are colleges racist?

[sewhappy]

<p>

I don’t disagree with this at all–but I don’t see how this translates into discrimination against Asians vs. whites in that pool of unhooked candidates. And I would just point out that these schools have the luxury of taking people who scored over 2300, had top GPA AND have something more “interesting” on the resume. And what would be the solution if this is the problem? Make the schools stop recruiting athletes and giving tips to hooked candidates? Why, exactly, should these private schools want to stop doing those things?

</p>

<p>I don’t know. But surely it is enormously awkward to be employing race as an admissions factor when that only adds to the inflationary score/gpa situation for other races. Again, I believe the intentions are pure of heart. I don’t believe there is any discriminatory policy. I think there is a process that has morphed into something bizarre and unpleasant. That is why this thread rages on and on.</p>

[mokusatsu]

<p>

The Espenshade study is evidence for this, although we don’t know what specific schools were included.

So the Espenshade study would not support the assertion “the average scores and gpa for Asians who get admitted to the top Ivy schools are consistently higher than for other applicants of other racial groups.” </p>

<p>Also, as I understand it, the Espenhade study is based on data from 1997 (and earlier years), so it is rather old evidence.</p>

[fabrizio]

<p><a href=“This%20was%20pointed%20out%20to%20Fabrizio%20over%20a%20year%20ago.%20He%20didn’t%20get%20it%20and%20posted%20nonsense%20about%20regression%20analysis%20that%20might%20have%20kept%20him%20out%20of%20grad%20school%20had%20the%20professors%20seen%20it.%20Lucky%20that%20CC%20is%20anonymous!”>quote</a>

[/quote]
</p>

<p>Hi siserune! How about the names of those labor studies, the “Berkeley NMF” data, and your 15+ years of hand collected panel data on competitors in U.S. math competitions? And while we’re at it, how about a link to a working paper explaining all these flaws in Espenshade’s research?</p>

[soomoo]

<p>^ If my son scored a 2250 (not sure about a 2320 though) on his first sitting, I would have him retest. Why not? I wouldn’t have him take it repeatedly, but certainly most people take it 2 times. It’s kind of a cheap shot, Limabeans, to accuse someone of being test happy in this situation. I’m just not buying the 2400 is just as good as a 2250. I know the counselors at my son’s school would have him retake a 2250 since his PSAT score was perfect. Maybe they are wrong too but hey it never hurts to take precautions.</p>

[Hunt]

<p>

Had he gotten into all the schools, people would say, “Why is he complaining?”

But he wouldn’t be complaining, would he? After all, he didn’t bring a complaint against Yale, which accepted him, or against Harvard, where he transferred.</p>

<p>I’m really sorry this thread has–as has happened before–devolved into accusations of anti-Asian bias against people who are questioning the premise that Asians are being discriminated against as compared to whites. It’s pretty unseemly, especially coming from people who don’t seem willing or able to articulate why it is that they believe the discrimination is actually occurring. Even putting aside any flaws in the Espenshade study, all it shows is that Asians have higher stats. That’s pretty much it. And that, I suspect, is why Jian Li’s case disappeared down a black hole–there is no other evidence of discrimination. If there is, let’s hear what it is. So far, in 116 pages of this thread, I haven’t seen it.</p>

<p>

So the Espenshade study would not support the assertion “the average scores and gpa for Asians who get admitted to the top Ivy schools are consistently higher than for other applicants of other racial groups.”

I would say it supports it somewhat weakly, since supposedly it contains highly selective schools, but not all the Ivies. I’m not aware of any study to the contrary, though.</p>

[texaspg]

<p>Bay - I know one ASIAN Athlete going to Harvard with a 35, top 2 percent of class, IB diploma, etc.</p>

[limabeans01]

<p>

C’mon. Drownings have nothing to do with ice creams sales. SAT scores ARE part of the college admissions process. To try to explain the fact that 2400s get accepted at a much higher rate than 2250s people here seem to be resorting to “soft” factors not being equally distributed among the range of test takers. This seems like a stretch to me.</p>

<p>This has nothing to do with Asians vs. Whites, but simply whether or a higher SAT score really matters past a certain point.

</p>

<p>Of course drownings have something to do with ice cream sales, they are very highly correlated. It is a fact that the most drownings occur in the months that have the highest ice cream sales. If you swim in August, you’re more likely to drown.</p>

<p>or wait—does the month you choose to swim in have little to do with the likelihood of you drowning? (Assuming of course you did not choose to swim in an icy pond in the winter–or had SAT scores below a 2100)</p>

[sbjdorlo]

<p>Bovertine,</p>

<p>My son’s had his 800 in math since 8th grade. But it would be nice to get an 800 on the writing portion of the SAT since he loves writing and is an excellent writer (though a 25 minute prompted essay is hardly the ideal writing situation) And you’re right; assuming he improves his score, at least he’d know if he’s rejected by certain colleges, it wasn’t due to a lower writing score.</p>

<p>Yeah, not sure about mentioning being PR but he can’t hide his skin color if he interviews.</p>

<p>My point is just that my expectations were so low compared to all of you on CC. I just didn’t know. And, we didn’t really know he’d be applying to highly selective colleges until this year when it became apparent that he’s had so many college math and physics courses, he’d run out of classes to take at smaller Christian colleges and that selective schools really might be a better fit.</p>

[texaspg]

<p>SBJ - If you are missing all the points in one section, it is worthwhile. If you have 770, 770, 780, then it can have odd consequences. Adcoms do mention that if one takes the test repeatedly with such numbers, they add a negative point (obsessiveness?).</p>

<p>If it was my kid, I would rather complete the SAT IIs in October before retaking SAT I (just in case you need to retake SAT IIs) or do ACT instead in September only because several schools seem to think ACT = SAT + 2 SAT IIs.</p>

[bovertine]

<p>

My point is just that my expectations were so low compared to all of you on CC. I just didn’t know.

They’re not low for me. I probably wouldn’t retake those scores. I know I wouldn’t have my kid retake them. I wouldn’t be around anyway, because I would die from shock if he got 2320.

My son’s had his 800 in math since 8th grade.

I guess I meant 800 on the CR then, whichever one he got the 790 on.</p>

[Bay]

<p>

Bay - I know one ASIAN Athlete going to Harvard with a 35, top 2 percent of class, IB diploma, etc.

</p>

<p>Our high school has sent 3 Asian athletes to Harvard in the past 3 years (3 different sports). One was Val, the other 2 were not, but I know that their gpas were above 3.7 (their names were on an award list for students with that gpa+). When I think about applicants like Jian Li, I want to shout at them: “You should have played Lacrosse!”</p>

[mokusatsu]

<p>

Even putting aside any flaws in the Espenshade study, all it shows is that Asians have higher stats. That’s pretty much it. And that, I suspect, is why Jian Li’s case disappeared down a black hole–there is no other evidence of discrimination.

Espenshade himself says that his study does not prove discrimination against Asians.</p>

<p><a href=“http://www.princeton.edu/main/news/archive/S26/34/81K42/index.xml[/url]”>http://www.princeton.edu/main/news/archive/S26/34/81K42/index.xml&lt;/a&gt;
‘Many times people will ask me, “Do your results prove that there is discrimination against Asian applicants?” And I say, “No, they don’t.”’</p>

[siserune]

<p><a href=“CollegeAlum314:”>quote</a> </p>

<p>I guess the big question is how Espenshade and Chung analyzed their data.

[/quote]
</p>

<p>They didn’t fully disclose the results, nor the methodology. </p>

<p>There are no confidence intervals (standard errors are usually given) for the regression coefficients. It means that the “50 point Asian effect” could have been a “20 to 70 point effect”, or whatever other range. The closer the smaller endpoint of the interval is to zero, the more likely it is that the findings overlap with the range that can be explained by race-neutral factors. Right now we don’t know what the intervals are.</p>

<p>They don’t disclose the classification rate: how many of the acceptances and rejections their model predicted correctly. This is pretty important for understanding whether their admissions model is realistic or not. Ideally they should have given the classification rate by SAT range, since even a bad model will fare well predicting the admission for very high and very low scores.</p>

<p>Nowhere is it explained how the 50 or 140 point figures were calculated.</p>

<p>

If the did it in a way that race-neutral selection could look like racial bias, then that would reflect poorly on their own abilities as researchers.

</p>

<p>It is almost impossible to perform (that is, obtain funding for) this type of study without race-neutral factors appearing as racial bias. You would need a much larger and richer set of variables in the data and on top of that to build a realistic statistical model of admissions.</p>

<p>The question, then, is not whether but how much of the Espenshade and Chung “50 points” or “140 points” (not!) comes from race-blind effects, including the construction of the study; and if there is a race-specific effect, whether any of it is discriminatory in any unfair way. Espenshade’s regressions also found that, after controlling for type of major, high school GPA, SAT scores and other factors, Asians underperformed academically in the sense of having substantially lower class rank than their academic credentials would predict. The effect was equivalent to more than 140 points, but whatever the correct number, underperformance of Asians vs whites (or males vs females, blacks vs whites, etc) would imply that under pure academic admission that tries to predict post-admission results, the underperforming group should suffer a penalty equivalent to some SAT points. Not imposing a penalty amounts to academic affirmative action for Asians, males, blacks or whatever the underperforming group. The underperformance analysis is more complicated than Espenshade’s class rank regressions, but the universities of course have much more data and are not limited to GPA or class rank as a performance measure. In these CC discussions I pointed out research prizes, Hertz fellowships and high scores on the Putnam competition as examples where Asian representation is far lower than the share of high school elite credentials would predict.</p>

<p>

However, these are both Princeton professors, right? So you would expect a certain level of competence. In general, appeals to authority are weak arguments. However, since two Princeton professors has conducted a study that alleges bias against Asians at their own institution, it is certainly worth looking into.

</p>

<p>There is one Princeton professor… of sociology, who received his PhD in 1972. The other coauthors are his graduate students and a “statistical programmer” at Princeton (who I think was another former grad student in the same department). Expertise on the methods Espenshade used for the regression studies resides in the fields of statistics and computer science (machine learning), and is generally more recent. Sociologists, demographers and policy economists are generally not very knowledgeable in these statistical methods and their limitations, whether they work at top schools or lesser ones. Looking at Espenshade’s publication history he seems to be a policy/demography/sociology guy and his graduate students don’t appear to have interests that are in any way more mathematical or statistical.</p>

<p>It also is not known which three top-tier schools were included in the analysis, and whether Princeton was one of them. It’s a reasonable surmise, but was never disclosed.</p>

<p>The study authors did not accuse the schools they analyzed, whichever those were, of discrimination against Asians. They raised the possibility, and said that their data set did not have enough additional variables to investigate the question of discrimination.</p>

[Shrinkrap]

<p>“My point is just that my expectations were so low compared to all of you on CC. I just didn’t know.”</p>

<p>For the record, a lot of folks on here now didn’t “know” until they found CC.</p>

[sbjdorlo]

<p>Do colleges factor in where students go to school and where students live when considering admissions? I just googled average SAT scores for California high schools and I see that our local high school is somewhere near the bottom of scores. It’s a high school that’s broken into 3 different subschools. The average SAT scores for the three schools (with the science school being the highest) are 1376, 1283, and 1279. I would guess most of these students would have no concept of the types of scores and activities discussed on CC. I can only imagine what a GC’s job description is at schools like this. I’m guessing no one’s encouraging kids to take and retake the SAT to get a 2400. They’re probably just encouraging them to stay in school. (Reason #4 why we homeschool)</p>

<p>Maybe colleges would be more likely to admit one of these types of students with a 2200 SAT over a student whose culture (whatever the culture, not particularly speaking about “Asian” culture) demands lots of test prep, expects extremely high scores, and expects its kids to apply to highly selective colleges?</p>

[Hunt]

<p>

It also is not known which three top-tier schools were included in the analysis, and whether Princeton was one of them. It’s a reasonable surmise, but was never disclosed.

My recollection is that it was disclosed that Princeton was not one of them.</p>

<p>

Maybe colleges would be more likely to admit one of these types of students with a 2200 SAT over a student whose culture (whatever the culture, not particularly speaking about “Asian” culture) demands lots of test prep, expects extremely high scores, and expects its kids to apply to highly selective colleges?

Here’s a theory that I could imagine to be true–could admissions committees be “discounting” Asian SAT scores because of a belief that Asian kids are more likely to have heavily prepped? I don’t think this would be fair, without knowledge of the particular kid, but it could be a plausible bit of stereotyping. I very much doubt if there would be any policy that would show this, though.</p>

[collegealum314]

<p>

I’m really sorry this thread has–as has happened before–devolved into accusations of anti-Asian bias against people who are questioning the premise that Asians are being discriminated against as compared to whites.

</p>

<p>Is this the only discrimination that matters? Specifically, if URM bias hurts ORMs, would it matter? Would an ORM that complained about be a “clueless jerk”?</p>

<p>

Even putting aside any flaws in the Espenshade study, all it shows is that Asians have higher stats.

</p>

<p>Well, their study also shows that legacy and athletic preference negligibly the demographics of admission. These were two of your reasons why you felt Asians numbers could be pushed down that are independent of race.
The quote is, "We also show that…despite the fact that athlete and alumni children admission bonuses are substantial, preferences for athletes and legacies do little to displace minority applicants, largely because athletes and legacies make up a small share of all applicants to highly selective universities. </p>

<p>While acknowledging sisserune’s criticism of Espenshade, I think it’s reasonable for Jian Li to file the civil rights complaint based on a peer-review study. There aren’t a lot of them out there on this subject, mainly because data is not freely shared. And I think the visceral hatred directed toward him is misplaced.</p>

[Hunt]

<p>

Specifically, if URM bias hurts ORMs, would it matter? Would an ORM that complained about be a “clueless jerk”?

I’d probably use a slightly different epithet. I wouldn’t say “clueless,” exactly. I think ORMs who complain about preferences for URMs are perhaps inadequately aware of our history (if they are young) or perhaps insensitive (if they are older). I think that some of them are misguidedly idealistic.</p>

<p>

largely because athletes and legacies make up a small share of all applicants to highly selective universities

I don’t this is true for the Ivies and the most selective LACs, is it? Added: Harvard sends out 200 likely letters to recruited athletes–the incoming class has about 1700 people in it. That doesn’t include other athletes who get recruited or just tipped but who don’t get likely letters. Apparently most of those who get the likely letters matriculate. That’s pretty many.</p>

<p>

But that does not make it a figment of the collective imagination.

So tell me why YOU don’t think it’s a figment of the imagination–what, specifically, are the facts that persuade you that Ivy League schools are discriminating against Asians vs. whites on the basis if race? I readily admit that it could be happening–but lots of things could be happening, or not happening.</p>

[Canuckguy]

<p>

Since I assume Professor Karabel understands the conclusions of his own work, and since he still supports race based diversity programs **am I to assume you also support **
programs which take racial diversity into account, provided they are designed to avoid quotas, and take all factors into account including ethnic underrepresentation (which is basically what the Karabel plan was)?

Not necessarily. **You can not assume I also support …**Is this better?

I posted that Professor Karable was an ardent supporter of AA. It is fairly well known he put a version of AA into his plan for Berkeley, and he has defended it in numerous public fora.

</p>

<p>Base on his official stand, you think he is an ardent supporter of AA. Based on the academic study I saw (47.7 percent is huge), I think you are jumping to conclusions. We don’t know what he really thinks because we can not get inside his head. The study, combines with the anecdotal evidence given by Sowell, tells me to be cautious. No more and no less.

Well, the world won’t necessarily exist tomorrow, but all existing evidence indicates it will.

I agree. To be rationally consistent, however, I can not say the world will necessarily exist tomorrow because tomorrow has yet to arrive. It is an important distinction in empirical philosophy when I was a student.

As far as your notion about this snippet not being a reviewer’s interpretation. Did you forget to add the quotation marks around the excerpts from “The Chosen”? Because otherwise I have no idea if this is the reviewer’s interpetation or not.

I was simply showing another poster that her daughter and I are not the only people that believe this. It was exactly as I posted it. Have a look for yourself:</p>

<p>[Jerome</a> Karabel’s The Chosen. - By James Traub - Slate Magazine](<a href=“http://www.slate.com/id/2128377/]Jerome”>Jerome Karabel's The Chosen.)</p>

[siserune]

<p>

Even putting aside any flaws in the Espenshade study, all it shows is that Asians have higher stats. That’s pretty much it.

</p>

<p>Setting aside flaws,</p>

<ul>
<li><p>what Espenshade’s studies DO show is that if Asians were converted into whites not only individually (which could be simulated by concealing race-revealing personal information from admissions officers), but also magically transformed into a white equivalents as a population – so that their math/verbal SAT split became the same as that of whites, distribution into states and towns and ZIP codes and high schools were the same as that of whites, selection of non-academic activities and sports and musical instruments were the same as whites, use of Early Decision was (in 1983-1997) the same as that of whites, and so on for all factors that make a difference – then Asians would gain in admissions. The gain is estimated by an undisclosed method to be about 50 SAT points out of 1600 M+V per applicant. </p></li>
<li><p>Espenshade’s studies DO NOT show that there would be any particular effect on Asian admissions from: converting Asians to whites individually (but not as a population); blinding admission officers to applicants’ race; using an admissions procedure that disregards Asian vs white distinction for individuals (if that is not already the case); or having admissions performed by computer programs that do not accept RACE as an input. In the event that these procedures would have an effect, it is not known whether the Asian “gain” would be 50 SAT points, or 3 points, or a loss of several points. Espenshade’s 2009 study claimed to find a <em>positive</em> Asian effect in the lower income ranges, with Asians much likelier than whites to gain admission, all other variables being equal, so both the magnitude and direction of the Asian individual-level effects are not clear if you try to use his studies not as summary statistics but as indicators of how the admission works.</p></li>
</ul>