CATEGORICAL-DATA ANALYSIS

This chapter reviews recent developments in the analysis of categorica l and contingency-table data. The first portion examines developments in model testing and selection. The second portion examines work on mo dels for the structure of dependence. These include log-linear paramet er models, models for latent classes, models for missing observations, numerical-scale-based association and correlation models (such as cor respondence analysis), the treatment of ordered categories, and models for marginal distributions.