This is a book in a tradition and so a review might well begin by briefly recapitulating that tradition. It probably begins in 1979 with the publication of David Moore’s Statistics: Concepts and Controversies. That book differed from most other introductory statistics textbooks of its day (or this) in putting almost all of its emphasis on statistical issues and reasoning, and very little emphasis on specific techniques. Of course, Huff’s *How to Lie with Statistics* could be said to have done that as well, but that was a short book for the layperson, while Moore’s was a full textbook with many exercises that could be (and was) used for college courses. Unfortunately, the techniques, rather than statistical thinking, are typically what departments serviced by an introductory statistics course are looking for. So about a decade later, and with George McCabe, Moore published the much fatter *Introduction to the Practice of Statistics*, which covered all the usual techniques, but retained the emphasis on statistical thinking. This satisfied another audience, but the reading level was very high, so in 1995 Moore published *Basic Practice of Statistics*. These books sold well, were highly influential, and stimulated the MAA to initiate a flurry of books, articles, and workshops on teaching statistics.

In addition to the influence Moore’s books had on college courses, they had an even larger influence on AP Statistics, where the official syllabus was very much in the spirit of Moore’s texts, and the first text explicitly written and widely used for AP Statistics was an adaptation of *Introduction to the Practice of Statistics* by Dan Yates. A descendant of that text is still widely used for AP, and the other common texts are also in Moore’s spirit. All are (or have related) college textbooks, with the main difference being that the high school texts emphasize graphing calculators as their technology, and have sturdier bindings.

Many of the principles embodied in Moore’s text have since been incorporated into the *Guidelines for Assessment and Instruction in Statistics Education* (GAISE) Reports published by the American Statistical Association. Those have now become so widely recognized that many textbooks claim to follow them, though often doing so more on the surface than in substance. By now the reader will not be surprised to hear that the book at hand is clearly in the Moore tradition, and implements most of the GAISE recommendations well. Beyond that, it offers little that many other textbooks have not already claimed to offer, and several that meet their claims. It is worthy of consideration, perhaps in the top 10% of such books, and adoption will probably depend on how well details fit local conditions.

Which brings us to a difficult issue. One of the Moore/GAISE principles is to use real data. Unfortunately, there are wide differences of opinion as to what that means. To your reviewer, it means using data gathered in real studies to answer a research question. This helps to convince students that statistics really is used outside the classroom, and introduces them to many practical applications that can be very motivating. It is also true that such data are more complex and more interesting than numbers made up for practicing calculating.

Unfortunately, very little such data appears in this textbook. The data in the examples and exercises are almost all tiny batches of numbers suited to computational practice. Few raise any interesting questions. Few include enough data to check assumptions (as Moore told us to do). Few illustrate real research questions or applications. They are real only in the sense of not being fictional. There are two clues as to why this might be the case. One is that graphing calculators are one of the technologies supported, and the other is that one of the authors teaches at a community college where graphing calculators unfortunately are often the technology of choice. (Note though that Stats Homework is but one example of free statistical software that is available.) This makes the text at hand a good recommendation in that environment, but is a serious shortcoming in any environment where more appropriate technology is available. So your reviewer has several words of advice to the authors.

- Market the current book to community colleges. It would be a standout there.
- Create a college edition where most of the data sets in exercises are of reasonable size and complexity and come from real studies answering real research questions.
- Adapt this book slightly to create an AP Statistics text. Keep the graphing calculator stuff but include the datasets from the college version as well as the ones small enough for the calculators.

Except when illustrating a formula, all the examples in all the versions should use the larger datasets. There should be an icon next to the minority of exercises with data sets small enough to fit in a graphing calculator.

In fairness, it should be mentioned that this advice applies equally to many other college textbooks and most AP Statistics textbooks. But there are options available, and so it is hard to recommend this text in its current form unless one supplements it heavily with data from real research studies, or is really and truly unable to use anything more powerful than a graphing calculator.

After a few years in industry, Robert W. Hayden ([email protected]) taught mathematics at colleges and universities for 32 years and statistics for 20 years. In 2005 he retired from full-time classroom work. He contributed the chapter on evaluating introductory statistics textbooks to the MAA’s Teaching Statistics.