ABSTRACT
In all of the previous chapters in this book, the examples involved only a single dependent variable. In this chapter, analyses that have more than one dependent variable are considered; such analyses are termed multivariate, whereas tests involving but a single dependent variable will here be called univariate. An analysis is only considered multivariate when multiple dependent variables are analyzed simultaneously. Although it is possible to analyze the multiple dependent variables univariately, that is, one at a time (say, with multiple sets of “MANOVA” commands), it is far more efficient to analyze all the dependent variables multivariately, that is, all at once. Doing this allows you to test with a single test (the multivariate ANOVA or MANOVA), whether the groups differ on the entire set of dependent variables taken together. Fortunately, MANOVA and GLM also automatically provide each of the univariate tests as well, the ones evaluating whether each specific dependent variable is significant.
