Damned Lies and Statistics: Untangling Numbers from the Media, Politicians, and Activists

Joel Best

Joel Best's lucidly written little book is a primer of critical thinking about the ways statistics are invoked in the mass media. He proposes that widespread innumeracy in the United States causes most people to be intimidated by numbers. Reporters and journalists share in this intimidation and further it. Repeating the numerical opinions of experts with respect to social problems is quicker and easier than doing investigative journalism. Big numbers seem to certify the existence of big problems, and big problems are sensational. Numbers are associated with science, and science is popularly associated with the truth. Best argues that statistics have become a fetish object in contemporary culture. His project is to demystify them for a popular audience.

He begins with some elementary lessons in the construction of social problems and with the ways in which measurement practices confound the portrayal of social conditions. Using the familiar examples of statistics of marriage and suicide, he shows how matters of definition, classification, and the mechanics of reporting and record keeping lie at the basis of numerical accounts. When faced with a statistical report, the astute reader must learn to ask by whom it was produced, for what reason, and by what procedure. The chapters that follow discuss the basic sources of bad statistics. These include people making guesses about numbers or defining problems in ways that are as broad as possible. They involve the posing of misleading questions and the selection of biased samples. Best next provides examples of what he calls "mutant statistics." He examines the logic of statistical comparison and considers some recent American debates over statistics. The book concludes by repeating a call for critical thinking when faced with media reports of numbers.

Best has clearly taken to heart one of his own aphorisms: "Never overestimate the understanding of an innumerate public" (p. 82), and, despite its generally well chosen and informative examples, the book frequently offers simple fare to anyone who has taken an introductory sociology course. The readers of this journal are not likely to be enlightened by being told that the questions asked in surveys influence the answers given, nor that summaries in terms of binary responses mask a more complex range of opinion. I assume most readers appreciate that international comparisons of educational test results have to be read in light of a knowledge of the tested populations. More interesting are Best's discussions of debates over crowd sizes or over census results, for here he provides more technical information. I particularly enjoyed his section on mutant statistics, that is, those usually bad numbers that take on a life of their own and circulate through a variety of media, changing shape as they travel, for example, that the suicide rate of gay and lesbian teens is three times higher than that of straight teens; that 150,000 young American women annually will die of anorexia nervosa; or that 90% of women have been victims of sexual abuse (to take a Canadian example, of which the book provides none).

On the whole, however, there isn't much here for an academic readership to sink its teeth into. The book seems aimed at the skeptical reader of the pulp press (2% of Americans have been kidnapped by aliens!!) or the high-school senior learning first lessons in media literacy (social problems are made, not taken), or the undergraduate who reads the newspaper as if it were a novel. These are all people Best considers on a par with the average journalist. On the other hand, when journalists do find themselves tempted to parrot official numbers as authoritative accounts, Best's work could encourage critical thinking.



  •  Announcements
    Atom logo
    RSS2 logo
    RSS1 logo
  •  Current Issue
    Atom logo
    RSS2 logo
    RSS1 logo
  •  Thesis Abstracts
    Atom logo
    RSS2 logo
    RSS1 logo

We wish to acknowledge the financial support of the Social Sciences and Humanities Research Council for their financial support through theAid to Scholarly Journals Program.

SSHRC LOGO