<p>I don't understand why no one else has seen this before, but I think this is a very logical reason as to why higher SAT scores (after a certain point) result in higher acceptance rates. After say 2200 or so, SAT scores really should not matter that much; SATs should be more of a cutoff. Now this does not explain why 28% of SAT scorers in the 2300-2400 range are accepted as opposed to 10% in the 2100-2290 range (for Princeton). It is pretty obvious: there are less 2300+ scorers and way more 2200-2290 scorers. Therefore, there are more 2200-2290 scorers to reject as opposed to 2300 scorers. It isn't that a 2300+ is way better than a 2200+, it is just that there are a lot more 2200-2290 scorers applying than 2300-2400 scorers.</p>
<p>^Yep I agree. However, that does mean you stand less of a chance scoring something several others score, so you technically still do have less of a chance unless you stand out. What's the acceptance rate for those who score 2000-2100? Does Princeton have any non-URMs, non-recruits, non-legacies that score in that range?</p>
<p>Well, it's not likely that they are 'accepting' a set percentage at each SAT range... Look at it a little differently. Assuming an academic [non-athlete, non URM] candidate has solid scores (maybe for the Princeton applicants, that's 2,100 and up), I think they'll move on to the other elements of the application. And it just seems logical that the pool of high achieving students with 2300+ SATs may have other credentials in common (high GPA, academic awards, strong subject tests, etc,) So that combination then produces a higher statistical admission rate for students between 2300-2400.</p>
<p>You're going in circles here. The higher acceptance rate for 2300+ means they do value 2300+s more highly, pure and simple. The sample size of each group, once large enough for the data to be statistically significant, has simply no effect on the probabilities involved. If one group shows a significantly larger acceptance rate than the other, then that group has a clear advantage, and there is clear bias.</p>
<p>A simple question: Say instead that the 2100-2290 group is smaller, say 100, while the 2300 group has 1000. Are you gonna say that 28% of 2200-2290 group will be accepted and 10% of the 2100-2290 group is accepted? </p>
<p>Your logic makes no sense; ask yourself, and explain to us, if 2200-2290s are no different from 2300+s, why doesn't Princeton accept equal percentages, say 15% of 2300+s and 15% of 2200-2290s? </p>
<p>If they really didn't distinguish between 2200-2290 and 2300+, and assuming that all other factors (ECs, essays etc) have no correlation to scores, then every one within the 2200-2400 range would have roughly the same probabilty of being accepted. Let's say that probabilty is about 15%. So no matter how you slice and dice the 2200-2400 group, and no matter how disproportionately you allocate applicants to sub-groups, if scores don't matter and people with outstanding ECs are evenly distributed, then the acceptance rate should be uniform throughout, for each and every person!</p>
<p>I already explained this...</p>
<p>Assume that there are more 2200-2290 applicants applying to Princeton than there are 2300-2400 applicants. If we also assume (though this is not true, but for simplicity's sake) that Princeton accepts the same number of applicants from each pool of SAT ranges, then they would have to reject more of the 2200-2290's than 2300+s. Now they probably do not accept the same number of applicants from each pool, but I can say for certain that there are not the same number of applicants from the 2300+ range as there are the 2200-2290 range. Last year only about 6000 people scored a 2300+ on the SATs. All 6000 of those applicants did not apply to Princeton. Also, I did not say that a 2200 is equivalent to a 2400. All I said is that SATs are not a huge deciding factor in admissions. After a certain score for non-hooked applicants, SAT scores really do not matter.</p>
<p>the point is, Princeton doesn't separate its applicants into separate pools by SAT ranges, and neither does it try to select the same NUMBER from each score range.</p>
<p>You're basically saying that 2200-2290 is not much different than 2300-2400. So what you're saying about the 2200-2400 range can roughly be equated to</p>
<p>1) For this scenario, we consider only 2200-2400 applicants
2) Cover the SAT scores
3) Pick applicants based on everything else, which we can assume is roughly equal throughout.</p>
<p>if SAT scores don't really matter from 2200 onwards, then naturally Princeton should accept the same PERCENTAGE of applicants from every part of the 2200-2400 score range. But it doesn't. So there's a bias for the 2300+ range.</p>
<p>You have to be retarded to make a statement like that. Just look at tokenadult's link. You'll see that there are way more 2200-2290 scorers than 2300-2400 scorers. I am not saying that Princeton separates by score range, rather they end up picking around an equal number of applicants from each pool if it were to be divided. There are more 2200-2290 applicants to reject as opposed to 2300-2400 applicants. That is why the admit rates are lower, not due to any bias.</p>
<p>Because the stats are given in % form, I can't see how a greater sample size affects what they are reporting.</p>
<p>screwitlah, go take AP statistics please....</p>
<p>
[quote]
"1) For this scenario, we consider only 2200-2400 applicants
2) Cover the SAT scores
3) Pick applicants based on everything else, which we can assume is roughly equal throughout."
[/quote]
</p>
<p>The problem is the 3rd point. There are not a roughly equal size in the 2200-2400 applicants, there would be more 2200-2300 than there would be 2300-2400. So if we do it as a blind sample then it should mean more 2200-2300 accepted.
The problem in your analysis is that you are looking solely from an SAT standpoint, correlation does not equal causation, in any statistic.</p>
<p>The OP's post shows that a person with a higher rather than lower SAT score has a higher expectation of being admitted, so it's not clear whom the OP is disagreeing with here. Whether the SAT scores near 2400 are desirable in themselves to admission officers is disputed </p>
<p>but for the purpose of outside observation of the black box of admission practices, it is enough to note that high SAT scores are correlated with higher expectations of admission. All colleges that show admission statistics stratified by SAT scores tend to show the same phenomenon.</p>
<p>Ok, I'm pretty sure this logic is flawed. Why would the larger number of 2200-2300's affect the percentage of accepted applicants?</p>
<p>Yes, if they are valued equally and there are more 2200-2300's than 2300-2400's, then more of the 2200-2300 would be rejected. HOWEVER, if your conjecture holds true, more of the 2200-2300 would also be accepted, and hence the end percentage should remain the same. Get it now? A different sample size would NOT affect the percentage of an outcome.</p>
<p>My question is, why does this matter? Are you just trying to convince yourself that a 2200 is just as good as a 2400? Because it's not. 200 points is a significant difference and yes, it will matter in the end. If you want to fool yourself, go ahead, but don't impose faulty logic unto others to mislead them as well.</p>
<p>And Dbate, your argument doesn't make sense either. A blind sample would yield more 2200-2300 accepts, but since the total possibilities is greater for the 2200-2300 group, it should be the same PERCENTAGE. Your point that we're only looking at the problem from an SAT standpoint is irrelevant, because that's the entire thread is about.</p>
<p><a href="mailto:lol@thisthread">lol@thisthread</a>. 2200 < 2300, pretty simple.</p>
<p>some of you are missing the fact that the students who scored in the 2300-2400 range could also be better overall applicants....the fact that the students in this range have a higher acceptance rate DOES NOT indicate with certainty that its the SAT scores that got them in....you have to look at the whole package...a student with a 2400 most likely has a very strong overall application, usually stronger than someone with a 2200 or less....Princeton doesn't need to boost their prestige and rankings with 2400s, they already have all the prestige they could ever want, which is another incentive for them to look at applications holistically and put less emphasis on the numbers</p>
<p>What bicyclekick is saying is basically what I am trying to say without saying. There are more unqualified 2200-2290 scorers just because it is easier for an unqualified student to hit that range than it is to hit the 2300-2400 range. So it is not the score that is getting those people through. Do not mistake cause and consequence. Although there is a correlation between higher scores and higher admissions rates, that does not mean that the higher score is the cause for the higher admit rate.</p>
<p>I agree that higher scorers are also more likely to be stronger in other areas. But Princeton admissions must be aware of this. Faced with the choice of two otherwise equal students (excellent, but Princeton can't fit every excellent person), wouldn't it make sense to pick the one that has a slight chance of being a little bit more qualified (i.e. there is something more than the application shows)?</p>
<p>i don't get why people bring up these hypothetical scenarios like, "what if 2 COMPLETELY equal applicants were only separated by 50 SAT points...who would they pick?"</p>
<p>This doesn't happen....each applicant is different and brings something different to the table, if their scores are both excellent, then a small score difference shouldn't be the tipping factor, it should be which applicant they think will contribute more to the university</p>
<p>Among many thousands of applicants, there are unquestionably must be similar people and borderline admits. And I'm not just talking about a small margin of error difference with the SAT. Maybe a 2250 that got waitlisted might've been admitted with a 2400.</p>