<p>Bravo Joe and Ben! What I find odd about this is that it is Caltech’s choice which is beleaguered. It’s not as if the vast majority of schools are unwilling to try AA and other such standard-adjusting ideas. It is Caltech which is the extreme minority. In one year, I heard that the accreditation agency actually threatened Caltech with non-accreditation because it wasn’t practicing AA hard enough. Yet Caltech is one tiny outlier in the world of academic behemoths.</p>
<p>One more thing: The claim that the scores of all admits to the top two or three schools is high enough that the pool is mostly indistinguishable is false. Even at these levels internal documents suggest that students scoring below a 700 on the SAT Math are more likely to do worse than those with a 780-800. The correlation may not be high, but it is there.</p>
<p>A more recent study of the top 5 PhD programs in econ found an amazing result. Even in a world where most everyone got between a 770-800 on the GRE Quant, GRE scores still had a significant correlation with first year course grades and likelihood of completion. It seems the results were heavily driven by the few students who got in with scores below a 750. The top 5 Econ PhD programs are probably more homogeneous academically than either Caltech or MIT partly due to self-selection. Yet the homogeneity of the student body, the clear focus of their work, and the extensive admission filtering should have rendered these miniscule differences moot. Apparently not. </p>
<p><a href=“http://www.aeaweb.org/annual_mtg_papers/2007/0105_0800_0603.pdf[/url]”>http://www.aeaweb.org/annual_mtg_papers/2007/0105_0800_0603.pdf</a></p>