Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!husc6!mit-eddie!ll-xn!ames!aurora!labrea!jade!ucbcad!zen!ucla-cs!coleman From: coleman@CS.UCLA.EDU Newsgroups: comp.edu Subject: Re: rankings of colleges Message-ID: <8359@shemp.UCLA.EDU> Date: Sun, 27-Sep-87 18:26:00 EDT Article-I.D.: shemp.8359 Posted: Sun Sep 27 18:26:00 1987 Date-Received: Mon, 28-Sep-87 04:37:32 EDT References: <1503@bloom-beacon.MIT.EDU> Sender: root@CS.UCLA.EDU Reply-To: coleman@CS.UCLA.EDU (Michael Coleman) Organization: UCLA Computer Science Department Lines: 78 Keywords: rankings, colleges In article <1503@bloom-beacon.MIT.EDU> hooner@athena.mit.edu (Dave Ko) writes: >The Gourman report is the premier source in the nation >on the top undergraduate schools in each field. They >use fourteen different factors to rate the programs. >I'll list a few fields and the top schools overall ... [various rankings included here...] I completely disagree. There have been many articles written which disparage the Gourman report as a ranking of college programs. In fact, at my undergraduate school (UMKC) the report was kept in ready reference and if requested, was always handed out with a disclaimer by the reference librarian and a manilla folder full of these commentary articles. But more significantly, even a cursory look at the figures which are supposed to justify these rankings should convince anyone that they are cooked and that some subjective criteria (probably the opinion of Mr. Gourman himself) is being used to create these rankings. If you scan the book, each page has a format similar to this: school overall Foo U. 4.59 4.56 4.53 ... Foo Tech. 4.58 4.54 4.52 ... Foo State 4.56 4.53 4.51 ... etc.... .... .... .... ... The criteria here are things like "faculty," "library," etc. (I don't remember exactly.) The thing that is immediately apparent here is that all of these columns of numbers are DECREASING, or at least non-increasing. Doesn't that seem odd? I mean, does any one that there are NO cases within the top 50 or so schools in each discipline in which, say school A has a better than school B, but school B has a better than school A???? Unless there is something very strange about the ways these numbers are being acquired and combined, as an objective ranking, this book REEKS! Now it just so happens that the rankings themselves are not too wildly out of sync with what most knowledgable people in the field would put forth, but I'm not willing to take the Gourman report's word for it... I don't think that too many people would argue that MIT, Stanford, and Carnegie-Mellon are the big three in graduate computer science programs, but the Gourman Report did not say so in previous years... So, where can the prospective grad student go for a more objective source of information? I would suggest a five-volume set called (approximately) An Assessment of Research Doctoral Programs in the United States. Something like that anyway. I think it's put out by the NSF or some branch thereof. It contains the results of faculty surveys of quality, as well as a great deal of objective data (library holdings, publication rate, grant monies, etc.) with which to evaluate your school in question. The schools are NOT actually ranked, but a little figuring will tell you the approximate rankings... I would have much more confidence in the stated results of this report than those of Gourman. A shortcoming of this report is it only ranks (graduate) programs of Ph.D. granting schools, but then, most student who are actually interested in the ranking of the school would probably not consider going to a non-Ph.D. granting grad school anyway... There are other such reports, I understand. I've heard of a ranking put out by the National Academy of Sciences. These would also be worth checking... And finally, it is worth considering that there is some variation in schools by specialization. For example, Purdue is considered (in my opinion) to be a pretty good computer science school (although I don't see it in the computer science rankings of Gourman), but if you wanted to study artificial intelligence, it would probably not be as good of a place to go because it's not their specialty (someone correct me if this is wrong). Look first, then leap. ** Mike Coleman ** ARPANET: coleman@cs.ucla.edu **