ABSTRACT
This chapter focuses on covariance component analysis (CCA) as a relatively new multivariate statistical model for analyzing the multitrait-multimethod (MTMM) matrix. Covariance component analysis (CCA) was proposed by Bock and Bock and Bargmann as a multivariate random model for factorial measurement designs. A successful early application by Bock, Dicken, and Van Pelt investigated effects of method-related acquiescence components in the Minnesota Multiphasic Personality Inventory (MMPI) scale scores. Covariance component analysis was developed along the tradition of main effects analysis of variance (ANOVA) models, in which the trait × method observations are described in terms of a linear function of a general term, trait effects, method effects, and residual error. Covariance component analysis employs multivariate analysis of variance (MANOVA) terminology to analyze factorial measurement designs. The method encompasses an additive, multivariate, random effects model—decomposing an individual's measurements into a general component, trait profiles, and method profiles.
