Statistics Aren't Science
Statistics are the lingua franca of junk science. They make good sound bites, adding a quantitative feel to otherwise "fuzzy" health scares. Credibility is added ostensibly by a statistic's neutral nature and authoritative source. The result is an inappropriate transformation of a likely meaningless number into conventional wisdom.
When you see statistical studies:
-
Demand Definitions: Former surgeon general Antonia Novello claimed in a May 1990 speech that 3,000 "kids" start smoking every day (13). The statistic originated from researchers at the U.S. Centers for Disease Control and was published in the Journal of the American Medical Association in 1989 (14). If you read the actual study, you'll discover that only persons aged 20 years and older are included. Surely you understand why anti-smoking advocates don't claim "3,000 twenty-something kids start smoking every day".
-
Know That Statistics Don't Prove Cause and Effect: EPA claimed in 1996 that fine particulate air pollution kills 20,000 Americans annually. The basis for the estimate was a statistical study comparing death rates among geographic areas with varying levels of pollution. There were no clinical evaluations of any of the deaths included in the study and the researchers didn't know whether air pollution caused or contributed to any of the deaths or have any idea of how much fine particulate pollution any study subject inhaled.
Look under the statistical rocks. You'll be surprised what crawls out.
-
(13) "Smoking is Too Often a Child's Decision," St. Petersburg Times, August 5, 1990
-
(14) J.P. Pierce, M.C. Fiore, T.E. Novotny, E.J. Hatziandreu, R.M. Davis, "Trends in Cigarette Smoking in the United States: Projections to the Year 2000," Journal of the American Medical Association, January 6, 1989, pp. 61-65
