Most of the important models and techniques for analysis of covariance structures are reviewed and illustrated with a few simple examples. Some general approaches to the problems of identification, estimation and testing of covariance structures are considered. The ACOVS-model and the LISREL-model are treated in some detail. Specific types of covariance structures described are variance and covariance components models, measurement models, path analysis models and simplex models. Simultaneous analysis of mean and covariance structures is also considered as well as the simultaneous analysis of fata from several populations.
Abstract