<p>50 Top Colleges has an interesting way of compiling the data. I can almost agree with some of the school’s rankings, albeit being a strong follower of USNEWS. Conversely, outside of CC if one were to state they go to Penn, Brown, Dartmouth, etc the average layperson would be impressed (providing they had heard of the school lol). Also, as many of you know from browsing CC, there is evidence that out of these three schools people have been rejected to one, accepted to another. People have also chosen these schools over Columbia (ranked higher). People have chosen Duke over Princeton, and I recently came across a posting of choosing Cornell over Harvard (eek!) lol. Anyway, my point is at this caliber, you are blessed to be able to make such a decision! :D</p>
<p>although splitting hairs of prestige is embarrassing and unhelpful, I find that US News college rankings are more indicative of popular opinion <em>among high school and college students</em> than counselor rankings are, and consequently they have far greater impact on college-bound seniors perception of general prestige (especially since they are anchored in contemporary admissions, financial and administrative data as opposed to the gestalt opinions of your 45 year old guidance counselor).</p>
<p>Nevertheless, this is sort of tautological - I’m arguing that the overall college rankings express de facto prestige because HS seniors consider them to do so, while the guidance counselor rankings express what the adult populace actually thinks. And I would argue that, while this might be true, it’s also a lagging indicator and completely unimportant. US News college rankings indicate what future counselor rankings are sure to, as well.</p>
<p>Also, and more relatedly, I think that Brown and Penn are equivalent. I chose Penn, as did Shaheiruden, but there are many who choose Brown. Fit is more important at this level.</p>
<p>^If you’re looking for popular opinion among the people who make college choices (i.e., high school seniors), then you should be looking at the revealed preferences rankings: [SSRN-A</a> Revealed Preference Ranking of U.S. Colleges and Universities](<a href=“http://papers.ssrn.com/sol3/papers.cfm?abstract_id=601105]SSRN-A”>http://papers.ssrn.com/sol3/papers.cfm?abstract_id=601105)</p>
<p>In that ranking, Brown is expected to win (according to the statistical model) 98% of contests with Penn and 90% of contests with Dartmouth.</p>
<p>In reality, at this level, I think that fit is more important than anything else.</p>
<p>^ ummm… no. in fact, the author warns on that document that 0.9 for dartmouth and .98 for penn is NOT the winning percentage… or else, harvard wins 100% against princeton, mit, stanford, penn, columbia, dartmouth, and brown, which is not true obviously.</p>
<p>I lost them at how they planned to handle the linear covariance estimators with such a (relatively speaking) small sample size.</p>
<p>But if I’m reading this correctly (in a rush so I only skimmed over a few of the paragraphs), the theta MLEs say that, in some number of simulations under the assumption of a Markov model, Brown had a higher than 50% chance of winning a cross-admit with Dartmouth in 90% of the simulations. For all we know, these chances were almost all 51% and 49%. Additionally, I’m also not sure I understand why linear covariance estimators are sufficient for this situation, but that’s okay, I guess?</p>