Sensitivity and specificity of WAIS–III/WMS–III demographically corrected factor scores in neuropsychological assessment

Abstract

This study explored the neurodiagnostic utility of 6 factor scores identified by recent exploratory and confirmatory factor analyses of the WAIS–III and WMS–III: Verbal Comprehension, Perceptual Organization, Processing Speed, Working Memory, Auditory Memory and Visual Memory. Factor scores were corrected for age, education, sex and ethnicity to minimize their influences on diagnostic accuracy. Cut-offs at 1, 1.5 and 2 standard deviations (SDs) below the standardization sample mean were applied to data from the overlapping test normative samples (N = 1073) and 6 clinical samples described in the WAIS–III/WMS–III Technical Manual (N = 126). The analyses suggest that a 1 SD cut-off yields the most balanced levels of sensitivity and specificity; more strict (1.5 or 2 SD) cut-offs generally result in trading modest gains in specificity for larger losses in sensitivity. Finally, using combinations of WAIS–III/WMS–III factors together as test batteries, we explored the sensitivity and specificity implications of varying diagnostic decision rules (e.g., 1 vs. 2 impaired factors = “impairment”). For most of the disorders considered here, even a small (e.g., 3 factor) WAIS–III/WMS–III battery provides quite good overall diagnostic accuracy. (JINS, 2001, 7, 867–874.)