Well, part way through writing a paper about statistical methods used to compare universities across Canada (the Macleans rankings), I've just about made it through discussing how Enlightenment skepticism a la René Descartes undermines the authoritativeness of such statistics no matter how accurate they become. Specifically, because the entire methodology is predicated on falsifiability and the reduction of doubt, and that we consumers of the rankings information did not collect or analyze the data ourselves.