Abstract
Howard Wainer of the National Board of Medical Examiners is a prolific author who specializes in books that help others understand statistics (especially but not only in medicine and education) and/or how to illustrate them. His latest book is well-written, generally easy (and often fun) to read, and should be very helpful to many audiences relevant to this journal: journalism professors and students, research methods professors and students, and journalists covering education, medicine, and other areas. He certainly has a knack for picking compelling case studies, keeping to the relevant facts, and making figuring something out seem easy in hindsight.
Wainer has divided the book into three sections composed of 17 numbered chapters, an introduction and a conclusion. The first section is supposed to help the reader “think like a data scientist.” The second is on “how data scientists communicate to themselves and each other.” The third is applying the thinking and communication skills to education (but actually he starts using education/testing examples in Chapter 5). The second chapter hooks the reader by explaining how athletes continue to break records and the number of piano virtuosos continue to increase if only because the world has more people than it used to: more outliers such as Hicham El Guerrouj with his 3.43-min mile.
Wainer continues with causal inference and happiness, and the same with medical treatment and death. On the latter, he shows the problems of making conclusions about cost-effectiveness and quality of life with four conditions: patients who had surgery lived or died anyway, and patients who did not have surgery also either lived or died. Readers may find this highly relevant in their own future and those of their loved ones. Chapter 6, on fracking, earthquakes, and Oklahoma oil wells, is relevant now (in case you had any doubt, yes, fracking is causing Oklahoma earthquakes, whatever business-overly-friendly Oklahoma officials claim).
Journalism professors, students, and professionals will find highly useful for their work Chapters 9 (“Improving Data Displays: The Media’s and Ours”), 10 (“Inside Out Plots”), and 11 (“A Century and a Half of Moral Statistics: Plotting Evidence to Affect Social Policy”).
Chapter 13 is a fascinating cost–benefit analysis of public schoolteacher tenure (it costs less than not having it), and Chapter 17 is a cost–benefit analysis of obtaining a lot of data over less data (being in the education section, Wainer frames it as a problem of “worthless subscores” from small parts of big standardized exams).
This book has something in it for everyone, and a lot in it for educators, journalists, and government bureaucrats of all types. But it should be used only as a supplementary text to books such as Phil Meyer’s Precision Journalism (4th ed.), a more comprehensive statistics text, a more comprehensive data visualization text, or a more comprehensive research methods text (depending on what course you are teaching).