Factor comparison across studies

The need to compare the results of factor analysis across studies has become more necessary as substantive and instrumentation studies are replicated across populations. The purpose of this article is to provide a review of four procedures, which utilize factor loadings, to compare factors across studies when the same design and variables are applied to different populations. The techniques described are the Salient Variable Similarity Index (S), the Coefficient of Congruence (CC); the Pearson Correlation Coefficient (r); and, Confirmatory Factor Analysis using latent variable modeling procedures (EQS). The disagreement over use of comparison procedures using factor loadings and those using factor scores is briefly described.