'Students, developers, and practitioners in this area will all want to have this thorough guide close at hand. The wealth of theory and extensive applications using 'real' data sets and contemporary software will provide a crucial resource for their research.' William Greene, New York University<br />'This is a well-researched practically oriented book on an important class of models relevant to over-dispersed count data. Recommended.' John Nelder, Imperial College London<br />'Every model currently offered in commercial statistical software is discussed in detail ... well written and can serve as an excellent reference book for applied statisticians who would use negative binomial regression modelling for undergraduate students or graduate students.' Yuehua Wu, Zentralblatt MATH<br />'I would recommend this book to researchers and students who would like to gain an overview of the negative binomial distribution and its extensions.' Fiona McElduff, University College London<br />'The text is well-written, easy-to-read but once started, is difficult to put down as each chapter unfolds the intricacies of the distribution.' International Statistical Review<br />'The second edition of Negative Binomial Regression is a unique statistical textbook. It is a very enjoyable read! It not only provides statistical fundamentals, but also provides historical perspectives and expert insights. This book is an excellent introduction for someone new to modeling count data, as well as an invaluable resource for the experienced practitioner grappling with complex overdispersed data.' Elizabeth Kelly, Statistical Sciences Group, Los Alamos National Laboratory<br />'As with all of Joe Hilbe's books this text is thorough and scholarly with an extensive list of references. Important theorems and other theoretical results are presented but are presented to be informative rather than to develop and teach the theory.' Michael R. Chernick, Significance<br />'... a valuable hands-on introduction to negative binomial regression and the analysis of count data in general. I am also pleased to see an advocation of the utility of the negative binomial distribution in applied work.' Psychometrika