This piece by Anne Price and Ariel Goodwin seems compelling at first glance, but I’ve grown more skeptical the more I read over. In it, they show a purported link between mode of commute (driving or carpooling vs. walking or biking) and obesity rate. The maps and figures they post are convincing enough, but here’s where I begin to take issue with their analysis:
As the Oct. 2011 article pointed out, the relationship between sedentary travel and health outcomes can be misleading when additional contributing factors are not taken into account. While it is not our intent to claim a direct causal link between transportation modes and obesity rates, it is hard to deny the existence of some geographic patterns.
And yet that seems to be exactly what they are doing. Here’s my question: Why didn’t they include those variables? That would have more clearly placed the role of commute mode in context.