http://www.nytimes.com/2011/02/01/health/01mind.html
The ecological fallacy is my Favorite public health concept. This New York Times article dispensing advice to Super Bowl fans illustrates it perfectly.
At its simplest, the ecological fallacy is when population level data is interpreted at the individual level. Here, the researchers take a city's mortality data post-Super Bowl and interpret it to mean that increasing stress coupled with high-rates of fatty-food intake common amongst Super Bowl fans means that a loss causes heart attacks etc when the city's team loses. Unfortunately, their study as it is cannot respond to their hypothesis. Their data does not tie those that died to the Super Bowl. There is no reason to know that those experiencing post-Super Bowl deaths had watched the game, had an emotional tie to the game, or even had any knowledge of the game.
That is fine. It could be important to know that mortality rates spike post-Super Bowl loss, but there is no grounding (yet) to say that any of these interventions will have an effect. This study relied on one city and one four year difference. Did something happen in LA around the beginning of January in 1980 or 1984 that could effect mortality rates? Is LA 1980 and 1984 the same as the rest of the country in 2011? We need more cities with additional win/losses to compare. That can't overcome the ecological fallacy of city data being used to issue individual fans' care plans, but we can at least have a surer more tested evidence base for recommendations.
I get that this article is a joke, but ecological fallacies just aren't funny. When you read a study authored by MD's, you must remember they likely haven't taken Epidemiology 101.
No comments:
Post a Comment