Applied Bayesian Statistics: With R and OpenBUGS Examples (Springer Texts in Statistics, 98)

Less of an introductory text to Bayesian statistics, and more of an example driven text on Bayesian statistics implementation in R and OpenBugs. Chapters One and Two are introductory covering what is Bayesian statistics and a quick review of probability. Most of the examples are simple, and similar to other online sources. The difference is, there is more explanation in the book as to why they are coded that way, than most of the online cases. The following chapters cover: estimating population proportions, considerations for Bayesian inference, conjugate priors, multiparameter models, the basics of MCMC, regression, convergence and model checking and hypothesis testing. Wherever possible, Cowles works and tries to illustrate the differences between frequentist and Bayesian approaches.My only real complaint about the book is that I never found the data sets online, so I had to type the data sets in (thankfully they were small).

Terse. Incomplete. Author is completely unresponsive to any requests for information.

As a theory book, it is superficial and doesn't go into much depth but still requires a calculus base. I wasn't really interested in a heavy theory book, so that was fine. The explanations are not the best, but not awful.As an applied book, it does give many examples, but the examples are light up till about chapter 8 (out of 11). Chapter 9 starts to get into hierarchical bayesian modeling, but it isn't very well explained. The example the author uses is a softball example and in table 1, the data is presented incorrectly. She labels the binary 1,0 as at bats, and then n as hits. How do you get 5 hits at 1 at bat? Later in the chapter, it is flipped back to 1 being there was a hit out of 5 attempts. That makes more sense. Little errors like that make it difficult for a novice like me to follow the model, notations, and what is trying to be estimated.As a software manual, it is poorly written. R is barely talked about in chapters 1 through 7, and she uses the LearnBayes package from another book. That made me wonder why I'm using this book and not the other one. Openbugs isn't introduced till chapter 8 (out of 11), and it is a very cursory look.I gave up after the middle of chapter 9, because it wasn't giving me what I wanted. I'm not sure what that is exactly, just not a good fit for me. I did not purchase this book, but got it from the library to look it over before deciding.This isn't a horrible book, just not very clear what it is suppose to be.

"Applied Bayesian statistics" is a decent, if unremarkable, introduction to Bayesian statistics - but it is just not competitive with "Doing Bayesian data analysis" by John Kruschke, which actually costs less if you get a used copy.UPD. Second edition of "Doing Bayesian data analysis" is now out, widening the lead. "Bayesian statistics for the social sciences" by Kaplan is another better-than-this option.