So I've read through all the posts and no one's responded to this comment yet, despite it's clear problems, statistically speaking. It is inaccurate to compare statistics from one group to another if the possibility of sampling bias exists. This is why Internet polls suck so bad. Here the possibility is a certainty--that's why admissions at Harvard et. al are more competitive than those at UF.
There have been other posts that have debated ad nauseum whether the boost you get from a top undergrad outweighs the loss in GPA from competing with really smart people, but here's my view: it doesn't even come close. I know someone at Penn who scored a 14 for the semester of intro bio, most decidedly an F. If you ask him about bio stuff, he actually knows quite a bit, it's just that in a world drawn from straight A in AP classes high school students, the overall level of the class is really quite high.
Furthermore, when people speak of grade inflation in the Ivies, they are neglecting to consider a form of Simpson's Paradox (no, not Homer, the other Simpson): while overall grades may be inflated, grades in hard science majors are definitely not, at least not by the standards of the last few years. At Penn most are curved around a C+; even bio courses like "Animal Behavior." Social sciences and small english seminars tend to be the worst. Language classes are fairly harshly graded, as are many history classes and literary theory classes. 'Curving around an A,' like I read in another post, is an absurdity. The worst that happens is a curve around a B+, and even that is truly rare.
--Ari
