ABSTRACT
Principal component analysis (PCA) is not factor analysis. PCA is a technique for data reduction, whereas factor analysis uses latent variables and can be used to identify the structure of measures/constructs or for data reduction. Many people use PCA when they should use factor analysis instead, such as when they are assessing latent constructs. Nevertheless, factor analysis has weaknesses including indeterminacy—i.e., a given data matrix can produce many different factor models, and you cannot determine which one is correct based on the data matrix alone. There are many decisions to make in factor analysis, including, (1) what variables to include in the model and how to scale them, (2) the method of factor extraction, (3) the kind of factor analysis: exploratory (EFA) or confirmatory (CFA), (4) how many factors to retain, (5) if EFA, whether and how to rotate factors, and (6) model selection and interpretation. These decisions can have important impacts on the resulting solution. Thus, it can be helpful for theory and interpretability to help guide decision making when conducting factor analysis.
