Consolidated thread about leaked US News 2006 Rankings

College Search & Selection
261

<p>Now USNEWS in 2001 or 2002 changed its ranking methodology. They started to count as expenditures per students even the research and medical expenditures. Now let's ceck these out...</p>

<p>If you calculate an expenditure per capita of the top universities, Duke tops the list, with Emory, Penn, Vanderbilt, MIT. Now we all know where the universities spend their money. In Duke's, Emory's, Penn's, Vanderbilt's and WUSTL's case, they spend more than 50-60% of their total budget for their medical centers (hospitals). MIT spend more than 40% of its budget for it's heavy tech research. Stanford spends more than 30% for its high-tech research. Now the top 10 schools for expenditures are:</p>

<p>Duke 218609.8655
Emory 190884.2241
Penn 186657.2972
Vander 180139.003
MIT 178336.7258
Yale 146774.4771
Stanford 142099.7978
Princeton 138325.7599
Dartmouth 112533.524
WUSL 111137.8554</p>

<p>Now if you remove the medical center and heavy tech research expnditures you would get these top 10 educational expenditures.</p>

<p>Yale 146774.4771
Stanford 142099.7978
Princeton 138325.7599
MIT 105918.8718
Dartmouth 112533.524
Harvard 102563.9393
Columbia 99004.03245
Duke 98374.43946
Emory 85897.90086
Chicago 85755.48304</p>

<p>PS:These figures are not 100% exact. I just used them to give an idea of how things are...</p>

262

<p>Rutgers(surprised you missed it on the RU football board)</p>

<ol>
<li> Ties with BU, OSU, Purdue, TAMU, Iowa</li>
</ol>

263

<p>St Marys 84</p>

264

<p>people selectivity isn't really shown in the admissions rate! Although the admission rate for school #1 might be higher than school #2 it doesn't mean it is less selective comparitively. For example, I keep hearing that penn shouldn't be ranked so high because it is the "2nd easiest ivy" to get into (admission rate-wise)...however this is solely due to the fact that penn's incoming class size (2500) is more than double most of the other ivies (avg. 1000). If all the ivies had the same class size (1000) and the number of applicants, & yield remained constant for those schools then you could actually compare how selective the schools are, and that would look like this:</p>

<p>CURRENT:</p>

<p>SCHOOL APPS ACPT RATE CLASS SIZE YIELD</p>

<p>Harvard 22,796 2074 9.10% 1650 79.56%
Yale 19,448 1880 9.67% 1325 70.48%
Princeton 16,516 1807 10.94% 1150 63.64%
Columbia 18,120 2250 12.42% 1000 44.44%
Brown 16,908 2463 14.57% 1375 55.83%
Dartmouth 12,615 2149 17.04% 1012 47.09%
Penn 18,823 3912 20.78% 2450 62.63%
Cornell 24,444 6384 26.12% 3000 46.99%</p>

<p>WITH SAME SIZE CLASS:</p>

<p>SCHOOL APPS ACPT RATE CLASS SIZE YIELD
Harvard 22,796 1257 5.51% 1000 79.56%
Yale 19,448 1419 7.30% 1000 70.48%
Penn 18,823 1597 8.48% 1000 62.63%
Cornell 24,444 2128 8.71% 1000 46.99%
Princeton 16,516 1571 9.51% 1000 63.64%
Brown 16,908 1791 10.59% 1000 55.83%
Columbia 18,120 2250 12.42% 1000 44.44%
Dartmouth 12,615 2124 16.83% 1000 47.09%</p>

265

<p>barrons: what abt other LACs? colgate? macalester?</p>

266

<p>D08, that's the type of ranking games institutions play. Many schools also count MEDICAL SCHOOL and LAW SCHOOL faculty when tabulating the student:faculty ratio. That is just plain wrong.</p>

267

<p>Some even count their medical staff as professorship even though none of surgeons ever give any clases...lol</p>

268

<p>LACs (school, acceptance, sat 25-75)</p>

<p>Williams 19 1330-1520
Amherst 21 1360-1550
Swat 25 1350-1530
Wellesley 37 1280-1460
Carleton 29 1300-1480
Bowdoin 24 1290-1460
Pomona 21 1370 1530
Haverford 29 1280-1460
Middlebury 26 1380-1500
Claremont 22 1310-1490
Davidson 27 1270-1440
Wesleyan 28 1310-1490
Vassar 29 1310-1460</p>

269

<p>I think the fact that so many people dispute Duke and Penn's ranking implies that those schools are overranked.</p>

270

<p>Also, check out the Revealed Preference Ranking published by none other than the WHARTON school of business.</p>

<p><a href="http://knowledge.wharton.upenn.edu/papers/1287.pdf%5B/url%5D"&gt;http://knowledge.wharton.upenn.edu/papers/1287.pdf&lt;/a&gt;&lt;/p>

<p>According to cross admit battles and the choices of the top students admitted to the top schools, the list goes like this:</p>

<ol>
<li>Harvard</li>
<li>Yale</li>
<li>Stanford</li>
<li>Caltech</li>
<li>MIT</li>
<li>Princeton</li>
<li>Brown</li>
<li>Columbia</li>
<li>Amherst</li>
<li>Dartmouth</li>
<li>Wellesley</li>
<li>U Penn</li>
<li>Notre Dame</li>
<li>Swarthmore</li>
<li>Cornell</li>
<li>Georgetown</li>
<li>Rice</li>
<li>Williams</li>
<li>Duke</li>
<li>UVA</li>
<li>Northwestern</li>
<li>Pomona</li>
<li>Berkeley</li>
</ol>

271

<p>Now bern even the stats you gave are not completely represantative.</p>

<p>You place too much importance on yield. However you forget that yield is affected by many other factors beside scholl reputation and school quality. The location of the school, the competitors of the school (for the same applicants), etc, etc...</p>

<p>Now let's say we have schools A, B, C. A and B compete together for the same pool of applicant and C is mainly their back up choice. So an applicant if admitted in both A and B he has to choose among them. However if an applicant is rejected from both A and B and admitted from C he will definitely go to C as that's his only choice. Therefore even if C ends up with a higher yield it does not mean it is more selective or more preferred. It simply is luckier...</p>

<p>Now there are a lot of third tier schools who have yiealds of over 75%. This does not mean that their are all more preferred than Harvard, but that they are all last and only choices...</p>

<p>Best way to determine the selectivity of a school is by looking at the stats. GPA, SAT, RANK and poolong...</p>

272

<p>The Revealed Pref ranking actually takes all those conerns into account.</p>

273

<p>Even though the report itself is dated 2004, it appears that the data used in this "study" is over 5 years old. I'm not sure how valid it is today.</p>

274

<p>Many college seniors adhere solely to US news and view it as the ultimate ranking on college education. That is how US news pays the bills. Penn and Duke will continue to receive more apps and MIT will continue to be a great school despite its "abysmal" rank. Moreover, why can no one move to the subject of grad/law/medical placement? If we were to traverse into that realm, then these rankings look just about right, don't they?</p>

275

<p>There are obviously flaws with the revealed preference ranking...Brown that high? No way.</p>

276

<p>OMG I know! Notre Dame and Wellesley ahead of Cornell/Duke? UVA above UCB? Georgia Tech above Middlebury? And my personal favorite....Brown over Columbia. In what world could that ever happen?</p>

<p>To quote Alexandre:</p>

<p>"...the Revealed Preferences is not a ranking of academic excellence or of reputation. It is popularity ranking. that explains why schools like Cal, Chicago, Johns Hopkins and Michigan do so poorly in it."</p>

277

<p>We KNOW people think Penn and Duke are overrated, but why they think so (gathered from reasons mentioned in this thread) is mainly based on perceptions of prestige and "selectivity" that, for many schools, is a product of decades of reputation. Ie: Everyone knows Harvard now, because everyone knew Harvard 30 years ago etc etc. For schools that are rising and improving recently that don't have the previous reputation to fall back on, there's more of a demonstrated bias. Which is why we're getting into important factors other than selectivity and admission and prestige which form a large part of a college's quality to analyze the controversy. </p>

<p>BTW: rooster08, the working paper from Wharton you posted is not a straight study of college rankings based on overall quality or in general, but rather, a study of around 3,400 students and their preferences for colleges (and I'd wager a considerable amount of those said students are influenced by at least some degree by perceived prestige and reputation).<br>
The study itself states that desirability from a student's perspective, while certainly interesting, should not be perceived as a end all product or measure of the school's quality.</p>

278

<p>
[quote]
For example, I keep hearing that penn shouldn't be ranked so high because it is the "2nd easiest ivy" to get into (admission rate-wise)...however this is solely due to the fact that penn's incoming class size (2500) is more than double most of the other ivies (avg. 1000). If all the ivies had the same class size (1000) and the number of applicants, & yield remained constant for those schools then you could actually compare how selective the schools are, and that would look like this:

[/quote]
</p>

<p>Again, "Lies, Damn Lies and Statistics"... My b-school prof told me there would be no end to this statement and if you keep your eye out for it you will see it every day.</p>

<p>1) How can you change one variable (class size) and then keep the other critical variables (# of app.) constant?</p>

<p>2) Why do you assume that those other variables are totally and completely independent of the one variable you are changing?</p>

<p>3) Have you ever thought that students consider class size when applying to schools? (I sure did when I applied) Have you ever thought that class size affects admissions decisions?</p>

<p>4) Class size affects applicant numbers, applicant numbers affect yield numbers, yield numbers affect applicant numbers, etc. etc. etc... Hello? These are all interlinked!</p>

<p>5) In other words, if you are "normalizing" school size across schools, why not make this "normalizing" adjustment for each category? I.E. why not just assume similar yield rates and applicants? A silly question you ask? An exercise in futility you say? Would render the outcome meaningless? YES.</p>

<p>6) You just can't assume to normalize school size and then keep everything else constant. When you change something as critical as school size, the other variables WILL BE affected. How? Positively? Negatively? Who is to say. Again it's all conjecture at this point. Which is to say that your analysis is built on a bed of sand.</p>

<p>7) Heck why not take applicant yield and divide by the year each school was founded and multiply by their distance from your ass? It would have just as much merit as your analysis.</p>

<p>8) Put another way, if Harvard turned around next year and said, "Harvard's next incoming freshman class will consist of 10 students, period.", would their applicant numbers stay the same? Why or why not? And if you agree that the applicant numbers would change, then you have just answered why your analysis is half baked at best.</p>

279

<p>bern700,</p>

<p>Do you think Wharton undergrad program can exist if UPenn shrink from 2500 to 1000? Sure you can play with class size number, but you'd better provide good explanation why other variables don't change. Oh how about just getting rid of other majors and just have Wharton? That would make it really competitive. Or would it??</p>

280

<p>Instead of just shrinking everyone to the same ratio and speculating as to how the other variables MIGHT change, why can't they all be taken at face value and evaluated in that manner? And then, try to understand why each variable is what it is and how it affects other variables.</p>