An information-theoretic framework is used to analyze the knowledge content in multivariate cross-classified data. Proposes measures based on the information concept, including the knowledge content of a cross classification, its terseness, and the separability of one variable. Presents applications for situations when classical analysis is…

Although RC(M)-association models have become a generally useful tool for the analysis of cross-classified data, the graphical representation resulting from such an analysis can at times be misleading. The relationships present between row category points and column category points cannot be interpreted by inter point distances but only through…

Covariance structures associated with linear structural equation models are discussed. Algorithms for computing maximum likelihood estimates (namely, the EM algorithm) are reviewed. An example of using likelihood ratio tests based on complete and incomplete data to improve the fit of a model is given. (SLD)