<p>I thought it might be interesting to see where the most students are going who score over 1500 on the SATs. The following is a list sorted by my estimate of how many full-time undergraduates have SAT scores over 1500 based on US News 2007. This is the raw number of students with over 1500 SATs, not the percent.</p>
<p>For schools that report ACT rather than SAT in US News, I converted the ACT to the SAT equivalent.</p>
<p>I was able to estimate the standard deviation because there are 1.36 standard deviations under a normal curve between the 25th and 75th percentiles. I estimated the mean by using the midpoint between the 25th and 75th percentiles. I calculated the z-score of a 1500 SAT using the estimated mean and standard deviation. Finally, I converted the z-score to the proportion of the area under a normal curve (one-tailed) above that z-score and multiplied by the number of full-time undergraduates.</p>
<p>school number over 1500
UC Berkeley 3668
Harvard 3333
U Penn 3110
Cornell 2805
UCLA 2791
U Texas Austin 2516
Yale 2513
Stanford 2435
Duke 2282
Columbia 2198
U Illinois Champaign-Urbana 2013
MIT 2007
Brown 1998
Northwestern 1998
USC 1977
Princeton 1953
NYU 1891
Washington U St Louis 1882
U Virginia 1723
U Michigan Ann Arbor 1645
U Florida 1639
U Notre Dame 1505
U Chicago 1501
Dartmouth 1486
Georgetown 1478
Georgia Tech 1470
UC San Diego 1349
Johns Hopkins 1278
Emory 1148
Tufts 1115
Rice 1030
Vanderbilt 1026
Carnegie Mellon 981
Boston College 893
UNC Chapel Hill 892
U Washington 844
UC Davis 721
College of William and Mary 711
UC Santa Barbara 708
Penn State University Park 687
Caltech 567
U Wisconsin Madison 561
RPI 543
Tulane 535
Case Western 514
UC Irvine 470
U Rochester 414
Brandeis 399
Lehigh 291
Wake Forest 278
Yeshiva 179</p>
<p>This would be interesting to see what % of class has a 1500 and to see LACs too.</p>
<p>The problem with your method is that you assume a normal distribution of scores around the 25/75th percentile. It may very well be that at some schools (particularly state schools) on your list, the top of the distribution may be truncated as almost all of the highest scorers end up at schools like HYPSM or SWAP, etc. In fact, do you even know if there are as many people total who scored over 1500 as you put on your list?</p>
<p>Haha… “SWAP” … what about WASP … :rolleyes:</p>
<p>I too would be concerned about a steep roll-off over the 75th percentile mark.</p>
<p>Illinois and Wisconsin have the same average ACt of 28 which most of their students take and are about the same size–how do you arrive at such different numbers?</p>
<p>Sanity check: The University of Florida has slightly higher SAT scores than UT-Austin, and the total number of undergraduates is within 7%, yet you report 53% more 1500+ scorers at UTA. How can that be?</p>
<p>This looks like a pointless argument waiting to happen.</p>
<p>This should be the college rankings</p>
<p>This shouldn’t be the rankings because it completely favors large schools, which offers a sub-par undergraduate experience for many. Now if you based these scores on % of the class that had a 1500 and included LACs, then I think it would be closer. </p>
<p>I like the acronym, ‘WASP’ better too.</p>
<p>obviously the pure numbers of make a difference- i find it hard to believe that texas is over mit…</p>
<p>barrons-
That is a good question. The ACT range for Illinois was 26-31 and for Wisconsin the range was 26-30, suggesting that Illinois has more high-end students. This converts to SAT ranges of 1180-1380 and 1180-1340, respectively. Illinois had 29,912 full-time undergrads. Wisconsin had 27,085.</p>
<p>To make a long story short, Wisconsin had a narrower estimated distribution, a slightly lower average SAT estimate, and three thousand fewer students.</p>
<p>It is just the way the calculation worked out. It has the do with the “spread” of the distribution, not just the average.</p>
<p>Interesting list by collegehelp. The large publics score very highly in sheer numbers.</p>
<p>“I was able to estimate the standard deviation because there are 1.36 standard deviations under a normal curve between the 25th and 75th percentiles. I estimated the mean by using the midpoint between the 25th and 75th percentiles. I calculated the z-score of a 1500 SAT using the estimated mean and standard deviation. Finally, I converted the z-score to the proportion of the area under a normal curve (one-tailed) above that z-score and multiplied by the number of full-time undergraduates.”</p>
<p>Lots of big words I don’t understand. Thus, it must be correct! :rolleyes:</p>
<p>rogracer-
The Texas mean SAT estimate is 25 points lower than Florida, so I understand why you are puzzled.</p>
<p>However the spread of the Texas SAT distribution is broader. If the distribution is symmetrical this suggests that it extends lower into the SAT range but also higher into the upper SAT range. In short, the broader distribution outweighs the lower mean. This plus the fact that Texas has 1500 more students leads to the result you see.</p>
<p>I think the number of over-1500 undergraduates is meaningful in its own way. If you collected all the over-1500 students together, Berkeley would boast the most. It indicates where the most top SAT scores choose to attend college.</p>
<p>The proportion of over-1500 SAT scorers is also interesting. It indicates where the over-1500 students are most highly concentrated. Here is a list sorted by the proportion. Keep in mind, these are estimates.</p>
<p>school proportion over 1500
Caltech 0.6213
MIT 0.5
Harvard 0.4697
Yale 0.4697
Princeton 0.4102
Stanford 0.3737
Dartmouth 0.3669
Duke 0.3527
Columbia 0.3416
Brown 0.3369
Rice 0.3369
U Chicago 0.3253
Washington U St Louis 0.305
U Penn 0.2984
Northwestern 0.2483
Johns Hopkins 0.2483
Georgetown 0.2272
Cornell 0.2053
Tufts 0.1946
U Notre Dame 0.1822
Carnegie Mellon 0.1822
Emory 0.1788
UC Berkeley 0.1645
Vanderbilt 0.163
Case Western 0.1383
Georgia Tech 0.1337
U Virginia 0.1286
College of William and Mary 0.1286
USC 0.123
Brandeis 0.123
UCLA 0.117
RPI 0.1105
NYU 0.0996
Boston College 0.0934
U Rochester 0.0934
Tulane 0.0869
U Texas Austin 0.0747
UC San Diego 0.0689
U Illinois Champaign-Urbana 0.0673
U Michigan Ann Arbor 0.0673
Wake Forest 0.0673
UNC Chapel Hill 0.0654
Lehigh 0.063
Yeshiva 0.063
U Florida 0.0513
UC Santa Barbara 0.0406
U Washington 0.0365
UC Davis 0.0349
UC Irvine 0.0243
Penn State University Park 0.0207
U Wisconsin Madison 0.0207</p>
<p>According to the UW Data Digest an average of 17% of each class enrolled has between 1350 and 1600 on the SAT (or ACt converted) (1475 midpoint). Applying that to the 27000 or so students equals about 4600 with over 1350 on the SAT. Does it fit the normal distribution to have only 12% (561) of that group over 1500?</p>
<p>barrons-
I used my formula to estimate the percent of students with SAT scores over 1350 at Wisconsin, pretending that I didn’t know the correct answer was 17%. My computation came up with an estimate of 22%, about 5% too high. So, this method is not exact but it wasn’t way off. Still, 5% of 27,000 is 1350 students (5940 versus the actual 4600).</p>
<p>I hope I get more known numbers like yours so I can test this method.</p>
<p>Do have have similar info for LACs?</p>
<p>Thanks for the rable collegehelp! Is it just me or is HYPSMC showing up again?</p>