Although Yale may have taken "more" cross admits than last year, it appears that Harvard's fraction of the total cross admit group is as large or larger than ever, which is one reason why the Yale yield rate will decline by a greater percentage.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9756989)

NB: "The Yale dean of admissions, Jeffrey Brenzel, said there would be about 45 wait-list offers this week and probably another round later this month."

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757191)

Based on Yale's methodology, Harvard's yield actually comes down to about 74%.

EDIT: poor mental math, the "74%" is actually 76.2%. Point still holds.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757482)

If Yale ends up with, say 1,892 initial + 65 WL admits for a class of 1320, the yield rate will be 67.5%

If Harvard ends up with 1,948 initial + 175 WL admits for a class of 1660, the yield rate will be: 78.2%

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757499)

If you add 45 WL admits with a 100% yield rate

(1286+45)/(1892-25+45) = 1331/1912 = 69.6

If Yale used Harvard's formula to determine its yield it would've come out to that figure.

Harvard, by refusing to release the # of admitted students, seems to be indicating that the WL admits are already included in the yield projection - better yield for better press, as was done with the admit rate. As of now, if H must go to the WL for 175 students, it's a fair assumption that they have about 1660-175= 1485 students at this point. Out of the 1948 initial admits, this would bring the current admit rate to 1485/1948 = 76.2%, the methodological equivalent to Y's 68.9%.

Harvard does not have 1948*.78 students matriculating at this point, because adding 175 to that would mean having a frosh class of about 1695, clearly overshooting.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757623)

There's no evidence here that H isn't doing the same with deferred admits, but in any case the effect of this (+/- 25 is minimal compared to the effect of arguably disingenuously including 175 100% yield rate admits in your original yield rate projections.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757677)

I would hope not. Harvard was really quite smart about this. By admitting based on unrealistic yield projections (85%!!!), it was able to claim the lowest admissions rate in recent history (a record which will be hard to beat), and now that H has to go to the WL en masse it's using this as a crutch to keep its yield rate en par with last year's.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757726)

1. The number of admits - including waitlist admits;

2. The number of matriculants - including matriculants from the waitlist.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757760)

Since WL admits have a very ED, 100% yield nature to them, schools that make greater use of the WL will have a better yield to report. In this case, H is using the WL more than any comparable school and benefiting the most from it by bumping its reported yield rate by 2%. Taking this to the extreme, if one school admitted the entirety of its class through the WL they could have a near 100% yield!!

The numbers before turning to the WL are better proxies for relative preference. In this case, Y's 68.9 compares to H's 76.2. H does significantly better, but the gap is smaller than last year's.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757928)

Moreover, they admitted enough people from the high-yield early pool to fill 67% of the class, if all enrolled!

And ... get this: they probably took another 250 from the high-yield early pool as DEFERRED ADMITS if the numbers from the past 3 years are any guide! Talk about yield manipulation!

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757976)

Also, it does seem to make sense to count the admit who defer as the matriculants (they did accept the offer, after all). What matters, of course, is that they not be double-counted (i.e., counting the admits who defer for both the current year and the previous year as matriculants clearly would be wrong).

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757694)

"If you add 45 WL admits with a 100% yield rate

(1286+45)/(1892-25+45) = 1331/1912 = 69.6 "

I have added 45 to both the numerator and the denominator. Yale is predicting taking 45 from the WL. Where are you getting 65 from?

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757743)

Have you considered that it is JUST POSSIBLE that a few of those potential 175 Harvard WL admits will come from the pool that has made deposits to YPS?

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757789)

My main point though is that before any WL action takes place, H's yield rate is actually closer to 76.2%,compared to Y's 68.9.

If one of the two schools' yield rate went down this year relative to last it's H more so than Y, although both seem to have been hurt by the greater incidence of cross-admits.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757821)

Thus, admitting 65 more people to reach the target class size at Yale will REDUCE the headlined yield rate.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757857)

look's like H's yield dropped this year after all.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757871)

There were 1,911 admits, and 1,320 matriculants.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757917)

I think the most legitimate numbers are H 76.2 and Y 68.9. Both schools were hurt by the larger cross-admit pools a little, arguably H losing out a bit more than Y.

If both schools use H's formula, the #'s are H 78 and Y 69.6. The H formula helps H disproportionately due to their higher reliance on the WL, but these 100% yield kids aren't the best proxy for relative desirability, or whatever we think is being measured here.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757757)

Kudos to Princeton for having more courage than either Yale or Stanford on this point, and being willing to go head to head with other schools in the open market.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757778)

My argument is that if Y and H had used more similar methodologies to report their yield rates, the difference in the school's rates would be smaller.

Whether or not EA (not ED) actually boosts Y's yield number is highly speculative - EA is not binding, kids are smart and can weigh their options - and also not the argument I'm interested in now.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757807)

That's why Levin was afraid to cast aside the early admissions crutch.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757823)

Y's SCEA pool in the past was comprised mostly of individuals whose true first choice was Y. This is less clear today because several of its rivals (H, P, UVA) did away with EA. Hence, the EA pool this year must have had a greater prevalence of students whose first choice was *not* Y, but rather P, H, UVA, etc. I'm almost positive the SCEA yield rate will show a sharp decline relative to previous years (obvious enough), as now it only indicates a hypothetical preference over say Stanford, but not necessarily Princeton or Harvard.

Looking at it from the perspective of a student whose dream school is P, she/he might apply to Yale SCEA and Princeton RD and get into both. The student would still go to P, same if the student had applied to both P and Y through RD.

Because what matters are the *options* students have at the end of the admissions cycle, you have to claim that the fact that the order in which students are admitted has a significant effect on formulating preferences - I dare to suggest kids of this caliber are smart enough to see beyond such frivolity.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757872)

Yale (and Stanford) retained their early admissions programs in order - hopefully - to gain an edge over Harvard and Princeton with cross admits by having a 3-month head start in recruiting them. Nothing hard to understand about this.

Yale positively drooled when its early apps rose 36% thanks to "what the hell" early applicants from those otherwise inclined to Harvard or Princeton.

For this same tactical reason, schools send out "likely letters" .

The idea is that you gain a clear recruiting edge by being the first school to tell a kid "we love you".

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757911)

If anything, I think that a lot of kids will be put off by a school that's begging them to come (You hear this a lot from H-Y cross admits go to H, if they feel the H bashing is too much). Surely some will be persuaded, but I doubt there's a definitive effect in the direction you claim.

A student's choice will come down to a lot of factors before "order of admission" is even considered. If you run a survey I doubt many would claim they chose one U over another because they were admitted to one of them first. Maybe there's an exposure effect, but this can work for or against the school (e.g. scandals), and kids at this level are generally pretty well informed about all their top choices anyway.

"I wouldn't want to belong to any club that would have me as a member." - Groucho Marx

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9757962)

Many studies and stats demonstrate this phenomenon.

Likewise, this explains the current trend to send hundreds and hundreds of "likely letters" or "early write" letters: the idea is to steal a march on the competition by getting to common admits first, and, hopefully, get their DEPOSIT CHECK! (Harvard is unique in not requiring a deposit to hold your place.)

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9758007)

Yale does not require a deposit check.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9763100)

A fringe benefit is that Yale identifies 1,500 of these early applicants as predisposed "buyers" and admits them at a very high rate after deferral. (The last several years, upwards of 250 nominal RD admits to Yale were deferred SCEA admits.)

If there were another 250 admits from the record deferred SCEA group again this year, added to the initial 885 SCEA admits, it would mean than an astounding 1,135 were admitted from the high-yield early pool for the Yale class of 1,320! Now THAT will help your yield rate, won't it!

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9764056)

Based on your logic, Princeton's yield this year should be whatever it gets through RD, or about 51%. Instead, it'll be closer to 60%, probably because a lot of those who would've applied ED applied RD instead, changing the composition of the RD pool. The Yale and Stanford EA pools changed in different ways as well, probably bringing the yield for those groups down significantly.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9763115)

Both, apparently, were afraid to take the overall yield rate hit that Princeton may take by giving up the early admissions crutch.

The yield rate hit will be far more modest for Harvard because its RD yield rate is far higher and it traditionally takes 3 of 4 common admits from its YPSM competitors.

To the extent that the Yale and/or Stanford overall yield rates decline this year, it will primarily be a function of additional cross admit losses to Harvard or Princeton, due to enlargement of the overlap pools.

The proven benefit of the early admissions tactic - binding or not - is to sign people up months before they hear from other schools, benefitting from the "bird in the hand" psychology.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9763160)

Anyway, I think it's important to differentiate between ED and EA policies. While there's a strong case to be made that ED hurts applicants by forcing them to commit to a school, there's no such consensus for EA, or SCEA for that matter.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9771935)

Any "benefit" flowing to applicants is purely coincidental.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9772426)

ED in particular is really bad news for students, because the 100% commitment means you can't reach for your reach schools or compare financial aid packages.

Comparatively, EA is much more benign. The way I see it, EA is a way for students to signal their top choice and universities may choose to reward those students; especially if their accomplishments are comparable to those who were rejected - who doesn't want to have students who *want* to go to your school? Part of the reason Y and S students seem to be so happy perhaps? It's not like Stanford or Yale are getting sub par candidates as a result anyway (which Princeton arguably was). Yale's average SAT > Harvard's after all.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9773113)

its an art that has been lost on this board for years.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9773152)

Is it because SCEA schools which don't get such "signals" (Yale? Stanford?) may reject or waitlist you if they think your "signals" may have been sent in another direction? (ie, Harvard? MIT? Princeton?)

Of Course.

That's what its all about.

Yield enhancement.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9773374)

But the value of the "signals" is much diminished when some of the schools decide not to participate in the game. This round, applying to either Yale or Stanford SCEA was merely an indication of preference for one over the other, but not necessarily over H or P (or UVA, etc. etc.).

A lot of P or H hopefuls jumped into the S and Y SCEA pools.

Though, here's where it gets interesting I think:

Y and S hopefuls who get in SCEA are probably? less likely to apply to P and H than the P and H hopefuls who also get into either S or Y SCEA. Therefore, the resulting cross-admit sample pool in the regular round will have an HP bias relative to the gen pop (this was pointed out in another thread recently). On the other hand, it's possible that most YS hopefuls and SCEA admits apply to HP anyway, so that this particular effect doesn't exist.

I'm having a hard time reconciling the fact that Princeton got hit so hard by the policy change whereas Harvard didn't (one way would be to acknowledge that H's yield actually did take a hit, another that ED was a much more significant yield crutch than EA)

In the end though, all the yield-maximizing strategies can only do so much, and I don't think admissions committees obsess over them either. If all the top tier colleges moved to RD you'd still see the same ranking, H > Y=S > P, but if everyone moves to SCEA you can arguably make the process more efficient by matching top-performing students with their first choice, and then have an open market competition for rejects/creme of the crop candidates.

Well, that's 10-15 minutes of my life that ain't never coming back.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9773641)

However I tend to think the "process" works best when schools avoid "enrollment management" maneuvers and applicants aren't forced to rely on "tactical applications" to maximize their odds of admission.

Let the marketplace sort things out.

http://www.educause.edu/ir/library/pdf/ffp9901.pdf

http://www.educause.edu/ir/library/pdf/ffp0001s.pdf

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9774434)

Before reading that however, my impression remains that:

"EA is a generally harmless policy. It allows top students to signal their top choice and gain a marginal advantage in the admissions process as a result. EA makes it significantly likelier that students will attend their first choice, rather than whoever chose them."

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9778667)

Moreover, early programs lead to tactical applications, where student frequently don't apply to their "first choice" but rather to a lesser school where they calculate their odds of admission are higher.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9779953)

interesting.

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9796478)

Mistaken?

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9761663)

The Yale Herald - Jan 19, 2007 - "The politics and policies behind early admission"

Quote:

"In 2003, after announcing Yale’s decision to abandon its binding Early Decision program in favor of non-binding early action, Yale President Richard Levin, GRD ’74, baldly stated at the annual meeting of the National Association of Independent Schools that early admissions programs “do not serve the interests of high school students.” Three years later, explaining Yale’s decision to retain its early action program in the wake of Harvard and Princeton’s elimination of their own, Levin seemed to reverse his earlier comments. Describing non-binding, single-choice early action as “popular with high school students and their counselors,” Levin said the program “solves the major problems inherent in the system we had before.” This disconnect between the University’s views only four years ago and Yale’s policy for tomorrow has prompted questions both from within and outside the Yale community regarding the right direction for college admissions in the twenty-first century."

Quote:

"Richard Ludlow, SM ’07, offered another interpretation. Since freshman year, Ludlow, an Economics major, has been developing an online database of student profiles at universities across the country designed to help prospective applicants find students similar to themselves in terms of grades, test scores, and demographics. Ludlow hopes the database will allow students to inform themselves of their chances for admission at school nationwide. “Keeping early action gives Yale a competitive advantage over Harvard and Princeton because basically, once a school says yes to you, you start to fall in love with it.” Ludlow said. With an early option no longer available at Harvard and Princeton, he said, some students for whom those schools are a top choice will now apply early to Yale, be admitted early, and enroll contrary to their initial preferences."

Quote:

"Yale, of course, has its money on Ludlow. Should the new policies at Harvard and Princeton encourage the strongest students from across the nation to send their first application to Yale, the University may very well return to the coveted spot atop the US News and World Report college rankings, an honor it hasn’t held since 1996."

(http://www.autoadmit.com/thread.php?thread_id=811062&forum_id=1#9764500)