The past quarter-century has seen an explosion in the development of methods for analyzing categorical data. These methods have influenced—and been influenced by—the increasing availability of multivariate data sets with categorical responses in the social, behavioral, and biomedical sciences, as well as in public health, ecology, education, marketing, food science, and industrial quality control. Categorical Data Analysis describes the most important new methods, offering a unified presentation of modeling using generalized linear models and emphasizing loglinear and logit modeling techniques. Contributions of noted statisticians (Pearson, Yule, Fisher, Neyman, Cochran), whose pioneering efforts set the pace for the evolution of modern methods, are examined as well. Special features of the book include: Coverage of methods for repeated measurement data, which have become increasingly important in biomedical applications Prescriptions for how ordinal variables should be treated differently than nominal variables Derivations of basic asymptotic and fixed-sample-size inferential methods Discussion of exact small sample procedures More than 40 examples of analyses of "real" data sets, including: aspirin use and heart disease; job satisfaction and income; seat belt use and injuries in auto accidents; and predicting outcomes of baseball games More than 400 exercises to facilitate interpretation and application of methods Categorical Data Analysis also contains an appendix that describes the use of computer software currently available for performing the analyses presented in the book. A comprehensive bibliography and notes at the end of each chapter round out the work, making it a complete, invaluable reference for statisticians, biostatisticians, and professional researchers.