In his preface, the author indicates that this book is not what is often called a “cook book” of statistics. With this the reviewer agrees, but he cannot agree with the implication that this is a text book, if he means a text book to be followed in a statistics course. Half of the book consists of discussions of ten groups of questions “that an investigator can ask himself regarding his projected, current or past research, and which a reader of journal articles can apply in the evaluation of others' work.” The second half discusses various statistical methods with minimum consideration of arithmetic but with emphasis on limitations and interpretation, frequently cross-referenced to discussions in the first half. Those who know the first edition will recognize from this description that the second edition is an entirely different book. It is now printed in letter press and technically well organized with an index containing 500 or more entries.