ERIC Number: ED395041
Record Type: RIE
Publication Date: 1996-Jan-27
Reference Count: N/A
Stepwise Methods Are as Bad in Discriminant Analysis as They Are Anywhere Else.
Moore, James D., Jr.
The serious problems associated with the use of stepwise methods are well documented. Various authors have leveled scathing criticisms against the use of stepwise techniques, yet it is not uncommon to find these methods continually employed in educational and psychological research. The three main problems with stepwise techniques are: (1) computer packages use the wrong degree of freedom in their computations, producing spuriously statistically significant results; (2) stepwise methods capitalize outrageously on sampling error and therefore yield nonreplicable results; and (3) they do not identify the best set of predictors. As the literature already contains several examples of the misuse of stepwise methods in the case of regression, the present paper explains the problems associated with their use in the context of discriminant function analysis. It is suggested that these methods are equally as bad in multivariate statistics as they are in a univariate context and therefore should be avoided entirely. (Contains four tables and nine references.) (Author/SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A
Identifiers: Research Replication; Stepwise Regression
Note: Paper presented at the Annual Meeting of the Southwest Educational Research Association (New Orleans, LA, January 25-27, 1996).