The intent of this paper is to explain nonmathematically the appropriate use and interpretation of canonical discriminant function analysis. The two primary purposes for which this multivariate statistical technique are employed are elucidated as (1) exploratory dimension-reduction, and (2) the testing of typological hypotheses. A heuristic discussion of the technique is provided emphasizing applied rather than theoretical issues presented via an empirical question relevant to social work. A detailed explanation of how the question was answered is provided in a discussion of the output from the SPSSx statistical package. Throughout the paper, the advantages and disadvantages of doing a discrminant function analysis on both SPSSx and SAS are explicated by an ongoing comparison of each package's features.
No takes yet. Share an insight, caveat, or question.
Carmelo Cocozzelli (1987) studied this question.