(h/b) bayesian models for categorical data

This book continues the themes in my two earlier Wiley books in seeking to make modern Bayesian methods accessible via a practically oriented exposition, with statistical computing and applied data analysis at the forefront. As before, I have focused on the WINBUGS package, which has now reached a wide degree of acceptance in application to the ever expanding corpus of methodological work and applications based on Monte Carlo Markov Chain techniques. I hope that the applied focus will help students and researchers alike, including those with primary disciplinary interests outside statistics (e.g. psychology, geography and epidemiology). Nevertheless a wide range of simple or more advanced modelling techniques are discussed and I hope the approach is also helpful for courses in Bayesian data analysis and statistical computing. I have sought to review recent Bayesian methodology for categorical outcomes (binary, count and multinomial data) but my take on this will obviously emphasize some themes more than others: particular aspects that I have focused on include non-parametric and non-linear regression models, model choice, time series and spatio-temporal models. Missing data models are also considered. As with my earlier Wiley books, a set of worked examples with documented code forms part of the book's presentation and can be downloaded from the publisher's website ftp:// www.wiley.co.uk/pub/books/congdon, the WINBUGS site or STATLIB. (from the publisher)