22 comments

[ 3.2 ms ] story [ 48.7 ms ] thread
The point of the SAT in the first place was to increase diversity. Back then, it meant smart kids who weren't part of the establishment. Now, they want to get rid of it to increase diversity. Except, now it means stupid kids who aren't part of the establishment. The only group that remains the same are the losers, and that's the whole point.
I'm not sure what you mean by "establishment" in this context, other than my guess that I'm not part of the establishment.

After edit: Quite a few scholars contend that for a long time the student characteristic that has set college availability even more than SAT scores is family income. Laura D'Andrea Tyson, one of President Clinton's economics advisors, called for taking student family income more into account in admissions back in 2003:

http://www.businessweek.com/magazine/content/03_27/b3840045_...

Several studies and journalistic reports have shown that in previous years, student income limited college choices more than student test scores:

http://www.tcf.org/Publications/Education/carnrose.pdf

http://harvardmagazine.com/2005/05/a-thumb-on-the-scale.html

http://www.theatlantic.com/doc/200511/financial-aid-leveragi...

http://www.jkcf.org/assets/files/0000/0084/Achievement_Trap....

Now some colleges claim to have outreach to increase socioeconomic diversity on campus:

http://www.boston.com/news/education/higher/articles/2008/11...

Based on what I have heard in anecdotes about this year's admission results, Princeton, which was the first college in the Ivy League to announce a no-loan financial aid policy, is in earnest about admitting more students from low-income backgrounds.

Establishment meant WASPs. Brandeis was set up as the Jewish Harvard because high-performing Jewish high schoolers were getting passed over for Ivy League admission for legacy WASPs. After World War II, the Ivies opened up considerably and the "need" for a Jewish school diminished although it maintains its identity (the same as HBCUs, for example).

The issue of the SATs still being a fantastic determinant of first year college success says something, but I'm not sure what. Perhaps the colleges would let in more minorities, but they'd not perform as well? Would those minorities do better since more of them would get in and they wouldn't be as isolated as they are now?

The submitted article says,

"These models suggest that any move away from the SAT or ACT in competitive colleges results in significant gains in ethnic and economic diversity. But the gains are greater for colleges that drop testing entirely, as opposed to just making it optional. (To date, only one institution -- Sarah Lawrence College -- has taken that step.)"

How does Sarah Lawrence College

http://collegesearch.collegeboard.com/search/CollegeDetail.j...

78% Women 22% Men

<1% American Indian/Alaskan Native 7% Asian/Pacific Islander 2% Black/Non-Hispanic 5% Hispanic 71% White/Non-Hispanic 4% Non-Resident Alien 10% Race/ethnicity unreported

37% in top 10th of graduating class 84% in top quarter of graduating class 98% in top half of graduating class

39% had h.s. GPA of 3.75 and higher 25% had h.s. GPA between 3.5 and 3.74 14% had h.s. GPA between 3.25 and 3.49 16% had h.s. GPA between 3.0 and 3.24 5% had h.s. GPA between 2.5 and 2.99 1% had h.s. GPA between 2.0 and 2.49

compare with Harvard

http://collegesearch.collegeboard.com/search/CollegeDetail.j...

51% Women 49% Men

1% American Indian/Alaskan Native 19% Asian/Pacific Islander 8% Black/Non-Hispanic 7% Hispanic 41% White/Non-Hispanic 10% Non-Resident Alien 15% Race/ethnicity unreported

95% in top 10th of graduating class 100% in top quarter of graduating class

or Princeton

http://collegesearch.collegeboard.com/search/CollegeDetail.j...

50% Women 50% Men

<1% American Indian/Alaskan Native 17% Asian/Pacific Islander 7% Black/Non-Hispanic 7% Hispanic 49% White/Non-Hispanic 11% Non-Resident Alien 8% Race/ethnicity unreported

97% in top 10th of graduating class 100% in top quarter of graduating class

or MIT

http://collegesearch.collegeboard.com/search/CollegeDetail.j...

46% Women 54% Men

1% American Indian/Alaskan Native 25% Asian/Pacific Islander 9% Black/Non-Hispanic 14% Hispanic 34% White/Non-Hispanic 9% Non-Resident Alien 7% Race/ethnicity unreported

97% in top 10th of graduating class 100% in top quarter of graduating class

in student diversity?

I love how diversi-crats treat college education as a fungible commodity - if we can just get more minorities into the Ivy League, we'll magically solve all of their problems. Putting someone into an academic environment which is too far beyond his capabilities is a bad idea.
Putting someone into an academic environment which is too far beyond his capabilities is a bad idea.

I agree with this statement of policy. How do you think a college admission office ought to judge what an applicant's capabilities are?

Maybe we could put together some sort of test...
Is the test score the sole issue you would consider, or would anything else at all make a difference?
What other things might legitimately make a difference?
should probably standardize that test, so that students across the country could be judged based on the same benchmarks :)
Is a student who took a very different set of courses in high school exactly as strong a candidate for college admission as some other student with the same test who took different courses? How about if one student sleeps on a sheet of cardboard under a bridge (working by day to support his family), while another lives in a mansion and has test-prep tutors come to coach him throughout his free time?
Agreed. This is the fundamental flaw with the affirmative action and other diversity advocates.

Artificially increasing diversity is institutional racism which only serves to mitigate a symptom, not eliminate a cause. Rather than forcibly put more minorities through college, we should improve schools in poor and underprivileged neighborhoods.

Unfortunately, this real solution may take 10 to 20 years to see results... which is far shorter than the average term in public office.

Schools can only do so much. If by putting more minorities through college you can increase the average level of education among minority parents (when those students have kids), you may be able to improve their kids' educations as a side-effect.
I'd much prefer that everyone get a proper primary education than a few people struggle through a secondary education to bet on their children.
Actually, a 1999 study showed that highly selective colleges made a bigger difference in the future income of students from low-income backgrounds than it did to students from high-income backgrounds. Putting dumb people in MIT isn't a good idea, but putting lower class people in is. http://www.mitpressjournals.org/doi/abs/10.1162/003355302320...
So there are at least two interpretations of this data (I'll admit I skimmed a bit torward the end).

1) The SAT is measuring things other than academic ability/achievement, particularly non-academic cultural differences between a significant portion of the cited minorities and those having more success getting into SAT-mandatory institutions. This seems to be implicit in many calls for diversity.

2) The SAT IS a good measure of academic ability/achiement/admissibility, and statistically speaking, the cited minorities just aren't as likely to have the requsite levels of whatever academic oomph the test is measuring, presumably for environmental reasons.

Right. This is what I find puzzling about the submitted article's report about the soon-to-be presented study

http://www.wfu.edu/provost/rethinkingadmissions/conference_p...

--very little information about how much student characteristics would change along other dimensions besides ethnic diversity if the SAT were not considered at all as an admission factor.

This is the same kind of crap that started the housing bubble... social engineers doing what they just feel in their heart is right for society.

The results will be tons of mediocre graduates unable to pay back their loans. That's OK... we'll pick up the tab; no fucking problem!

well, duh. minorities and underprivileged individuals get lower sat scores. colleges that take into account sat scores prefer students with higher sat scores, all else equal. why did they need to do a "study" to come to this conclusion?
I'd have felt a lot better about the article if it wasn't for the word "model" sprinkled liberally through the piece. It sounded an awful lot to me like they could have just looked for the evidence of the effect they are claiming, and I find myself wondering if they used a model because it was easier, or if they did in fact look and failed to come up with the morally-correct evidence.

It's easy; go look at those colleges that didn't consider the test, then compare the statistical outcomes predicted by the SAT for those who took it and what actually happened. (There may be some correlations to sniff out ("I know I'm only applying to non-SAT colleges so I won't take the test"), but I bet it's managable.)

Diversity's great and all (though when I say diversity, I mean real diversity and not a whole bunch of differently-colored people all believing the same thing, but that's a topic for a different discussion...), but it's evil to skew statistics to increase the number of "diverse" students taking out increasingly expensive loans who then do not end up with a degree just so you can feel good about how many happy rainbowed diversities your college has. If that is what is going on here. I admit I can't prove it but I don't like the smell.

(Yes, there are serious issues with educational opportunities that should be addressed, and I don't know what the solution is. I just know the solution isn't forcing the already-behind to take on tens of thousands of dollars of useless debt.)

It's easy; go look at those colleges that didn't consider the test

The submitted article says there is exactly one such college, which I think from other sources I have read is correct. And the figures for that college are already posted to this thread.