<p>No study shows that, because it isn’t true. At Caltech where math SATs are 780-800, the math SAT doesn’t predict the performance of enrolled students because their scores are virtually the same. It does not mean that math SAT doesn’t predict performance at Caltech or that they should stop using it to evaluate applicants. </p>
<p>
</p>
<p>Both the SAT and test scores as a whole are predictive when taking into account the pre-selection of the students using those scores, and the selection of fields of study being influenced by the scores. Asians may be twice as likely to choose low-graded STEM majors, but students with top quartile math SAT are <em>three</em> times likelier than others to choose those majors.</p>
<p>
</p>
<p>Grades are a measure that is strongly biased in favor of females, and has a lower ceiling than the SAT where gender discrepancies (especially for the verbal) appear mainly near the top of the scale. I think there is some overperformance by females on harder measures, and for some of the same reasons that whites overperform Asians (again, on top-level measures, not baseline outcomes such as graduation). Females start later and the male advantage dissipates over time. But this overperformance is much more limited than the studies of college grades would suggest.</p>
<p>limabeans - for medical school, there is a boost from the prestige of the ug institution. You will find a gazillion threads on CC claiming this isn’t so but it’s just not the case. I think less for law school, but honestly – who wants to be a lawyer these days? Tons of money to go to law school and no jobs out there.</p>
<p>First, I’d like to emphasize that you have a misconception. Opposing racial preferences DOES NOT MEAN support for “numbers only” admissions.</p>
<p>Second, you stated “In order to admit the best students across the board, adcoms must look beyond standardized test scores.” I agree. But looking beyond standardized test scores doesn’t mean we have to consider racial classification, now does it?</p>
<p>No, it just means you should consider subjective criteria, like extracurriculars, essays, recommendations, and so forth. So what kind of a system am I advocating? The same one we have now, except it doesn’t consider racial classification (or gender).</p>
<p>As a compromise, if we just HAVE to have racial preferences for "URM"s, I’m fine with assigning everyone a unique identifier and modifying the “race box” to “Are you a ‘URM’? If yes, please check one or more of the following (optional):”</p>
<p>Since private colleges are also businesses, gender imbalance affects competitiveness with peer universities. Nevertheless, as I pointed out earlier from Princeton’s CDS’s, the committee strives to admit relative to approximate percentages of applicants by gender, even if unbalanced. Those figures are apparent from the data. </p>
<p>Of course, the joke about assuming that trying to micro-manage private universities for them (since “we all know” they are incompetent at selection :rolleyes:) will result in some “blind” system, is that publics such as UC and UVA have target selection quota (vaguely) by region. If all the Southern CA applicants have 2300+ and UW 4.0’s, whereas all Northern CA applicants have 2200+ and UW 3.9’s, Berkeley and UCLA are nevertheless going to consider geographic proportionality in their offers, and CC posters have nothing to say about that unless you’re on the Board of Regents. </p>
<p>There have been threads and lots of posts about this regarding UVA as well. The committee strives to ‘spread the wealth around.’ Shall we open an investigation about that, too, based on regional “discrimination”?</p>
<p>Both public and private universities manage their enrollments by various means, some of which supercede what anyone on this thread would call “merit.” That includes institutions which have been forbidden to use race/ethnicity as selection criteria.</p>
<p>Back it up or take it back. If you think I’ve made a post that shows anti-Semitism or prejudice toward blacks, quote it. If you can’t do that, then take it back.</p>
<p>I find it hilarious that I am accused of prejudice toward blacks. Why? Because I don’t support racial preferences? Funny how supporting a policy that is paternalist toward blacks is a sign of non-racism…</p>
<p>Mind you, that sentence does not mean “People who support racial preferences are racist.” I’m just pointing out how funny it is when believing that blacks can be successful without racial preferences is construed as racism.</p>
<p>Unlike Asian discrimination which is hypothetical, or hard to observe, and denied by colleges, gender balancing is openly proclaimed. MIT worked hard to get to 50-50 and proudly announced every annual increase. Lower-tier schools with a surplus of females don’t hide that they give admission preference to males.</p>
<p>Espenshade et al’s statistical studies, although debatable when it comes to Asians, are extremely strong findings of a preference for females in elite admissions. The female effect is the same no matter how the study is performed, unlike the Asian results that fluctuate wildly (or are reversed in some income ranges) when you add more variables.
For Asians some of the effect comes from geographic clustering, selection of fields of study and other race neutral effects. But for females, no amount of additional controls for socioeconomics, geography, type of high school, etc can break the 50-50 gender ratio. The male to female ratio of high-end test scores or equivalent objective credentials is about 1.5, and much higher as one goes toward the top of the scale. Espenshade’s calculations were, consistently, that females were (in odds-ratio terms) advantaged by a factor of 1.3 to 1.6 depending on which year and which variables were included. It is hard to imagine how to make this effect disappear by adding more selection factors or how the universities could get 50-50 gender neutrally by increasing the weight of grades or of verbal over math.</p>
<p>If somebody wants to file a Jian-Li type complaint based on the same study for gender discrimination, it is much harder to defend against. The results, however, would be a case of “be careful what you wish for”.</p>
<p>Fabrizio, obviously, if you make it your business to assume every reference to right-wing bigotry on a public forum means, they are posting about you – then, I defer to you. Maybe, you’re a right-wing bigot.</p>
<p>This kind of post from you frustrates me. A lot.</p>
<p>You’re writing as if I believe that 2300+/4.0 is more qualified than 2200+/3.9. If it’s all else equal, then yes, but is it all else equal? Most likely not. Under the system I propose, there is NO reason why a 2200+/3.9 cannot shine more than a 2300+/4.0.</p>
<p>Your posts are really fascinating. I did’t know that females tend to score lower on the SAT than males. I’ve assumed that my DD with her 2340/35 will have a harder time than my DS with his 2400/36. </p>
<p>How do you account for the success of the girls at getting admitted to the top schools? I would wonder if it is gpa. Don’t the girls tend to have the higher gpa?</p>
<p>Fabrizio and the rest of you – just settle down and play nice.</p>
<p>Quoting from #2333, “FWIW, trying to play wedge politics between blacks and Jews is a time honored right-wing tradition. Otherwise, it would make little sense for Fabrizio…”</p>
<p>Right, johnwesley. Your reference to a bigoted right-wing tactic had nothing to do with me. It just happened that you referred to me in the NEXT sentence. No, I was supposed to read that sentence by itself, without any reference to the previous one. :rolleyes:</p>
<p>You can disagree with me. That’s fine. What you cannot do is claim that I’m a “right-wing bigot” without even trying to back it up.</p>
<p>Lemaitre,
One other possible for the high scores in Ma (to wit also true for PAST/NMS cut-offs) could be the presence of a good number of excellent Boarding Schools.
To wit again, the NMS cut-off in 2010 for New England BS was based on the MA score, so they are the bar - the highest in the country.</p>
<p>I’m assuming, yes, that everything else is equal. And I’m telling you, knowing precisely how selection works at the U.C.'s, geographic balance is a factor that can ‘edit’ (I will repeat myself)</p>
<p>
</p>
<p>A minor student cannot do anything about where he or she resides, just as that student can’t do anything about accidents of gender and personal origin. But he or she will be affected by a variety of unknowns when it comes to college admission even for publics, so your effort to engineer the “fair” system is an ivory-tower exercise.</p>
<p>limabeans - anyone who wants a good GPA that is needed for medical or law schools is usually advised against majoring in engineering. It is not to say engineering GPAs can’t be 4.0 but the harder the schools are, the tougher the GPA gets. It is generally assumed going to UT Austin for example, people may be able to graduate with a 4.0 GPA in order to support their applications for medicine/law with a little bit of work. I was quite surprised to read a graduation announcement in the local paper about a kid graduating in biomedical engineering from UT A and going onto MIT for a Ph.D. (white if anyone is wondering) state that he was the only one in his graduating class with a 4.0 GPA (makes you wonder about privacy laws but I guess it must have been mentioned in dept announcement or something?). Here is a class of 80 or 100 or something like that and one kid has perfect GPA.</p>
<p>sewhappy - I don’t know if your daughter is already in college or applying this year. Based on what I know, Caltech would love to have her because they are trying their best to gender balance (as siserune mentions about MIT) and they are still at 60:40.</p>
hmm … I beleive that. In addition there are a bunch of folks from a scattering of ethnic groups that do share a relgious heritage that might help explain those high test scores … MA has one of the highest concentrations of Jews in the US; especially the greater Boston area.</p>
<p>No, that analysis makes no sense at all, because you’re making the assumption that an equal portion of whites-with-5’s and Asians-with-5’s would apply to elite schools (and hence if selected “fairly,” would result in a 3:1 ratio). You’re completely missing the fact that there are a good chunk of white students who aren’t interested in applying to elite schools at all - who are going to be siphoned off by large state schools, especially in the midwest (Big 10) and southeast (SEC).</p>
<p>In the years we’ve discussed this, when have I ever expressed opposition to geographic preferences? Last time I checked, geographic location wasn’t a suspect class. I also don’t see any ballot initiatives in California asking for the removal of geographic preferences in UC admissions.</p>
<p>And have you ever known anyone to complain that he didn’t get in because his spot was “taken” by someone who matched him on X - 1 categories and bested him on the OBJECTIVE criterion?</p>
<p>The other side of comparing scores of 5 is that most schools seem to ignore the AP scores in their decisionmaking process since it is not a required component of the application process. AP comes into play mainly after an admission.</p>
<p>Is there a corresponding “good chunk of Asian students who aren’t interested in applying to elite schools at all”? </p>
<p>And could you elaborate on your statement “I also have to say, you’re in a southern state and I’ve got to believe the experience of an Asian-American in a rural southern area is not like the experience in the north or in larger cities (regardless of whether such cities have high Asian pop or not)”?</p>
<p>johnwesley has told me that I may have misunderstood what you were getting at. If so, my apologies.</p>