Generalized Linear Models and Extensions, Second Edition

by James W. Hardin

Generalized Linear Models and Extensions, Second Edition provides a comprehensive overview of the nature and scope of generalized linear models (GLMs) and of the major changes to the basic GLM algorithm that allow modeling of data that violate GLM distributional assumptions. Deftly balancing theory and application, the book stands out in its coverage of the derivation of the GLM families and their foremost links, while also guiding readers in the application of the various models to real data. This edition has new sections on discrete response models, including zero-truncated, zero-inflated, censored, and hurdle count models, as well as heterogeneous negative binomial, generalized Poisson, and generalized binomial models. The book also includes a substantially expanded discussion of both proportional-odds and generalized ordered models, making it easy for readers to use these models in their own research.

387 pages, Paperback

Published February 20, 2007

Reviews

[Stephen Cranney · Rating 2 out of 5]

Why are stats specialists usually incapable at explaining their knowledge to human beings? If you're going to explain something rather statistically simple, it doesn't help to start it off with a solid page of Greek characters. The fact that I knew how to explain the stuff I did know in simpler (and more concise) terms made me much less willing to try to slog through the stuff I didn't know.