Admission stats for perfect SAT scorers?

<p>

</p>

<p>[Sorites</a> Paradox (Stanford Encyclopedia of Philosophy)](<a href=“http://plato.stanford.edu/entries/sorites-paradox/]Sorites”>Sorites Paradox (Stanford Encyclopedia of Philosophy))</p>

<p>A 10 point differences has to have some non-negligible impact or else we would be reduced to accepting that a 600 is no different than a 2400. Obviously a 10 point differences is very, very small but it is still significant to some extent. It is tempting to make this conclusion due to the high variability of the test, but none of the “objective” measurements of ability that colleges use are completely exact.</p>

<p>

</p>

<p>This is true of any score. The number of 2030’s will be lower and more rare than it could be, because of all the people who get 2020’s and the like and decide not to retake. This effect probably increases as scores go up because students less likely to be dissatisfied with their scores (presumably), but it is not uniquely relevant to the 2400 score.</p>

<p>

</p>

<p>This is an extremely good point. Once the ceiling is reached, the test no longer gives any exact gauge of the student’s ability. That’s the disadvantage of the 790: you might be just about equal to the 800 scorer, but you’re not equal to the “800+” scorer whose ability doesn’t show on the test.</p>

<p>The real difficulty is that there are plenty of 750-790 scorers who are the equal or the better of the 800 scorers. I know of several extremely advanced math students (the least of whom took the equivalent of BC calc as freshmen) who scored under 800 on the math section. The SAT tests are not precise - in some cases a sub-800 score indicates lesser ability/skill/talent, in other cases it does not. Someone who is merely tired, slightly ill, distracted, bored or annoyed by lousy test conditions can easily fall from the supposed perfection of 800. (To be clear - yes 800 is ‘perfect’ in that it is the highest score. But , it certainly does not indicate a perfect knowledge/ability/talent in math.)</p>

<p>

</p>

<p>Yes, this has been acknowledged. However, such people are decidedly in the minority. Exceptional circumstances that lead to achievement that is less than the potential are just that: exceptions.</p>

<p>And, as I’m sure you know, being more advanced in math than another student does not necessarily mean that that student is more mathematically talented than the less advanced student; often, it merely indicates that the student had more opportunities to get ahead.</p>

<p>But we’ve gone over all this before, nemom. :)</p>

<p>Ah, but my dear silverturtle, you toss in irrelevant points.
The key question here is how much weight the rare 2400 actually carries. I am arguing that one component (at least - I have said nothing of the others, from which you can, of course, conclude nothing) is flawed because I know of a considerable number of very talented high school aged math students whose math SAT scores vary ,randomly , between about 750 and 800. Exceptionally talented math students are rare. 750-800 math SAT scorers are rare as well, although less so. The mapping between the two is imperfect. This is something well-known to many in math (some of who, in fact, scored below 800 and yet went on to great careers).
Certainly, there are undiscovered math talents. Some of them even score below 750 on the math SAT and yet do well in math-based careers.
The difficulty is that the SAT is a relatively crude instrument with fine markings. Here’s an analogy. I can have a plumb line for measuring the depth of the river (lets use the Mississippi here just for fun). It can be marked in tenths of inches. Does that mean that when I toss the lead and haul it up, I actually have a measurement that is good to a tenth of an inch? Of course not. I may not toss perfectly each time. The mud may be soft or hard. The line may stretch. Is my measure perfect? No. Is it of some value? Yes. The SATs are much the same.</p>

<p>How, silverturtle, do you know that “such people are decidedly in the minority”? And, what do you mean? That exceptionally gifted math students are in the minority? Well, of course. That’s why we call them exceptional, after all.
I’ve known a lot of very talented math kids. If the SAT math test was perfect, wouldn’t they all get 800s? If even one does not, would it not be wise to consider that the SAT math test is not perfect and certainly doesn’t find exceptional math talent. (Of course, the aim of the SAT is not to find exceptional math talent - at least when given to high school juniors and seniors. It works fairly well at that when given to youth under age 13.)</p>

<p>

</p>

<p>I mean that most people who miss questions on the SAT do so for question-related reasons, not because of fatigue, illness, or annoyance with test conditions. If a student is truly great at math, he or she should probably be able to get all the questions (which are, after all, very simple if one has a good ability to problem solve) correct. You present anecdotes which belie this, but I can’t see why that would be the case. Do they misbubble? Do they misread questions? If they can solve the questions quickly, the potential for these problems can be rendered so low as to be negligible. The reality is that when most people miss questions, it’s because they didn’t know how to solve them in the given time. </p>

<p>If admissions officers gave everyone the benefit of the doubt (i.e., assuming that the fact that someone didn’t score better was due to meaningless mistakes), there could be no meaningful objective differentiation and scores would become irrelevant.</p>

<p>

</p>

<p>Even if the test were perfect, students could still make occasional silly mistakes, resulting in some rare cases of exceptional students’ not getting 800’s. And, indeed, I’ve never called the SAT perfect. I fully acknowledge that the correlation between aptitude and scores is imperfect, but you take this too far by seemingly arguing that the correlation goes away at 750+.</p>

<p>Sorry 'bout all that, then. My inner statistician has been wrong twice tonight (luckily didn’t post the other one) and is off in a corner feeling inadequate. Bah…oh well, my bad.</p>

<p>Wait, actually, I have some data to contribute!
Here are the presidential scholar nominees from my high school (which is a fantastic! public magnet, and has an average SAT score that is out of this world. It may be over 2200) and their college destinations. Presidential scholars are hand-picked from the best students in the country, and so are a thoroughly biased sample. The nominees are a much better picture, but nobody publishes where they’re going to college. There are a lot of nominees from my HS, so here they are, and hopefully they’re a good sample of all nominees. The amount I know about some of them is limited.
Our HS traditionally does best at Princeton…Harvard was a slaughter this year, waitlisting or rejecting all of these people who applied except me ><…I feel incredibly lucky and humbled by my acceptance. (Harvard also took two guys with lower SATs) Many did just apply to Yale and Princeton of HYP, though. Some of the rejections here were huge scandals (Stanford SCEA! O.o) I’ve noted who made semifinalist, but not all of these people applied for the program, as it was a massive time commitment on the essays.</p>

<p>Boys:
Accepted ED to Dartmouth
Rejected by the upper Ivies, WL’d at Chicago, likely letter from and attending NYU Tisch
WL’d at Harvard, attending either Yale or MIT.
Attending Stanford
Rejected from a bunch of Ivies, accepted at NYU, attending State Flagship U (semifinalist)
Attending Cornell
Accepted to MIT and Caltech EA, didn’t apply elsewhere. Attending MIT
Rejected from HYPSM etc., going to Cornell (semifinalist)
Rejected from MIT, Yale, WL’d at Caltech, attending State Flagship U
Attending Stanford
Attending Princeton</p>

<p>Girls:
Going to State Flagship U
Rejected from MIT and Princeton, accepted at Johns Hopkins, attending State Flagship U
Accepted at Pomona ED
Rejected at Stanford SCEA, WL’d at Harvard, accepted at Yale, MIT, attending Jerome Fischer at Wharton. (state scholar)
WL’d at Harvard, accepted at MIT, attending Yale, likely letters from Columbia, Dartmouth (semifinalist)
Rejected from Yale and Princeton, WL’d at MIT, attending Cornell
Asian name of indeterminate gender that might be my female acquaintance who’s going to Penn
Attending WUSTL
Attending Second Best State U, possibly ED
Rejected at the upper Ivies, WL’d at Brown, accepted at Johns Hopkins and attending Second Best State U
Rejected from most if not all Ivies to which she applied, attending Duke
Attending Amherst
Me, attending Harvard (semifinalist)
Rejected from Stanford SCEA, all Ivies to which she applied, attending Wellesley
Accepted at Cornell, Duke, attending Princeton (semifinalist)
Attending State Flagship U</p>

<p>^ That’s some high school you’ve got there.</p>

<p>Grape1: “A 10 point differences has to have some non-negligible impact or else we would be reduced to accepting that a 600 is no different than a 2400. Obviously a 10 point differences is very, very small but it is still significant to some extent. It is tempting to make this conclusion due to the high variability of the test, but none of the “objective” measurements of ability that colleges use are completely exact.”</p>

<p>I would argue that the paradox doesn’t apply as much in the upper brackets. Top colleges take the SAT averages and extrapolate how students perform in college. Typically, we see that those with higher SAT scores do better in college. However, I think the variance is simply too high once you start creeping into the upper percentiles due to fewer data points, which is why we see so many admissions officers claiming that a 750 versus 800 is not seen as hugely different. I would assume that the college performance rates for 750’s isn’t much different from the 790’s, for instance, and so this is why the entire “750+” score bracket is treated as its own equivalency threshold.</p>

<p>The paradox may make more sense in the 600’s, where a difference in one score to the next implies a greater number of questions missed (due to the nature of the test curving) and the correlation between score and success is perhaps more direct.</p>

<p>

</p>

<p>Treated by whom? MIT is the only school that claims this policy, and other schools’ admissions data indicate that this is not their practice. </p>

<p>

</p>

<p>I’m not sure that I’m following you. People who scored 750 missed more questions than people who scored 800.</p>

<p>Right, but if I recall the SAT curves, the difference between, say, a 620 and 630 or even a 650 and 660 is greater than a 780 versus 790.</p>

<p>

</p>

<p>Well, it’s difficult to say. The difference between 620 and 630 is usually one question. Often, 780 and 790 are not both possible on a given administration because the curve comes into play a lot at the high end. If that was what you meant, then I see your point.</p>

<p>So to clarify: people who score 780 and people who score 790 generally both missed the same number of questions; the people who scored 780 simply had an easier test.</p>

<p>Sorry – yes, that is what I meant.</p>

<p>

There are way more “good colleges” than this comment would suggest. Many more. Harvard, Yale, Princeton, Stanford, and MIT have this acronym because they suit the interests of many top students on this website. The Ivy League has been considered to hold many of the very best colleges for a long time. US News has a very flawed system. Harvard, Yale, Princeton, Standford, MIT, Brown, Dartmouth, Cornell, The University of Pennsylvania, and Columbia are among some of the best colleges. They are not the only ones who could qualify for “the best colleges”, and they are most certainly not the only good colleges.</p>

<p>Anyway, on topic: There are no official statistics on acceptance of those who score 2400. One would assume that it would be high, but a study of CC users is not an unbiased and/or random sampling (though if all 2400 scorers could be identified, the sample is small enough for comprehensive results). Assuming that the acceptance rate to top schools is high, it could of course be said that this is due to all the parts of the application that are also likely to be excellent (grades, ECs, essays, recommendations). Also, the acceptance rate would vary depending upon whether you mean how many get into all top schools of their choice, or just one top school of their choice. </p>

<p>Also, there is obvious a difference between lower scores and higher scores. But remember, admissions officers for top schools are not robots. What matters is their perception. They may perceive a greater difference between 2390 and 2400 than 2370 and 2380, or they may not. Almost certainly, they will perceive a significant difference between 2200 and 2300 or 2300 and 2400. However, they will likely perceive all of these scores as excellent, just to varying degrees. A student with a higher SAT score will not always be favored, given that a student with a lower SAT score may have better grades or a superior essay. However, when it comes to the score itself, a higher SAT score will always be better than a lower SAT score. An exception being “Tufts Syndrome”, that Tufts and Washington University in St. Louis have both been alleged to have. Also, I wonder if one four digit number above, say, 2200 may appear more psychologically attractive than one above it. This, naturally, would be void if the lower score was physically being compared to the higher score, and is overwhelmingly likely to be completely not the case anyway. Just an interesting thought. But, again, higher score > lower score. However, this does not necessarily entail higher scorer > lower scorer, considering that other parts of the application could make up for the inequality.</p>