The science of statistics can help us think critically about a wide range of issues covered in the media, said statistical scientist Dr. Nancy Reid, who provided a statistical critique of several recent news stories at the university professor lecture series on Jan. 31.
One Globe and Mail article, titled “Asthma now hits one in 10 children, study says,” claims that “the rate of asthma has soared fourfold over the past 20 years.” The article’s subtitle read “Fourfold increase linked to air quality.”
Reid, however, noted that the author relied on an environmental report that referred only to the prevalence (the number of people with asthma) and which made no findings about the asthma incidence or rate.
“We have no data in Canada on the incidence of asthma, so although there is a higher percentage of people who now have asthma, according to the study, we don’t know if the rate is increasing or not,” she said.
Reid also pointed out that the environmental report contains no data on the prevalence of asthma 20 years ago, or on the link between asthma and air quality: “It’s very possible that asthma has been increasing, and [that] it’s directly linked to air pollution, but there’s nothing in the report that says that.
“There doesn’t seem to be much basis in the report for the kind of language that’s in the article,” she concluded.
Other articles included statistics on the health effects of mammograms, airbags, acetaminophen, and whether eating chocolate is good for you. Despite the flaws she identifies in these articles, Reid does not believe we’re experiencing a statistical epidemic. “I think by and large the media do quite a creditable job,” she said.
Reid also attempted to dispel some myths about her chosen field: “Most statisticians I know have a story…of being at a party and being asked what they do and they say they’re a statistician and it kind of brings the conversation to a halt.”
A common misconception is that statistics simply involves collecting numbers or doing math. She preferred a broader definition: “Statistics is a way of thinking about things that vary or that are unpredictable. The most important thing is thinking about what causes things to vary.
“Statistics is mathematics that is meant to be used,” she added.
Reid noted that technological advances, such as supercomputing, are having a dramatic effect on statistical science. One consequence is an increasing demand for statistical analysis in other fields, resulting in more collaboration between statisticians and other scientists. “The other sciences are collecting data at an unprecedented rate…it can be overwhelming.”
Is there anything we can do to better assess the statistics we encounter in the media? “I think the best guideline when reading statistical items is not to lose your common sense,” Reid prescribed.
Nancy Reid’s recommended reading:
J. Allen Paulos, A Mathematician Reads the Newspaper. (Toronto: Random House, 1995)
J.S. Rosenthal, Struck by Lightning: The Curious World of Probabilities. (Toronto: HarperCollins, 2005)
Chance, published quarterly by the American Statistical Association
Significance, published quarterly by the Royal Statistical Society
