostic category, either identifying an exceptionality (e.g., learning disability [LD], emotional disturbance [ED], or simply representing “normal” variation. This article familiarizes readers with the importance and meaning of base rates. It presents several univariate and multivariate base-rate procedures useful for identifying unusual IQ subtest variation. It compares the various base-rate procedures with the statistical significance-testing approach routinely used by psychologists. The mathematical superiority of one base-rate procedure is highlighted (i.e., the nonlinear multivariate baserate method), and its practical and scientific benefits are discussed. The nonlinear multivariate base-rate method is used to address the more important question of whether subtest analysis has validity for differential decision making. Specifically, the nonlinear multivariate method is employed to determine whether children with LD (N = 925) and ED (N = 100) are more likely to show unusual subtest patterns than children from the normative sample of the Wechsler Intelligence Scale for Children-Third Edition (N = 2,200). Results are discussed and recommendations are provided for improving future research on subtest analysis.