Another in the genre that began with the Darrell Huff’s 1954 best-seller, How to Lie with Statistics. If history is any guide, it will likely be ignored by those who do the lying.
In his first book for nonacademic readers, Smith (Economics/Pomona Coll.; Essential Statistics, Regression, and Econometrics, 2011, etc.) delivers an entertaining primer on his specialty, packed with figures, tables, graphs and ludicrous examples from people who know better (academics, scientists) and those who don’t (political candidates, advertisers). “We live in the age of Big Data….Sometimes these omnipresent data and magnificent computers lead to some pretty outlandish discoveries,” writes the author. We hear that children who play competitive sports are confident, so sports must build character. Selection bias makes nonsense of this if only confident children choose to play competitive sports. Enthusiasts tell us how to live to the age of 100, run a profitable business or enjoy a lasting marriage. However, all examine those who have succeeded, ignoring the losers, so survivorship bias renders their advice worthless. Few can resist the fallacious law of averages. If a coin flip turns up 10 heads in a row, the 11th flip is not more likely to be tails. If you fly regularly, the odds that your plane will crash do not increase. Good and bad luck do not even out. Chance is just chance. The Texas sharpshooter peppers the side of a barn and then draws a bull’s eye around the densest clump of holes. In other words, even honest observers find patterns in random data and can’t resist explaining them. We believe these stories if they seem reasonable and love them if they’re provocative—see Freakonomics, whose authors have admitted some mistakes.
“We are too easily seduced by explanations for the inexplicable,” writes the author in this amusing, informative account of how many arguments are backed by meaningless statistics.