MODELING MTMM DATA FROM ADDITIVE AND MULTIPLICATIVE COVARIANCE-STRUCTURES - AN AUDIT OF CONSTRUCT-VALIDITY CONCORDANCE

The purpose of this study was to determine the extent to which finding s derived from four approaches to MTMM analyses were consistent in pro viding evidence of construct validity related to the measurement of fo ur dimensions of perceived competence (social, academic, English, math ematics) across four maximally dissimilar rating methods (self, teache r, parent, peer). MTMM methodological approaches included that of Camp bell-Fiske (1959), the general confirmatory factor analytic (CFA) mode l (Joreskog, 1969), the CFA (Correlated Uniqueness) model (Marsh, 1988 ), and the Composite Direct Product model (Browne, 1984). Procedures w ere applied to data from 158 grade 11 high school adolescents. Advanta ges, disadvantages, findings, and caveats related to each procedure ar e discussed.