This book has already been reviewed in the Archives of General Psychiatry,1 the previous reviewer having been more favorably impressed by it than I was. Chassan has much to say that might be useful in clinical psychiatric research, but also much which is misleading and wrong.
The first four chapters are excellent. Chapter 1 surveys the logical and probabilistic foundations of statistical inference; chapter 2 presents some of the simpler and more commonly applied statistical techniques; chapter 3 explains the need for randomization and for adequate control groups; and
chapter 4 describes the rationale for, the execution of, and the pitfalls in the doubleblind clinical trial. In the latter chapter Chassan in effect says, and I agree with him, that unless a clinical team is prepared to put up with the scientific rigor and personal sacrifice demanded by double-blind trials, it should not even attempt such a study. The