ABSTRACT
Factor analysis is a widely used set of techniques in the behavioral sciences. It is also a primary technique for many researchers, especially those who conduct assessment-related studies. The basic logic and mathematics of factor analysis were first described by Charles Spearman (1904b), and many variations of factor analysis were developed over the following century. Factor analysis is unique among multivariate statistical procedures in that it was developed mainly by psychologists in order to test hypotheses about the correspondence between scores on observed (manifest) variables, or indicators, and hypothetical constructs (latent variables), or factors, presumed to affect those scores. Spearman and his contemporaries (e.g., Thomson, 1920) used factor analysis to evaluate models about the nature and organization of intelligence. Factor analysis is still widely used today in mental test studies as it is in many other research areas as a means to discover and identify latent variables, given initially only sample covariances among a set of indicators (Mulaik, 1987).
