Exploratory Factor Analysis
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
Wedel, M. and Shi (Savannah), W. 2010. Exploratory Factor Analysis. Wiley International Encyclopedia of Marketing. 2.
- Published Online: 15 DEC 2010
Factor analysis is a group of methods that enjoys high popularity in marketing research. We either use it as a data-reduction tool in the domain of data mining, or as a latent structures detection tool based on subjective judgments, and in perceptual mapping. In this article, we introduce two basic factor analysis methods: principle component analysis (PCA) and factor analysis in the strict sense, using satisfaction survey data of an administrative computer system as an example. Specifically, PCA is a formative approach that “forms” the components from the observed variable. It is a data analysis or data-mining tool that reduces the number of variables and explores meaningful patterns. We discuss “loadings,” “eigenvalues,” “communality,” “component score,” and “biplot” in the output of PCA, and also provide some guidelines to determine the number of principal components. Factor analysis, as a reflective approach, assumes that the unobserved factors are reflected in the observed variables. It differs from PCA in that it is not a data-reduction tool, but a statistical model. It identifies latent factors under observed variables, and specifies a relationship between them with error terms. We briefly introduce the estimation and rotation methods of factor analysis. Software available for PCA and factor analysis is also discussed at the end of this article.
- factor analysis;
- principal components;
- biplot, scree plot;
- latent variables;