bruno123

Which suggests in turn that USNWR is basically a "selectivity ranking", or, at least, is a ranking that is based on criteria that are strongly correlated with selectivity.

Readers should be aware also that, in some of the universities listed above, the percentile SAT scores for individual colleges/programs within the university may be actually much higher than the global scores for the university as a whole, giving a misleading impression that university is less selective than it actually is.

noimagination

Test scores are the best measure of selectivity.

ewho

Now I understand why people got in Cornell but Harvard simply because they did not have high SAT scores.

I used to have a notion that no matter how high your SAT score is, you can not get into HYPSM.

xiggi

Not only are the schools not comparable, but the multi-tiered application process of the military academies obscures the real admission rate. The 11 percent acceptance rate is just as bogus as lumping the academies with the LACs.

xiggi

Well, should it not be expected that the best colleges are culled from a list of most selective schools? Duh!

On the other hand, the selectivity index is also obscured by the tools used by the USNews to "level the playing field." In addition, all what one needs is to compare the selectivity of Harvey Mudd to the all-female schools to understand that selectivity does not correlate with the final rankings. The impact of the abject geographical cronyism of the Peer Assessment is much larger than the mere selectivity index.

And last but not least, schools with high selectivity indexes that also have a tough curriculum (read heavy STEM) are penalized through the expected graduation rate.

All in all, although the basic correlation exists, the USNEWS is NOT a selectivity ranking.

Kei-o-lei

collegehelp said: "I think SAT is the best indicator of quality."

Given the SATs strong and abiding link to SES, shouldn't that be "indicator of quality people" (wicked grin here)?

csdad

http://cshe.berkeley.edu/publication...ons.php?id=265

....gives credence to the old adage ..."the best predictor of future behavior, is past behavior" ....

bclintonk

Which is why acceptance rate is such a pointless statistic for ranking colleges. UCLA gets a little over 60,000 applications a year, in part because there are so many Californians, and in part because it's so easy to apply to all the UCs with a single application. Which also means it's easy for a lot of people to apply who are real longshots but may be competitive for lower-ranked and less selective UCs, but figure, "Heck, all I have to do is check the box for UCLA to give it a shot, so why not?"

Michigan used to get about 30,000 applications a year until it switched to the Common App last year and that number shot up to 40,000. The figures you're looking at for its admit rate are from the pre-Common App era. It will be significantly lower in the next edition of that book But even with the increased applications, it's likely the quality of Michigan's applicant pool has probably remained slightly higher than UCLA's. Most of the increase will be in OOS applicants whose stats tend to skew a little higher than the in-state applicant pool at Michigan. So you should see both a lower admit rate and slightly higher stats for Michigan. I know their 75th percentile ACT score has gone up from 31 to 32, which is the equivalent of a 40-point jump in SAT CR+M.

mom2collegekids

The publics tend to use GPA for admitance to give some advantage to instate students who've attended lesser-quality high schools. If state schools heavily used test scores, too many kids from the very best high schools would be admitted, shutting out those who attended "lesser" schools....which are often the lower-income students.

bclintonk

This is exactly how US News does it. It ranks schools by, as you say, "a fictional SAT score that doesn't statistically represent anything."

Even apart from that, I'd take these scores with a huge grain of salt. Some schools superscore, others don't. Also, some of these schools are test-optional or test-flexible, which is going to skew their reported SAT scores higher because the applicants who have an incentive to submit SAT scores will be, naturally, those with high SAT scores. See, e.g., Bowdoin (test-optional; 70% submit SAT scores in a SAT-dominant region where most colleges get SAT scores from 85-90% of applicants); Colby (test-flexible; 64% submit SAT scores, same region as Bowdoin). Also, it's rumored some schools in the Northeast encourage their recruited athletes to submit ACT scores instead of SAT so as not to mess with the schools' reported SAT medians; no one pays much attention to ACT medians at SAT-dominant schools.

For all these reasons, you're not really comparing apples-to-apples.

BobWallace

!!! I did not know that. The more I learn, the more of an absolute buffoon Mr. Morse appears to be. And yet these rankings are viewed with grave reverence by the rank-and-file.

bclintonk

OP, any particular reason you chose to rank by 75th percentile rather than by 25th? People often make the argument on CC that it matters more where the bottom of the class is, because that's what determines how much and how fast things can be taught.

Of course then Harvard wouldn't come out on top, which would mean that in the minds of many it couldn't be a true ranking . . . .

Ranked by 25th percentile

1470 1580 California Institute of Technology 93
1410 1560 Harvey Mudd College 93
1410 1560 Massachusetts Institute of Technology 93
1400 1580 Yale University 91
1400 1580 Princeton University 91
1400 1560 University of Chicago 73
1390 1590 Harvard University 91
1390 1540 Washington University in St Louis 63
1380 1560 Pomona College 80
1370 1550 Columbia University in the City of New York 96
1360 1550 Stanford University 89
1360 1530 Vanderbilt University 46
1360 1530 Northwestern University 75
1350 1550 Dartmouth College 75
1350 1530 University of Pennsylvania 87
1350 1500 Tufts University 71
1340 1540 Amherst College 69
1340 1540 Rice University 87
1340 1530 Swarthmore College 87
1340 1530 Duke University 87
1330 1530 Brown University 85
1320 1510 University of Notre Dame 49
1310 1530 Williams College 96
1310 1500 Bowdoin College 70
1310 1460 Reed College 88
1310 1460 Vassar College 74
1310 1460 Washington and Lee University 82
1300 1500 Georgetown University 94
1300 1500 Carleton College 68
1300 1500 Cornell University 90
1300 1500 Carnegie Mellon University 94
1300 1490 Haverford College 88
1300 1480 Claremont McKenna College 77
1300 1480 Emory University 84
1300 1470 Hamilton College 56

I only went down to 1300, to capture all the Ivies . . . since they seem to be the object of special attention on CC.

mom2collegekids

People often make the argument on CC that it matters more where the bottom of the class is, because that's what determines how much and how fast things can be taught.


that is a nonsense argument, especially at larger schools that offer 100s of majors. Those in the lower quartile are rarely going to be found in the same majors as the top quartile. The lower quartile are filled with athletes and special admits whose majors often barely register above basket-weaving.

The idea that the Complex Calculus class or Thermodynamics class is being taught at a "slower pace" because there are kids in easier majors on the other side of campus with lowish scores is just ridiculous.

xiggi

Unless your name is Greg McElroy.

sylvan8798

The fact that there are schools missing that should be here makes the list rather less useful, IMHO. Is there a link for this data?