ABSTRACT
Discriminant Function Analysis (DFA), often just called discriminant analysis, is related to various multivariate techniques and could be covered just about anywhere in the first three parts of this book. However I have included it under analysis of variance since it is a logical follow-on from MANOVA. If MANOVA enables us to explore differences between groups, DFA allows us to predict group membership on the basis of the same dependent variables. In effect, it is MANOVA in reverse.
