R. Barker Bausell is a biostatistician who worked for the NIH's Complimentary Medicine Program, which was designed to test the efficacy of Contemporary and Aleternative Medicine (CAM). As a biostatistician, Bausell is the one who designs studies so that they are as fair and unbiased as possible. His big "beef" with CAM? That the less biased the study, the less effective CAM seems to be.
This book has several strenghts and several weaknesses. I will go into the strengths first.
First, while the book suggests that it is primarily about 'debunking' alternative medicines, the bulk of the book is spent talking about how effective studies are designed and different things that can undermine the validity of studies (small sample sizes, shoddy control/placebo treatments, attrition). In short, this book offers a VERY good explanation of how science works. (Only after explaining how good studies are designed does our author go on to suggest that most CAM studies are quite poorly designed.)
This book spends a lot of time talking about the 'placebo effect,' a large player in CAM research. The placebo effect is a (generally) psychological effect where the person experiences betterment SOLELY from having any kind of treatment at all (even a sugar pill). Our author's point with explaining the placebo effect is to suggest that well-designed CAM studies point to one conclusion: that most CAM treatments are only as effective as any other placebo (incorrectly performed accupuncture is as effective as 'legitimate' acupuncture, not because accupuncture works, but because the subject wants or expects it to work).
The author is very far from biased. Despite its outragous title, Snake Oil Science is not a 'gotcha' book written by a mean-spirited and fun-poking author. The discourse is very professional and fair. The author never 'slams' CAM, but only suggests that CAM has ALOT of work to do in order to prove itself, assuming that it can.
For those wanting a comprehensive discussion 'debunking' CAM treatments and remedies, this book - again, despite its title - will not be satisfying. The author, a biostatistician, spends so much time talking about how to design a good study, how to spot a bad one, and adding caveat after caveat, that only one (and a half) chapters really discuss what the research actually saya. Really, the book should have been subtitled, "A primer on the methodology of clinical studies."
For those who want a somewhat friendly and relatively non-academic read, this book probably is not it. The author certainly tries to bring it down to non-specialist language, but when talking about statistics, controls, variables, and confounds, technical jargon and dry verbiage ls unavoidable. While this book is certialy informative about how clinical trials are designed, the placebo effect, and explaining why most CAM studies are poorly and hastily done, it is a somewhat dry read.
So, there you have it. If you want to become more familiar with how the medical profession tests their treatments (and compare it to how CAM proponents 'test' their treatments) this is a very good and exciting book. If you are looking for a good old-fashioned Shermer and Randi style 'debunking' of CAM, there are several other books you are better to read than this one. (Try "Trick or Treatment: The Undeniable Facts About Alternative Medicine.")