Construct Validity in Personality Assessment

Abstract

In recent years, the conceptualization and assessment of validity in psychological measurement has been transformed. Where one could formerly denote various types of validity (i.e., construct, content, and criterion validity), there is a now widespread understanding that construct validity subsumes all that was previously seen as disparate (Guion, 1980; Messick, 1980). Where convergent and discriminant aspects of construct validity were inferred from inspection of a correlation matrix (Campbell {&} Fiske, 1959), there are now various quantitative procedures, including factor analytic (e.g., Jackson, 1975) and structural-equation models (e.g., Judd, Jessor, {&} Donovon, 1986) to aid in the interpretation of research outcomes.