Doing Bayesian Data Analysis: A Tutorial with R, JAGS, and Stan

This book is outstanding. The author covers Bayesian analysis starting with the assumption that you know virtually nothing about it and builds to the point that you can do actual, meaningful analysis, interpret the results and communicate them to people that are not aware of Bayesian techniques. (I bought the book because of the "See inside" feature. Page 5 sealed the deal. If you do any type of statistical analysis check it out.) The writing is clear and there are numerous examples that are typically interesting which really helps. The author has a good sense of humor as well which is rare in a book that covers advanced material like this. The book is long. (>700 pages) But there is a LOT of material being covered. The "Doing" part of the book is done with R, JAGS and Stan, so if you aren't familiar with any of those, it's a lot to take in. I wasn't familiar with any of thee and I did fine. I had to read some parts multiple times but that might just be me. I did most of the exercises which really helped. (notable exception: 13.1) The book seems expensive at first. It's a textbook so there's that. However, you also get a very large number of R scripts to demonstrate the concepts. The scripts are useful and in my opinion worth as much as the book. (All of the software is free.) I have already used the scripts to suit analysis I needed to do. The book also really covers multiple topics so once I got into it I realized I got a great deal. I have received more value that I paid. Don't buy this thinking you are going to breeze through it and be able to actually *do* Bayesian analysis well. The book is true to the title but only if you put forth the time and effort. You absolutely can learn an enormous amount from this book. To the author, if you're reading this: Thank you! I am better at what I do because of this book.

Edit: I've updated the rating from 1-star to 5 stars to properly reflect the quality of the content. Anything else would be unfair. I returned my original copy which was in an unacceptable condition (see text below plus the comments). I didn't mind the hazzle much, but I did incur some additional costs. I had to send the original copy back to Amazon which cost more than the 15$ refund, and I also had to pay for the (quick) delivery cost. Nonetheless, my impression is that the book is possibly the best introduction to Bayesian statistics on the market. And not only the best, but also very good in its own right. The book by Gelman et. al. is a leading textbook on the subject, and for a good reason, but the authors assume their readers have mastered intermediary statistics and have received a thorough prior introduction to Bayesian statistics. John Kruschke, in contrast, assumes very little knowledge of the former, and none of the latter. All expositions are intuitive rather than technical. The chapter on R taught me that I still have much to learn on the language, and makes me wonder how inefficient and cumbersome my R code has been thus far. I'm trying to think of any real complaints I have that are not merely a reflection of my eccentric nature, but I keep coming up short. Recommended. ------------- Please be aware that my review concerns the quality of the binding of the pages in the book, and not the actual intellectual content. I've only recently begun reading it, and I have no reason to believe that the other positive reviews are anything other than accurate. But yes, good as the exposition of the subject must be, the binding of the pages is poor in regular patterns. Although no pages have fallen out, I can't help but wonder if the book will stay in one piece for very long. I don't know if you can see the picture that I uploaded, but it shows only one opening, on page 82-83. Other openings are either similar or share the same fate.

Over the past couple years, I've been trying to learn Bayesian statistics, both for theoretical understanding and for practical use in my job. That has lead me to read 5-10 different books on the subject (with a range of scopes and focuses), which lead me to read the first edition of this book. Some of the books that I read are better than others, but I can easily say that Kruschke's was the best introductory book I found. That is NOT to say that it lacks rigor. What it does is start off with the basics, and it communicates in a clear, readable, and often humorous approach. What it does not do is assume that you have an advanced degree in statistics. In addition, the book gives you TONS of useful programming help in R, all in downloadable files. Even better (for me at least) were the programs that helped access OpenBUGS from R (in the first edition). That is a tricky process, and I found the book's insight and programs to be very valuable. I would also like to thank the author for reading my mind. After having worked with OpenBUGS for a little while, I was hearing good things about STAN, and I've been wanting to give it a try. Right about the time I decided that, this second edition came out, and this time it includes STAN. Woot! I'm reading through the second edition, and I'm enjoying it just as much as the first. Heck, this book is probably worth the price for the programs alone.

Great book This is a first review as I proceed through the book; I plan on writing others as I delve deeper into it. The second edition is a significant improvement over the first. I am using the electronic version, and there are a few glitches: the Kindle reader for IOS does not display the equations correctly; whereas the Kindle reader on my PC does display the equations properly, but does not increase the size of the equations as the text size is changed. This is distracting.

This author is the model ever educator and STEM author should strive for. This book makes the topic very approachable and equips you to understand and apply Bayesian analysis. There is just right level of mathematics to enable you to understand the concepts without needless confusing the reader with proofs they do not need if they are not pursing a doctorate in stats. The author has created his own R packages so the student can focus on Bayesian analysis and not programming. The author has created nice problems to help you solidify your knowledge of important topics with worked solutions. Really thankful for the effort put into this work of knowledge. The author’s time invested really shows. The wit is appreciated as well.

I have an undergraduate degree in Statistics however I never learned Bayesian statistics as this is typically taught to graduate students. Even so, this book is very easy to learn from. The author puts together a recommended order of reading the chapters depending on how much time you have. I read 3-4 chapters and I skipped ahead to the model with one predictor (essentially the linear regression model with one predictor). He gives the code for programs with comments in R that are easy to understand and run. His website has all the programs written in R that use JAGS and Stan and solutions to the exercises to the first 15 or so chapters (more are on the way). Loving this book, style of reading is also very conversational. It does get into the math at times but generally it is light and the technical parts can be skimmed over if you're more interested in the "doing" part as I am. Very practical book! Explains very well with plenty of graphs.

The author does a fantastic job of providing concrete illustrations of the abstract principles a non-statistician must grapple with to understand and apply Bayesian analysis. The equations are there but unlike many books on the subject the meaning of each is presented in a clear and concise manner. He seems to know where a reader might stumble and takes great care in these trouble spots to get you over the hump. The R scripts are a valuable addition and I found them easily adaptable to my engineering research.

Beautiful explanations. R code could be more modern, but that issue is addressed by some academics (like professor Randall Pruim with his notes on the book). The concepts are priceless and Dr. Kruschke also makes clear that he has a knack for teaching. The first two parts of the book are dedicated to explaining the underlying ideas that build Bayesian analysis. Part III goes through applied aspects and crucial aspects of model interpretation. 12/10, would buy again

Éste fue mi libro en la carrera para la asignatura de Probabilidad e Inferencia Bayesiana. Si entiendes inglés, es una joya.

A mon avis, c'est un livre unique qui remplie le Gap un gap entre un ouvrage tel que Bayesian Data Analysis de Gelman et les livres classique de statistique fréquentiste. Ce livre est plus qu'une introduction sur l'analyse Bayesian. J'y ai trouvé les réponses bien argumentées sur les débats tel que: -les problèmes des testes séquentiels dans le NHST, -le shriankage; -comment le Bayesian réponds au erreur de type I (false alarme rate); -"model selection vs parameter estimation"; -etc. Il nous guide pas a pas pour apprendre a faire une modélisation hiérarchique. Les codes et les figures sont aussi très utiles. Le sol inconvénient est que j'aurai aime un peu plus de matière mathématique et un peu plus de théorie qui soit formulé. Cela dit, il rends accessible des notions qu'on ne peut que trouver dans des ouvrages très technique et c'est tout son intérêt. Bonne lecture

An exceptional book. Even if it is presented as a "Bayesian" book, it actually has way more stathistical insight than you can expect only by the (cute) cover. Firstly: since bayesian analysis is based on probability, the book will you teach probability from its basis. Second: since it would be useless to only know probability theory without knowing how ot apply it, the book also teach you regressions theory and how to develop in a bayesian way. Third: the book also ansewer to the big question that any statistician or future statistician may encounter: "Why shoul ìd I bother with Bayesian analysis?" by giving you examples on what are the main differences with the classical frequentist approach. A must have for any researcher who need to apply statisthic in his every-day job or research.

Este libro rebasó mis expectativas. Es una excelente introducción a la estadística bayesiana con una pedagogía notable. Al autor de este libro, si alguna vez lees este comentario, gracias por este hermoso trabajo.

produto excelente, material muito bom.