Strictly speaking, this book is not about computing, and so should not necessarily be reviewed in Computing Reviews. I assume, however, that many readers of <CR> perform statistical data analyses and need to know how to interpret computer output from statistical programs. Those readers will find the book interesting, and this review is for them.
The book was written as a text for an introductory college course in statistics for non–statistics majors. It uses no mathematics beyond high school algebra. The emphasis is on ideas rather than on formulas and computations. The authors provide a plethora of interesting examples from a wide variety of fields. The examples in the body of the text are worked out by hand or by computer. In the latter case, a photocopy of the computer output is embedded in the text with handwritten explanatory comments. More examples appear in the problems at the end of each section, and even more at the end of each chapter. Answers to some of the chapter problems appear in the appendix.
The topics are those covered in standard texts for introductory probability and statistics courses, but the presentation is much more user-friendly. In fact, the style might be criticized as too casual. The authors are careful to present statistical techniques in a critical light. For example, in the chapter on nonparametric statistics, they teach the sign-test and confidence intervals using coverages, but they do not teach the Mann-Whitney-Wilcoxon test because of its restrictive assumptions, which they explain. I am pleased that they present confidence intervals first and hypothesis testing afterwards. This order is much more intuitively appealing to the non-statistician.
The last three chapters in the book contain programs written in SAS, MINITAB, and BMDP respectively. The programs are accompanied by the output they produced, again with handwritten annotations. The examples for these programs are taken from the various chapters in the text, so the student can run examples during the school year on the program available at his/her college.
If the style of the text is too casual, the index is downright glib--in fact, infantile at times. I found entries in the index for “chickenpox,” “Chinese fortune cookies,” and “cholesterol and death,” but no entry for “chi-squared” or “contingency tables,” although an entire chapter is devoted to that topic. An appendix entitled “Hints on How to Succeed in Elementary Probability and Statistics” contains the warning that each week’s material is built upon the previous weeks’ material. This warning is referred to in the index under the entry “week-building in statistics.”
Another point that bothered me is that the authors use the word “they” instead of “he/she” and “their” instead of “his/her.” Their word processor does not hyphenate properly at the end of a line; on page 272 the word “assumptions” is split between the t and the i. I also found a few typographical errors. Problem 1.3.9 is erroneously labeled 1.3.7 (p. 28). An example on page 132 says “find the snowfall amount that we have a 5% chance of exceeding,” but the authors solve the problem using 95 percent instead of 5 percent. On page 130, in an example on testing psychic ability, they use a two-sided procedure. I think a one-sided procedure would be more appropriate.
On balance, I welcome this textbook, which tries to make statistics more enjoyable and more comprehensible to non-statisticians.