Bayesian Statistics

An Introduction

Second Edition

PETER M. LEE

Preface to the Second Edition

I have been gratified by the reception of the first edition of this book and have tried to respond to comments from reviewers and readers in preparing the second edition. An additional chapter has been added which deals with hierachical methods in Bayesian statistics and gives a fuller treatment of empirical Bayes methods. Some of the material in this chapter originated in lectures given at a summer school organized by the Societá Italiana di Statistica in Varese in September 1995. Another chapter deals with real numerical methods, especially the EM algorithm and Gibbs sampling. Minor alterations have been made in a number of places, including the addition of a brief description of Bayes linear methods, and the number of exercises has been considerably increased. Solutions to these exercises can be found on the World Wide Web at

http://www.york.ac.uk/depts/maths/histstat/pml1/bayes/book.htm

and facilities are provided there for comments on any aspect of the book to be sent to me. Among statisticians the Bayesian approach continues to gain adherents, but it can be seen that there is still some way to go before we live in a fully Bayesian world when we read in The Times (9 May 1996, p. 36) that, “The Bayes Theorem might be an appropriate and useful tool for statisticians but it was not appropriate for use in jury trials or as a means to assist the jury in their task” (Court of Appeal, Criminal Division, Regina v. Adams Before Lord Justice Rose, Mr Justice Hidden and Mr Justice Buxton [Judgment April26]; cf. The Times (3 November 1997, p. 40), Regina v. Adams (No 2). Before Lord Bingham of Cornhill, Lord Chief Justice, Mr Justice Potts and Mr Justice Butterfield [Judgment October 16]). Hope, as always, lies in the next generation.

Peter M. Lee

24 February 1997
Click on the required option below: