ABSTRACT
This chapter reviews two broad categories of statistical methods for multitrait—multimethod (MTMM) data: the analysis of variance (ANOVA), and factor analysis. The MTMM matrix assembles the correlations among different traits, each measured using the same set of methods. By factorially combining traits and methods, Campbell and Fiske hoped to learn about the convergence of trait influences across methods, and the dominance of method influences across traits. The broad impact of Campbell and Fiske attests to the appeal of this idea. Any MTMM investigation begins with some notion about method influences. The ANOVA was one of the first methods proposed for the analysis of MTMM data. MTMM data form a factorial design with three factors: person, trait, and method. As many as seven effects are estimable in this three-way design (three main effects, three two-way interactions, and a three-way interaction). Only a few of these are of interest in practice.
