USNWR ('07-91) Avg Rank + WSJ Feeder + Revealed Preferences

College Search & Selection
1

<p>**Results:</p>

<p>USN('07-'91) Avg Rank / (USN Wgt.) / WSJ Rank / (WSJ Wgt.) / Rev. Prf. / (RP Wgt.) / Tot. Wgt. Avg.**
1 Harvard University 1.41 0.56 1 0.3 1 0.3 1.16
2 Yale University 2.53 1.01 2 0.6 2 0.6 2.21
3 Princeton University 1.82 0.73 3 0.9 6 1.8 3.43
4 Stanford University 4.65 1.86 4 1.2 3 0.9 3.96
5 Massachusetts Inst. of Technology 4.82 1.93 8 2.4 5 1.5 5.83
6 Dartmouth College 8.47 3.39 7 2.1 10 3 8.49
7 Columbia University 10.06 4.02 11 3.3 8 2.4 9.72
8 Duke University 6.06 2.42 6 1.8 19 5.7 9.92
9 Brown University 13.41 5.36 12 3.6 7 2.1 11.06
10 University of Pennsylvania 8.65 3.46 16 4.8 12 3.6 11.86
11 California Inst. of Technology 5.71 2.28 28 8.4 4 1.2 11.88
12 Cornell University 12.12 4.85 25 7.5 15 4.5 16.85
13 University of Chicago 11.47 4.59 14 4.2 27 8.1 16.89
14 Rice University 15.12 6.05 20 6 17 5.1 17.15
15 Northwestern University 12.65 5.06 21 6.3 21 6.3 17.66
16 Georgetown University 21.71 8.68 17 5.1 16 4.8 18.58
17 Johns Hopkins University 14.18 5.67 24 7.2 28 8.4 21.27
18 University of Notre Dame 19.00 7.60 35 10.5 13 3.9 22.00
19 University of Virginia 21.06 8.42 33 9.9 20 6 24.32
20 University of California-Berkeley 20.47 8.19 41 12.3 23 6.9 27.39
21 University of Michigan-Ann Arbor 23.71 9.48 30 9 42 12.6 31.08
22 Tufts University 26.50 10.60 45 13.5 40 12 36.10
23 Emory University 17.86 7.14 36 10.8 61 18.3 36.24
24 Carnegie Mellon University 22.94 9.18 51 15.3 46 13.8 38.28
25 Washington University 15.94 6.38 47 14.1 62 18.6 39.08
26 University of California-Los Angeles 25.38 10.15 61 18.3 36 10.8 39.25
27 U. of North Carolina-Chapel Hill 26.29 10.51 90 27 31 9.3 46.81</p>

<hr>

<p>**USNWR 2007-1991 Average Ranking<a href="Weighting%20=%2040%">/b</a>
**Wall Street Journal Feeder Ranking<a href="Weighting%20=%2030%">/b</a>
**Revealed Preferences Ranking<a href="Weighting%20=%2030%">/b</a></p>

2

<p>Wow, this is good stuff, though I still think the revealed preferences ranking is bunk since it doesn't give school v school comparisons...</p>

3

<p>Well, none of the rankings are "perfect"... given that USNWR seems to be the default standard - I gave it the highest weighting of the three.</p>

<p>The average USNWR ranking over the 17 years is also more useful (IMO) since this tends to "normalize" dramatic drops and rises in any given year.</p>

4

<p>I don’t really trust the RP that much. I mean, how many people apply to both Caltech and Georgetown? Also, the rankings assume that there is something “revealing” about the behavior of the students responding to the survey. Moreover, most preferences are made before applications are sent, not after people are accepted.</p>

5

<p>
[quote]
I don’t really trust the RP that much

[/quote]
</p>

<p>...and I don't trust the Peer Assessment, but whaddyagonnado? gotta take the good with the bad... nothings perfect.</p>

6

<p>very interesting. Thanks for doing this. I think this ranking makes a lot of sense. I think the Revealed Preference method has merit. Vanderbilt and Wake Forest didn't make the top 27?</p>

7

<p>In News Week, this one, US News, WSJ Feeder, the THES, the "Prestige Rankings" based on peer assesment...the same group of schools is at top of all of them, glad to know theres some sort of consistency.</p>

<p>I think this one, which weights all of these, is one that can be cited for future reference.</p>

8

<p>
[quote]
Vanderbilt and Wake Forest didn't make the top 27?

[/quote]
</p>

<p>The short answer is "no". The list of 27 are the way the total weighted averages sorted out.</p>

<p>However, Vanderbilt and Wake Forest would likely have rounded out the top 30 (the other being U. of Rochester - more on that later) - although without "complete" information, it was impossible for me to create a rank that would have been 100% bona fide - 100% consistent.</p>

<p>Why? Both Vanderbilt and Wake Forest (the only schools - besides U. of Rochester - again more on that later - that did not have bona fide rankings in all three lists)</p>

<p>Specifically, Vandy and Wake did not rank (or at least not in the published Overall Top 50) in the WSJ Feeder survey. However, you'll note from the data above I was able to use the separate "WSJ Publics Feeder List" to ascertain the overall ranks for some of the publics that did not make the original WSJ 50 Feeder list (for example, the separate WSJ Publics Feeder list showed that UCLA was no. 60 and UNC was no. 90 - figures which I did use for the rankings above).</p>

<p><a href="http://wsjclassroom.com/pdfs/wsj_college_092503.pdf%5B/url%5D"&gt;http://wsjclassroom.com/pdfs/wsj_college_092503.pdf&lt;/a> (TOP 50)
<a href="http://wsjclassroom.com/pdfs/wsj_college2_092503.pdf%5B/url%5D"&gt;http://wsjclassroom.com/pdfs/wsj_college2_092503.pdf&lt;/a> (TOP PUBLICS)</p>

<p>Also, you'll also note that in the comment section for no. 50 (Reed College) WSJ made a reference to Carnegie Mellon ranking no. 51 on the list - a ranking I did use for CM in the above ranking.</p>

<p>So we can definitely assume that there was an overall LARGER (i.e. more than 50 schools) WSJ list but it wasn't released as far as I know. Therefore I was reluctant giving Vandy or Wake a "phony", "guesstimate" or even "controlled" number. Though, Vanderbilt and Wake Forest did have rankings in the Revealed Preferences survey (Vandy no. 34, and Wake Forest no. 50).</p>

<p>As for U. of Rochester, they were ranked no. 25 in the 1991 USNWR list and did not appear again. U. of R also did not feature in the WSJ feeder list and ranked no. 96 in the Revealed Preferences survey.</p>

<p>For those reasons, I stripped out those three schools: Vandy, Wake, U. of R. I strongly suspect that Vandy and Wake would have made the list, although I can't say with any degree of confidence which spots specifically they would have ranked.</p>

9

<p>this is great stuff. i agree with the methodology (esp. the averaging of the previous US News rankings).</p>

10

<p>I like this ranking. It is skeletal, but very promissing. </p>

<p>I would have added three more sections (the peer assessment score, alumni effectiveness and professional placement) and replaced the revealed preferences rank with a quality of student body rating and the USNWR rankings with a rating of university resources. Once thos 6 components are properly measured, I would give a 16.66% weight to each of those 6 sections to come up with an overall ranking. </p>

<p>Here's my reasoning:</p>

<p>1) The peer assessment score is very important, whether we agree with it or not, it is an accurate gauge of what the academic world thinks of peer institutions. </p>

<p>2) Revealed preferences is meaningless. A university does not have to appeal to a million mega talented students, so long as it can attract enough mega talented students to fill its freshman class. Does it really make a difference if 20,000 incredibly gifted students instead of 5,000 incredibly gifted students dream of enrolling into a university that can only enroll 5,000 students? That is why I think a rating of student bodies is far more effective than the revealed preferences ranking. When evaluating the quality of a student body, it is important to calibrate certain statistics to come to an accurate conclusion. For example, some universities report SAT scores according to the best result in one sitting whereas others mix and match best scores in individual sections. Also, some universities report average unweighed GPAs whereas other schools report average weighed GPAs. % of students graduating in the top 10% of their class is important, but it is important to understand the makeup of the student body. If the majority of the students at a university come from elite high schools, chances are, fewer students will have graduated among the top 10% of their class. On the other hand, a university that attracts mostly students from mediocre high schools could easily have a very elevated portion of students graduating in the top 10% of thei class. An accurate rating of student bodies must take all of those factors into consideration.</p>

<p>3) The graduate school placement conducted by the WSJ a couple of years ago was a very good start, but it could have been improved upon by expending the list from the top 5 to say the top 15 and by adding separate sections for Engineering graduate schools and placement into graduate programs in the traditional disciplines.</p>

<p>4) Professional placement is a very important and is never concidered. Universities should work closely with companies and recent graduates to compile a detailed report on the types of jobs (industries, regional distribution, average pay, job titles and the number of students joining "exclusive" companies like the zippy dot.coms, MCs and IBs etc...).</p>

<p>5) Resources are an important part of the equation. How good are the facilities at a university? Are they open to all students? How are the laboratories, libraries, computer centers? How good is the faculty? Are professors readily accessible to those who seek them out? Are there opportunities for cutting edge research? How much does a university spend on its students? Size of endowment and endowment per student (adjusted properly for state schools).</p>

<p>6) Alumni effectiveness. By that, I mean how loyal and connected is the alumni base? This is important because an alumni network can greatly enhance a person's future success. It is also important because alumni loyalty and success is an indication of quality of life and education at a university. This said, alumni giving rates are meaningless as they do not measure alumni loyalty or success but rather, alumni size and the university's ability to reach out to its alums.</p>

<p>Obviously, we will not be able to have an accurate measure on most of those 6 cirteria anytime soon. The day we do, we'll be able to come up with a very accurate rating of universities.</p>

11

<p>Peer assessment score means nothing and favors publics.</p>

12

<p>Sternman, the PAS measures the quality of the faculty and of the overal academics of a university. It does not always translate into great undergraduate education, which is why it must be taken in concert with several other factors.</p>

13

<p>Alexandre-
The revealed preference rating is based on where students choose to attend who are accepted at both schools. So, if a student is accepted at Michigan and Stanford, and chooses Michigan, then that is counted as a "win" for Michigan. It isn't directly related to the number of applications as your above post suggests. Or, did I misunderstand you?</p>

14

<p>
[quote]
Wow, this is good stuff, though I still think the revealed preferences ranking is bunk since it doesn't give school v school comparisons...

[/quote]
</p>

<p>
[quote]
I don’t really trust the RP that much. I mean, how many people apply to both Caltech and Georgetown?

[/quote]
</p>

<p>Uh, these points were addressed in the paper. You don't HAVE to have direct school cross-yield comparisons. </p>

<p>I can detect no methodological flaws in the paper. More importantly, nobody else in the academic world has ever debunked the methodology, and the easiest way to debunk any academic paper is to present flaws in the statistical methodology. The paper is still not only academically valid (until debunked) but has actually become one of the more highly cited papers in academia.</p>

<p>Now, if you want to say that revealed preferences should not be important, then that's a different story. But there is no reason at this time to not believe that the paper supports exactly what it says it supports. That is, of course, unless you can find a methodological flaw. I know I can't. </p>

<p>
[quote]
2) Revealed preferences is meaningless. A university does not have to appeal to a million mega talented students, so long as it can attract enough mega talented students to fill its freshman class. Does it really make a difference if 20,000 incredibly gifted students instead of 5,000 incredibly gifted students dream of enrolling into a university that can only enroll 5,000 students?

[/quote]
</p>

<p>I'm afraid I can't agree with that because what you just said inherently encapsulates revealed preferences. In your scenario, those 5000 students carry the revealed preference of wanting to go to that school that enrolls 5000 students. Hence, methodologically speaking, according to the paper, that would represent 5000 wins (and zero losses), which would necessarily mean that this school would rank very highly using the revealed preferences methodology. In fact, it would probably be #1, unless there was some other school out there that also had no losses, and for that, you would then perform the statistical cross-comparison that the paper discusses. </p>

<p>In reality, what happens is that no school the #1 choice of all or even most students who go there. Every school has some students who would rather be going somewhere else, but didn't get in. Even Harvard has some students who would rather be going elsewhere. I know a Harvard student who said that he'd rather be studying engineering at MIT, but didn't get in. That is what the paper measures. </p>

<p>The more I learn about statistics and research methods, the more respect I have for the revealed preferences paper. It's a metholodogically sound paper. Most rankings actually become less respected the more you learn about statistics and research methods. But the RP paper actually becomes MORE respected. </p>

<p>The only "flaw", if you will, of the paper, is that revealed preferences are just that - they are revealed preferences, and that human beings sometimes "prefer" things that are not good for them. That's why some people prefer to smoke, drink, not exercise, and make other bad choices. Heck, I sometimes "prefer" bad choices like eating a bag of chips instead of getting a real meal. Hence, I can agree that one could make a case that perhaps people should not be preferring the schools they are preferring. But that doesn't take away from the fact that certain schools are preferred.</p>

15

<p>The flaw in the revealed preferences is that a school that has many cross-admits with the highest ranked schools loses out...it does not contain information on head-to-head so you can't judge where a school is among its peers</p>

<p>The current ranking made is probably the best one I've seen on this board</p>

16

<p>I think my point was misunderstood. I was merely saying that popularity does not equal quality. I can see how the quality of the student body matters to the overall quality of a university. However, the RP and the quality of the student body aren't the same.</p>

17

<p>I don't think WSJ feeder should be considered either as not many kids WANT to attend grad school if their ugrad experience gives them the job/career they want.</p>

<p>Example: Stern kids usually accept their jobs because they are good. Other business school universities are also hurt by this such as UPenn and Carnegie Mellon. Upenn should be top 5 in the rankings.</p>

<p>Also the revealed preference is a bit flawed. I think using past us news AVERAGED is the best plan. revealed preference is flawed because kids have different reasons for attending college. This helps PUBLIC SCHOOLS AGAIN BECAUSE STATE TUITION MAKES IT AFFORDABLE FOR MANY KIDS WHO MIGHT CHOOSE UMICH OVER A 40K IVY. </p>

<p>This ranking OVERRANKS publics by the full 30% and even more as Us News peer ranking helps Publics as well.</p>

18

<p>Edit: Sorry about the caps.</p>

19

<p>Well, for pre-law and pre-med students, their goal is to attend a top med or law school...I mean, you HAVE to attend grad school to become a doctor or lawyer pretty much</p>

<p>Business school is a bit iffy...I mean, if you are succesful you might not need to go, but its still useful</p>

<p>These are for professional schools, not grad schools, so entrance into them is more necesary than grad school when set on a certain career</p>

20

<p>sakky et al:</p>

<p>With regards to Revealed Preferences, check out Simpson's Paradox. But, the main flaw in the RP is the statistical bias to begin with -- check out the footnotes.</p>

<p>Alexandre:</p>

<p>While professional placement may help the Unis, not sure its valid for a LAC. But, I do agree, that the WSJ feeder ranking is just plain silly; more east coast statistical bias. Only ONE grad school west of the Mississippi River? Perhaps for Law Schools, the Top 5 is important, but for med schools, there are two, H&H, and then the next 8; for med gravitas, attend either one of the big two, or top 10. I have no doubt that many Calif kids would go to H law, but might prefer UCLA or Boalt over Columbia or Chicago, for example. OTOH, what the feeder ranking shows is that kids who attend east coast colleges tend to attend east coast grads schools! Duh!</p>