[OT] Sarah Hughes and Harvard?

[jeh25]

Well, foreit or not, I was rather pissed in HS when I applied ED and got deferred and someone else that applied ED got accepted, only to bail on Cornell and go to Emory instead.  Add to that the 3 or so people that got in to Harvard and kept their Cornell app's active even after sending checks to Harvard "just to see if they'd get it" and I got shafted by people taking up admissions slots from my HS that never intended to go to Cornell.

[Adam \'01]

The process to get into Cornell ought to be the same as for hockey season tickets.  First come, first serve, and you need to camp out several days for it.

If you want to be an:

AAPer--You'll only be seen on line really really late at night.
Aggie--You'll show up way early to tend to the land around the field house.
Artsie--You'll start filling out job applications while waiting on line...4 years early.
Engineer--You'll come up with a way to manipulate the ticket computer, thus never actually having to show your face in person.
Hotelie--You'll want to wait in line,  but instead find a nice 5 Star resort nearby.
HumEcer--You'll bake cookies and bribe your way to the front of the line.
ILRie--You'll sign a contract with the guy next to you to stand in line for you, so that you can go read somewhere quiet.

Extras:

Frat Boy--You'll just cut.  And then have your ass firmly booted back to Lawwwng Island.
Sorority Girl--You'll "bake cookies" and bribe your way to the front of the line.



::Plants tongue firmly in cheek::

[Erica]

They aren't going to do anything. What can they do? It's more of like a word type of contract. I applied early decision, however, I had to wait for my aid package. Since I had also applied to UMich on a rolling admissions type of basis, I knew I had already been accepted to that school. But I was still planning on going to Cornell had I been able to pay.  I also still went ahead and applied to three other schools. You don't really have to withdraw your applications. What are you going to do if you find out your aid package sucks? (like I did.) People might say that I shouldn't have applied to the other schools, but if Cornell really wanted me to attend, knowing it was my first choice, they should give me enough money to go. (I ended up with only loans, lots o' them.

[CUlater \'89]

I did my homework on this one last night, Al, albeit through anecdotal evidence, since one of my parents has been a high school math teacher, program coordinator, coach of the NYC math team and annual writer of numerous letters of recommendation for many, many years. And from what I've been told, although you may believe that the percentage of candidates extended offers may be higher in December than in April (and that guy from Yale may have said something to that effect), generally speaking the quality of candidate to whom the December offer is made is higher than the quality of the April candidate to whom an offer is made.

[Al DeFlorio]

All homework is welcome, CUlater.  This is excerpted from Jim Fallowes's article in The Atlantic:

"The real question about the ED skew is whether the prospects for any given student differ depending on when he or she applies. Last fall Christopher Avery, of Harvard's Kennedy School of Government, and several colleagues produced smoking-gun evidence that they do. The authors analyzed five years' worth of admissions records from fourteen selective colleges, involving a total of 500,000 applications, and interviewed 400 college students, sixty high school seniors, and thirty-five counselors. They found that at the ED schools an early application was worth as much in the competition for admission as scoring 100 extra points on the SAT. For instance, a student with a combined SAT score of 1400 to 1490 (out of 1600) who applied early was as likely to be accepted as a regular-admission student scoring 1500 to 1600. An early student scoring 1200 to 1290 was more likely to be accepted than a regular student scoring 1300 to 1390.

"The equivalent of a 100-point increase in SAT scores makes an enormous difference in an applicant's chances, especially for a mid-1400s candidate. Indeed, the difference is so important as to be a highly salable commodity. A gain of roughly 100 points is what The Princeton Review guarantees students who invest $500 and up in its test-prep courses. The Avery study's findings were the more striking because what admissions officers refer to as "hooked" applicants were excluded from the study. These are students given special consideration, and therefore likely to be admitted despite lower scores, because of "legacy" factors (alumni parents or other relatives, plus past or potential donations from the family), specific athletic recruiting, or affirmative action."

Now I suppose we can discount the Avery study because of the institution at which it was conducted;-), but it appears to be more than anecdotal.

[CUlater \'89]

That is very interesting, Al, thanks.  I wonder if the same holds true for "early action" candidates.  

I would also be interested to know how they determined whether any two particular candidates were equivalent, so that they could conclude that the only differences were SAT scores and time of application.  As we all know, high school grades across the country are not based on equivalent standards and students from certain high schools are looked at more favorably despite having lower test scores or grades (often, these are the very schools that push hard for students to apply early decision/action).

[nshapiro]

Effectively, what the article and this Avery study are saying is that an ED candidate is making himself a "hooked" candidate.  It gives anyone a chance to be looked at as favorably as the legacy, athlete, and under-represented minority candidate.  This seems to allow a student, who through no fault of his own  is un-remarkable, to level the playing field.

[ugarte]

I don't know the controls the study used, but with a large sample size, I would imagine that it is a reasonable assumption that the peripheral numbers of the typical 1200SAT/early-decision applicant are the same as they are for the typical 1200SAT/regular-schedule applicant.

[Al DeFlorio]

Well, if you really want to check the methodology, you can go here:

http://ksgnotes1.harvard.edu/research/wpaper.nsf/rwp/RWP01-049/$File/rwp01_049_avery_rev1.pdf

I'll confess I haven't read it--and probably won't.

[marty]

You mean unlevel Neil.  Or at least that is how many of us view what you are trying to justify.

And just for the record all four preference categories are given different weights and of course vary from school to school.  The law suit that is being fought by the University of Michigan has forced some of their criteria to surface and nothing but nothing is anything but uneven.

[Adam \'04]

Al,
   
     I think you need to read the paper with a fine-tooth comb.

[Q] By contrast, at ED schools, the credentials of early applicants are nearly equal on average to those of regular applicants. Early applicants and regular applicants were within 10 points in average SAT's at 6/10 ED colleges. Across the ED schools, early applicants averaged 9 points less on the SAT (1329 vs. 1338) and 0.3% lower on class rank than regular applicants (90.1% vs. 91.4%).[/Q]

Later in the article...

[Q]A comparison of Admission Office ratings suggests that EA receive higher Admissions Office ratings than regular applicants with the same SAT scores and class ranks. If EA are systematically more attractive than regular applicants in non-academic or other unmeasured attributes (some of which are captured in the Admissions ratings), then the classifications based on SAT scores and class ranks will underestimate the percentage and number of EA who would have been admitted at regular decision standards.[/Q]

Also, when you have the ability to define your own rating criteria it is much easier to filter out portions of a population because they do not meet a criteria you have set, or exclude/include particular criteria in an equation that makes your statistical analysis a sure bet. Statistics are nice and all, but they never show you what they didn't use. Hell, if you gave me the numbers, and enough time, I could tailor some evaluation system that counters nearly every point in this paper.

[Al DeFlorio]

Or perhaps you should.  

If, as the first paragraph you cite states, "...at ED schools, the credentials of early applicants are nearly equal on average to those of regular applicants," then why is the acceptance rate so much higher for early applicants?

I'm not sure I understand the point you're making with the second quote, but I hope you noticed the big If in the middle of it.  Perhaps the EA applicants are "systematically more attractive" to Admissions Office staff simply because they applied early, or perhaps even because more early acceptances means higher yield, which means a higher USN&WR ranking.  [Worth reading on that topic is the Thompson article cited in the references at the end of the Avery et al paper.]  You also conveniently didn't mention the immediately following paragraph, suggesting "quite possibly there is a significant bias cutting in the opposite direction."

Regarding the third point, if, for whatever reason, you don't want to accept what they've concluded, then don't.  Matters not at all to me.  Sounds like you think the principal investigators had an axe to grind.

[Adam \'04]

It is not at all unreasonable that the universities view the ED with higher regard. This is a type of reciprocal relationship; The student commits to only applying to their number one choice, and in return the university reciprocates this action with possibly an easier eye, or a more favorable view of a student that REALLY WANTS TO GO TO THEIR SCHOOL. Also, it is most likely that students that apply ED have higher quality application as a whole. This may not seem like much because we usually think of the larger aspects of the acceptance process. I know someone that works in an admissions office that said, "I read through applications cover to cover with a red pen marking all grammar and spelling errors. I tabulate the total number of errors, and write that number on the first page of the application." ::twitch:: Lets just agree to disagree.:-P  Although the SAT scores may be higher in the spring pool, that is not an indication of the quality of the person/student. In fact, the SAT is only correlated with present knowledge base, although it is supposed to be correlated with potential to learn. It is not highly correlated to future academic, personal, or professional success. With that said; the SAT is CRAP. Why don't they just make everyone take an IQ test?  ::nut::

[Al DeFlorio]

Adam '04 wrote:

Also, it is most likely that students that apply ED have higher quality application as a whole.

I'm delighted to agree to disagree.;-)

[jeh25]

Adam '04 wrote:

In fact, the SAT is only correlated with present knowledge base, although it is supposed to be correlated with potential to learn. It is not highly correlated to future academic, personal, or professional success. With that said; the SAT is CRAP. Why don't they just make everyone take an IQ test?  ::nut::

Not quite true. The SAT is NOT designed to predict how smart you are, how well you will do in life or what sort of work ethic you have. Maybe it has changed as they recentered the scores, but historically, the SAT *DID* correlate with academic performance during the 1st 2 years of school. (And If somebody busts out with correlation =! causation, so help me god I'll come to your house and smack you upside the head with a large trout.)

In fact, to the best of my knowledge, this is how MIT justified lowering their standards and admitting women with SATs that were 150-200 points lower than otherwise comparable male applicants.  They were able to rationalize it by showing that, at least in their sample, for men, an SAT score of X correlated with a freshman GPA of Y, while for women, an SAT of X-150 correlated with the same GPA Y, or so I was told by a feminist engineer at MIT while discussing gender quotas in the MIT coffee house.