<p>I thought 26% of applicants were Asian and 30% of the accepted students were Asian?</p>
<p>^^ I was being sarcastic in case nobody noticed…</p>
<p>
</p>
<p>The percentage of Asians in California is much higher than for the rest of the country. This is part of the reason why the percentage of Asians at UCB and UCLA is so high. I know it’s in the 40-50% range at Berkeley. As for Caltech, I don’t think the percentage of Asians is significantly different there from MIT. </p>
<p>Go ahead and look at the names and pictures of the physics and math olympics team. There are plenty of white people there, enough to suggest that there doesn’t need to be discrimination in their favor in order to admit them to MIT.</p>
<p>The only obvious differences between MIT and Caltech demographics is the minority population.</p>
<p>“Go ahead and look at the names and pictures of the physics and math olympics team. There are plenty of white people there, enough to suggest that there doesn’t need to be discrimination in their favor in order to admit them to MIT.</p>
<p>The only obvious differences between MIT and Caltech demographics is the minority population.”</p>
<p>Here are the websites for USAMO winners, US physics team, US top 20 chemistry Olympiad finalists, USABO team. In every category, Asian Americans made up at least 50% or more at each of the rosters ([USA</a> Mathematical Olympiad: 2009 USAMO Winners](<a href=“http://www.maa.org/news/051209usamo.html]USA”>USA Mathematical Olympiad: 2009 USAMO Winners | Mathematical Association of America), [United</a> States Physics Team Blog (AAPT)](<a href=“http://aapt-physicsteam.blogspot.com/]United”>http://aapt-physicsteam.blogspot.com/), [2009</a> Chemistry Olympiad Top 20 Finalists](<a href=“American Chemical Society”>American Chemical Society), <a href=“http://www.cee.org/sites/default/files/IBO_09_final.pdf[/url]”>http://www.cee.org/sites/default/files/IBO_09_final.pdf</a>). Even in this year’s Putnam math competition, which MIT won, 4 of 5 fellows are Asians. The top three math competitors from MIT are of Asians (or Asian decent). However, I am not saying than Asians are superior. But you claims that Asians are not discriminated in MIT, and I think the otherwise. Here is Caltech racial makeup for class 2009 ([Caltech</a> Undergraduate Admissions: Facts & Stats](<a href=“http://www.admissions.caltech.edu/about/stats]Caltech”>http://www.admissions.caltech.edu/about/stats)), whose admission has minimal political agenda, and here is the MIT student body, which throws in a bundle of populist politics in its admission process besides enforcing affirmative action ([MIT</a> Admissions: Incoming Freshman Class Profile](<a href=“http://www.mitadmissions.org/topics/pulse/incoming_freshman_class_profile/index.shtml]MIT”>http://www.mitadmissions.org/topics/pulse/incoming_freshman_class_profile/index.shtml)). As you can see, Asian applicants suffer most of the blunt from affirmative action, dropping by 13% (39% in Caltech to 26% in MIT), while whites by only 4% (40% in Caltech to 36% in MIT). Both schools have very similar caliber of applicants. One has to find a reason why Asians are not favored when affirmative action is implemented. I have to conclude that either Asians are just not as good as whites in personal qualification or simply that Asian applicants are subconsciously discriminated in MIT admission. The latter conclusion is actually supported by the finding by Thomas J. Espenshade of Princeton University that Elite colleges favor whites over Asians by 3 fold (<a href=“http://www.nacacnet.org/EventsTraining/NC10/Baltimore/educational/Documents/C313.pdf[/url]”>http://www.nacacnet.org/EventsTraining/NC10/Baltimore/educational/Documents/C313.pdf</a>).</p>
<p>I’m familiar with the Eppenshade study, although I haven’t read it. Did it specifically show that Asians with high SAT scores (say above 1500 or 1550) had a lower admission rate than whites in the same SAT regime? That would be a stronger indication of outright discriminatory bias then indirectly showing that differences in admissions rates are equivalent to that expected by 20-30 SAT points. The demographics of recruited athletes may skew things.</p>
<p>I was surprised by the demographics of MIT, and I went back to look at old Tech articles. It appears I’ve become out-of-date. The URM population was 15% when I entered (mid-nineties), and is now 25%. I can’t find the Asian percentage, but I think it was around 35% or so, about where Caltech is. Also, the percent of USAMO qualifiers that were Asians used to be a lot lower. I counted about 30 out of 150 for 1995, so that’s why I made that earlier statement. (As an aside, future MIT Putnam fellow Reid Barton was one of the qualifiers–as a 7th grader!) I guess things have changed.</p>
<p>I know that MIT’s former dean made some insulting remarks about Asians being “math grinds.” Also, I’ve heard that Harvard did an internal study which showed a bias, particularly in how well-rounded Asians are perceived in comparison with whites with similar activities. So in general I am receptive to such arguments. However, one point I will make is that Caltech has no such discrimination and their demographics aren’t close to 50% Asian, which is what I think UCB and UCLA has. In the case of UCB and UCLA, the fact that Asians make up a greater percent of California does partly explain their demographics. So I think Caltech is a better model of what the Asian population would be without affirmative action. </p>
<p>However, I cannot explain your main point as to why Caltech’s Asian population is 14% over MIT’s while there is only a 4% difference in caucasians.</p>
<p>
</p>
<p>I think both your and harvardfan’s views on what makes UCB have a high percentage of Asians combine to give the picture I have in mind. I think it’s generally been the case that UCB, as a public school, admits people with very high standardized numbers consistently – a 2300+ SAT/4.0 UW GPA with a solid schedule is very likely to be admitted, and certainly folks with slightly lower stats are too. I think it’s reasonable to say that among my Asian acquaintances, there is a more prominent “pure academic stats and nothing else” culture, which is almost exactly what gets one very high chances at UCB.</p>
<p>
</p>
<p>East-Asian Americans and “Asians” (esp. Indians) are rather different and should not be lumped together. As you go up the selection ladder, Indians perform spectacularly well – their representation rises, similar to or better than the same phenomenon for Jews – but the East Asian share steadily declines as the selectivity level rises. Other Asian groups such as SE Asians, Arabs, etc are in very small numbers at the USAMO level and beyond and there is not enough data to include them in the analysis. The real issue is what happens to US Chinese/Koreans versus US whites in these contests, and the answer is that the relative East Asian performance is mediocre at the upper levels, though absolute representation is of course quite high. </p>
<p>Among American students, East Asians are currently at about 55 percent of USAMO qualifiers, 40-50 percent of IMO training camp selectees, 35-40 percent of IMO team members, 30 percent (if that) of IMO gold medals, and 20-30 percent of the higher scores in the Putnam competition. They continue to get decimated at the further selections for undergraduate prizes, scholarships to graduate school, elite postdocs, tenure competition, and professional prizes. The pattern has been obvious to anyone in upper tier academia for some time now. The statistical effects are as strong or stronger as anything found in Espenshade’s studies, i.e., they imply odds ratios higher than what he measured, except that these are blind-graded math exams or other pure performance selections. Good luck selling “discrimination” as an explanation for this, outside of Chinese internet discussion boards.</p>
<p>
</p>
<p>The 2009 US Putnam cohort is drawn from a pool that had 50-60 percent East Asians (about 55 percent of USAMO qualifiers during 2006-2009, plus summer science camp attendees and other math/science/CS specialists). The “discriminatory” effect of the Putnam competition grading was then to reduce the East Asian American figure below 30 percent. It is similar within the MIT population after you subtract the foreign IMO gold medalists. After making this calculation it would be hilarious to compare the odds ratio with Espenshade’s numbers.</p>
<p>“Among American students, East Asians are currently at about 55 percent of USAMO qualifiers, 40-50 percent of IMO training camp selectees, 35-40 percent of IMO team members, 30 percent (if that) of IMO gold medals, and 20-30 percent of the higher scores in the Putnam competition. They continue to get decimated at the further selections for undergraduate prizes, scholarships to graduate school, elite postdocs, tenure competition, and professional prizes. The pattern has been obvious to anyone in upper tier academia for some time now. The statistical effects are as strong or stronger as anything found in Espenshade’s studies, i.e., they imply odds ratios higher than what he measured, except that these are blind-graded math exams or other pure performance selections. Good luck selling “discrimination” as an explanation for this, outside of Chinese internet discussion boards.”</p>
<p>This guy has a habit not providing the sources of the evidences but love to fudge facts. For those curious, you can check on the debate on CC here (<a href=“http://talk.collegeconfidential.com/college-admissions/790609-do-elite-colleges-discriminate-against-asian-students.html[/url]”>http://talk.collegeconfidential.com/college-admissions/790609-do-elite-colleges-discriminate-against-asian-students.html</a>).</p>
<p>
</p>
<p>Certainly there are posters here who, as a familiar and all too predictable gambit, claim data fraud when confronted with math competition results that “hurt the feelings of the Chinese people”, insofar as those competitions show the East Asian numbers being decimated by blind-graded math exams. The data itself is much harder to dispel with such flimsy and transparent dismissals.</p>
<p>If you actually check that earlier thread (start at page 50 or so), you will find that my comments about the math competition data were confirmed by a count of over a thousand 2006-2009 USAMO, MOSP, and IMO qualifiers performed by a Chinese CC user named “NCL”. NCL initially thought the data refuted the declining East Asian representation, but he was then shown to have improperly handled (let me not say “fudged”) the data, and the corrected numbers supported my assertions. After this was seen, numerically, to be the case, NCL then did fudge the statistical significance levels downward, was caught doing so, and switched to personal attacks. The thread was locked due to length (1000+ postings, fewer than 10 of which were about the East Asian USAMO data) but if NCL, harvardfan or others from the China Pride faction have any remaining statistical disputes I will be happy to continue the public airing of the numbers.</p>
<p>By the way, I just posted in this MIT forum the US-only Putnam competition data for 2009, which (of course) qualitatively and quantitatively bear out the numbers I posted above, i.e., the US East Asian share being sliced from about 50-60 percent of USAMO qualifiers each year from 2006-9 (NCL’s data that Harvardfan doesn’t dispute) to 20-30 percent of the upper Putnam scores last year.</p>
<p>Also, US Asians aren’t over-represented in MIT’s high scores compared to whites (here including Indians as “Asian” to match MIT’s enrollment and admission data): none in the top 5, 1 of 3 in the top 15, 1 of 4 (25%), in the top 25, 7 of 18 (39%) in the honorable mention or above. The numbers excluding the anomalously high-performing Indians (i.e., comparing whites to East Asians) are: 1 of 3, 1 of 4, 4 of 15. These are at or below the Asian-to-white ratios for the MIT undergraduate population.</p>
<p>More data for harvardfan, NCL and others claiming discrimination against Asians.<br>
2010 national olympiad qualification results are now available.</p>
<p><a href=“http://www.unl.edu/amc/e-exams/e8-usamo/e8-1-usamoarchive/2010-ua/10-Qual_list.pdf[/url]”>http://www.unl.edu/amc/e-exams/e8-usamo/e8-1-usamoarchive/2010-ua/10-Qual_list.pdf</a></p>
<p><a href=“http://www.unl.edu/amc/e-exams/e8-usamo/e8-1-usamoarchive/2010-ua/10-JMO-Qual_list.pdf[/url]”>http://www.unl.edu/amc/e-exams/e8-usamo/e8-1-usamoarchive/2010-ua/10-JMO-Qual_list.pdf</a></p>
<p>A quick count using NCL’s methodology (US students only, count Chinese/Korean/Japanese/Vietnamese names as “East Asian”, ignore SE Asians and Indians), shows that:</p>
<p>67.2 percent (148/220) of US qualifiers for the national Junior Math Olympiad (USAJMO grades 10 and under), are East Asian.</p>
<p>57.4 percent (151/263) of US qualifiers for the much more advanced USAMO (grade 12 and under, selector for the IMO team training camp), are East Asian. </p>
<p>Let’s express this in Espenshade’s terms, using odds ratios. Switching the selection level from the grade 10 to grade 12 lowers the East Asian to non-East-Asian ratio from 148/72 to 151/112, for an odds ratio of 65%. That is, the relative chances of East Asians to get past the selection hurdle go down by about 35% when the bar is raised. This is a stronger discrimination effect than the 29 percent (0.712 odds ratio) that Espenshade-Chung-Walling quantified as “equivalent to 50 SAT points” in their study of 1997 admission data. I encourage anyone interested to repeat the counts and the calculations for themselves.</p>
<p>Adjusting this for the time difference between 10th and 12th grade, with 2 years of small increases in the US Asian/white population ratio, would not change the story, as population doesn’t change all that quickly. It would be very amusing to see whether those posters, such as NCL, who claim a rising tide of Asian demographics are skewing the numbers, could cite population figures that would even reduce the effect down to the Espenshade 50 point level, i.e., mitigate a 35 percent underperformance down to 29 percent.</p>
<p>The bottom line is that, in a large national sample, going from the 10th to the 12th grade level of competition reduces the relative chances of US East Asians by (at least) 25-30 percent. And this continues with the IMO medals, the Putnam contest, etc, etc all the way down the line. I’m sure that harvardfan, NCL and the rest of the posters asserting Asian discrimination (at MIT or other top schools) have a convincing explanation of all this. Admissions offices have a mountain of data on past performance of admitted Asian students, and the admission itself looks to anticipated future performance, not past decorations on the resume.</p>
<p>Further context for the numbers:</p>
<p>
</p>
<p>Separating the data by grade level sharpens the pattern. </p>
<p>46 percent (46/99) of US grade 12 qualifiers for the USAMO are East Asian. (There were 44 clearly E.A. names and another 1-2 that I considered ambiguous but marked as Asian for present purposes.).</p>
<p>As a comparison, the winners of the grade 8 national competition (“Mathcounts”) are, easily, 70 percent East Asian (or other such figure enormously higher than in the high school competitions for the same cohort) and this has been the rate for many years.</p>
<p>We see, again, the steady decline in the Asian representation as the grade level and selectivity level increase. The main explanations are clear, as they have been for decades:</p>
<ol>
<li><p>the advantage of starting early is progressively nullified as students age. </p></li>
<li><p>the more stringent the selection for talent, the more the population of hard workers (however talented they may be) will tend be decimated. Hard work will of course increase the absolute number of hard workers who reach any level of performance, but the harder they work compared to others, the more their objective performance will appear to be discounted (Espenshade-style statistical signs of discrimination will appear) under a pure merit selection.</p></li>
</ol>
<p>Additional factors enter at university and beyond, but the two listed above are enough to indicate that any selection process that tries to infer talent or predict future performance, will tend to under-select Asians compared to their numbers in the pool, and also in comparison to their numbers at any given level of measured performance. </p>
<p>Of course, it could be that universities are being too generous to Asian credentials, perhaps in fear of discrimination lawsuits. Statistical signatures of this would include academic underperformance by Asians, which is exactly what Espenshade found and quantified (but the media have ignored) in his recent book. His methodology was identical to the one he used to quantify the Asian-versus-white admissions “disadvantage”.</p>
<p>Student Body for mit…</p>
<pre><code>* 45% Women
-
55% Men
-
1% American Indian/Alaskan Native
-
26% Asian/Pacific Islander
-
9% Black/Non-Hispanic
-
14% Hispanic
-
36% White/Non-Hispanic
-
8% Non-Resident Alien
-
6% Race/ethnicity unreported
</code></pre>
<p>that 1800 something sat score that a prev poster mentioned was well under the 25 pct of the class. Not a urm, but obviously had something going for him that mit liked. typically at other schools, people under the 25 pct ile mark of tests are athletes, legacies, donor-kin, or urm. </p>
<p>maybe founded a charity or built some kind of invention? (hint, op, build something , an invention, a product, a company - that’ll get their attn.)</p>
<p>does mit have sports , and does it care enuf about sports to offer slots to athletes?</p>
<p>SAT Math 751 average
720-800 range of middle 50%
SAT Critical Reading 702 average
660-760 range of middle 50%
SAT Writing 702 average
660-750 range of middle 50%</p>
<p>(that wd be a 2040 on the low end, right? i am more familiar with act)</p>
<p>ACT Composite 32 average
31-34 range of middle 50%</p>
<p>
</p>
<p>I’m not sure if this is PC or not, but the core curriculum for CalTech is a lot more demanding than that at MIT. Like, they require 5 quarters of physics and math compared to just 2 semesters each at MIT, and I’m pretty sure each quarter of physics/math is the same as a semester at MIT. And not only that, but the beginning calculus class is proof-based, so you basically need to have prior exposure to calculus before coming to CalTech, which isn’t really true at MIT. So CalTech needs to discriminate a lot more based on high school background than MIT, or otherwise people won’t pass their classes. I have a feeling this kind of explains part of the differences in racial distribution between MIT and CalTech.</p>
<p>^ As a Californian, I have a different hypothesis regarding the higher proportion of Asian students at Caltech. I think self-selection plays a role:</p>
<p>California’s population has a relatively large proportion of Asian families, and the top high-schools in California have large populations of Asian students. I attribute this here in Silicon Valley to socioeconomic status, since many families are middle-class and highly educated, having arrived initially from abroad to work in the technology industry. Such families seek out neighborhoods with the strongest schools. Many of the families I have come to know prefer their children to stay in California, at a school like Stanford, Berkeley, Caltech, or UCLA. Not all, but many. I believe California now sends more students to MIT than any other state, but my guess is that the desire to keep families closer may play a role in the higher levels of matriculation at Caltech.</p>
<p>@Siserune, have you published your data as a paper?</p>
<p>
</p>
<p>They have three quarters per school year. So while that is more physics and math than MIT requires, it’s not as much as you’re implying.</p>
<p>Unless you’re saying that the courses themselves are condensed - in which case, I’d like to see you present evidence of that, as I don’t feel like digging around on their website to prove your point :)</p>
<p>"@Siserune, have you published your data as a paper? "</p>
<p>No, he is fudging his claim and data. Just for the argument of the Putnam math competition: the US East Asian share being sliced from about 50-60 percent of USAMO qualifiers each year from 2006-9 (NCL’s data that Harvardfan doesn’t dispute) to 20-30 percent of the upper Putnam scores last year. For that matter: Asians make 4/5 fellows, 13/25 top 25, 37/81 on top 81. Quite the opposite what he claims, Asians are actually top heavy. He tries to separate Asians into East Asians or non-east Asians (probably as non-Asians, and I wonder whether anyone can separate Italian from European). When he counts East Asians, he eliminates East Asians non-Americans but not non-east Asians non-Americans. Substantial numbers of USAMO qualifiers, Putnam top scorers, or even US IMO team members are not US citizens (This is quite unfortunately a reminder how far we have fallen). They are immigrants’ children or immigrants. By counting citizenship, he can eliminate a quarter of MIT professors as well. All he did are counting last names sounding like Chinese or Korean and taking off names with non-American sounding first names. He has no chance for a peer review process with such a dubious analysis. Here is the Putnam results of 2009 ([2008</a> Putnam Competition Results](<a href=“American Mathematics Competitions | Mathematical Association of America”>American Mathematics Competitions | Mathematical Association of America)).</p>
<p>“Of course, it could be that universities are being too generous to Asian credentials, perhaps in fear of discrimination lawsuits. Statistical signatures of this would include academic underperformance by Asians, which is exactly what Espenshade found and quantified (but the media have ignored) in his recent book. His methodology was identical to the one he used to quantify the Asian-versus-white admissions “disadvantage”.”</p>
<p>Again, you need to provide the sources of evidences for your claim. You are doing a wild speculation here.</p>
<p>Here is a MIT website of awards and election of Phi Beta Kappa for students in course 7, I don’t see any lack of representation of Asian students (<a href=“http://mit.edu/biology/www/undergrad/awards.html[/url]”>http://mit.edu/biology/www/undergrad/awards.html</a>). Seven of 9 PBK are Asians (2007). Seven of 10 awardees of Asinari awards are Asians (2007). I have no way to know their nationality (which I think it quite silly to count nationality in such a way in international institutes such as MIT). Where is the “Asian student underperformance”? Here is another website on Dartmounth, one of the two valedictorians in Dartmouth is Chinese ([Dartmouth</a> to honor two valedictorians, three salutatorians at 2009 Commencement](<a href=“http://www.dartmouth.edu/~news/releases/2009/06/13.html]Dartmouth”>http://www.dartmouth.edu/~news/releases/2009/06/13.html)). Also, the chinese val appears to have English as second language but double majors in history and math.</p>
<p>
</p>
<p>
</p>
<p>As another California resident, I can say what CalAlum says is likely right on. There are indeed lots of Asians in the top high schools, and a good number of them would like to attend the best schools in technical disciplines IN CALIFORNIA. </p>
<p>I think it might be a little more subtle than matriculation issues though. It is probably a combination of likelihood of being admitted to one of Stanford, Berkeley, Caltech, etc vs. MIT and <em>then</em> the likelihood of matriculation. Schools like Berkeley and Caltech are comparatively friendlier to students who were purely very good math and science students in a traditional sense. And the very kind of student Caltech caters to just may attract a more lopsided percent of Asians, even just in terms of admission.</p>
<p>As for shravas’s words – I will say that upon a glance, the main thing Caltech seems to do a little more of is math and physics, in terms of what is strictly required for every student. However, I think having 5 quarters as opposed to fewer is not necessarily so relevant in regards to high school preparation, as the preparation should be for the classes you have to do freshman year. Presumably if you can survive a whole year of Caltech’s classes with just high school preparation, the sophomore year classes should not be bad, having had Caltech-level preparation. So the real question is if more is required of frosh. </p>
<p>I think it might be less a question of who can handle what, and more a question of the culture they’re trying to cultivate, and what kinds of students they’re catering to…which may be in part reflected in the nature of the core requirements.</p>
<p>
</p>
<p>PBK data are worthless because grades are a very noisy, non-uniform measure that can be manipulated by students aiming for high grades. This is especially true for Course 7 pre meds, whose incentives are stronger: <a href=“http://talk.collegeconfidential.com/1064609387-post6.html[/url]”>http://talk.collegeconfidential.com/1064609387-post6.html</a> ; <a href=“http://talk.collegeconfidential.com/massachusetts-institute-technology/906815-pre-med-mit.html[/url]”>http://talk.collegeconfidential.com/massachusetts-institute-technology/906815-pre-med-mit.html</a>). </p>
<p>
</p>
<p>We were discussing US East Asians and their underperformance. The correct figure is 3 (US East Asian) out of 9, which is at or below their representation among MIT biology students. Your seven-of-9 Asian PBK include 1 international student from Hong Kong (Derek Chu), 3 US Indians, and 3 US Chinese. </p>
<p>We get it that you want to pump up the US East Asian results by adding in higher performing groups such as US Indians and Chinese, Chinese-Canadian and Taiwanese IMO gold medalists. But the hypothesis was about US East Asians only, and if the data apparently support that hypothesis, you would be better off dealing with numerical reality than trying to beautify it. </p>
<p>
</p>
<p>MIT publishes the list of PBK electees by hometown, and for the entire university (Course 7 is an invalid sample for reasons explained below). </p>
<p>2009 US students 71; 16 East Asian, 1 Indian, 54 other
2008 US students 67; 21 East Asian, 5 Indian, 41 other
2007 US students 50; 17-18 East Asian, 7 Indian, 26-25 other
2006 US students 53; 11 East Asian, 4 Indian, 39 other</p>
<p>You can do the math, but the US East Asian numbers indeed appear lower than their share of US non-URM students (there are no identifiably black or Hispanic names that I noticed in the quick runthrough of the Phi Beta Kappa list).</p>
<p>
</p>
<p>The underperformance is in US East Asians, as I said. Of the val/sal listed at that page:</p>
<p>valedictorian: 1 US white, 1 Chinese international
salutatorians: 2 US whites.</p>
<p>for an “Espenshade ratio” (representation of US East Asian vs USEA-plus-USwhite): 0/3 = 0%.</p>
<p>Harvard valedictorian data is available for years 1992-1995 and 1997-2008; I just posted it in the Harvard forum with request for further information. The number of highest-GPA prizes Harvard awarded in those years was 18, of which one went to a Chinese-American (David Riu in 1994), and one to a US Indian female in 1995. The rest were white and frequently Jewish. That’s 1 out of 18, or 2/18 including the Indian female.</p>