<p>Does 2003 really qualify as “a few years ago”?</p>
<p>Technically speaking, no one is stomping on new grounds here. The Cass-Birnbaum college guides were mixing research universities with LACs nearly a half-century ago; I assume your mention of the Atlantic Monthly issue is for footnote purposes only since they too separate (is segregate too harsh a word?) their rankings and have been doing so ever since they resurrected them four years ago.</p>
<p>stateuniversity.com has another ranking that has been mixing research universities and LACs for years. Its major criteria are ACT/SAT scores, student retention, faculty salary, and student / faculty ratio. So apparently it is a largely (or completely) data-driven ranking with no subjective elements (such as the Forbes “rate my professor” or USNWR’s peer assessments). This year’s results? 4 LACs (but only 1 Ivy) make the top 10. 10 LACs (and 3 Ivies) make the top 20. Harvard, at #22, is sandwiched between Bowdoin (#21) and Grinnell.</p>
<p>I’m not saying that any of these rankings reflect Ground Truth better than any other, but it’s not as though only one off-the-wall set of criteria shows relative strength in liberal arts colleges.</p>
<p>“But I’m suspicious of the urban legends of the kids who regularly turn down Harvard for … if it were as frequent as we’re led to believe, Harvard’s yield numbers would be half of what they are”</p>
<p>But more to the point, what does it MEAN that some kid turned down Harvard for (x)? If the goal here is to measure how good of a quality various institutions are, the actual choices made don’t matter. The choice of Harvard over x or x over Harvard tells you zero about either Harvard’s or X’s caliber of education.</p>
<p>Hey, that’s a pretty impressive site that doesn’t receive nearly the notice from CC it deserves (perhaps, because it’s not connected with a magazine?). It also illustrates what I was referring to in post #44: thirty years ago, six or seven LACs would have made the top ten easily.</p>
<p>tk - Interesting link and perspective. (#63)</p>
<p>Also, regarding ranking places like Harvard, we cannot underestimate the influence of historical gravitas despite transiencies in “current” capabilities.</p>
<p>Yes, stateuniversity.com is an interesting site. The methodology appears to be simple and, in my opinion, focuses on 3 areas that really should count for a lot: quality of the students, quality of the faculty, and quality of engagement among students and faculty. Whether their specific metrics accurately assess those 3 areas, I’m not so sure.</p>
<p>Their ranking does seem to have a few problems:
fairly extreme year-to-year swings
a lack of transparency in the methodology (with little or no detail about how, for example, the S:F ratio is calculated)
some specific results that seem rather far off the mark (e.g. Columbia at #71, showing a 14:1 student-faculty ratio… compared to USNWR, which reports a 6:1 ratio)</p>
<p>As far as I can tell from this and other sites, Chicago has smaller classes, higher faculty salaries, higher student retention, and higher SATs than Cornell. So how is it that according to such criteria, Cornell winds up at #11 and Chicago #40? Perhaps it has something to do with idiosyncracies of how S:F ratios are calculated, or something like that. This is why the criteria need to be more clearly described.</p>
<p>^^On quick glance, it appears that faculty/student ratios are driving the stateuni rankings among the top.</p>
<p>Dartmouth claims to have a better faculty/student ratio than what is available on stateuni.com. Perhaps that site is including grad faculty as well? (If so, I’m not convinced that the number of med/law school faculty should matter to undergrads.)</p>
<p>“But I’m suspicious of the urban legends of the kids who regularly turn down Harvard for … if it were as frequent as we’re led to believe, Harvard’s yield numbers would be half of what they are”</p>
<p>What if they were? What if Harvard’s yield was only half of what it was, and comparable to universities that are lower on the food chain? All that would mean is that Harvard admitted students who, in weighing all their options, decided to go elsewhere for various reasons. It wouldn’t mean ANYTHING with respect to “and how good is Harvard.” All yield reflects is popularity among 18 yos. BYU has a high yield because Mormons want to go there and don’t want to go elsewhere, and won’t go elsewhere if admitted. I surmise ND probably has a pretty good yield for similar reasons. Great; but so what? It provides no data for the question at hand, which is some attempt to measure quality.</p>
<p>-student faculty ratio
-classes taught by TAs (none or few)
-class sizes
-job placement success (fickle number due to WHAT job but looking for high number employed or in grad school)
-consistent advisor throughout
-racially and economically diverse student body
-availability of lots of sports at club/intramural level
-financial and merit aid availability
-focus on undergrads, if uni</p>
<p>…probably some other things but those would be a great start for me personally.</p>
<p>In addition to having missed that data list of cross-admitted students, I admit to also draw a blank here. Are you saying that the Atlantic Monthly has published college rankings for the past four years, or should I write the past few years? </p>
<p>That statecollege site is wack. Whatever rating method they use is completely nontransparent, since colleges with the same metrics for the things they say are most important can be ranked very differently, and colleges with very different metrics can be ranked the same. The information they present seems to include graduate and professional programs – they have a chart that helpfully informs us that the vast majority of Harvard students major in business administration, law, education, or public policy. I have no idea where the student-faculty ratios come from, but it clearly isn’t the same place anyone else uses.</p>
<p>ohiobassmom, your list is very close to what I would like to see as well. That is why the intent that Forbes has is good, but the methodology is still poor. I would like to see graduate success, not just number of graduates immediately after graduation, but success 10 and 20 years out. I wish there was a way to measure student engagement, but the largest student engagement surveys don’t release enough data to make that feasible.</p>
<p>No, you were correct. I was thinking of the <em>Washington</em> Monthly. I remember The Atlantic Monthly poll; it had its possibilities. To their credit (and, James Fallows’) they had too much integrity to keep recycling the same basic information year after year as though it were “news”.</p>
<p>Well, yeah, that’s the basic idea: if you accept the premise that such a poll would be a good thing, then you pay attention to it and point out obvious mistakes. You don’t toss the baby out with the bath water.</p>
<p>Here’s why I think cross-admit “data” (and xiggi’s right, even the figures reported on places like parchment.com are essentially unverified, self-reported anecdotes, therefore highly suspect) are even less informative than you think. Almost invariably, the school that wins the cross-admit preference in any pairwise comparison is the school that has the lower admit rate. I just ran 100 pairwise comparisons using cross-admit “data” from parchment.com, and in 82 out of 100 cases the school with the lower admit rate won. In 2 of the remaining cases, the schools had identical admit rates and cross-admits (or people claiming to be cross-admits) expressed a slight (<60-40) preference for one of the two schools. In four cases, cross-admits expressed no preference as between the two schools despite differences in admit rates. In the 12 remaining cases, cross-admits preferred school A even though school B had a lower admit rate, but in those cases the preferred school was more highly ranked in US News and/or had higher middle 50% test scores, and therefore was arguably the harder school to gain admission to despite a lower admit rate.</p>
<p>What does this tell us? Well in the main, cross-admit comparisons give us little or no additional information beyond what we can learn from selectivity data: it tells us which schools are hardest to get into. Now it could be selectivity really is the key to quality and that these are just smart consumers making smart consumer choices. Or it could be that selectivity is highly correlated with 18-year-olds’ sense of “prestige” (a word you’ll see frequently on CC). </p>
<p>But I think there’s actually a simpler and more straightforward explanation. If your top choice is Michigan State (and it actually is for a lot of kids in Michigan, including a fair number who could get admitted to Michigan), then you typically don’t apply to Michigan as a back-up. You line up schools that are easier to get into than Michigan State (70% admit rate) in case your dreams don’t come to fruition. But if you’re a Michigan resident dreaming of attending the University of Michigan (51% admit rate in 2010, less than 40% now that they’re on the Common App), you might very well also apply to Michigan State because you’ll think it’s an acceptable Plan B if your first choice doesn’t come through. Consequently, the Michigan-Michigan State cross-admit pool will be heavily weighted toward applicants whose first choice is Michigan; those whose first choice is Michigan State by and large won’t apply to Michigan. And, no surprise, Michigan trounces Michigan State in the cross-admit preference, 78-22. (Anecdotal confirmation: I grew up in a small town in Michigan back in the day; 5 kids in my HS class had Michigan State as their first choice, all 5 were accepted, not a single one of them applied to Michigan. Two of us had Michigan as our first choice, we both applied to Michigan and Michigan State and were accepted at both schools, and we both chose Michigan. This is perfectly consistent with the reported cross-admit preferences.)</p>
<p>Similarly, if you’re a Virginian and your first choice is UVA (33% admit rate), chances are you’re not going to apply to Duke (16% admit rate) or Princeton (9% admit rate) as a back-up; that would be just plain silly. You’re going to formulate a back-up plan that might include VaTech (67%) and some other schools that are as easy or easier to get into than UVA. But if VaTech is your dream school, you’re not likely to apply to UVA as a backup, because if you don’t get into VaTech your chances of getting into UVA are slim to none. And sure enough, UVA trounces VaTech 93-7 in cross-admit preferences, but UVA in turn is soundly thrashed by Duke, 86-14, and by Princeton, 80-20, because it’s perfectly reasonable for applicants whose first preference is Duke or Princeton to also apply to UVA as a back-up, but not vice versa. For the very same reason, however, Duke is hammered by Princeton, 72-28, in cross-admit preferences; not surprising given that Princeton’s admit rate is roughly half that of Duke, making it reasonable for applicants who aspire to Princeton to apply to Duke as a back-up, but not vice versa.</p>
<p>Cross-admit preferences are always going to be heavily skewed–dominated, I would argue–by considerations like this. Consequently, the school that has the lower admit rate or is otherwise perceived to be harder to get into is almost always going to come out on top. All this tells us is that applicants are rational enough to know that it’s kinda dumb to apply to Duke as a back-up if your first choice is UVA, because Duke is harder to get into than UVA; but it’s sensible to also apply to UVA if your first choice is Duke, because UVA is easier to get into than Duke. That is, it tells us virtually nothing except which school is harder to get into, which we can more easily and more reliably determine by looking directly at admit rates and entering class stats.</p>
<p>bclintonk and Pizzagirl have interesting points about yield but the preferences that matter are those of the students that are actually Harvard caliber that decide to stay in-state and attend UMich or UVA. Based on the number of National Merit Scholars that these three schools enroll, its clear that Harvard or <em>insert elite private school</em> is getting a fair number of the super smart kids.</p>
<p>Yeah, there are a lot of Michigan high school students that don’t even bother applyting to Harvard and are perfectly content going to UMich, but they wouldn’t have gotten in to the big H anyway and you bet they know it.</p>
<p>Oh bclintonk, I think you are far oversimplifying the dynamics of college selection.</p>
<p>
I would argue that most Virginians, certainly not all, but most who are smart enough students to be viable candidates to get accepted into Duke or Princeton in the first place don’t prefer UVA to these schoools though. Students aren’t that attached to particular institutions and intend to go to the most prestigious school that they can afford. You are wrong if you think Virginians have some sort of special affection for UVA that they won’t let the school go even if they get into Duke or Princeton and these schools are affordable</p>
<p>Those admit rates are outdated by the way with Princeton having an 8.5% acceptance rate and Duke boasting an 8.9% acceptance rate.</p>
<p>You said in a previous post that Virginians aren’t all that enamored by private schools and far more choose UVA than their neighbor Maryland when in reality Virginians are just as well represented at Duke and Princeton as MD residents.</p>
<p>The student you are describing who just simply prefers UVA to Duke and Princeton even though he/she’s a viable candidate for the latter two schools and thus doesn’t apply to the two privates is very rare, far rarer than I suspect you are willing to admit.</p>