The numbers may be against us because reporters and readers too often fail to dissect the statistics in news reports.
A large audience at the MAA's Carriage House Conference Center on Oct. 28, 2008, absorbed this message when mathematician Rebecca Goldin of George Mason University illustrated how the media miss the mark in the use and presentation of statistics in stories about the economy, health, science, and education. Goldin titled her talk "Spinning Heads and Spinning News: Statistics in the Media."
But that doesn't let readers and consumers off the hook. They need to be aware that inaccurate representations of science shape public policy and legislation and affect people's choices, Goldin warned. Everyone should have some understanding of statistical concepts and their use in such fields as epidemiology and toxicology.
Key concepts include: the meaning of statistically significant; causation versus correlation; relative risk versus absolute risk; scales and orders of magnitude; and margin of error.
In her role as research director for the Statistical Assessment Service (STATS), Goldin described journalists' failures to grasp these essential concepts in dubious stories about drugs in drinking water and the risks of infant formula (versus breastfeeding).
In some cases, however, the problems may originate in the scientific studies themselves. "Bad science has consequences," Goldin noted.
In one example, Goldin showed an ad contending that hot dogs cause cancer. A CNN report on the ad did point out that the ad was sponsored by the Physicians Committee for Responsible Medicine, which is "an animal rights group that wants us all to be vegans." Nonetheless, CNN went on to affirm the apparent link between processed meats (nitrites, in particular) and cancer, citing research published in reputable journals. However, it did not mention that other organizations, including the American Medical Association, had looked into the possible cancer-nitrites link and determined that the available evidence did not confirm any such association. The studies all had significant flaws.
Politics can also play a role in science, Goldin argued. She described several political actions seemingly based on science, including the recent move to ban phthalates from toys, even though there is little evidence that the route to exposure is through toys. Goldin also discussed coverage of estimates of the number of "excess" deaths in the Iraq war; responses to reports of apparent differences between men and women in "intrinsic" math ability; and flawed polling.
Goldin noted that she is aware that journalists face deadlines and do not have, or don't take, the time to research every story. But that doesn't justify inaccuracy in print, on the airways, or online.
STATS is trying to increase numeracy in the media, encouraging a higher standard for journalists working on stories that deal with science and statistics. Members of the organization write op-eds, speak on radio shows, fire off letters to the editor, advise journalists, run workshops, provide educational materials, and testify before Congress. The organization even presents yearly "dubious data" awards.
Goldin indicated that journalists welcome a resource tool such as STATS, so much so that it has been featured on NBC's "Nightly News," "The NewsHour with Jim Lehrer," ABC's "20/20," the New York Times, and the Wall Street Journal. —H. Waldman
This MAA Distinguished Lecture was funded by the National Security Agency.