<p>Is that why their SAT 50% range is lower than those three? Does its location influence more people to apply as their dream school? Or do they, like U of Chicago, just look for a different kind of applicant?</p>
<p>I think it's definitely because they look for a different kind of applicant. soo many people from my school last year got rejected from columbia and got into harvard. you have to be really clear and definite about your passions and your personality should fit new york city.</p>
<p>"soo many"</p>
<p>I bet.</p>
<p>"Is that why their SAT 50% range is lower than those three? Does its location influence more people to apply as their dream school? Or do they, like U of Chicago, just look for a different kind of applicant?"</p>
<p>i do believe on average columbia gets worse applciants, i think the accepted applicant pools are also marginally worse, nyc is a pull factor, but heck a larger applicant pool never hurt anyone. they do look for different things, i feel columbia is more willing to take a less academically inclined person, harvard has been known to do the same though. that being said i feel like columbia is converging, so as years go on, columbia's applicants increase in number and competitiveness more so than other schools. in my experience slightly more get into columbia and rejected by harvard than vice versa, and almost always an applicant accepted to both will take hyp, even though i've met many at columbia who've turned those schools down to be here.</p>
<p>I'm willing to bet that despite the raw score differences, the academic ability differences between the cohort are probably close to zero. </p>
<p>But I find it hard to believe that there's a significant difference in ability between someone who earns a 1400 and someone who earns a 1500.</p>
<p>"But I find it hard to believe that there's a significant difference in ability between someone who earns a 1400 and someone who earns a 1500"</p>
<p>there often isn't, it's just that when someone with a 1400 is smarter than someone with a 1500, one takes notice, so people tend to think the SAT is a useless indicator, on average it's a good indicator, but individually there's a LOT of variance in 'smartness/IQ' at a given score. on average though a 1500 student body will be smarter than a 1400 one.</p>
<p>At what inflection point, however, do you see diminishing marginal returns? I mean, on the margin, is it going to even be noticeable? How different, in reality, is the 99.9th, and 99.999th percentile as far as ability is really concerned?</p>
<p>I question whether or not we should be concerned with this distinction, even if admissions has to be (in order to play its assigned role.)</p>
<p>^on the SAT, there is a difference between a 1400 and a 1500, but sure 1550 vs 1600 is negligable. in most tests, and real contexts though there is a massive difference between the 99.9th percentile and the 99.999th percentile. there's practically no difference between 59-60th percentile, but as you go towards the edges of a distribution distinctions can be made ebcause there are few people there in the first place, that's the reason why you come to college and see many smart people but then there are some freaken geniuses, that's the difference between the 99th and the 99.99th percentile. you can see it in einstein like professors vs. normal profs, the sat however has no such scope to distinguish that close to the top, i agree with you that's it's more of threshold.</p>
<p>putting the last post in math sense,</p>
<p>assume ppl's IQs are normally distributed ( which is the same assumption SAT and many other intelligence test makes) like the previous poster stated, in such a distribution the tail end thins out. the difference between 99 and 99.999 may be a few standard divinations where as 50 to 70's not even 1.</p>
<p>confidence interval suggested by CB is about 80 points (god knows where they came up with that one) which means every score should be taken as +-80 points. notice that 1400 and 1500 including their respective interval over laps each other allowing for the case that a 1400 might actually have a higher intelligence's score...theoretically..</p>
<p>but SAT has been proven to be a pretty bad measurement even when set against the standards for what attempts to measure.</p>
<p>I don't think that the SAT is an accurate measure of the applications. I got the general impression that Columbia receives STRONGER applications because of their focus on who are you and what you do not just your grades. It's assumed that you're smart and have good SATs. </p>
<p>I think it's their emphasis on the individual applicant's experiences, personality and general "being" is the key.</p>
<p>As the Midwest Admissions Officer pointed out at the information session I went to, "Columbia appeals to the quirky students. We look at the experiences that someone has been through in their life to fully have a diverse student body."</p>
<p>I suppose she means diverse in the truest sense of the word instead of applicable only to race, ethnicity, gender, sex, or class.</p>
<p>confidentialcoll,</p>
<p>But remember that a 1600 is not necessarily indicative of hyper intelligence, anyway. I'm betting that the ability to measure that is lost on the test, since that's not what it aims to find.</p>
<p>"I suppose she means diverse in the truest sense of the word instead of applicable only to race, ethnicity, gender, sex, or class."</p>
<p>i agree.</p>
<p>"On average though a 1500 student body will be smarter than a 1400 one."</p>
<p>Well said.</p>
<p>I still disagree.</p>
<p>For one, we have to be willing to assume that the SAT is an accurate measure of intelligence. That's unlikely at best. Then, taking into account the confidence interval of 80, and you're really in trouble saying that the 1500 group is somehow smarter.</p>
<p>I don't think you understand the principles of statistics. =/</p>
<p>Flippy,</p>
<p>I'm an applied economist by training. I understand well enough.</p>
<p>Using the SAT as a proxy for intelligence is, at best, a difficult exercise. You have so much to control for it's absurd.</p>
<p>It should only be taken as part of a larger P-OLS regression to control for SOME of the variance. Assuming that any of these variables (the SAT, GPA, etc.) can capture some of the variance of "intelligence," the SAT should only be considered part of a well-specified model.</p>
<p>But then, you certainly have an interaction effect between income and the SAT, as well as ethnicity and the SAT-- though then you have issues with collinearity, assuming you do an ordinal variable for ethnicity...</p>
<p>And then you have to control for year taken...and...</p>
<p>Huh, what were you saying about me not understanding stats?</p>
<p>
[quote]
I'm an applied economist by training. I understand well enough.
[/quote]
</p>
<p>Then what are you doing on CC?</p>
<p>I don't work/study every single hour of every single day, you know.</p>
<p>Besides, you guys need SOME adults to provide context for all this madness...</p>
<p>Two points:</p>
<p>1) Nobody really knows sufficient detail about the applicant pools of the top schools for a meaningful comparison to be made. Columbia gets apps from a lot of losers who want to go to college in NYC. But I'm sure Harvard gets a lot of apps from a lot of losers who want to go to Harvard or figure they'll throw out a "hail mary" and see if it sticks.</p>
<p>2) UCLAri is right</p>