No they didn't! One more "new" ranking

College Search & Selection
1

<p>In a tangible departure from the unscientific voodoo and stealth rankings a la Forbes' Vedder's Crap and WM's Mother Teresa, or the utterly irrelevant Sino-British rankings of graduate schools, the renewed efforts by Christopher N. Avery, Mark E. Glickman, Caroline M. Hoxby and Andrew Metrick will survive criticisms of the lack of qualifications and integrity of the researchers. Those folks know what they are doing! </p>

<p>Obviously, this hardly means that the methodology is beyond reproach, as the source of the basic information for the "tournament results" might remain as suspect ever.</p>

<p>But that will not stop to fuel plenty of discussions and dissecting on College Confidential, including the ubiquitous unveiling of the silly tiered rankings and the necessary manipulation to bring the fav's to the top! </p>

<p>Let the "panem et circenses" begin! </p>

<p>
[quote]
A Revealed Preference Ranking of U.S. Colleges and Universities*</p>

<p>We present a method of ranking U.S. undergraduate programs based on students’ revealed preferences. When a student chooses a college among those that have admitted him, that college “wins” his “tournament.” Our method efficiently integrates the information from thousands of such tournaments. We implement the method using data from a national sample of high-achieving students. We demonstrate that this ranking method has strong theoretical properties, eliminating incentives for colleges to adopt strategic, inefficient admissions policies to improve their rankings. We also show empirically that our ranking is (1) not vulnerable to strategic manipulation; (2) similar regardless of whether we control for variables, such as net cost, that vary among a college’s admits; (3) similar regardless of whether we account for students selecting where to apply, including Early Decision. We exemplify multiple rankings for different types of students who have preferences that vary systematically. JEL Codes: I2, I23, C35, D11.

[/quote]
</p>

<p>A</a> Revealed Preference Ranking of U.S. Colleges and Universities*</p>

<p>PS No, I won't share "free" links!</p>

2

<p>I just think this whole “revealed preferences” thing is a bunch of hooey. It doesn’t tell you about the preferences of students generally, only about the preferences of students who are cross-admits to two schools. And that’s going to be a skewed sample. Where does it account for the “revealed preferences” of students like my D1, who applied to her first choice school ED, was accepted, and never looked back? She doesn’t end up in any cross-admit pools because she only applied to one school, her top choice; you can’t have a much stronger “revealed preference” than that, but she’s excluded from the population being studied. At some highly selective schools ED admits make up nearly half the entering class, but they’re completely ignored in any study that looks only at cross-admit data. In fact, at any school that has binding ED, you’re going to exclude from the study the very people who had that school as their #1 choice, those who applied ED; the people who end up in the study are almost by definition people who didn’t have that school as their #1 choice, because if they had, they’d have applied ED. (Of course, the RD round admits will always include some people who were deferred in the ED round, but most are not in that category).</p>

<p>Sure, these cross-admit studies always show Harvard as the most-preferred, but so what? In part, that’s because most of the people who apply to Harvard have it as their #1 choice already, and they apply to a bunch of other schools as back-ups since they can’t be certain they’ll get into Harvard. That’s, what, 35,000 people a year? Sure, most of them would choose Harvard if that have that choice. But there are another 3 million who don’t apply to Harvard; no doubt some of those (1 million or so?) would love to go to Harvard if they thought they could get in, but they don’t bother to apply because they have no realistic chance. But another very large number (the remaining 2 million?) have no interest in Harvard even if they thought they had a chance at admission. What about their “revealed preferences”? They just voted with their feet, or rather with their applications, and rejected Harvard. My D1 was certainly in that category. I’m not saying she’d have been admitted to Harvard, but she certainly would have been a plausible candidate, with credentials as strong as most who apply and many who are admitted. But she had no interest in Harvard or any other Ivy except Brown, which she didn’t like as well as her ED LAC, the only school she actually applied to.</p>

<p>In the end, all the cross-admit data can tell us is the preferences of cross-admits, and if you do those comparisons–surprise! surprise!–it almost invariably comes out that the school with the lower admit rate wins. Why might that be? Hmmm, maybe it’s because applicants aren’t stupid. In general they’ll apply to their #1 choice school, and then to a bunch of slightly to substantially less selective schools as back-ups, fall-backs, or outright safeties. If they get into their #1 choice, it’s a no-brainer where they’ll enroll. But very few people are dumb enough to apply to Harvard as a back-up in the event they don’t get into a less-selective first choice. Someone whose first choice is Harvard might well apply to Tufts as a back-up. Someone whose first choice is Tufts is highly unlikely to apply to Harvard as a back-up. So the cross-admit pool between those two schools is going to be heavily skewed toward those whose first choice was Harvard, because those whose first choice was Tufts will be screened out of the sample at the application stage.</p>

3

<p>Bunch of hooey, I agree…</p>

4

<p>Here’s the 2004 version by the same authors:</p>

<p>[A</a> Revealed Preference Ranking of U.S. Colleges and Universities](<a href=“http://www.nber.org/papers/w10803]A”>A Revealed Preference Ranking of U.S. Colleges and Universities | NBER)</p>

5

<p>Do I really have to pay $32 to access this article? :rolleyes:</p>

6

<p>Excellent analysis by bclintonk. I would add that a study like this only reflects the opinions of high school seniors. It does not reflect the opinions of professors or others who might actually have some first hand knowledge about these schools. The opinions of high school seniors are mostly based on the US News rankings and the opinions of their high school classmates.</p>

7

<p>Someone with a lot of time should analyze the preferences from this CC thread:</p>

<p><a href=“http://talk.collegeconfidential.com/college-confidential-cafe/959717-would-you-rather-attend-310.html#post15517546[/url]”>http://talk.collegeconfidential.com/college-confidential-cafe/959717-would-you-rather-attend-310.html#post15517546&lt;/a&gt;&lt;/p&gt;

8

<p>A lot of time indeed. And a lot longer than it would take to count the handful of students who account for the revealed preferences of Cal over Stanford. What was it again the last time we added them up? Nine or eleven? I can’t remember. </p>

<p>Have a superb weekend, my cubby friend. And send in that alumni donation to … UCSF!</p>

9

<p>The authors did discuss ED and the “manipulation” of yield and acceptance rate extensively. The regional analysis is also quite interesting. The paper, if not the ranking itself, is worth a read.</p>

10

<p>Haha! UCSF doesn’t need my money. :)</p>

11

<p>I just wanted to say a few words about early decision, since it seems to be the main reason people immediately dismiss any study on revealed preference. In the section titled “Early Decision and Strategic Self-Selection of Applications” (P. 460), the authors listed two other rankings that took ED into consideration. In fact, those rankings may be subtly biased in favor of ED schools. “(some students)…were engaging in self-protective psychological behavior in which they convinced themselves that the college where they would be committed to enroll (if admitted) was the only one they liked.”</p>

<p>Obviously, these are just student preferences. They are probably more useful to spectators and school administrators than to prospective students.</p>

<p>Below are the three rankings. You’ll just have to read the paper to find out what they mean :)</p>

<p>Harvard University 1 1 1
Caltech 2 7 6
Yale University 3 4 2
MIT 4 3 8
Stanford University 5 2 3
Princeton University 6 5 4
Brown University 7 6 5
Columbia University 8 8 7
Amherst College 9 13 10
Dartmouth College 10 11 9
Wellesley College 11 33 29
University of Pennsylvania 12 12 11
University of Notre Dame 13 14 18
Swarthmore College 14 10 17
Cornell University 15 15 14
Georgetown University 16 9 13
Rice University 17 21 20
Williams College 18 22 15
Duke University 19 16 12
University of Virginia 20 25 23
Brigham Young University 21 47 53
Wesleyan University 22 24 22
Northwestern University 23 17 19
Pomona College 24 20 26
Georgia Inst. of Technology 25 43 41
Middlebury College 26 29 16
UC Berkeley 27 42 30
University of Chicago 28 19 25
Johns Hopkins University 29 30 21
U. of Southern California 30 37 58
Furman University 31 49 71
U. North Carolina at Chapel Hill 32 39 38
Barnard College 33 26 27
Oberlin College 34 45 51
Carleton College 35 23 34
Vanderbilt University 36 55 59
Davidson College 37 31 37
UCLA 38 44 54
University of Texas–Austin 39 53 79
University of Florida 40 84 105
New York University 41 36 31
Tufts University 42 40 32
Washington and Lee 43 61 67
Vassar College 44 28 49
Grinnell College 45 112 45
University of Michigan 46 46 50
U. Illinois Urbana-Champaign 47 41 76
Carnegie Mellon University 48 27 28
U. of Maryland–College Park 49 86 102
College of William and Mary 50 69 64
Bowdoin College 51 62 33
Wake Forest University 52 58 47
Claremont Mckenna College 53 85 36
Macalester College 54 70 43
Colgate University 55 83 60
Smith College 56 57 40
Boston College 57 59 55
University of Miami 58 73 82
Mount Holyoke College 59 82 56
Haverford College 60 32 24
Bates College 61 74 35
Connecticut College 62 75 70
Kenyon College 63 65 85
Emory University 64 35 42
Washington University 65 50 46
Occidental College 66 96 90
Bryn Mawr College 67 51 63
Southern Methodist University 68 90 113
Lehigh University 69 79 75
Holy Cross College 70 52 48
Reed College 71 18 44
Rensselaer Polytechnic Institute 72 56 74
Florida State University 73 99 112
Colby College 74 64 39
UC: Santa Barbara 75 106 108
Miami University: Oxford 76 94 93
George Washington University 77 89 77
Fordham University 78 113 85
Dickinson College 79 63 89
Sarah Lawrence College 80 34 66
Catholic University of America 81 111 109
Bucknell University 82 88 87
U. of Colorado at Boulder 83 101 88
U. of Wisconsin at Madison 84 71 69
Arizona State University 85 91 98
Wheaton College 86 48 61
Trinity College 87 72 68
Rose-Hulman Inst. of Tech. 88 54 73
UC–Santa Cruz 89 104 111
Boston University 90 76 83
UC–San Diego 91 81 86
Tulane University 92 93 80
University of Richmond 93 60 65
Case Western Reserve 94 95 81
Colorado College 95 68 57
Indiana Univ.–Bloomington 96 98 101
Penn State–University Park 97 87 96
American University 98 100 99
Hamilton College 99 97 72
University of Washington 100 80 95
University of Rochester 101 67 92
Michigan State University 102 107 114
Lewis & Clark College 103 109 91
Clark University 104 110 103
Skidmore College 105 77 78
Purdue University 106 66 106
Colorado State University 107 103 100
Syracuse University 108 105 97
University of Vermont 109 92 104
Scripps College 110 38 52</p>

12

<p>Interesting to note among those 3 rankings that the same 8 colleges share the top 8 spots in all 3 rankings. That does seem to set them apart, as a group, from the others.</p>

<p>What any preferential ranking doesn’t tell you farther down the list is what drove the preference. Was there a difference in preference between lower/middle/upper class students, depending on whether they needs and/or qualified for aid? Were there regional differences?</p>

13

<p>^ There is a regional analysis in the appendix. BYU ranked #6 in Region 8 (AZ, CO, ID, MT, NM, NV, UT, WY) :)</p>

14

<p>Unlikely that anyone deciding would a) find, b) care or c) understand the ‘new’ rankings. </p>

<p>For those who want to have fun with inaccurate data, there are dozens of more user-friendly preference tables available, like this one:</p>

<p>[Compare</a> Colleges: Side-by-side college comparisons | Parchment - College admissions predictions.](<a href=“Compare Colleges: Side-by-side college comparisons | Parchment - College admissions predictions.”>Compare Colleges: Side-by-side college comparisons | Parchment - College admissions predictions.)</p>

15

<p>^This is NOT a ranking of who’s “better”. The study was never meant to be a measurement of school quality, no more so than a survey of Coke vs. Pepsi drinkers is a measurement of beverage quality. It’s just a research on consumer behavior. Like I said, it’s probably more useful to spectators and school administrators than to prospective students.</p>

<p>That said, I would not compare a peer-reviewed paper by four professors from Harvard, Boston U, Stanford and Yale to the like of Parchment, which is pure entertainment.</p>

16

<p>The data here is really old. They’re looking at high school seniors from the class of 2000. The methodology may be interesting, but this data is quite outdated, and doesn’t reflect changes in revealed preference that could’ve happened over the past 10-12 years.</p>

17

<p>

</p>

<p>That is probably the most charitable term one could use to describe the works of that company.</p>