Misconceptions and misunderstandings (M & M) of exploratory factor analysis: Some clarifications

Exploratory factor analysis (EFA) is a popular statistical technique in research studies. In recent years, structural equation modeling (SEM) has become more popular. To distinguish itself from the measurement model of the SEM (i.e., confirmatory factor analysis), factor analysis is always referred to as EFA. However, a review of published articles using EFA demonstrates that some of the researchers, and even the reviewers, are bewildered with its usage and applications. For example, EFA (also known as common factor analysis) is always confused with principal component analysis (PCA). Henson and Roberts (2006) commented that PCA was often misused as a substitute or variant of EFA. Though both PCA and EFA are exploratory techniques that can be used to summarize the data and to test hypotheses (Haig, 2006), their usage and application are quite different in nature. The central idea in PCA is summarization. It is a data reduction procedure (i.e., to simply reduce a large number of items to a smaller number of underlying latent dimensions). Strictly speaking, PCA should be considered as"component analysis"- (Garson, 2012), but it is frequently mistaken as a form of factor analysis. In contrast, EFA is used to examine the factor structure or the pattern of relationships among variables. The main purpose of current article is to provide an overview of these two analytic methods and their applications along with some recommended practices.