Discriminant Analysis

by William R. Klecka

These procedures, collectively known as discriminant analysis, allow a researcher to study the difference between two or more groups of objects with respect to several variables simultaneously, determining whether meaningful differences exist between the groups and identifying the discriminating power of each variable.

72 pages, Paperback

First published August 1, 1980

Reviews

[Lawrence Linnen · Rating 5 out of 5]

Klecka presents an introduction to several related statistical procedures known as discriminant analysis, which is a technique for examining differences between two or more groups of objects with respect to several variables simultaneously. The book introduces canonical discriminant functions, classification functions and procedures, and selection criteria for the inclusion of variables in discriminant analysis. Klecka derives canonical discriminant function coefficients, provides a spatial interpretation of them, and discusses the interpretation of canonical discriminant functions. Unstandardized and standardized coefficients are discussed, as well as procedures to determine how many discriminant functions are significant. He includes a discussion of the violation of the assumptions which underlie discriminant analysis.

[Steven Peterson · Rating 3 out of 5]

In the late 1980s and early 1990s, I used discriminant analysis quite a bit. It was a helpful method of finding out which variables "discriminated" between two groups. One simple example (Page 5) is the value of this statistical technique to "isolate variables which discriminate among citizens who will vote for Democreats versus Republicans. . . ." Since then, I have come to prefer logistic regression. However, while using this technique I did come to appreciate its value. This book was a useful aid to me as I worked through interpreting results. For those who choose to use discrimnant analysis, this volume--while dated--can be a welcome resource.