<p>
</p>
<p>I hope you’re joking. He wasn’t trying to “help” the Jews by keeping them from being admitted.</p>
<p>And anyway, how can you not see the difference between wanting to include more diversity, and wanting to “cut down” representation of a specific group? It’s inconceivable to me that so many smart people don’t see the distinction.</p>
<p>Here’s a test - suppose, tomorrow, all Asian Americans were to boycott / not apply to the Ivies. At all. Not a single one. </p>
<p>Would the Ivies be concerned with this and want to do outreach? That tells you whether there’s “discrimination” or not.</p>
<p>Because you know darn well if no Jews had applied to Harvard in the 1920’s, Lowell would have been secretly pleased as punch.</p>
<p>eh, collegealum314, if I did Nobel Prize winning research today and the End of the World came tomorrow, I wouldn’t win the Prize anyway.</p>
<p>Here is what I would do, as a first investigation of the possibility of discrimination at a school that does release the racial composition of the set of admitted students, but provides nothing else: With a bit of statistical analysis, one could estimate the expected magnitude of fluctuations in the numbers of admitted students. If the composition in a given year is j% group P, k% group Q, m% group R, n% group S, and g% group T, then assume that those are the mean numbers, and that selection is random, conditioned on those means. It should be possible then to estimate the range of the number of admitted students in any group P, Q, R, S, or T, at (say) a 95% confidence level. If the fluctuations in the numbers of admitted students in the different groups are much less than predicted, it would suggest to me that some sort of soft-quota system is in place.</p>
<p>What I have suggested in post #1123 is analogous to the way that people deduced that Gregor Mendel had “cooked” his results–they kept coming too close to the predicted percentages. Or at least so I have heard. I suspect that Mendel was a keen observer, and selected the subset of traits that fit Mendelian genetics quite well. Whether he then also threw out a few plants from his data set to make the numbers prettier, I don’t know.</p>
<p>
The commenters on Harvard’s statement at the NYT site are mostly critical, to put it mildly.</p>
<p>
</p>
<p>Using this method, I suspect the first thing you would notice is that there is a soft quota system in place by gender at a lot of schools.</p>
<p>
</p>
<p>But selection ISN’T random. If I’m reading this right, your analytical approach would work if admissions claimed they used a lottery and you suspected that they “cooked” the lottery somehow (e.g., instituted a quota for or against a group instead of letting the lottery fall as is).</p>
<p>You’d probably also get a soft-quota on kids from New York and New Jersey and California, too, who are “crowding” the admissions fields. Is that “discrimination” against them? They can no more change their state of residence than they can their race. (Well, yeah, adults can move, but practically speaking, high school seniors cannot.) </p>
<p>What if you found a soft-quota on number of math majors or violin players?</p>
<p>What if you were to run this statistical experiment only base it on the number of people who score over 2300 on the SAT each year by ethnicity. I wonder what the size of the variation in number of examinees above that threshold would be? Of course, I have no idea.</p>
<p>
That’s what I was thinking as well. Assume you found quotas in place for several groups. I understand some people believe any sort of quota like this is repugnant, but set that aside for a moment. </p>
<p>What criteria would you employ to decide which group was being discriminated against?</p>
<p>is the 2300 in the new way or old way before the super score?</p>
<p>I think there is a soft quota by gender at some schools. There have been clear statements to that effect by admissions personnel at some of the liberal arts colleges–Kenyon comes to mind.</p>
<p>Pizzagirl, you are certainly right–selection isn’t random (at least not entirely random). I believe that the fluctuations in the numbers of each group should be <em>larger</em> than expected on a random basis (from year to year) as a result of the non-randomness of admissions. My test focuses on smaller than expected fluctuations (just like the examination of Mendel’s results).</p>
<p>I suspect that one might find geographical quotas. Actually, that would be very interesting to explore. Given the small numbers of students from some of the states (not the “disfavored” ones), statistically one would expect rather wild fluctuations in their numbers from year to year. </p>
<p>Would one find a soft quota on violin players? Personally, I would guess not. However, there might be a quota on top-notch violin players. Math majors? I don’t know. At Princeton, I think that only about 20% of those who initially intended to major in math wind up doing so.</p>
<p>bovertine, #1129, you are absolutely right that one can’t tell from my suggested test which group is being discriminated against (or groups).</p>
<p>Also, I like your suggestion in #1128. Those numbers probably are available. That would provide a useful check on the estimate of expected fluctuations (although if the percentages are provided only to whole figures, it might obscure the magnitude of the fluctuations).</p>
<p>
</p>
<p>If HYP decides that they just don’t want too many violin players, is that discrimination against violin players? Is it discrimination against an ethnicity if that ethnicity for whatever reason is overrepresented in violin players? If a certain region of the country happens to disproportionately tend to play violin, is that discrimination against that region of the country?</p>
<p>
I knew I could count on you to throw a fly in the ointment TPG. :D</p>
<p>@PCHope, 1103,
</p>
<p>Adcoms have to try to enroll a class which will take advantage of all the university’s resources. That means to a certain extent they have to admit people who have varied talents and interests. No one can develop all their talents to unusual levels in high school. The applications to colleges reflect the students’ interests.</p>
<p>To take an extreme example, if the 1,000 highest scoring students are valedictorian premeds from the New York metropolitan area, who have studied French, play the flute, and don’t play contact sports, admitting only by test score would lead to an imbalance in the demands placed upon the university’s resources. The Glee Club, Spanish department, Political Science department (etc) will be underutilized. </p>
<p>@JHS,
</p>
<p>More lab space, but the same number of undergraduates may translate to a higher percentage of the student body interested in majoring in the sciences.</p>
<p>I think the issues that you raise are very good in #1133, Pizzagirl, but I don’t actually think that a soft quota of violin players is likely. I would guess that there are too many people who play the violin reasonably well (thanks to Suzuki), but who will not be in the running to be concertmaster in any of the Harvard orchestras. It is probably a minor EC plus to play violin pretty well, but doesn’t give any “credit” beyond that. I doubt that there is any cap on violinists.</p>
<p>Re #1130 and #1134: CB doesn’t publish superscore statistics that I have ever seen. So I’m saying single sitting.</p>
<p>Holy Sheesh.<br>
Anyone who has followed this thread throughout must see a whole lot of hypothesizing based on very little info available. </p>
<p>Btw, in my experience, there is no paper trail. </p>
<p>Quant, just a question: have you ever volunteered to read STEM apps for your school?</p>
<p>
</p>
<p>Why people would be troubled by discrimination against people based on their race and not troubled by discrimination based on their developed abilities – academic, musical, or athletic – ought to be obvious.</p>
<p>I’m really torn here (not that it matters, I get that) because there’s all the supposing and smelling dead fish- and nearly everyone here is outside the walls. I jokingly want to suggest maybe these colleges really like Asian Americans and take as many as they can til the seats fil up, leaving just a bit for URMs and Whites- after all, they take 4x the representative sample.</p>
<p>In a world where there is a stats bar (and much lower than assumed, depending,) and after that a tsunami of applicants-- in a world where stats are only that first bar and not the determining factor-- why assume Asian Americans are being discriminated against based on race? Maybe, in the swarm of applicants, not every single high stats Asian Am candidate has the “it factor.” Is that so hard to imagine, when you get past stats? (Equally applies to many, many high stats kids.)</p>
<p>Why assume all CA’s are somehow magically equal? (In order to base an argument on AA stats or even the number of freshmen, you have to make that assumption.) </p>
<p>I know, I know, party pooper.</p>