What is Harvard's "real" admit rate?

Applying to College
1

I came across this article:
http://news.harvard.edu/gazette/story/2015/04/harvard-college-admits-1990/

and I was struck by the specificity of the following excerpt.
(apologies for the lack of HTML skills)

“About 13,500 students scored 700 or above on the SAT critical reading test; 16,100 scored 700 or above on the SAT math test; 13,900 scored 700 or higher on the SAT writing test”

Hmmm, says I, I wonder if you can use those numbers to estimate the number of applicants who scored a 2100 overall SAT? That would be an interesting thing to know, because overall, we know very little about the applicant pool at Harvard (or any other top school.)

So, using the data from here:
http://research.collegeboard.org/content/sat-data-tables

I came up with an estimate of about 11,000 applicants with at least a 2100 SAT. (Not going to bore you with the details here, but if anyone is actually interested, I can post the details of the two methods I used to develop the estimate.)

Now, coincidentally, 2100 also happens to be pretty close to dead on the 25th percentile number for Harvard (not exactly, but close enough for a back of the envelope type estimate of this type) - so that let us calculate an admit rate for applicants with SATs of 2100 and over:

3/4 of the Harvard admits is 1500 students. If there are 11,000 applicants in that range, that puts the admit rate for those students at roughly 14%. Of course, that number doesn’t really apply equally to the whole group (I think we can all agree that chances of admission rise with higher test scores, right?), so the applicable score is the population midpoint of that group, and that is right around a 2200 score.

There is one caveat here - so far, I’ve completely ignoring the ACT. True enough, but only 15% of Harvard applicants in the last few years submitted ONLY the ACT, so we can ballpark that by inflating the 11,000 by 15%, and that drops the admit rate to 12%. Also, it ends up meaning this high scoring group is almost dead-on 1/3 of the total applications.

I think that is fairly reasonable to assume that’s probably not too far off - if you’ve got a 2200 SAT, your estimated admit rate to Harvard is probably pretty close to 12%. We can also turn the whole thing around, and say with at least the same amount of confidence that if your test scores are below 2100 (or the equivalent ACT score of 32), you have less than 2% admit rate.

Anyway, I thought it was kind of interesting, because it is a topic people speculate about this all the time - how many applications to the top schools are “no-hopers” and I think this kind of analysis provides a framework for estimating that kind of thing.

2

(geez, I suck at proofreading…sorry)

3

Or you can look at Princeton’s admissions page where they supply this info:
SAT SCORES % ACCEPTED
2300-2400 14.5
2100-2290 8.1
1900-2090 5.2
1700-1890 2.1
1500-1690 0.3
Below 1500 0
No SAT Scores 5.9

4

This would be a really valuable data set to look at. It would be interesting to compare the average SAT score of Harvard’s reject pool with the average SAT score of Harvard’s acceptees. However, I don’t think these sort of statistics are released on their website.

5

Yeah, there is similar data provided at Brown, and a few other places. It tells you some interesting stuff, but lacking a detailed breakdown of the admitted class, it’s hard to make solid guesses at what the applicant pool looks like. The Harvard story happens to line up almost perfectly with the 25th percentile data, that’s what made it interesting to me.

6

@Studious99 I did do an estimate of where the median is for the Harvard applicants, and I’m pretty sure it’s close to 2000 SAT…

7

12% is still slim. And in holistic, despite great scores, your other choices and your own words can slay you.

Brown’s is also revealing. http://www.brown.edu/admission/undergraduate/explore/admission-facts

50% of those accepted may have scored 700+, but 12035 (of 13666) in that group were denied. The two aspects to consider are stats and the what-else. I don’t think a casual observer can get to the what-else. And that includes myriad considerations.

8

@lookingforward

Scores and grades are certainly “necessary, but not sufficient” and a big part of the admissions process at these schools is more subjective (or holistic or whatever you want to call it), but there is something to be learned from looking at the objective side. That doesn’t mean the “what-else” isn’t important - clearly it is, no matter what the level of the applicant, but even so, a large number applicants, even at these schools, are killed off by their lack of stats.

9

Can’t disagree with that.

Lightning rarely strikes the kids without the stats. It’s just that the kids who do have them need to fully understand what all it takes.

10

Agreed.

11

My D is applying to Brown, probably, though I calculate her chances at less than 5%. Why? Although her test scores put her above the 50% mark, her GPA and all the other stuff will do her no favors (to be clear, I am specifically talking about Brown. Most colleges in the country would be delighted to have her.) It is the “what else”, as you call it, that does her no favors. Specifically, being female, white, from a wealthy suburb of NYC, unhooked, plus the fact that Brown already accepted ED a boy from our school. It just isn’t going to happen, but it would be great if it could. Thing is, she really loves the school. But so do lots of other people. Because she really loves it, she is probably going to apply, and I guess it’s the tiny possibility of the dream school that makes kids apply despite the odds.

The point I wanted to make though, is that I am very sure that the Ivies and Stanford, MIT, CalTech, et al, most likely immediately chuck out 50% of the applications they get. Not literally throw them away, but I am sure there is a basic algorithm that, unless there is a “flag” (URM, first gen college, etc…), means that apps that don’t make the basic cutoff for GPA and test scores will not be considered. Do they really have the time and resources to read through 30,000+ apps, when at least 50% literally have no chance?

12

@Lindagaf Did you see the Reddit AMA with the Cornell adcom? It’s been linked here a few times, but I’ll give the short version here:

Every application gets read independently by two staffers (they do use temps for this step) - if both decline to recommend, the application is dead.

That’s obviously Cornell-specific, but I’m sure most other top schools do something similar.

13

Ok, humans do the “algorithm” instead of computers, but even so. I am willing to bet without the the basic markers of grades, test scores, and a “flag” they probably just get a cursory glance before getting an automatic deny. I understand that a new application format is going to be coming out for the top colleges next year, and I think it must be largely to help make life easier for the HYP type colleges that receive silly numbers of applications. I also understand it is to help students create a more streamlined “package” to present come application time.

14

Agreed, for the most part (although I think that most applications get a little more human attention than you think, but the outcome for most is essentially the same - rejected very early in the process.)

And I definitely agree that the new initiative is another method (as is ED or REA) to attempt to put the most promising applicants into a fast track pool. Like I said on another thread today, I think the top schools have probably reached a point of diminishing returns with increased applications - they already have a surplus of qualified candidates for the number of offers they can make - and the yearly increase in applications probably contains very few qualified applicants.

15

I see a major problem with trying to base admit percentages on the tables provided by SAT. The tables are based on scores made per single test while Harvard and many other colleges decide admission and report admission figures based upon superscoring SAT. I recall seeing reports that superscoring results in a much larger number of students having any particular test score than the one that would be indicated by those SAT tables, particularly when you are dealing with the higher test score levels.

16

The 25/75 scores reported aren’t based on super scoring, nor are the figures given in the article I linked to.

I did use those numbers to estimate actual SAT cumulative scores (which are not going to match up with reported “mean scores” if they are based on superscoring, but I’m not using that information), and they are certainly reasonable approximations of what the groups look like in terms of percentiles.

Moreover, the way that most people actually use the 25/75 numbers is completely inaccurate, because they aren’t just superscoring individuals, they are effectively superscoring across the entire class. 25% of incoming Harvard students don’t have a 2400, or even close to it, but that seems to be what many people seem to think.

17

The app is key, it’s how you present yourself; stats, rigor, and the activities and writing matter very much. The latter two show your thinking, how you make choices, etc. You can’t apply to a tippy top with fierce competition and expect being a top dawg in your own high school to be the “all.” Nor just match yourself against CDS stats. Nearly all the kids past first cut are tops in their hs, with scores and what kids typically see as leadership (clubs and a little service.) Not all show what the individual colleges value and see as fit. That’s the “more.”

First cut can be the area rep. The better look at holistic is the Duke commentary.

So, Lindagaf, you’d need to assess more than where her stats fall and her personal demographics.

18

@lookingforward , of course. Personally, I think she would be a great fit for the school. She is 100% aware of everything discussed here. She has no illusion of her chances. She is going to make her essays stand out, but at the end of the day, she knows that even if they think she is a great fit, so might another 1000 girls be. And her ECs are not packed with leadership, etc… She is shy and quiet. The kid who got accepted from her school is a truly deserving kid who is a wunderkind and we are very happy for him. Brown has only ever accepted 7 kids from our school. Let’s face it, they are not going to accept a non-wunderkind from our school again this year, even if they do think she is a great fit.
Can I ask for a link to the Duke thing you mentioned? Not at all clear what you are referring to.

19

@nickflynn your analysis does not compute.

80% of Harvard test takers for each part of the SAT scored over 700.

So if your number holds, a huge percentage of Harvard applicants would have to be ACT only.

(And the 14% acceptance rate you imputed is for which range of applicants? - it certainly is not for the middle 50%)

20

Lindagaf, it’s late in the game and I don’t want to throw a wrench in. And I know you’re trying to be rational, which is good. Plus, what matters is that any kid feel they did their best and have that satisfaction when they hit Submit. If you feel it’s a long shot, then she can just enjoy throwing her hat in the ring.

My ongoing point is so many kids and adults focus on stats and forget this isn’t simply hierarchical. How a kid presents him/herself matters very much. It shows judgment, perspective, and many other qualities- and in many cases, what the kid really knows about the school. Best wishes.

http://talk.collegeconfidential.com/parents-forum/1632160-christoph-guttentags-letter-demystifying-admission-process-at-duke-p1.html