The Analysis of Cross-Classified Categorical Data by Stephen E. Fienberg is a comprehensive book on the analysis of categorical data using loglinear models. The second edition, published in 2007, is a reprint of the 1980 edition originally published by MIT Press. The book is intended for nonstatisticians and provides an introduction to the analysis of cross-classified categorical data using loglinear models. It includes a summary of theoretical statistical terminology for nonstatistical readers. The book covers topics such as two-dimensional tables, three-dimensional tables, selection of a model, four- and higher-dimensional contingency tables, fixed margins and logit models, causal analysis involving logit and loglinear models, and fixed and random zeros. The book also includes problem sets at the end of each chapter, expanded discussions of linear logistic response models and polytomous response models, and a new appendix on the small-sample behavior of goodness-of-fit statistics. The book is supported by various statistical methods and includes references to statistical literature. The book is written for nonstatisticians and includes a variety of new methods based on loglinear models that have entered the statistical literature since the author's previous work with Yvonne Bishop and Paul Holland. The book also includes a discussion of contingency tables with ordered categories and several new applications of the methods associated with incomplete contingency tables. The book is supported by various statistical methods and includes references to statistical literature. The book is written for nonstatisticians and includes a variety of new methods based on loglinear models that have entered the statistical literature since the author's previous work with Yvonne Bishop and Paul Holland.The Analysis of Cross-Classified Categorical Data by Stephen E. Fienberg is a comprehensive book on the analysis of categorical data using loglinear models. The second edition, published in 2007, is a reprint of the 1980 edition originally published by MIT Press. The book is intended for nonstatisticians and provides an introduction to the analysis of cross-classified categorical data using loglinear models. It includes a summary of theoretical statistical terminology for nonstatistical readers. The book covers topics such as two-dimensional tables, three-dimensional tables, selection of a model, four- and higher-dimensional contingency tables, fixed margins and logit models, causal analysis involving logit and loglinear models, and fixed and random zeros. The book also includes problem sets at the end of each chapter, expanded discussions of linear logistic response models and polytomous response models, and a new appendix on the small-sample behavior of goodness-of-fit statistics. The book is supported by various statistical methods and includes references to statistical literature. The book is written for nonstatisticians and includes a variety of new methods based on loglinear models that have entered the statistical literature since the author's previous work with Yvonne Bishop and Paul Holland. The book also includes a discussion of contingency tables with ordered categories and several new applications of the methods associated with incomplete contingency tables. The book is supported by various statistical methods and includes references to statistical literature. The book is written for nonstatisticians and includes a variety of new methods based on loglinear models that have entered the statistical literature since the author's previous work with Yvonne Bishop and Paul Holland.