How should journalists cover advocacy groups that promote allegedly scientific "studies" but whose results have not undergone peer review and grind that particular group's ax? Boyce Rensberger, must have known what to expect from the panel of speakers that he and Doug Levy lined up for the heavily attended session, "When Advocacy Groups Do Science," they presented to the District of Columbia Science Writers Association (DCSWA) September 17, at the Freedom Forum, Arlington, VA. His first act as ringmaster of ceremonies was to place a kitchen timer next to the microphone on the speaker's lectern.
The panel consisted of highly articulate leaders and critics of advocacy science: Michael Jacobson, executive director of the Center for Science in the Public Interest(CSPI); Kenneth Cook, president of Environmental Working Group, a nonprofit environmental research organization; David Allison, a Columbia University statistician who has studied media coverage of alleged health risks; Sylvia Rowe, president of the International Food Information Council, a food-industry group; and Steve Milloy, statistician, lawyer, president of the Environmental Policy Analysis Network, and proprietor of the "Junk Science" homepage on the Web.
Rensberger addressed several general questions to the panel:
It was clear from the initial presentations why each panelist had been chosen. As backgrounder, Sylvia Rowe described her industry-supported organization as one designed to "communicate science-based information on food safety and nutrition issues to health and nutrition professionals, educators, government officials, journalists and consumers." Food-safety information, once primarily conveyed by doctors to patient-consumers, she noted, has become "wellness information" conveyed directly to the consumer by the media and health professionals (who get much of their information from the media as well). Running against the timer, she listed several anomalies in media coverage: Although diets low in fiber and high in fat have both been implicated in cancer incidence, a three-month survey of a range of newspapers yielded 47 stories on dietary fat and cancer and none on low fiber. Consumers, she complained, would have difficulty making intelligent changes in diet based on available media coverage, often because the reported study was not placed within the context of a body of knowledge required for an intelligent lifestyle decision.
Michael Jacobson, representing an extraordinarily successful advocacy group, claimed that history has confirmed many CSPI concerns: sulfites, sodium nitrite and nitrate, and baby food fillers. After much derision from the meat and dairy industries, he continued, not only have federal dietary guidelines caught up with CSPI studies on the relationship between diets and chronic diseases, but even the industries themselves have come around. Publishing in peer-reviewed journals, he scoffed, may be laudable, but it is hardly an accurate test of quality. CSPI, he said, did not need to wait a year for its study on the fat content of typical restaurant meals to be published in a peer-reviewed journal; "We just buy Fettucine Alfredo from 12 different restaurants." One complaint to the media: CSPI studies with positive results get too little coverage.
David Allison, as skeptical statistician, suggested paying less attention to credentials and more attention to data, less attention to conflicts of interest and more to the strength of the evidence and the method that generated it. Acting as a sort of fulcrum for the panel, he recommended that since all scientists are biased to some degree, journalists must learn how to test the internal validity of a study and take no opinion on faith or, contrariwise, through a presumption of bias. He especially warned against accepting a sampling of clinical opinion since physicians were, in his experience, notoriously incapable of evaluating statistical significance. Finally, he urged, ask if an effect is significant from a public-health context, or only as an fascinating statistical blip?
After blaming a lack of scientific initiative decades ago for many of the extraordinarily expensive clean-up projects now coming due, Kenneth Cook, representing environmental concerns, suggested that science journalists might also blame their own lack of initiative for a shrinking science newshole. You're spending too much time rewriting Science and Nature, he said, and missing the big story: Why are cancer incidence rates rising? (His remarks preceded the November 15 issue of the journal Cancer, which showed that the overall cancer mortality rate, adjusted for age, had dropped each year from 1990 to 1995, for a total decline of about 3.1 percent.)
Steve Malloy, supported primarily by the Department of Energy with some industry support, billed himself as a skeptic's skeptic and was notably scornful of the scientific community and its product. He shared Jacobson's disdain for peer-reviewed journals, but for the opposite reason. For example, he questioned whether any conclusions about the hazards of pesticides in drinking water are significant absent significant morbidity data. He also questioned recent assertions of global warming and increasing incidence of prostate cancer and directed his listeners to his appropriately named URL: http://www.junkscience.com.
Boyce Rensberger, a reporter and editor for The Washington Post, and Doug Levy, a science reporter for USA Today, are both members of the DCSWA board. The host Freedom Forum also provided food and drink.
-- H.J. Lewis