Elite Colleges Still Favoring Kids from Private Schools?

[JuniorMint]

<p>

Stanford has certainly embraced holistic admissions, but no more so than any of its competitiors. I think there’s probably something wrong with that chart. It seems completely unlikely to me that Stanford’s 1966 average (more likely, median) SAT was way above Harvard, Princeton, Dartmouth, all of which were meaningfully more selective than Stanford then.

</p>

<p>Ah, that does make intuitive sense. I tried looking for historic admissions data from the 1960-80 period, but couldn’t find anything relevant, although it did make for some enlightening reading (I was born in 95). I apologize for posting the chart.</p>

<p>

But the Brown data suggest that there’s no meaningful difference between, say, 28 and 31, or 27 and 30. Or 630 vs. 680 on the SAT tests. That’s really surprising, but the numbers don’t work otherwise.

</p>

<p>It’s surprising, but also encouraging. For you high scorers reading this, there’s always Caltech! ;)</p>

[Hunt]

<p>

But the Brown data suggest that there’s no meaningful difference between, say, 28 and 31, or 27 and 30. Or 630 vs. 680 on the SAT tests. That’s really surprising, but the numbers don’t work otherwise.

I think this suggests that there are some applicants who are not being judged primarily on their scores, but on some other characteristics that they have–as long as they have some minimum level of academic ability. As Data10 suggests, these are probably not all “standard hooks,” but I don’t think the existence of these types of candidates means that scores don’t matter–I think there are many, many applicants who are being compared to others primarily based on grades, scores, and relatively similar ECs.</p>

[JHS]

<p>Sure, Hunt, I agree with that. I note, by the way, that not paying attention to differences in test scores below 700 does not mean that all academic credentials get ignored. Brown seems to emphasize grades more than most other schools – something like 95% of its students coming from high schools that disclose class ranks were ranked in the top 10%, and the admission rate for valedictorians is only a tad lower than the admission rate for kids with 36 ACT. Applicants from schools that do not provide class rank are 55% of the applicant pool but receive only 48% of the offers of admission. I assume that lots of the kids with relatively low test scores have great grades, great ECs and other great qualities. </p>

<p>Also, it’s pretty clear that Brown DOES pay attention to relatively small score differences over the 700/33 level and below 600/26. So it’s not like test scores don’t matter. It looks like someone with a 35 ACT has double the chance of admission of someone with a 31 ACT, and a 36 ACT has double the chance of admission compared to a 35. But someone with a 31 ACT has roughly the same chance of acceptance as someone with a 27 ACT (which is more than twice the chance of someone with an ACT below 26).</p>

[bluebayou]

<p>

I only meant to guess that for average applicants, it would be hard not to differentiate between a single-sitting applicant with a 2250 and a five-time retaker whose best unique score was a 2100.

</p>

<p>Adcoms just don’t care. Why should they? Just because you had a great day as a first time taker, doesn’t mean that someone could not have a bad day that Saturday morning. Perhaps they were coming off a Friday night football game after getting their bells rung. Or, perhaps having to work until 2:00 a.m. at Taco Bell to make money.</p>

<p>But more importantly, they might not even know if a student had one take or multiple takes. In the not-so-distant old days, a clerk in Admissions would write the highest test scores on the outside of the Applicant’s envelope – to save the adcom time from having to slog thru the CB and ACT reports. Today, I’m guessing the colleges have computers that post the high score in the appropriate slot.</p>

[Hunt]

<p>

Adcoms just don’t care. Why should they?

Well, they might think that a person who had to take the test five times to cobble together a superscore of x is not as impressive as a person who achieved that score in a single sitting, or over two sittings. As to whether they do care, or whether they see that information, I wouldn’t assume that all colleges are the same in that respect.</p>

[bluebayou]

<p>

Also, it’s pretty clear that Brown DOES pay attention to relatively small score differences over the 700/33 level and below 600/26. So it’s not like test scores don’t matter. It looks like someone with a 35 ACT has double the chance of admission of someone with a 31 ACT, and a 36 ACT has double the chance of admission compared to a 35. But someone with a 31 ACT has roughly the same chance of acceptance as someone with a 27 ACT (which is more than twice the chance of someone with an ACT below 26).

</p>

<p>Gotta be careful here. Brown made its cut at 32, which ~700 (SAT) for admissions purposes. Thus, the SAT/ACT tables have different cohorts.</p>

<p>More importantly, we don’t know the distribution of the admits from 29 to 32…and I’d guess that it is much higher for the 32’s than the 29’s, but yet that total cohort average is 7.1%. It’s possible that the 32 has an admit % in the 13% range (as the higher cohort), and that the 29’s are less than 5%. That leaves the next cohort at 7%, which is only 108 students, and could easily be filled by hooked applicants.</p>

<p>The other thing we don’t know is whether Brown has quite a few “lopsided” admissions: 750+/750+/680.</p>

[JuniorMint]

<p>

Adcoms just don’t care. Why should they?

</p>

<p>Well, I just did a quick cruise and found this thread, <a href=“Does super scoring hurt Ivy League chances - Test Preparation - College Confidential Forums”>Does super scoring hurt Ivy League chances - Test Preparation - College Confidential Forums, in which you had this to say:</p>

<p>

As far as focusing on one test at the expense of the other two, I’ve seen posters say this would hurt, but haven’t seen any real evidence of that.

A couple years ago, an adcom of an Ivy told me in no uncertain terms she tossed an app from someone who did just that. The applicant submitted an 800/600 and then an 550/800. (Forgot the writing score.) The point to the adcom was clear – the student was not competing fairly, and was obviously resting during certain sections (or worse).

</p>

<p>I’m sure the vast majority of adcoms could care less about superscoring for admissions purposes, and for institutional purposes, they obviously stand to benefit from it. My point is that at HYP, a fiple retaker wouldn’t have me smiling.</p>

[JuniorMint]

<p>

But more importantly, they might not even know if a student had one take or multiple takes.

</p>

<p>Valid point. :)</p>

<p>

Well, they might think that a person who had to take the test five times to cobble together a superscore of x is not as impressive as a person who achieved that score in a single sitting, or over two sittings.

</p>

<p>That’s my thinking. </p>

<p>

Perhaps they were coming off a Friday night football game after getting their bells rung.

</p>

<p>Does the football player have a Friday night game before every SAT sitting?</p>

[Hunt]

<p>I would note that even some schools that say they superscore nevertheless require you to submit all SAT scores. This tells me that they may look at how many times you took it, and how you got to that superscore. Maybe this is primarily to smoke out weird and suspicious cases like the one referred to in #287, but who knows?</p>

[bluebayou]

<p>

Does the football player have a Friday night game before every SAT sitting?

</p>

<p>Think about it: you raised the idea that a single sitting was better than multiple. If a HS senior takes the first SAT in the fall, it will likely be an October Saturday morning after a Friday night football game. (Whether that person attends the game is another issue.)</p>

<p>

I would note that even some schools that say they superscore nevertheless require you to submit all SAT scores. This tells me that they may look at how many times you took it, and how you got to that superscore. Maybe this is primarily to smoke out weird and suspicious cases like the one referred to in #287, but who knows?

</p>

<p>Perhaps, but more likely it is to avoid ‘cheating’ in the Subject Tests – which is rather easy to do with Score Choice. (plenty of online descriptions on how to accomplish…)</p>

[JHS]

<p>

More importantly, we don’t know the distribution of the admits from 29 to 32…and I’d guess that it is much higher for the 32’s than the 29’s, but yet that total cohort average is 7.1%. It’s possible that the 32 has an admit % in the 13% range (as the higher cohort), and that the 29’s are less than 5%. That leaves the next cohort at 7%, which is only 108 students, and could easily be filled by hooked applicants.

</p>

<p>That was my first thought, too, but then I noticed that the same pattern appeared with the (roughly) equivalent SAT scores. So the “next cohort” is not just 108 students, it’s over 11% of all students admitted. And there’s a next cohort under that, too, which almost certainly IS 100% hooked students. Between the two of them, they represent almost 15% of all students admitted, or about 480 students. I don’t think there are that many “hooked” students admitted (unless your definition of “hooked” is very broad), much less that many with relatively low test scores. (Not all hooked students have relatively low test scores.)</p>

[Pizzagirl]

<p>

Well, they might think that a person who had to take the test five times to cobble together a superscore of x is not as impressive as a person who achieved that score in a single sitting, or over two sittings. As to whether they do care, or whether they see that information, I wouldn’t assume that all colleges are the same in that respect.

</p>

<p>I would think that the issue is less than the person “isn’t smart enough” and needed multiple sittings to superscore, and more that the person clearly had nothing better to do with their time than re-take a test in the hopes of getting a higher score. That such a person needed to be doing something else with their Saturday mornings other than trying to eke out a few more points. I am quite sure if I were an adcom, I’d mentally roll my eyes at someone who took the SAT or ACT more than, say, 3 times max. And I’d probably mentally ding them for it, unless there were some compelling reason. There’s just such a suck-up/desperation factor to taking it multiple times, IMO.</p>

[JuniorMint]

<p>

I would note that even some schools that say they superscore nevertheless require you to submit all SAT scores. This tells me that they may look at how many times you took it, and how you got to that superscore. Maybe this is primarily to smoke out weird and suspicious cases like the one referred to in #287, but who knows?

</p>

<p>Precisely, although it appears of HYPSM, only Stanford and Yale do not accept ScoreChoice. Columbia, Penn, Caltech, Dartmouth, and Duke allow it. Penn and Cornell don’t.</p>

<p>If the school doesn’t use ScoreChoice and you are submitting four or more tests, then obviously, you are going to look insane. </p>

<p>Of course, there are many places to trip up on the Common App, so this is a trivial detail compared to more meaningful weaknesses like, er, missing essays.</p>

[JuniorMint]

<p>

Think about it: you raised the idea that a single sitting was better than multiple. If a HS senior takes the first SAT in the fall, it will likely be an October Saturday morning after a Friday night football game. (Whether that person attends the game is another issue.)

</p>

<p>I don’t really want to dwell on a hypothetical, but if this football player is taking the SAT for the first time in October of his senior year, and he is going to superscore, then he has to take at least one more sitting either in November and December, at which time the football season should be over. </p>

<p>I was actually suggesting that when the player does get an unfettered test date, why wouldn’t he just max out on his score? </p>

<p>Don’t most HYP applicants take the SAT in the spring of their junior year? That’s when I took both of my sittings.</p>

[Hunt]

<p>I wonder if taking the SAT a lot of times might suggest that the student is doing a lot of prep–something which might cause some schools to discount the ultimate score. I think it’s an interesting question whether schools try to correct for the likelihood of intense test prep–they might look for clues, or they might (consciously or unconsciously) profile people from some localities, or even ethnic groups, and assume scores are based on intensive prep.
Of course, it’s also possible the schools just don’t bother to try to figure all this out, and just look at the superscore, period.</p>

[YZamyatin]

<p>I would not be shocked if there were a higher acceptance rate at Stanford for an ACT score of 24 vs 29. I would speculate that the number of apps goes down in some proportion to the test scores thereby leaving a higher percentage of apps at the lower scores belonging to applicants with, essentially, a 100% acceptance rate. With low n it would not take very many recruited athletes at all to have a significant affect on the percentages in question.</p>

<p>We do not and never will have the numbers to come to any certain conclusion, but I think we can safely say that the number of apps coming into these schools below the 95th percentile is small enough that conjectures presented with no consideration made whatsoever for how the percentages are affected by athletic recruits have a high probability if being erroneous. </p>

<p>I would classify all other hooks, as well as unique cases that fall outside the strict definition of “hook”, differently. Recruited athletes are the only sizable group of applicants who are, essentially, guaranteed acceptance. If the schools were interested they could easily strip out their affects on the numbers and release much more accurate and reliable statistics on acceptance rates for non recruited applicants, but they are obviously not interested in doing so.</p>

<p>Clearly it is not ‘impossible’ to get into these schools without very high grades and test scores or a well defined hook, as data10 has illustrated. However, most students who have extraordinarily unique and/or exceedingly (to peers) accomplished backgrounds know who they are. I guess by that I mean its not like winning the lottery. For the great majority of unhooked applicants, there is no point applying to these schools without very high grades and test scores. If someone is not sure whether they are unique or accomplished ‘enough’, well… then the colleges have done their job… they should submit their app but curb their expectations (which is probably good advice for all applicants anyhow).</p>

[Hat]

<p>For a data point of 1 on the subject of triple taking the SAT (or ACT) and flipping scores to get a high superscore (Post 287), I offer my son. He took the SAT three times scoring 800 or near 800 on Math on each occasion. The first time his SAT score was 2090, but with Math at 800, he scored 1290 on CR and W (mid to high 600s on W and low 600s on CR). The next time he took the SAT, he concentrated on CR and raised his score to 720 (Math dropped slightly and W also dropped slightly). The last time he took the SAT, he concentrated on W and raised his score to 780, but CR dropped to mid-600s. His superscore was 2300 (a 210 point improvement over the first SAT). He was accepted to Harvard as a white male without any traditional “hooks” (although he did have other things that probably helped his application).</p>

[Pizzagirl]

<p>

For the great majority of unhooked applicants, there is no point applying to these schools without very high grades and test scores.

</p>

<p>Isn’t this kind of a no-brainer, though? Why WOULD you bother without very high grades and test scores as your basis for consideration? Are there THAT many people who are stupid enough to think they could take their 3.2’s and 1800’s (absent something else remarkable) and apply?</p>

[QuantMech]

<p>Hope springs eternal . . .</p>

[Data10]

<p>

Gotta be careful here. Brown made its cut at 32, which ~700 (SAT) for admissions purposes. Thus, the SAT/ACT tables have different cohorts.</p>

<p>More importantly, we don’t know the distribution of the admits from 29 to 32…and I’d guess that it is much higher for the 32’s than the 29’s, but yet that total cohort average is 7.1%. It’s possible that the 32 has an admit % in the 13% range (as the higher cohort), and that the 29’s are less than 5%. That leaves the next cohort at 7%, which is only 108 students, and could easily be filled by hooked applicants.</p>

<p>The other thing we don’t know is whether Brown has quite a few “lopsided” admissions: 750+/750+/680.

Regarding Brown, they claim to not be as focused on test scores as most on this forum appear to be. The Brown CDS indicates that Brown does not include test scores (or any stats) in their “very important” criteria category, but they do put various holistic, non-stat criteria in this “very important” category including things like talent/ability, character/personality qualities, and level of applicant’s interest. The Brown admissions FAQ specifically states that many applicants have strong academic potential with modest test scores, as quoted below.

While we do receive many applications from students with high test scores, we consider standardized test results in the context of all the other information we have about a candidate. We find that many of our applicants demonstrate strong academic potential in spite of relatively modest test scores.

Intuitively this type of focus makes sense to me. Does it really matter if an applicant is in the 96th percentile or 93rd percentile on one particular test that has a different focus from his college major? If the rest of the app is equal, is the 96th percentile scorer going to have a much better chance of success in college and beyond than the 93rd percentile scorer? Instead of a freshman class full of near perfect test scores, I’d expect they want a freshman class full of great people who will be a good addition to the campus and are likely to do amazing things both in and out of the classroom, as well as after graduating. Identifying this group has more to do with the other criteria (similar to the criteria Brown marked as “very important”) than minor differences among test scores. Sure, the chance of an applicant meeting the description increases as test scores increase, but you’ll also find many who fit the description without stellar test scores.</p>