<p>A new college ranking system (cuz we all know how badly we need one of these) was just released. It is based ONLY on the preferences of colleges.</p>
<p>I’m not explaining this well, but if someone chooses to go to Syracuse over Harvard, Syracuse gains a “point” and Harvard loses a “point” (and these things are all weighted by some superscientific profs at Penn, Harvard and someother ivy or maybe Stanford). So, here the site: <a href=“http://ssrn.com/abstract=601105[/url]”>http://ssrn.com/abstract=601105</a></p>
<p>and here are the top 20 for all you lazy bums:</p>
<p>So seventeen year olds are the best at ranking the quality of schools? This is a good indicator of what highshool kids THINK are the best schools (and maybe they're right) but there is no evidence to back it up. The general public is not always right. Britney Spears might sell more records than Bob Dylan, but would anyone seriously argue that she is a "better" musician?</p>
<p>
[quote]
We should let distinguished faculty and experienced individuals rank colleges rather than 17 year old kids.
[/quote]
</p>
<p>I have disagree with this. The fact is, after you start doing PhD and know your profs better, you'll see that distinguished faculty often have very biased and indifferent perception on colleges, in fact no better than good 17 year old students. For example, on the average MIT professors would give higher score to Caltech, Stanford, Berkeley and less high score to Brown, Yale, etc. On the average Dartmouth profs would give higher score to Harvard, Berkeley, Stanford, and less high score to John Hopkins, Caltech and MIT. The peer assessment is biased because it fails to take into the account many crucial factors such as school specialty, student quality, and recruiter's assessment.</p>
<p>It would be better if we combined the peer assessment from the recruiters, faculty, and top cream students with the weights of each depts based on the number of their students to formulate a more accurate number. I guess they should hire some MIT mathematicians at USNEWS to get a more sensible measurement system. But of course, it may lead to less appeal to the average consumers who likes hot stuff and less sensible information ;)</p>
<p>Peer assesment might be a bit biased, but at least they've been to college and know what they're ranking. I'm not saying that there's anything wrong with the above rankings in terms of showing what highschool kids think are the good schools, but I don't see how you could argue that 17 year olds are more qualifed to rank schools that people that have actually made careers out of academics.</p>
<p>Many of these professors have different experiences and have visited many colleges as well as attended different colleges.</p>
<p>At least they aren't biased like high school students. I know idiots who think UVA is an Ivy and is > Harvard. The same goes for kids who think Virginia Tech is comparable to MIT.</p>
<p>I doubt a Professor's bias would extend to such extremes.</p>
<p>FWIW (which likely isn't much), here's the most current version of the revealed preference study (published Dec. 2005; revised methodology, but using same data, which is several years old):</p>
<p>Harvard
Cal Tech
Yale
MIT
Stanford
Princeton
Brown
Columbia
Amherst
Dartmouth
Wellsely
Penn
Notre Dame
Swarthmore
Cornell
Georgetown
Rice
Williams
Duke
UVa
Brigham Young
Wesleyan
Northwestern
Pomona
Georgia Tech
Berkeley
Chicago
JHU
USC</p>
<p>Back in the real world, MIT reports that the only schools to which it loses more cross-admits than it wins are Harvard (which wins about 2/3) and Yale; not Cal Tech. CalTech's own reports confirm it loses out to MIT most often. So, as to the study: grain of salt.</p>
<p>The best example is Berkeley, which has peer assessment score 4.8, which is higher than any other lower Ivies, even higher than Caltech. Now you should be able to see the bias that is introduced by faculty. We know that Berkeley's faculty are the best of the best. We also know for sure that NONE beats Berkeley's faculty profile. And according to peer assessment ranking, Berkeley is better than anyother but HYPSM. But we can be sure that neither recruiters nor the top students see that way. Hence its measurement is far from flawless.</p>
<p>
[quote]
Back in the real world, MIT reports that the only schools to which it loses more cross-admits than it wins are Harvard (which wins about 2/3) and Yale; not Cal Tech. CalTech's own reports confirm it loses out to MIT most often. So, as to the study: grain of salt.
[/quote]
</p>
<p>Well, the fact that on the average Caltech students are better academically than MIT students may make you think twice.</p>
<p>That seemingly incredible data is resulted from the fact that weight factor used in self-selective pool of applicant is perhaps too high. That is why I said, that this students' assessment are to be combined with the faculty + recruiters' assessment. And to get more accurate data, some weights should be included for each departments based on the number of students in that depts.</p>
<p>Nobody is saying that peer assesment is flawless (at least I'm not), but just because its not perfect doesn't mean highschool kids know more than faculty at top schools. And is this ranking done by the top 1-2% percent of highschool students? Btw, why would professors and faculty rank berkeley higher than Caltech unless they thought it was better? Peer assesment is a glorified prestige ranking, and these academics think berkeley has more prestige than Caltech. Caltech is a great school, but many, many people don't know caltech from devry or some internet tech school.</p>
<p>
[quote]
Nobody is saying that peer assesment is flawless (at least I'm not), but just because its not perfect doesn't mean highschool kids know more than faculty at top schools. And is this ranking done by the top 1-2% percent of highschool students?
[/quote]
</p>
<p>It is not unfortunately. I was saying that top faculty, while better at the average perception on college ranking, may give measurement that is far from exceptional.</p>
<p>The top students are not qualified to rank colleges either. They usually overrank Ivy Leagues and apply to such instead of other top notch schools like Caltech/Stanford. </p>
<p>These top students still have state biases as I knew a kid two years ago who got into Princeton and thought the next best schools were UVA/William and mary.</p>
<p>Uh, my point was only that the Revealed Preference study is supposedly telling us who wins cross-admit battles, on the basis of a sample of HS students and a methodology; that the study has CalTech ahead of MIT; that the actual admission data from MIT (and CalTech), at least as reported by the schools, shows that, in fact, MIT, not CalTech, wins most of the cross admits to those two schools; and so even on its own terms the revealed preference study is to be taken with a grain of salt. I was saying nothing about whether MIT or CalTech has smarter students (jeez) or which school is better.</p>
<p>
[quote]
So just because its not perfect its garbage? Is the highschool perspective perfect?
[/quote]
</p>
<p>I didn't say that the highschool perspective is perfect, as a matter of fact far from it. Nevertheless, I wouldn't consider it any less than USNEWS peer assessment ranking because at least the people who published the preference ranking use a more serious/scientific methodology than USNEWS peer assessment ranking.</p>
<p>
[quote]
The top students are not qualified to rank colleges either
[/quote]
</p>
<p>It depends on how you measure top students, if I would say top students are top 1-2% in the top 20 high schools in the States, may be it would be very different, isn't it?</p>
<p>
[quote]
Uh, my point was only that the Revealed Preference study is supposedly telling us who wins cross-admit battles, on the basis of a sample of HS students and a methodology
[/quote]
</p>
<p>And my point is: the fact that MIT actually wins over MIT vs. Caltech yield battle, but the result is not shown in the preference ranking, may not be because it failed to incorporate that data, but instead it put some weights on self-selectivity that resulted in the published ranking. For example it is possible that most students who went Caltech had already determined to go to Caltech, where most students who went to MIT would have arbitrarily chosen MIT/Caltech whichever accepted them with higher preference on MIT. Now the more dominant number of the later group may give a higher yield to MIT while not necessarily make its preference ranking higher.</p>
<p>"It depends on how you measure top students, if I would say top students are top 1-2% in the top 20 high schools in the States, may be it would be very different, isn't it?"</p>
<p>No matter how you measure it, no top student as a generalization is able to rank colleges correctly. That Princeton acceptee was top in almost everything and he would've told you UVA > Stanford and other Lower Ivies. Even the best students have state biases, local biases, Ivy biases, and other biases that are much more extreme and common than masters of academics.</p>
<p>Quote: "That seemingly incredible data is resulted from the fact that weight factor used in self-selective pool of applicant is perhaps too high."</p>
<p>The study itself admits as much, in almost the same words. Cal Tech's ranking is "problemmatic" because of the degree to which its students are self selected -- far more than MIT's, whose students routinely cross-apply to HYP and other top schools. In fact, the authors conclude that Caltech's ranking vis a vis the rest of the list is "largely through MIT". [pp 26-27, Hoxby, pdf]</p>
<p>Also, Wesdad's list omits Middlebury which ranks just below Georgia Tech and Berkeley, I think.</p>