<p>
</p>
<p>Because the numbers of ORM’s and majority applicants who are well-qualified to enter such universities hugely exceeds the numbers of URM’s who are both qualified and who apply.</p>
<p>
</p>
<p>No, those are five fantasy votes, which didn’t happen. Because you want something to be law and think you “know” how people “would have” voted, does not make it when it comes to legal principle.</p>
<p>I’d love to see SATs eliminated as a college-admissions evaluation tool. In our experience, they are a source of intense stress that takes away from useful learning in high school. As far as I am aware, they offer no predictive value with regard to future “success” in life. Any test that produces different results among different races surely cannot be considered “standardized.”</p>
<p>
</p>
<p>While this may be true, I don’t think this is the test’s fault. I think it has a lot to do with racial “culture”. And I put “culture” in quotation marks because the “culture” does not apply to all people in the racial group, for sure. Asian students tend to be driven to succeed (and therefore study hard and test well) whereas African American students are less likely to go to college and are less likely to emphasize education as part of their “culture”. I’m with Bill Cosby on this one - there’s nothing wrong with blacks in America, but there is something wrong with the black community.</p>
<p>
</p>
<p>OK, so this answers the second question, but what about the first (ie. what is the impact)?</p>
<p>
</p>
<p>What was the point of this? Why do I think I prefaced my comment with “At least on paper,…” and used the word “can,” which you yourself italicized for emphasis?</p>
<p>
</p>
<p>How are you defining “standardized”?</p>
<p>According to the honourable folks at [url=<a href=“http://www.fairtest.org/facts/whatwron.htm]FairTest[/url”>http://www.fairtest.org/facts/whatwron.htm]FairTest[/url</a>], “Standardized tests are tests on which all students answer the same questions, usually in multiple-choice format, and each question has only one correct answer.” Thus, equality of result is irrelevant to whether a test is standardized or not.</p>
<p>
</p>
<p>I don’t know how you’re defining future success in life because you didn’t define it. However, taken literally, first-year university grades occur in the “future” (ie. after you take the SAT). According to [The</a> College Board](<a href=“http://professionals.collegeboard.com/profdownload/Validity_of_the_SAT_for_Predicting_First_Year_College_Grade_Point_Average.pdf]The”>Higher Education Professionals | College Board),</p>
<p>Hezlett et al. (2001) performed a comprehensive metaanalysis of approximately 3,000 validity studies, with more than one million students, and found that the SAT is a valid predictor of FYGPA, with multiple correlations corrected for range restriction and attenuation ranging from 0.44 to 0.62.</p>
<p>If there were no predictive value, then the correlations should be zero or close to it. But, they weren’t, were they?</p>
<p>Here I agree with fab. 
Adcoms are well aware of the limitations of these tests (racial/ethnic/culture/language), just as they’re aware of the limitations of all other criteria. A few students have only high test scores to attest to their abilities.</p>
<p>
</p>
<p>Well, good luck finding ANY test that will, unless you’re working with a very small group of people.</p>
<p>Standardized tests do have a lot of problems, but racial disparities is hardly one of them.</p>
<p>An astute lawyer will read the dissents just as carefully as the majority opinion when the Supreme Court rules on a controversial issue. It’s regrettable that the majority opinion in [Plessy</a> v. Ferguson](<a href=“{{meta.fullTitle}}”>http://www.oyez.org/cases/1851-1900/1895/1895_210/) was ever written. (Perhaps we wouldn’t be having this discussion but would be enjoying agreeable discussion with one another on some other issue, if it had not been written.) And it took altogether too long for Plessy v. Ferguson to be overruled. But it was overruled, and the dissent in that case showed the way to better race relations in the United States. The structure of the Grutter case is such that it could be largely overruled in effect just by subsequent cases being distinguished by their facts, respecting stare decisis but sending a signal that actual cases will turn out differently in the future.</p>
<p>
</p>
<p>That may or may not indicate anything. Their stats are quite close to those of whites and there are all sorts of non-discriminatory (or even meritocratic) factors that can drive the Asian numbers higher. Examples include: </p>
<ol>
<li>If there are fewer Asian athletes and legacies, the Asian SATs will be relatively higher.</li>
<li>If Asian matriculants are more concentrated in SAT-intensive admission tracks (engineering, math) than low-SAT majors typical of whites (sociology), they will tend to require higher scores for entrance.</li>
<li>If class rank is weighed heavily, groups such as Asians that are more clustered by high school will be disfavored and this is equivalent to an SAT penalty (on average, higher score needed to get same admission result).</li>
<li>If geographic diversity is considered, Asians and other geographically clustered groups will suffer and this too is equivalent to an SAT penalty.</li>
<li>If Asians are less prone to gain admission based on non-numerical factors (essays, EC), their scores will on average be higher for the same admission outcome. </li>
<li> If Asians prefer a given university more than whites, there will be a higher yield on the high-SAT Asian applicants than on the whites. Such preferences can arise for all sorts of reasons that may operate in the Asian applicant pool, such as finances (state school), high Asian enrollment (UC’s), prestige, location (urban/coastal over rural LAC), or any number of others.</li>
</ol>
<p>That’s with Asian-blind admissions, where whites and Asian are treated identically.</p>
<p>It is also possible that being Asian is correlated with factors that specifically lead to discounting SAT scores, such as enrollment in a competitive high school, evidence of test prep, math team participation (SAT prep by another name), a group-specific academic track (afterschool academies), the appearance of parental pressure, and many other such factors. Detecting and accounting for exogenous advantages that don’t come from the applicant’s own abilities is useful if the goal is to enroll the smartest students. This can have the dreaded “disparate impact” upon Asians.</p>
<p>Post 311:</p>
<p>I understand your point, tokenadult, and the legal basis you’re elucidating regarding dissent. I’ve studied this for many years, actually, and am aware of the point you articulated well. But that didn’t seem to be the context here, at least as it was being expressed.</p>
<p>Naturally it’s possible that there would be a different decision in the future. Of course. But right now there isn’t. Right now the courts have not held that using personal origin as an element of enrollment proportionality is an example of using personal origin in a (legally applied) discriminatory way, as discrimination is properly understood in legal channels. It would be discriminatory if there were substantial or wholesale exclusion of a particular group or groups, or if only one group experienced limits on enrollment (despite high qualification) while all other groups did not. </p>
<p>I think the point that you have earlier tried to make, and that fabrizio often wants to make, is that AA (as opposed to proportional enrollment) is discriminatory because indeed a different standard is allowed for URM’s than either for majority (Caucasian Anglo) or ORM’s (various Asian subgroups, except for Southeast Asians). (‘Different standards’ being another element of the technical legal understanding of ‘discrimination’) It is true that this is an aspect of AA, since the philosophy behind it is that in the context of previous lack of opportunity for certain minorities, vs. the option of opportunity in higher education, ‘discrimination’ in this sense, against the majority, is a by-product of special opportunity and is a legal possibility in principle, although (and we have seen such litigated cases), such situations are sometimes open to legal challenge.</p>
<p>Asians with < 2400 scores are admitted (by the hundreds), while many Caucasians with 2400 scores are not admitted, and vice-versa. Default caps on enrollment operate across the board among the highly qualified, since that self-selected pool is disproportionate to those who are not in the same range of qualification. (Colleges talk about ranges, not numerical levels.) Those not in the range (which would include some but not all admitted URM’s) are capped de facto, given their small numbers – numbers not large enough to result in significant impact (“harm”) on a different group or groups of applicants.</p>
<p>I assume folks have sufficiently calmed down for me to continue?</p>
<p>Another phenomenon I noticed is the adjustment of the standard of proof requirement depending on whether one agrees with it. Here are a few just off the top of my head:</p>
<p>I have briefly commented on the use of statements from admission officers as evidence. This seems to me no more than the “Jesus loves me because the Bible said so” variety of argument (circular). We know these folks may be very nice people, but they represent the schools and it is their job to defend those institutions, and telling the truth to applicants may or may not be to the institutions best interest. </p>
<p>The suggestion that there is no bias because the %tage of Asians accepted by the elites are greater than their %tage in population. If this is not an insult to statistics, I don’t know what is. I know innumeracy is a problem, but I am hard pressed to believe it has come down to this. Trying too hard to make a point?</p>
<p>Using the pass rate of URM in the elites as the gauge of success of AA. We all know that the rate can easily be manipulated by a combination of high grades and fluff courses. Students too can “help” themselves by picking easier majors and/or profs etc. etc. … My feeling is that if the topic is the elimination of AA instead, the standard for success will immediately jump to something such as the graduation rate of URM in say, pure math. My standing joke with friends is that my great, great, great grandchildren may have to struggle with this problem, certainly not me.</p>
<p>I am sure people in behavioral finance will have a field day here, simply by taking a longitudinal look at this threat. Status quote bias or cognitive framing?</p>
<p>Fascinating nonetheless.</p>
<p>
</p>
<p>I know some posters have said that (here or elsewhere). I have not proposed such reasoning. Rather, I have stated, in different ways, a couple of the things that siserune has stated. His statements do accord with statistical principles. The essential statistical principle that does apply, moreover, is that the greatest adversity will be experienced by whatever group or groups are most largely represented when efforts are put in place to correct imbalances. Switching gears now from overall balance to AA in particular: when AA was inaugurated several decades ago, Caucasian Anglo applicants were the largest racial/ethnic group in elite college admissions, and were by that fact the most affected by the new policy.</p>
<p>Fast forward to today, when many factors of balance, including but by no means limited to “race” are introduced, some of the aspects of overlap siserune refers to come into play both for Asians and for Caucasians, and I do agree that for Asians the effect will be greater. Locale, socioeconomic status, department or school-within-college/U applied to (or evident in the student’s profile), even types of e.c.'s & awards that manifest along lines of academic interest – will add to a cluster effect for Asians with such duplicate elements. </p>
<p>It has nothing to do with “percentage in the U.S. population.” </p>
<p>And again, as I have said many times, percentage enrolled < percentage admitted. This is particularly true for U’s other than Harvard. I can’t help it if people disbelieve this; they simply have never worked in any admissions office.</p>
<p>As for the rest of Canuckguy’s recent post, it is less mathematical and statistical than much of the hard information offered on any of the 5+ threads (which may not include the previous AA threads of several years ago). For example, information on graduation rates (links which have been posted). Fine, you doubt that URM’s have had great graduation rates in “pure math.” The question is, how many did apply as “pure math” majors? Possibly not many to be statistically significant. This is what is called manipulating statistics.</p>
<p>As to statements from admissions officers, it depends: often in the past they have provided hard data to inquirers. Often they have not done so in areas that violate privacy rights.</p>
<p>One book that was based on information made available to researchers from admission offices is The Early Admissions Game. </p>
<p>[Amazon.com:</a> The Early Admissions Game: Joining the Elite (9780674010550): Christopher Avery, Andrew Fairbanks, Richard Zeckhauser: Books](<a href=“http://www.amazon.com/Early-Admissions-Game-Joining-Elite/dp/0674010558]Amazon.com:”>http://www.amazon.com/Early-Admissions-Game-Joining-Elite/dp/0674010558) </p>
<p>That book, as its title implies, mostly has to do with the advantage of applying to college in the early round (an advantage Harvard and Princeton have removed by no longer having an early round). But the book also mentions the effect of ethnicity on student chances for admission. It isn’t the last word on any of its subjects, but it did cause some serious rethinking in some admission offices about practices admission officers were engaging in unawares. That’s why I think information spoken by admission officers on some issues needs confirmation from other sources. I think the admission officers honestly desire to promote honest admission practices, but their job so often gets them so busy with handling individual cases at just one college that they lose sight of broad practices and how those compare across multiple colleges. That’s why letting independent researchers have access to college admission files (with suitable safeguards for applicant privacy, of course) is a good idea: it results in opportunities for improvement of the system that might be missed by insiders.</p>
<p>
</p>
<p>Which minorities? Socioeconomic arguments don’t support racial preferences; they support socioeconomic preferences. Using race as proxy for disadvantage gets you [West</a> Indian and African immigrants or their children](<a href=“http://www.nytimes.com/2004/06/24/education/24AFFI.final.html?pagewanted=all]West”>http://www.nytimes.com/2004/06/24/education/24AFFI.final.html?pagewanted=all). What kind of opportunities were these students deprived of? Most likely, they came from a “highly motivated, self-selected group.”</p>
<p>Now, lest I be misunderstood, there is nothing “wrong” with admitting lots of these students in the slightest. I’m merely pointing out my disdain for using socioeconomic arguments to defend racial preferences.</p>
<p>If it’s “lack of opportunity” that you’re concerned for, don’t use race as a proxy for poverty. You won’t get what you ostensibly want.</p>
<p>^ Well said, fab. Colleges (clearly?) invoke racial AA when they want to increase racial diversity above the levels produced by invoking socioeconomic and other admittance criteria.</p>
<p>Like Fabrizio said, African and Caribbean-born blacks are overrepresented among the black population in top colleges, and maybe overrepresented among the general college population. To supporters of AA, do you think we need to balance this group out too?</p>
<p>When international and racial diversity overlap, it seems (to me) like too much micro-management to warrant a policy. It would seem that the number of such applicants at any given school would be small enough that the individual characteristics of the applicants would override any attempt at policy.</p>