It is not only double counting but also contrary to the way all the other schools report their RD numbers; it’s comparing apples to oranges. Sounds like a lot of funny math coming out of UChi to me.
It’s Chicago. What else do you expect?
@spayurpets You asked me a question. I answered it. Chicago did not officially release ED vs RD rate. I was just trying to help. You are free to interpret the answer whichever way you like. Technically anything you post on ED vs RD rates for Chicago is speculation anyway, since there is no official word. But that is your choice. I really don’t care, but it is interesting that you will accept the ED rate but not the RD rate, when both are coming from the same source, namely from admit weekend
@veryluckyparent, It’s either 4%RD/14%EA or 6%RD/10%EA based on the numbers Chicago published, there’s not any math that gives you 4%RD/10%EA. What would you have me do? If you want to compare apples to apples, Chicago’s 4% RD rate means that Harvard’s RD rate goes down to 2.9%.
Even based on my rough 50-50 split of EA/RD acceptances, that already makes Chicago an outlier. Only one school in the country (GA Tech) reports more EA acceptances than RD acceptances. I think this is the real secret sauce to Chicago’s numbers.
@spayurpets I don’t really care. Its your table, do whatever you like, but that does bring me to one question. Lets keep the math simple here.
Total Apps 200. 100 apps in EA/ED round. 10 accepted. 10 rejected. 80 deferred. Lets say 4 out of those deferred were accepted in RD round, with total admits in all rounds as 30 (So as per your calculations, EA/ED rate is 10% (10/100), your RA rate would be what? 20/100 i.e 20%(because you don’t like to include the 80 deferred apps as you believe that is double counting). Since you don’t want to include those 80 applications in the RD pool. Where would you put those 4 who got accepted? Because that number is in the 20 you used to calculate the RD rate. They were not accepted in the EA/ED pool, so they should not belong there. You don’t count their apps in the RD pool so the admits should not go there, otherwise your numerator is wrong. So you are overestimating the RD rate. In this example the real RD rate is 16% not 20% as you calculate.
What happens to those students in your calculations? If you want to do it your way, for it to be accurate you need to back out the percentage of deferred students who got accepted from the admits in the RD round, but you will never be able to do that, since most schools will not release that number.
Remember this number is not a small number. For a school that defers lets say 75% (a lot of the elite schools) of its application and gets around 6,000 applications as an example, that could be 450 to 500 admits (assuming lets say a 15% admit rate). How are you calculating those. In which bucket are those acceptances?
Either way you skew the data. Your calculations are not any more accurate. So pick your poison, if you want to report ED/EA rate and RD rate separately, you must know for sure that there were no deferred applications or must know how many got deferred and how many of the deferred students got accepted, Otherwise its all just funny math.
I agree with the @VeryLuckyParent analysis. The only way to get an accurate RD acceptance rate is to include all the students deferred from the ED round as they are part of the pool being considered in the RD round. If the data to do that isn’t available, an RD calculation isn’t possible.
I know some may disagree here, but given that schools generally don’t release the number of applications deferred and the percentage of those deferred who get accepted, even the EA/ED rates are inaccurate. Under those circumstances, Chicago is actually doing the most reasonable thing here. They are only releasing the total admit rate, the total number of applications and the number of applications that applied EA and not the EA admit rate or the RD admit rate. Saying anything more would be inaccurate if they did not release any more data officially. For schools that accept 10 to 15% of their deferred pool and defer a large pool of their applicants( and those are a lot of universities, including a lot of the Ivies) the EA/ED rate is actually artificially low and technically meaningless. To be more accurate EA/ED rate should be increased to reflect the folks who got in after being deferred.
Everybody is gaming the system. :))
I have seen a lot of schools release the specific ED pool size and number or percent accepted, which seems like a true stat for the ED acceptance rate. It’s irrelevant to that ED/EA calculation how many who were deferred are later accepted RD, because even if accepted RD they were not accepted ED. And most schools seem to list their blended total acceptance rate inclusive of both ED/EA and RD. What seems less common is schools releasing their specific RD acceptance rate. If you know a school defers 100% of applicants not accepted ED you could figure out the RD rate with simple math. If you don’t know whether they reject rather than defer some in the ED pool and they don’t specifically list the total RD pool size, you don’t have all the information you need for an RD calculation.
@VeryLuckyParent I don’t disagree with your logic, it’s just not the way every other school did it, so you can’t report Chicago differently than the others. Also, in your method, you are counting certain applications twice, one as an EA applicant and then again as an deferred RD applicant. Yes, it tells the RD applicant a better sense of what the “odds” are of applying RD because you are competing against all the deferred applicants as well, but it screws up all the total application numbers.
Republishing the list with UChicago ED and RD removed:
Harvard RD 1119 out of 32868 (3.4%)
Stanford RD 1318 out of 36175 (3.6%)
Yale RD 1177 out of 26793 (4.4%)
Princeton RD 1109 out of 25074 (4.4%) (1237 waitlisted=4.9%)(rej=90.6%)
Columbia ED/RD 2193 out of 36292 (6.0%)
Penn RD 2326 out of 33156 (7.0%)
MIT RD 829 out of 11253 (7.4%) (437 waitlisted)
Brown RD 2250 out of 29360 (7.7%)(~133 deferred accepted=7%)(~1000 waitlisted=3.4%)
University of Chicago EA/RD 2482 out of 31,411 (7.9%)(yield=66%)
Pomona RD ~566 out of 7190 (~7.9%)
Northwestern RD 2690 out of 32077 (8.4%)
MIT EA 656 out of 7767 (8.4%) (4776 deferred=61.5%) (2175 rejected=28%)
Duke RD 2501 out of 28600 (8.7%) (49 deferred accepted=2.9%)
Vanderbilt RD 2526 out of 28700 (8.8%)
Dartmouth RD 1682 out of 18748 (9.0%)
Stanford REA 745 out of 7822 (9.5%)
Johns Hopkins RD 2539 out of 25188 (10.1%)
Harvey Mudd RD 421 out of 3716 (11.3%)
Bowdoin RD 687 out of 5918 (11.6%)
Tufts RD ~2168 out of 18152 (~11.9%)
Amherst College RD 969 out of 7943 (12.2%)
Cornell RD 4939 out of 40084 (12.3%) (4572 waitlisted=11.4%)(rej=76.3%)
Swarthmore College ED/RD 963 out of 7,717 (12.5%)
Georgetown EA 892 out of 7027 (12.7%) (remainder deferred=87%)
UC Berkeley (OOS) 2734 out of 21213 (12.9%)
Pitzer College ED/RD (12.9%)
Notre Dame RD 1955 out of 14,178 (13.8%)
Carnegie Mellon University ED/RD ~5270 out of 37,247 (14.1%) (enrollment 1503=28.5% Yield)
Middlebury RD 1042 out of 7866 (14.2%)
Boston University ED2 ~245 out of 1721 (~14.2%)
Harvard SCEA 918 out of 6173 (14.9%) (4673 def=75.7%) (464 rej=7.5%)
Williams College RD 960 out of 6397 (15.0%)
Barnard College ED/RD ~1131 out of 7071 (~16%)
Georgetown RD 3276 out of 20002 (16.4%)
USC RD 8920 out of 54100 (16.5%)
Harvey Mudd ED1/ED2 ~77 out of 464 (16.6%)
Yale SCEA 795 out of 4662 (17.1%) (53% def) (29% rej)
Colby College ED/RD ~1720 out of 9822 (17.5%)
Grinnell College ED/RD ~1326 out of 7368 (~18%)
UC Berkeley (IS) 8363 out of 45,773 (18.3%)
Princeton SCEA 785 out of 4229 (18.6%)
Middlebury College ED2 60 out of 318 (18.9%) (40 def=12.6%) (218 rej=68.6%)
Pomona ED1/ED2 ~177 out of 914 (19.4%)
Georgia Tech RD ~3206 out of 15,659 (~20.5%)
Brown ED 669 out of 3030 (22.1%) (1905 def=62.9%) (456 rej=15.0%)
Scripps RD ~632 out of 2743 (23%)
Penn ED 1335 out of 5762 (23.2%)
Duke ED 813 out of 3455 (23.5%) (1663 def=19.2%)
Vanderbilt ED1/ED2 ~800 out of ~3390 (23.6%)
UVA EA (OOS) 2955 out of 12308 (24.0%) (3005 def=24.4%) (6348 rej=51.6%)
Hamilton College ED/RD 1317 out of 5434 (24.2%)
Lehigh University ED/RD 3420 out of 13,408 (25.5%)
Skidmore College RD ~2200 out of 8608 (~25.6%)
Dartmouth ED 494 out of 1927 (25.6%)
Kenyon College ED/RD 1688 out of ~6400 (~26.4%)
Vassar College ED1/ED2/RD 1943 out of 7306 (26.6%)
UVA RD (IS/OOS) 4166 out of 15658 (26.6%)
Cornell ED 1338 out of 4882 (27.4%) (1153 def=23.6%) (2391 rej=49.0%)
Pitzer ED1/ED2 ~118 out of 423 (27.8%)
Wellesley ED/RD ~1368 out of 4888 (~28%)
Lafayette ED/RD 2291 out of 8121 (28.2%)
Georgia Tech EA 4424 out of 14861 (29.8%)
Bowdoin ED2 ~77 out of 256 (~30.1%)
Notre Dame EA 1610 out of 5321 (30.3%) (818 def=15.4%) (2893 rej=54.4%)
Johns Hopkins ED 584 out of 1929 (30.3%)
Boston University ED1/ED2 ~1050 out of 3421 (~30.7%)
Boston College EA ~2700 out of 8600 (~31.4%)
Tufts ED ~663 out of 2070 (~32%)
Bowdoin College ED1 207 out of 614 (33.7%)
UNC EA 6948 out of 19842 (35.0%)
Northwestern ED 1061 out of 3022 (35.1%)
College of William & Mary ED/RD 5095 out of 14380 (35.4%)
Amherst College ED 180 out of 454 (39.6%)
Middlebury College ED1/ED2 398 out of 954 (41.7%)
George Washington RD 10101 out of 24168 (41.8%)
Williams College ED 246 out of 585 (42.1%)
University of Florida RD 13,624 out of 32,000+ (~42.5%)
Dickinson ED1/ED2/EA/RD 2636 out of 6171 (42.7%)
Fordham ED/EA/RD ~19,650 out of 44,697 (~44%)
Occidental College ED/RD ~2884 out of 6409 (~45%)
Davidson College ED 207 out of 458 (45.2%)
Boston University ED1 ~805 out of 1700 (~47.4%)
Scripps ED 113 out of 235 (47.9%)
UVA EA (In-State) 2237 out of 4460 (50.2%) (1060 def=23.8%) (1163 rej=26.1%)
University of Georgia ED 7500+ out of 14516 (51%+)
Middlebury College ED1 338 out of 636 (53.1%) (74 def=11.6%) (224 rej=35.2%)
George Washington ED 841 out of 1373 (61.3%)
Skidmore ED1/ED2 ~337 out of 542 (~62.1%)
University of Maine RD (OOS) 7803 out of 10,062 (77.5%)
University of Maine RD (In-state) 3600 out of 4134 (87.1%)
This may be obvious to others, but it occurs to me that it would be possible for a school to improve its yield numbers by “over-accepting” from the EA pool (i.e., admitting a relatively large number of students with a high likelihood of accepting an offer, as evidenced by their decision to apply early) and deferring all other applicants except those who are clearly unqualified. They could then “under-accept” in the RD round by taking slightly fewer students than they need and filling out the class by creating a large waitlist, seeing who elects to stay on it and demonstrates continuing interest, and then managing that list aggressively.
Similarly, it could make sense to favor students in the RD round who applied early and were deferred (and who therefore presumably are relatively more likely to accept an offer than RD applicants who hadn’t demonstrated the same interest). As was noted above, though, because schools don’t typically disclose how many are admitted after being deferred, it’s impossible to tell from the outside if this is happening.
The elite schools seem to avoid filling more than half of the class from the EA pool, and I guess there are at least two reasons for it: (i) it’s generally acknowledged that the EA pool contains a disproportionately high number of higher-SES kids who know how to play the college admissions game, and a correspondingly low number of URM/first-gen/less-advantaged applicants; and (ii) the situation ultimately reaches an equilibrium because if there’s too much of a perceived advantage to applying early somewhere, more kids will do it and the early admit rate will fall. That said, if a school defers a lot of applicants and then disproportionately admits them RD, that’s a way of favoring EA applicants without seeming to do so.
This too may be obvious, or maybe my math is wrong… In the chase for yield, with schools now admitting half the class in ED, those schools are getting a 50% floor. The interesting thing is then that the process of a school like Penn, admitting 1332 through ED which gives near 100% yield, and filling approx another 1113 places later (for an expected enrollment of 2445) from the 2329 later admits (for a total admit of 3661 in both rounds) implies a RD yield of below 50%. The overall yield rate of 69%, which is the headline number, seen in this context is not really as impressive.
http://www.thedp.com/article/2016/03/regular-decision-release-class-of-2020
http://www.thedp.com/article/2016/05/penn-yield-rate-record.
The yields of schools with EA, SCEA and REA are not comparable, and arguably far superior in terms of real yield. Chicago and Princeton also have yields of 69% but do not bind their EA admits. Furthermore, Harvard and Stanford with 80%+ yields and SCEA and REA respectively have a true yield rate (effectively a RD yield rate) of 80%+
I am also left wondering if the schools do care that much about yield from a competitive standpoint, why they would not all switch to ED from EA (for Notre Dame, Georgetown, MIT, BC, etc). Or perhaps they don’t care that much beyond a certain point? I am not a fan of gaming the system, but understandably these numbers do serve a marketing purpose.
I assume yield management is a complicated business that has a lot of repercussions besides just bragging rights and gaming the rankings. I know from my D’s experience that getting yield wrong at a small LAC has definite negative repercussions. The Pomona Class of 2018 was over-enrolled due to an unexpected jump in yield. They were targeting a class of 400 and got 450 acceptances. My D is Class of 2019 and the large size of the rising junior class has impacted her class and will continue to do so until D is a senior and finally “free” of the oversize class ahead of her. Having an over-enrolled class negatively impacts class selection, class size and housing, especially for the class that’s right behind the oversize class. As freshmen they were protected in housing and course selection but as rising sophomores they are in the back of the line. It makes sense for LAC’s like Pomona to fill a large portion of the class with ED admits, be conservative with RD admits and utilize a WL.
@Corinthian yes, how true! When a class overshoots by that much, does it also force future classes to be slightly smaller for the next year or two to rebalance and therefore disadvantage applicants in those years?
I would think SCEA would effectively be close to ED in terms of yield as these students are forgoing all other comparable early apps.
The high yield that I find most impressive is the straight EA as those applicants would/could apply to all other EA schools, as well as an ED school. I’ve wondered if schools like UChicago with EA only, purposefully defer students that they expect may be applying ED elsewhere (UPenn legacy kid for example), then potentially accept in the RD round once it’s clear the kid did not get in ED elsewhere. Also whether a tippy top kid that appears in the RD round would be assumed to have unsuccessfully SCEA/ED’ed elsewhere and thus get a bit of an askance glance.
@ihs76, some very insightful thoughts! SCEA is only offered, to my knowledge, at HYPSM level universities. Therefore, I would expect the only additional applications made by accepted students would be to shop for better FA or to collect scalps, and, practically speaking, the yield on those accepted students would be very similar to the ED yield at other elite colleges. In other words, those universities have the advantage of appearing benevolent at little to no cost to their yield.
My guess is, assuming UChicago is able to identify and read signals that suggest an EA applicant is not serious about matriculating at UChicago, they would defer them. My counter is that given how competitive admission to UChicago has become (7.9% overall) and how important wise usage of the SCEA/EA/ED option has become, it is only wise to use the SCEA/EA/ED option if UChicago is the college the student is most interested in attending. That said, once an applicant has been accepted ED to UChicago, they have a free option to try to upgrade to HYPSM.
Your last question is the most interesting, and the one where it is likely the universities have become duplicitous in their efforts to increase relative standing. The consistent umbrella message propagated by colleges has been that students should only use the SCEA/EA/ED option if they are certain of the college they wish to attend, and otherwise apply RD. However, in their on-campus briefings, many of the elite (outside of HYPSM), often very explicitly, state that there is a substantial advantage to applying early (just 4 years ago the differential at Harvard was roughly 21% SCEA admit rate vs. 4% RD with roughly 58% of the class taken SCEA - that year was a slight aberration but interesting none the less). These same universities, probably, do wonder if the RD applicants are rejects from the SCEA/EA/ED round, just collecting scalps or shopping for better financial aid. That said, all RD round applicants are in the same boat so it probably doesn’t disadvantage any of them.
@DeepBlue86, there is no question there as been a substantial amount of gaming done to increase the applicant pools (through heavy marketing - often to attract applicants who are marginal at best) and increase yield (by increasing SCEA/EA/ED acceptance rates and % of class size, and, historically, in the cases of Tufts and Princeton (there is an interesting academic paper supporting this), among many many others, rejecting the most qualified applicants who were likely to go elsewhere) to reduce the admit rate which then makes the university appear more selective and improves the ranking. Most elite universities try to limit the SCEA/EA/ED portion of the class to just under 50%, while Harvard, Yale and Princeton are all currently at 55-58%. There is a tipping point where having the SCEA/EA/ED admits represent too large a portion of the matriculated students will have the number of RD applicants decline, which will hurt the admit rate. In other words, if potential RD applicants believe their chances are too low, because the percentage of students taken in the SCEA/EA/ED round is too high, they may not bother to apply. Part of the reason the tier of elite universities just below HYP may be keeping their SCEA/EA/ED portion of the class below that of HYP may also be to have the opportunity to accept the best of the HYPSM rejects. And, an important component of why the HYP SCEA/EA/ED portion of the class is so high is that experience tells them these will be the best applicants, and there is little downside for those schools as they will still get the overall applicant numbers. Many years ago the admissions director at one of the Harvard graduate schools told me that of the four dates they listed as accepting applications, by far and away the most qualified applicant pool was the first. I imagine it is the same for undergraduates.
Your suggestion that accepting deferred SAEA/EA/ED applicants should further improve yield makes sense, but I don’t see it happening in huge numbers. I imagine having been deferred helps an applicant all else being equal, but too often there are too many more desirable candidates.
For my kid, Chicago was effectively SCEA because the next two preferences (Harvard and Princeton) on DC’s list were SCEA, while the most-liked ED school (JHU) came in at #5. Personally, I think that if you’re talking about HYPSM, you might as well switch to CHYMPS.*
MIT and U of C look very similar from an Admissions standpoint. Both are EA, both are somewhat niche schools that appeal to kids who may well have HYPS alternatives. Each has a significantly smaller than HYPS gap between early and regular admissions rates. And both have sufficient courage of their convictions/understanding of their target audience to know that they’ll prove choice-worthy to a critical mass of highly-qualified students regardless of what other options end up in the mix.
All that said, I was surprised to note (on CC) how many Penn/Columbia ED applicants were part of the UofC EA pool. Though lots seemed to be deferred from both schools, so (to me) it looked less like Chicago deciding to wait and see and more like kids at the margin looking for early app advantages from more than one school.
*Which not only has the advantage of being easy to pronounce, but would also give Cal Tech and Columbia (as well Chicago) a chance to stake their claims, LOL!
@am61517 I agree with pretty much all of what you’re saying. I have a minor quibble with one point, though, which is the statement that EA admits account for 55-58% of the class at HYP. Yale, for example, admitted 795 students SCEA this year, and I understand has a targeted class of 1,360. If every one of those SCEA admits accepted their offer, they would account for ~58% of the class, but surely that won’t happen, because students may get better aid elsewhere or change their minds for various reasons. Put another way, it would take 94% of the SCEA admits to accept their offers to account for 55% of the class. That sounds like an ED yield, rather than an EA one (to me, anyway). If we assume that 85% accept, that would account for almost exactly 50%, which sounds more plausible to me.
Related to something else you said, I think there are certain SCEA/REA applicants a school feels almost obligated to accept for fear of losing them. Although they’re not obligated to matriculate, the school can do its best to make them feel welcome, starting by accepting them early, and make it more likely that they won’t bother applying anywhere else. In contrast, certain students well within range but not off the charts can safely be deferred.
@DeepBlue86, the rule of thumb on ED is that at an elite university can expect 95%+ yield. I would expect SCEA at Yale would generate similar results given it is arguably as good an admit as a student can achieve, it has the 2nd largest endowment per capita (after Princeton) to apply to generous FA packages and will match the FA from similar universities, and the student is, presumably, already strongly predisposed to want to attend. Using 95% puts the Yale SCEA % of class at 55.5%.
On a separate note, it is interesting how the perceptions on CC of UChicago and Columbia have changed as their extensive marketing and gaming of the admissions process has brought down their admit rates. Even within this relatively educated community the admit rates are taken at face value as indicators of true selectivity (the cross admit data between UChicago and Columbia and HYPS on Parchment.com indicates they are actually far less selective than HYPS) and prestige, with little appreciation of how they have been derived. Among elite universities, none have done more marketing (extensive mail & email blasts to even margins candidates to encourage applications) and none have been less transparent in the provision of admissions related data - which makes you wonder what they are hiding.
Any word on Northeastern University in Boston EA?