How do yields compare at the top colleges?

College Search & Selection
41

<p>Collegehelp:</p>

<p>

</p>

<p>There is a big distinction between claiming that the study incorporates matriculation decisions which is true and making it a survey of cross-admit results which it is NOT. The Revelead Preference Study is a MODEL based on some very questionnable assumptions. </p>

<p>An actual cross-admit survey is what has been attempted on CC where a number of students report the actual schools they were admitted to and where they matriculated. Despite its clear selection bias, that is a cross-admit survey. You only base the cross-admit battles on head to head results.</p>

<p>The Revealed Preferences Survey, provides HYPOTHETICAL matchups between colleges based on its model without in many cases any ACTUAL matchups. Let’s face it, only 17 schools even have more than 50 matriculants in the database., many far less. How could ever derive any meaningful results from such a limited set. The authors themselves agree that the predictive value of their model is very limited based on the very small data sample. </p>

<p>They use a highly questionnable “Swiss system” ranking for colleges based on sparse data. There is no evidence to support the assertion that such Swiss ranking model designed for ranking chess players is applicable to college choices which are far more complex. </p>

<p>

</p>

<p>As an example, your claim that Cornell beats Duke in cross admits is not based on tallying actual cross-admits between the two schools. There may actually be none in the database. Rather, they will look at finding some common college against which both schools have cross-admits and see who loses more or less against that third college. So, for instance if Cornell does better against Columbia than Duke, it is inferred that Cornell would beat Duke which may be completely false based on actual cross-admits. Cornell may do well against Columbia because it is close geographic proximity or because it has a stronger engineering program than Columbia, factors that would weigh differently in a Cornell, Duke matchup. The transitivity assumption, (A beats B, B beats C and therefore A beats C) may have some validity in sports but is highly dubious with college rankings especially between schools with very different programs. </p>

<p>My point is that constantly referring to the Avery/Hoxby study as some objective measure of cross-admits is highly misleading. The study itself makes no such claims. It is a model, that’s all.</p>

42

<p>I wonder whether Middlebury and Bowdoin have relatively high yields in part because they are SAT optional… could that be a factor somehow?</p>

43

<p>“I wonder whether Middlebury and Bowdoin have relatively high yields in part because they are SAT optional… could that be a factor somehow?” lol, what? Being SAT optional would lead to more application but it would have nothing to do with yield</p>

44

<p>

It should be noted that Middlebury is not test optional – it simply allows 3 SAT subject tests to be substituted for the SAT/ACT.</p>

<p>Yes, I agree that the best test-optional schools could get a slight boost in yield. Admitted applicants with strong transcripts but weak test scores might matriculate at Bowdoin and similar schools in greater numbers because those are the most selective schools they’re likely to get into.</p>

45

<p>cellardweller-
The Revealed Preference Study is based on actual crossadmits when they exist. The model you refer to comes into play when there are no crossadmits, or very few. A has no crossadmits with C but A has plenty of crossadmits with B and B has plenty of crossadmits with C so you make the transitive inference that A wins against C. It is a sound statistical technique and, otherwise, there would be no way to tell if Harvard wins against Podunk.</p>

<p>IBclass06-
College preferences have not changed very much since 2000, despite doing away with ED at some schools. The basis for college preferences is much more complex than that.</p>

46

<p>collegehelp, what the revealed preference survey doesn’t do is tell you which students preferred A over B and did not even apply to B.</p>

<p>Also, the survey works backwards. Chooses the colleges it wants to look at, then sees which high schools send students to those colleges, and then asks those students what they preferred. </p>

<p>It doesn’t ask students from high schools that don’t send kids to the prepicked schools which schools they prefer.</p>

<p>Also, the way the revealed study ranks schools is puzzling. How many students want to go to UCLA but prefer Furman? Yet, the survey ranks all the schools together and you get results like more students prefer Furman than UCLA.</p>

<p>I don’t think so. ;)</p>

<p>Maybe there is one student per year that applies to both Furman and UCLA. Now if that person chooses Furman, it gets a point or whatever and UCLA doesn’t. Even though more students want to go to UCLA than Furman. It’s not even close.</p>

<p>Come on.</p>

47

<p>I am also not a fan of Revealed Preference, overall. It is far from perfect. The most profound dispreference of a school is to not apply at all; However only the results of people who liked schools well enough to apply to them in the first place became data points, subject to extrapolation, in the RP study. Hence anomalous results elevating Notre Dame, BYU, etc. beyond all reason. Many of the [mormons] who applied to BYU indeed may prefer it to most other colleges, but a broader swath of people would not have voted that way, had they cared enough for BYU to even apply to it (so they could then reject it, whereupon this would have become a data point influencing the results).</p>

<p>In this case, though, the RP findings do have some value, generally, in backstopping an evaluation of the likelihood that the 85-15 assertion is plausible. Not to me it isn’t. But in any event, as has been pointed out already, it would be unlikely that the guy had all the data needed to support this claim as being actually valid, since many people don’t return cards, and evidently he didn’t even say it anyway.</p>

48

<p>

</p>

<p>False, again. Every single cross-admit outcome is based on the Revealed Preference model, not on actual data.</p>

49

<p>While yield is an overrated category, I do find it amusing the ends that hawkette goes to “explain away” Duke’s relatively low yield…</p>

<p>

</p>

<p>Stanford competes against HYP (as well as the other Ivies and top privates), has a no. of really strong public universities on the West Coast for which it has to compete against, as well as having a no. of highly regarded LACs close by - and yet, its yield rate is pretty much at the top.</p>

<p>

</p>

<p>C’mon hawk - common sense should tell you that such a claim is dubious at best.</p>

50

<p>I wonder how the yield at hawkette’s alma mater compares with other top schools. Anyone want to bet that I’d never get an answer to this question?</p>

51

<p>

Thanks. Does yield factor into the rankings?</p>

52

<p>I wish I could see more “enrolled” stat ranges, rather than applicant pool data. I’m guessing that when students are admitted to 7 to 10 schools, finances being equal, the student will enroll in the reach school, not one of the lower ranked 5 or 6. Those lower schools have the students higher states in their applicant pool data, but not in the enrolled data. Therefore the enrolled stats would be much lower then the applicant stats.</p>

<p>Don’t be discouraged by applicant profiles or applicant statistics - with yeilds in the 50% range, half of the students - those with those “profile” stats - are going elsewhere, leaving lots of seats for you.</p>

53

<p>Georgia College & State University, the state’s public liberal arts university, has had one of the most dramatic rises over the past dozen years in all of higher education. Its yield rate reached 59% last year, which would tie it at the top of Hawkette’s public university list.</p>

54

<p>Toadstool,</p>

<p>As a parent of two kids on FA who applied to multiple schools, I can tell you that finances are never equal, and that FA was a major determining factor in their school choice.</p>

55

<p>I don’t know why a lot of threads started with something, and after 10 posts, it turned to how good Duke was. </p>

<p>Duke admitted 4065 including 547 in EA in 2009, and estimated to enroll 1716. That means they admitted 4065 - 547 = 3518 in RA, and only 1716 - 547 = 1169 will enroll. This translates into a yield of 100%*1169/3518 = 33%.</p>

56

<p>An example of what a determining factor finances are, can be seen with UF’s high yield. The school is the strongest academically from all the Florida state schools. Most students in Florida who can not afford going out of state to a top private or ivy, (or if they can not attend UM) will choose UF, therefore in that respect the students are self-selective. They apply to UF knowing that they will attend. If it weren’t because of finances or if the other states schools in Florida were up to par academically, the yield would not be nearly as high.</p>

57

<p>

</p>

<p>I think Duke suffers from a major inferiority complex because it competes with all the ivies, MIT and stanford. The fact is that is loses in cross admits to all of them. I have met very few people that have chosen Duke over any of the ivies and they are all from the South. It is certainly a top regional school as far as 'recognized prestige" is concerned.</p>

58

<p>MyOpinion,
The issue of cost is a MAJOR factor in a college’s yield and a potentially major advantage to a school with a very high percentage of IS enrollees. </p>

<p>For example, the cost differential is vast for a U Florida IS student and the vast majority of other college options, public and private. Consider the following Tuition & Fees:</p>

<p>$3,790 U Florida IS student</p>

<p>$38,664 Tulane
$22,342 U Georgia OOS student
$23,401 Clemson OOS student</p>

<p>For a school with 90% IS enrollment, it is easy to understand why the yield is so high. Heck, I’d even expect it to be higher, but I suspect other factors like their 21/1 student/faculty ratio and their relatively high percentage of large classes (20% are over 50 students) dissuade students who might prefer a more intimate educational experience and can afford it. </p>

<p>Cost and the IS/OOC difference can also influence yields at even more prominent public Consider the publics of U Virginia and U Michigan, each of which enrolls about a third of its incoming class from OOS. Here is their IS and OOS yields and their IS and OOS cost of attendance. </p>

<p>63%, $20,409 U Virginia IS (T&F = $9300)
33%, $40,709 U Virginia OOS (T&F = $29,600)</p>

<p>69%, $22,403 U Michigan IS (T&F = $11,111)
28%, $43,693 U Michigan OOS (T&F = $32,401)</p>

<p>I don’t think anyone would dispute that both U Virginia and U Michigan are fine state universities, but their cost of attendance is approximately double for their OOS students and that’s not even taking into account their relative amounts of need-based, financial and merit aid. This cost differential is not chump change to most families and clearly will impact the college selections of some students. </p>

<p>As for your Duke comments, I don’t know if you’ve ever talked to many students who attend/graduated from Duke. I haven’t seen much of an inferiority complex, but maybe we’re talking to different folks. Anyway, if you’ve interacted with them, you will learn that, while HYPS will clearly dominate them in cross admits, you will find many Dookies who chose Duke over other selective universities, including the non-HYP Ivies. BTW, 64% of Duke’s first-year class came from outside of the Southeast/Texas.</p>

59

<p>^ Yes, Duke is a fine private university.</p>

60

<p>Why is princeton’s yeild only 59%</p>