ABSTRACT
Latent variables analysis is a powerful and useful tool which has languished too long in the shadowy regions on the borders of statistics. It is now taking its place in the main stream, stimulated in part by the recognition that it can be given a sound foundation within a traditional statistical framework. To express all latent variable models within a common mathematical framework appears to have been that of Anderson. Fielding used Anderson's treatment in his exposition of latent structure models but this did not seem to have taken up until the present author used it as a basis for handling the factor analysis of categorical data. This work developed in Bartholomew by introduction of exponential family models and the key concept of sufficiency. One of the most general treatments, which embrace a very wide family of models, is also contained in Arminger and Küsters.
