17 of 17 people found the following review helpful
on 16 May 2010
I had looked at a couple of other books on Bayesian statistics and noticed that they all focus mainly on all kind of very technical aspects. Sivia's book on the other hand deals almost exclusively with basic applications of Bayes theorem. It does this by discussing a number of examples in detail. This makes it much more useful people like me (I am a physical chemist), as I can often simply copy what Sivia does. I rarely feel the need to consult the more advanced books.
In the second edition of the book a chapter on nested sampling has been added written by John Skilling. I have seen that this technique can be very useful, but only because I found a paper in which it was described much better than in the book. It is not just that a completely different notation is used by Skilling, also the whole approach is different. Whereas Sivia aims to write an introductory textbook, Skilling seems to write for specialists. This really lessens the quality of the book.
12 of 12 people found the following review helpful
on 25 August 2009
I bought this book with the aim of improving my data analysis skills and also to try to figure out what it meant to do things the Bayesian way. In both cases this book did an admirable job. Due to the understandable explanations from first principles that this book offers it is possible to get a really intuitive feel to what is going (perhaps this is due to the Bayesian approach, as a physicist I felt that statistics was no longer just a complicated mix of formulas). In terms of getting a better grasp of data analysis I found that after reading and on occasion re-reading relevant chapters I have been able to apply it to actual problems in the field. The least squares extension chapter is particularly good in that it first highlights how some of the normal assumptions aren't appropriate before discussing how to proceed in those cases.
Fantastic book that I have used countless times over the last year.
6 of 6 people found the following review helpful
on 5 August 2010
This is a _great_ book. The early chapters which introduce the broad concepts underlying Bayesian reasoning are particularly strong. Although it's aimed at students of physics, it would be useful to a much broader range of disciplines (I'm a psychiatrist which is about as far from physics as you can get...).