ABSTRACT
Applications of factor analysis arise in contexts where there are p dependent response variables and we believe that the dependencies among these variables can be explained by q latent variables where q ≪ p. By a latent variable we mean that the values of the variable are not observed. A factor analysis is a methodology for identifying a value of q that accounts for a meaningful fraction of the variation in the observed responses, provides estimates of the dependencies of the response variables on the latent variables and also estimates of the values of the latent variables. The latent variables may or may not have meaning in light of the application.
