This textbook encompasses full-fledged methodologies in categorical data analysis which, together with the supplementary online material, can be readily used in the classroom. The book includes motivating and illustrative examples, a full array of exercises and solutions to a fair selection of these, historical notes about the statistical theory at the core of the data analysis tackled in the text (a prominent founding figure of the underlying theory being R.A. Fisher) and many references. Bayesian inference methodologies are also addressed whenever appropriate. Hence, this textbook is virtually self-contained.
Clarity of the presentation, the gradual in-depth progression of concepts and the sensible chapter organisation, all conduct the reader into a natural flow of theoretical and applied contents. Parametric inference techniques have particular emphasis on maximum-likelihood estimation and testing. Non-parametric procedures regarding small sample inference are also addressed. Moreover, the author offers a sound assessment of the strengths and weaknesses of the methodologies in each chapter, often by comparison, either via interesting and pertinent illustrative examples or, in many occasions, by pointing out accurate references supporting the author's claims. This provides the reader with a good insight about the comparative performance of estimation and/or testing procedures tailored for the practical situation at hand, which I find a major and powerful strength of this book. Bayesian methods also intervene. Modern issues such as high dimensionality and meta-analysis are all tackled in the book. In my opinion, the shortcoming of this book concerns the graphical methods. I agree that more graphics would bring more pages to an already thick book, but one or another example might really benefit from an ancillary interpretation via a proper plot. Perhaps the latter could be a good way of bringing out the supplementary online contents. The website of the book provides guidelines, many examples and exercises for users of different statistical software which in my opinion is not stressed enough in the text. Another criticism is that, since this is such a dense and complete textbook, sometimes it is difficult to find some definitions due to lack of cross referencing. Although the book contains an exhaustive subject index, a list of symbols and notation in the beginning of the book would be appreciated.
Overall, this textbook constitutes an important reference for those with a basic knowledge in probability who wish to engage in statistical methods for categorical data at several levels, and to statisticians who seek acquaintance with relevant concepts and practical problems in the fields of biostatistics or epidemiology; it is a comprehensive and reliable manual for categorical data analysis.
This textbook alludes to a lifetime of dedication to the analysis of categorical data, gathering an impressive set of analytical tools and an immense collection of information. After reading the book, one is left with the distinct notion that a colossal task was seriously undertaken and rigorously accomplished by the author.
I would strongly encourage libraries who own the second edition to buy this edition as well.