Customer Reviews

Statistical Power Analysis for the Behavioral Sciences (2nd Edition)

5 star:		(6)
4 star:		(1)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

Cohen does a masterful job of taking the guesswork out of statistical power estimation. This text provides procedural guidelines for determining power for many designs, and can be quite helpful in determining proper sample sizes. Not for the casual reader, but a necessary addition to any serious researchers statistical library.

In the design of clinical trials particularly in the regulatory environment of the FDA the determination of the appropriate number of subjects to recruit for a well-powered statistical hypothesis test ot tests is extremely important. This topic is often neglected in statistical texts, even the elementary ones. It has also been neglected in many of the statistical software products that are available. Recognizing this need Janet Elashoff and Ralph o'Brien were among the first to construct tools for sample size determination and calculation of statistical power. There are now many software products available including nQuery Advisor (Janet Elashoff's product), Power and Precision, PASS, and StatXact. For group sequential and adaptive design methods thare are products such as East, PEST, SeqTrial+ and ADDPLAN to mention a few. Ralph O'Brien's package has now been modified and implimented in SAS with the procedures PROC Power and PROC Glmpower.

This book by Jacob Cohen preceded these developments and became a classic reference for researchers particularly those in the social sciences including behavioral scientists as mentioned in the title. I know of only one other competing book from that time period that being "How Many Subjects" by Helen Kraemer of Stanford University.

Clearly, a must for every statistical library. This book is considered the authority on power analysis.

Absolutely the main text but check it out from the library - you will use it approximately 10-15 times in your research life.

Classic stats book for the scientific grant writer. Excellent info and tables concerning power analyses for experimental effects. If you don't include a power analysis in your proposal, you're wasting your time...

Supposedly this is THE book to read on power analysis. The book touches on power calculations very quickly and then provides the relevant statistics in a large set of tables; you could probably just borrow it from the library when needed rather than pay to have it, but a good reference text nonetheless.

This is a very comprehensive review of power analysis from one of the early authors in the field. It is very applicable to fields outside of the behavioral sciences.