Arizona Republic reporter Michelle Ye Hee Lee reports:
As the U.S. Department of Veterans Affairs struggles to redeem its reputation after a data-manipulation scandal, it faces new scrutiny over a book used to train up to 500 employees a year: “How To Lie with Statistics.”
Of course those of us in statistics education know that, as Lee puts it, “The title is intended to be ironic and reflects the cheeky, almost farcical, tone [author Darrell] Huff carries through the book to show readers the many ways statistics are manipulated.” So it kinda misses the point to consider the use of this book as a scandal.
At another level, though, I think there is a problem with the near-universal acclaim in which the “How to Lie” book is held within the statistics profession, and I in fact would not again use it in a class. My problem is not with the book’s message (as Lee notes, the “how to lie” message is ironic) but because of its quality. It is no slam on Darrell Huff — he was who he was — but he had neither training, nor expertise in statistics. He was a journalist who wrote books in the 1940s through the 1960s with titles such as “Pictures by Pete,” “The Dog That Came True,” “How to Take a Chance,” “Score: The Strategy of Taking Tests” and “How to Lower Your Food Bills.”
“How to Lie With Statistics” was a good concept for a book, and Huff pulled it off well. But we have to remember that, ultimately, he didn’t know what he was talking about. The book contains a mixture of good and bad advice and its efforts of criticism are scattershot with little relation to the substance of the material being criticized.
One way to see this is to look at a lesser-known project of Huff: his work doing publicity for cigarette companies, something for which there is ample documentation in the Legacy Tobacco Documents Library and that I discussed a few years ago in an article, Statistics for Cigarette Sellers. At one point in the late 1960s, Huff came close to writing a book called “How to Lie with Smoking Statistics,” funded by the cigarette companies. At this point I’m guessing that the irony of the title was getting bitter. I think he dodged a bullet by scuttling this book. If “How to Lie with Smoking Statistics” had come out, I expect it would have destroyed his reputation — remember, this was 1969, five years after the surgeon general’s report — and taken a big bite out of the later sales and reputation of his 1954 bestseller.
My point is not that Huff should be tarred (as it were) for his work for cigarette companies. The list of actual distinguished statisticians who have worked for the cigarette industry is long, including such luminaries as Joseph Berkson, Ronald Fisher, Joseph Fleiss, James Heckman, Donald Rubin, and Ingram Olkin.
My point is that the exact approach used in “How to Lie”– picking out apparent absurdities in uses of statistics and encouraging broad-based, undifferentiated skepticism — is what Huff did in his paid work for the cigarette companies. As historian Robert Proctor described it:
Huff testified at hearings on cigarette labeling and advertising, accusing the recent surgeon general’s report of myriad failures and “fallacies.” Huff peppered his attack with amusing asides and anecdotes, lampooning spurious correlations like that between the size of Dutch families and the number of storks nesting on rooftops—which proves not that storks bring babies but rather that people with large families tend to have larger houses (which therefore attract more storks).
So, no, I won’t be recommending “How to Lie with Statistics” to my students. Statistical communication is important, and so is criticism, but it becomes empty if it is used indiscriminately.