Statistical Analysis Of Categorical Data

Accessible, up-to-date coverage of a broad range of modern and traditional methods. The ability to understand and analyze categorical, or count, data is crucial to the success of statisticians in a wide variety of fields, including biomedicine, ecology, the social sciences, marketing, and many more. Statistical Analysis of Categorical Data provides thorough, clear, up-to-date explanations of all important methods of categorical data analysis at a level accessible to anyone with a solid undergraduate knowledge of statistics.

Featuring a liberal use of real-world examples as well as a regression-based approach familiar to most students, this book reviews pertinent statistical theory, including advanced topics such as Score statistics and the transformed central limit theorem. It presents the distribution theory of Poisson as well as multinomial variables, and it points out the connections between them. Complete with numerous illustrations and exercises, this book covers the full range of topics necessary to develop a well-rounded understanding of modern categorical data analysis, including:
* Logistic regression and log-linear models.
* Exact conditional methods.
* Generalized linear and additive models.
* Smoothing count data with practical implementations in S-plus software.
* Thorough description and analysis of five important computer packages.

Supported by an ftp site, which describes the facilities important to a statistician wanting to analyze and report on categorical data, Statistical Analysis of Categorical Data is an excellent resource for students, practicing statisticians, and researchers with a special interest in count data.

CHRIS J. LLOYD, PhD, is Associate Professor at the Australian Graduate School of Management, University of New South Wales.

The Tools of Statistical Inference.

Distribution Theory for Count Data.

Binary Contingency Tables.

Binomial Regression Models.

Smoothing Binomial Data.

Poisson Regression Models.

Conditional Inference.

Index.