TEACHING ARTICLES Repairing Tom Swift’s Electric Factor Analysis Machine

Proper use of exploratory factor analysis (EFA) requires the researcher to make a se-ries of careful decisions. Despite attempts by Floyd and Widaman (1995), Fabrigar, Wegener, MacCallum, and Strahan (1999), and others to elucidate critical issues in-volved in these decisions, examples of questionable use of EFA are still common in the applied factor analysis literature. Poor decisions regarding the model to be used, the criteria used to decide how many factors to retain, and the rotation method can have drastic consequences for the quality and meaningfulness of factor analytic re-sults. One commonly used approach—principal components analysis, retention of components with eigenvalues greater than 1.0, and varimax rotation of these compo-nents—is shown to have potentially serious negative consequences. In addition, choosing arbitrary thresholds for factor loadings to be considered large, using single indicators for factors, and violating the linearity assumptions underlying EFA can have negative consequences for interpretation of results. It is demonstrated that, when decisions are carefully made, EFA can yield unambiguous and meaningful results. exploratory factor analysis, principal components, EFA, PCA Exploratory factor analysis (EFA) is a method of discovering the number and na-ture of latent variables that explain the variation and covariation in a set of mea-sured variables. Because of their many useful applications, EFA methods have en-joyed widespread use in psychological research literature over the last several decades. In the process of conducting EFA, several important decisions need to b