Multivariate Analysis of Variance (MANOVA) and Discriminant Analysis

Multivariate analysis of variance (MANOVA) can be considered an extension of the analysis of variance (ANOVA). MANOVA can feature more than a single independent variable, and the researcher can also hypothesize interactions among categorical independent variables on the hypothesized dependent linear combination. Moreover, researchers often wish to include one or more covariates in a MANOVA in the same spirit as one would do in analysis of covariance (ANCOVA), making the model a multivariate analysis of covariance. The chapter demonstrates how to run and interpret a MANOVA using SPSS. It then demonstrates how to perform a discriminant analysis, which is the “reverse” of MANOVA. The researcher can obtain Box's M test for the MANOVA through Homogeneity tests under Options. The chapter discusses Box's M test more extensively in the context of discriminant analysis shortly.