I will synopsize this article.
Its very clear from the article that most of the academically superior students of both sexes are passed over for kids at least a standard deviation down the scale. The author begins with analyzing the admission of Asian and Jewish students relative to the general population and contrasts this with the NMSF composition. For example, just 1000 of the NMSF students were Jewish yet 3000 got into the top 3 Ivies. That means that 2000 of the non-Jewish NMSF were denied. A comparison to Caltech to NMSF shows that Caltech admits solely on merit and the relative numbers by race from Caltech deviate by several orders of magnitude from the Ivies relative numbers.
Its also clear that most admissions officers lack the high caliber academic credentials that the average student applying possesses. Thus they are unable to develop a clear, objective strategy, and then apply it to the kids they see. The result is a very skewed process that allows too many unqualified kids and greatly reduces the number of very superior students. The author cites books by the same admissions folks and statisticians.
The author then asks what is the solution. The author looks at several alternatives then discards each because each would then leave out kids with many talents other than just academics and notes that many kids will do well above a certain threshold. EO Wilson is a notable example. The author comes up with a two-tier system. First tier is to admit 10-20% solely on merit. Ie rank kids solely by academic merit and take the top x%. The remaining seats are filled by lottery from the list of kids who meet the threshold.
Think what this means. NMSF is the 99.5 percentile. 1000 kids at 99.5 means there are 200,000 students in that group and if you pull 2000 from that group that means you pull from the 98th percentile. There are huge differences between the 99.5 percentile and the 98th in terms of performance at the highest levels. Its a huge disparity.
Last edited by Austin; 11/29/12 01:03 PM.
