Principal component analysis (PCA) is a term that is often used interchangeably with factor analysis. While both approaches aim to simplify the structure of a set of variables and the analyses are structured in similar ways, they are not exactly the same thing. PCA performs data reduction by using a linear combination of a set of variables, in order to create one or more index variables (components). Factor analysis is modelling the measurement of a latent (i.e. unobserved) variable.

To make it even more confusing, many statistical programs (e.g. SPSS) apply PCA as the default estimation method for factor analysis. In Stata, PCA is not default (but an option). Rather, Stata uses the principal-factor method (pf) to analyse the correlation matrix. When the principal-component factor method (pcf) is used, the communalities are assumed to be 1.