Introduction

In this brief introductory chapter, the material covered in the book is summarized. The material centers around multiplicative models for the probabilities or intensities associated with the cells of a contingency table and some of the generalizations of these models. Also, more general sample spaces are considered. The coverage includes log-linear models, graphical models in the directed and undirected versions, marginal log-linear models, and relational models. These simple structures make it possible to model effects and associations, conditional, marginal, and even coordinate-free. Special attention is paid to the applicability of some of the models to investigate causality and also to the smoothness (regularity) properties of the parameterizations and of the models. A number of research problems where these models may be useful are also discussed.