Confirmatory factor analysis of four general neuropsychological models with a modified Halstead-Reitan battery.

Abstract

Four theoretical factor models for a modified Halstead-Reitan battery were formulated, drawing from previous work by Swiercinsky, Royce and co-workers, Christensen and Luria, and Lezak. The relative explanatory power of these four models for this particular battery in an adult neuropsychiatric population was examined using confirmatory factor analysis. None of the models was shown to fit adequately in an absolute sense, but three of them represented substantial, statistically reliable improvements over a null model of mutual independence, and a clear pattern of relative fit was observed. Further improvements were achieved by modifying the best fitting initial model in several ways. A cross-validation with an independent sample supported the results of the model development step. Tentative theoretical and clinical implications for the overall organization of the neuropsychological abilities measured by this battery were drawn, and recommendations were made for further application of this method in neuropsychological research.