Categorical Data Analysis

This book is the standard for this material. It is clear, concise and comes with enough examples to assure the reader of the empirical importance of all the theory.

The text is comprehensive in covering categorical data. Other reviews make this clear so I wanted to focus on the following. I was able to understand more general topics in statistics because of Agresti's depth of coverage on CDA. For example, for repeated measurements, Agresti clearly explains marginal models, conditional models, and generalized estimating equations. When I needed to understand these topics, I used this text because I have not found clear explanations elsewhere. In addition, SAS code and R code is available for the examples presented.

When this book came out in 1990 it was the first book to provide a truely modern treatment of categorical data analysis for both ordinal and nominal data. It provides an excellent treatment of the asymptotic theory for binary and multinomial data. It is extremely well written and is still a favorite of statisticians and practitioners. Because of its popularity and continued value, it should soon be added to the Wiley Classic series. This is the first book to take the regression approach to categorical data analysis tieing the subject to the methods and theory of the generalized linear models. It also was one of the first to show the modern practicality of exact permutation methods. The only drawback of this book is that it is 11 years old and there have been many interesting and relevant research developments in computer-intensive methods, analysis of missing data and mixed effects linear models to make a revision useful. Some of the latest developments can be found in Lloyd's new book "Statistical Analysis of Categorical Data" that was recently published by Wiley. Agresti provides clear advice and also gives a nice historical perspective on the development of the subject. The book is authoritative and includes numerous relevant references. Each chapter contains many exercises and a wealth of practical examples for illustration of the techniques. This is a good text from both practical and theoretical perspectives. It is excellent for a graduate level course on categorical data analysis.

If you want one book on Categorical Data analysis, this is the one. But there are others that are easier to read, if your math is not great (including the same author's book with an almost identical title)

This is a very demanding, thorough, and clear description of just about everything anyone could want to know on the subject. The second edition is considerably more rigorous than the first. Agresti stresses that logistic models are one kind of generalized linear model. This offers solid connections to many other models, but places corresponding demands on the reader. In particular, Chapter 4 is difficult going, but might be skipped or skimmed on first reading. Given the mathematical level and rigor, this is a remarkably clear book. Anyone who analyzes categorical data on a regular basis should read it and have it on his or her shelf.