Best Books for Social Scientists on Bayesian Analysis
I do a little Bayesian work as it makes sense in my research (okay, always) and, since Bayesian work is in vogue, I sometimes get asked what Bayesian books I would recommend. I haven't come close to reading them all, but these are my three recommendations for social scientists who have taken coursework though maximum likelihood estimation. All three books use R and BUGS, the standard software for Bayesian work.
- Jeff Gill's Bayesian Methods. Jeff's book is fantastic. It is one of the few advanced and thorough statistics books that is fun to read. He goes into detail on philosophical issues, gives lots of pedagogical examples, and discusses computational issues. He gives a lot of pedagogical R code. Start here.
- Andrew Gelman and Jennifer Hill's Data Analysis Using Regression and Multilevel/Hierarchical Models. This book is more implicitly Bayesian and the most applied, but provides a lot of real examples and plenty of R code to estimate models, perform simulations, and create nice graphs. They also discuss BUGS and provide code for many different models.
- Simon Jackman's Bayesian Analysis for the Social Sciences. This book discusses computation in detail and provides lots of real examples with plenty of code and explanation. It introduces a wide range of libraries useful for Bayesian work in R and illustrates their use on real political science data sets. This book is an indispensable reference for many problems, though too dense to read cover to cover.
These are my favorite books on Bayesian analysis. If you feel I've left something out, feel free to add it in the comments.