Robert P. Hirsch and Richard K. Riegelman. 409 pages. Boston: Blackwell Publications; 1992. $29.95.

Over the past three decades, using statistical methods in the clinical and basic medical sciences has increased, creating a need for all health professionals, including physicians, to understand basic statistical concepts and methods. Without this understanding, it is extremely difficult to read and evaluate intelligently the medical literature. Teaching statistics to medical students and practitioners, however, has not been entirely successful. Medical schools and specialty programs have often given it short shrift in their curricula, and its presentation is often not tailored to the gestalt of this audience. The text by Hirsch and Riegelman offers a systematic flow-chart approach with heuristic and graphic rather than mathematic demonstrations, which the authors believe is more compatible with the way this audience is accustomed to learning new material.

After introducing the basics of probability, estimation, and inference in the first three chapters, the authors present their flow-chart approach to statistical methods by guiding the reader through a decision tree with branches defined by characteristics of the dependent variable and by the number and characteristics of the independent variable or variables. These branches eventually lead to the appropriate method of estimation and hypothesis testing. For example, in a parallel study of two treatment plans for weight control, the dependent variable, weight change, is continuous and the independent variable, treatment plan, is nominal. The corresponding flow chart would then specify the difference between the two treatments, with mean weight gain as the appropriate estimate and the t-test as the appropriate statistical test. Using this approach, the authors cover the basic statistical procedures generally included in a one-semester course.

Although the title suggests a cavalier approach, the book is carefully constructed and well written and contains many illustrations that help clarify the material. The reader is given the opportunity to obtain from the publisher a free companion diskette containing Stat-Aid-Plus, a computer-assisted learning aid to be used with the book. Although the book contains practice problems with detailed solutions, it would have benefitted from a greater number and wider variety.

The book's approach should appeal to the health professional who wishes some understanding of statistical concepts without an educational overload and who is comfortable with a "road map" approach for choosing the correct statistical procedure.