Discriminant Analysis for Marketing Research Applications
Part 2. Marketing Research
Published Online: 15 DEC 2010
Copyright © 2011 John Wiley & Sons, Ltd. All rights reserved.
Wiley International Encyclopedia of Marketing
How to Cite
Feinberg, F. M. 2010. Discriminant Analysis for Marketing Research Applications. Wiley International Encyclopedia of Marketing. 2.
- Published Online: 15 DEC 2010
Among marketers' main tasks is segmentation: breaking consumers, products, and firms into meaningful groupings. Marketing data often appear in discrete buckets, like “light,” “medium,” and “heavy” users, and marketers need to understand and predict which consumers (or products, or firms, etc.) fit into which group. That is, they would like to use available information (predictor variables) help explain how, and perhaps why, these groupings come out the way they do. Common examples include distinguishing new versus returning customers, explaining which stores different consumers chose to shop at, using demographics to predict various kinds of shopping behavior, or how to entice “loyal” versus “switcher” customer groups. Multiple discriminant analysis (MDA) allows marketers to do several important things: distinguish among two or more known groups, using available predictor variables; classify new items into those known groups; verify whether there actually are significant differences across the groups; and test for which specific predictor variables best account for between-group differences. We illustrate the meaning and use of the linear discriminant function in marketing applications, as well as how managers can interpret DA output to make better segmentation decisions. Connections to other methods included in this volume are highlighted throughout.
- cluster analysis;
- discriminant analysis;
- marketing research;
- statistical methods