A Statistical Guide for the Ethically Perplexed

by Lawrence Hubert, Howard Wainer

Lauded for their contributions to statistics, psychology, and psychometrics, the authors make statistical methods relevant to readers' day-to-day lives by including real historical situations that demonstrate the role of statistics in reasoning and decision making. The historical vignettes encompass the English case of Sally Clark, breast cancer sc

592 pages, Paperback

First published August 1, 2012

Reviews

[Edd Simmons · Rating 5 out of 5]

The fact the author used Supreme Court cases and statistics was a nice effect to read upon. I would like to read more on Howard Wainers thoughts on the SAT’s.

[Mark · Rating 4 out of 5]

While reading this dense very wide-ranging mathematically-detailed (for a book for lay readers anyway) book, I kept forgetting what it was ostensibly about. It is supposed to "try to reconcile the role of statistics and behavioral...sciences through the standards of ethical practice", but it just seems to be every interesting problem in statistical practice with several classical or current examples for each and abundant associated comments. It's an excellent source of material for teaching and a gateway to many fine examples and references. It would be a 5 star book, but explanations are about as terse as they can be - compare the author's one or two line definition of regression to the mean (it's almost as if they assume you already know what they are talking about) with the chapter given by Kahneman in Thinking Fast and Slow. Obviously, you couldn't have a book this size with this much in it and still give long explanations of key concepts, but they could be a little longer. In the same vein, how hard would it be to include a figure of the Fagan nomogram if you are going to discuss it? I still recommend it highly. I've already used it a few times for some excellent examples, which are often placed in an outline that places the phenomenon among similar statistical concepts (e.g. what other phenomena are related to Simpson's paradox).