ABSTRACT
The expansion from a single dependent variable to multiple dependent variables necessitates a Multivariate Analysis of variance (MANOVA) analysis. The relatively simple MANOVA has one independent variable, with three levels, and three dependent variables. The assumptions for multivariate ANOVA are analogous to those for univariate ANOVA, except that they apply to the multivariate situation. With three dependent variables, the null hypothesis is a bit different than when a single dependent variable is involved. Morrow and Frankiewicz present strategies for the post hoc procedures in studies of human performance. Other post hocprocedures that a researcher might pursue after finding a significant lambda value include discriminant analyses and stepdown analyses. If two or more independent variables are involved, the researcher must examine interactions among the independent variables before interpreting main effects. For ANOVA, the final test statistic is the F-ratio. For MANOVA, several test statistics result.
