[A] “counterintuitive” finding is more interesting, and thus more likely to get published, than an intuitive one. But maybe lots of our intuitive ideas are correct and the “counterinuitive” selection bias is obscuring that. Certainly this is a problem in punditry and (especially) magazine writing, where the key to getting a lot of column inches is to have an interesting idea rather than a true one.
Sullivan, we’re gunning for you . . .