Saturday, August 27, 2005

Matt Yglesias: Not a Bayesian

In the course of Matt's discussion of The Bell Curve, he mentions this:
"Everyone knows" that women are less competitive than men, so mere scraps of evidence can convince people that evolutionary psychology explains this fact, while a tenuous ev-psych account of why women are genetically more competitive than men would face a lot of skepticism since "everyone knows" the conclusion is wrong. ... But just as real science produces results and not just carping, one should expect a serious research program into the genetic bases of human behavior to produce some surprising conclusions along with some banal ones.
Perhaps, but this is simply looking at scientific evidence in a Bayesian manner. If your prior is that women are less competitive, you're going to need a lot of evidence to the contrary to change your posterior. This explains what Bayesian statistics is and why it is useful.

0 Comments:

Post a Comment

<< Home