250 doesn't sound like a large enough statistical sample to eke out even the most trivial performance differences in a single person when you consider the size of the companies being discussed.