I just glossed the article and thought seems far-fetched to me and moved on. Over at the Freakonomics blog, Steven D. Levitt did what he does so well: Examine the data and explore how it might yield false positives. I don't always agree with Levitt and Dubner's conclusions, but I love the way they approach data—they suspect it of trying to trick them, which John Atkinson assures me, is the only way to approach testing.
Read…