<p>Case Western Reserve 1430
William & Mary 1440
Brandeis 1460
Southern Cal 1460
Tufts 1480</p>
<p>are ahead of Virginia in that list.</p>
<p>Cornell is always discussed on this kind of thread because it is anomalous in this group.</p>
<p>I think this is an interesting list, because it tries to answer the question: how smart are the smartest students at this school? That is obviously a very different question from overall quality, or quality of education, but it is a factor that probably matters to a lot of people.</p>
<p>The University of Florida is waaaay underrated using this criteria, as the 75th percentile of entering 2008 freshman is 1410, putting it on equal footing with Virginia, Case Western and William and Mary.</p>
<p>^^^Please clarify if you will, but I think that U Florida number is for ADMITTED students. The numbers for enrolled students won't be available until November or later at most colleges.</p>
<p>U.S. World News & Report(2008)</p>
<p>Average Sat Scores </p>
<p>Notre Dame 1290-1500
USC 1280-1460
Berkley 1200-1450
Virginia 1220-1430
UCLA 1180-1410</p>
<p>Top 30 based on top 25% SAT</p>
<p>1 Princeton 1590
Harvard
3 Yale 1580
4 Cal Tech 1570
5 MIT 1560
6 Dartmouth 1550
7 Stanford 1540
Duke
Columbia
Rice
11 Penn 1530
Chicago
Wash U
Brown
15 Northwestern 1500
Notre Dame
17 Cornell 1490
Johns Hopkins
Georgetown
Carnegie Mellon
21 Tufts 1480
22 Emory 1470
Vanderbilt
24 Southern Cal 1460
Brandeis
26 Cal- Berkely 1450
27 William & Mary 1440
28 Virginia 1430
Case Western Reserve
30 Tulane 1425</p>
<p>I don't know why your selection of 75% has any more utility than looking at the 25%. Is there a way in which you and the other top 25% can practically segregate yourselves from the rif-raff bottom 75%? </p>
<p>Caltech's 25% is 1470, while Harvard's is 1390. Harvey Mudd's is 1420.</p>
<p>Because one cannot, practically speaking, choose which quartile one spends time with in class or out of class, if I need for some reason to look at one number, it is the average of 25% and 75%. Harvard is 1490, Caltech is 1520.</p>
<p>Assuming one could decide to spend the majority of his/her time with, say, the Top 100 students in a University, then it is fair to say thase Top 100 students in virtually all of your top thirty schools would be between 1590 and 1600 (with a much larger number than 100 having been rejected with 1600).</p>
<p>but what is the point? You cannot create your own subset within your University, so it's a moot point.</p>
<p>A school's 75th percentile is generally more indicative of the student body than the 25th percentile because the latter depends on the school's policies regarding athletes, legacies, etc.</p>
<p>Correct, lgellar.
Harvard is more apt to enroll the nation's top figure skater or someone with experience as a Hollywood actress than Cal Tech or Mudd.</p>
<p>The error in this comparison is assuming a student who scores in the top range on math is the same individual who scores in the top range on critical reading. That's actually not likely. But thinking about this procedure of adding section score interquartile ranges that are actually reported separately suggests a procedure for finding colleges with the smartest bottom group: add (invalidly, again, but this time the error goes the other way) the 25th percentile score for each section at a lot of colleges. Then you could produce a list of colleges with the highest-scoring bottom-level students.</p>
<p>It's an error in all cases to assume that the individual student who scores above the 75th percentile on math is the same individual who scores above the 75th percentile on critical reading. (It's actually more likely that "high math" students are "low reading" students within any particular college.) It might be more informative to look at this issue the other way around. Look at the sum of section scores (again, this procedure is actually invalid, because the section score percentiles are reported separately for each section) defining the 25th percentile for each test section. This procedure is erroneous, but it is erroneous in the other way, overstating the number of low-scoring students rather than overstating the number of high-scoring all-around students. </p>
<p>Here is an example of that procedure, for a few colleges (with links provided for you to check my arithmetic and transcribing information): </p>
<p>[Caltech[/url</a>] </p>
<p>700+770=1470+680=2150 </p>
<p>700+700=1400+700=2100 </p>
<p>700+700=1400+690=2090 </p>
<p>690+700=1390+690=2080 </p>
<p>660+720=1380+660=2040 </p>
<p>660+680=1340+660=2000 </p>
<p>[url=<a href="http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=3693&profileId=6%5DCarnegie">http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=3693&profileId=6]Carnegie</a> Mellon](<a href="http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=4214&profileId=6%5DCaltech%5B/url">http://collegesearch.collegeboard.com/search/CollegeDetail.jsp?collegeId=4214&profileId=6) </p>
<p>620+690=1310+620=1930</p>
<p>
[quote]
I don't know why your selection of 75% has any more utility than looking at the 25%. Is there a way in which you and the other top 25% can practically segregate yourselves from the rif-raff bottom 75%? </p>
<p>Caltech's 25% is 1470, while Harvard's is 1390. Harvey Mudd's is 1420.</p>
<p>Because one cannot, practically speaking, choose which quartile one spends time with in class or out of class, if I need for some reason to look at one number, it is the average of 25% and 75%. Harvard is 1490, Caltech is 1520.</p>
<p>Assuming one could decide to spend the majority of his/her time with, say, the Top 100 students in a University, then it is fair to say thase Top 100 students in virtually all of your top thirty schools would be between 1590 and 1600 (with a much larger number than 100 having been rejected with 1600).</p>
<p>but what is the point? You cannot create your own subset within your University, so it's a moot point.
[/quote]
The top students generally hang out together. For engineering major especially, they stick together in a group doing scary projects together. They bond together because let's face it, most do not shower regularly and get used to others' smell.</p>
<p>lgellar, danas,</p>
<p>Are you suggesting that at least 25% of the students admitted to these schools are admitted for other than academic prowess?</p>
<p>I don't think Harvard has 400 slots (from 0-24%) for students who underperform academically. Let's say that number is 12%. It does not affect the 25% number at all.... doesn't touch it.</p>
<p>So long as a school does not have AT MINIMUM 25% of admitted students through athletic/legacy/etc., the test scores of the 10% or 12% that are those students is irrelevant to the 25% score. they could be 800 (Math + CR) SAT scorers and it doesn't touch the 25th% student's score. These are not averages, they are ordinals.</p>
<p>For those who don't know, the OP is using data for freshmen class ENROLLED (not admitted--big difference!!) in 2007. The "2008" USN is using 2006 data.</p>
<p>DunninLA...
Princeton's student body for one has 20% varsity athletes and a typical 14% of the class as legacies. The 20% is likely a good proxy for recruited. A very few may be walk-ons, but some recruits drop out because of injury and others from disinterest- it doesn't affect their pocketbooks in a no athletic scholarship world. Then add in the accomplished, world class musicians, dancers, actors, political and celebrity offspring, faculty and administrator kids, and the occasional big time donor's kid, and IMO you are over 40% already. Not even counting URMs.</p>
<p>Data for varsity athletes can be found at the US Government Office of Postsecondary Education. On the web-</p>
<p>ope.ed.gov/athletics</p>
<p>Varsity athletes as a percentage of the student bodies at some Ivies-</p>
<p>Princeton 20.79%
Dartmouth 18.76%
Yale 15.49%
Brown 14.97%
Harvard 12.54%</p>
<p>I don't know if the "top" students necessarily hang out together. I have no idea what any of my friends' SAT scores were, and I don't need to find out. I'm sure some were below and above me, but it doesn't matter because SAT scores do not really predict performance (A math prof even showed that verbal SAT scores were inversely correlated to GPA). Really, a school is only as strong as its weakest students.</p>
<p>14.7% of the Princeton Class of 2011 are legacies.</p>
<p>
[quote]
I don't think Harvard has 400 slots (from 0-24%) for students who underperform academically. Let's say that number is 12%. It does not affect the 25% number at all.... doesn't touch it.
[/quote]
Based on how people I know fared in the admissions game this year...you might be surprised. They were all very interesting individuals, but to call them the academic cream of the crop would be misleading.</p>
<p>
[quote]
I don't know why your selection of 75% has any more utility than looking at the 25%. Is there a way in which you and the other top 25% can practically segregate yourselves from the rif-raff bottom 75%?
[/quote]
</p>
<p>Yes, there is a way. It's called majoring in science or engineering.</p>
<p>
[quote]
So long as a school does not have AT MINIMUM 25% of admitted students through athletic/legacy/etc., the test scores of the 10% or 12% that are those students is irrelevant to the 25% score. they could be 800 (Math + CR) SAT scorers and it doesn't touch the 25th% student's score. These are not averages, they are ordinals.
[/quote]
</p>
<p>You're forgetting URM's, who are held to much lower standards than other non-athlete, non-legacy students. Your argument is valid, but it is based on an incorrect assumption (namely, that athletes and legacies are the only ones dragging down the 25th percentile SAT averages).</p>
<p>
[quote]
I don't know if the "top" students necessarily hang out together. I have no idea what any of my friends' SAT scores were, and I don't need to find out. I'm sure some were below and above me, but it doesn't matter because SAT scores do not really predict performance (A math prof even showed that verbal SAT scores were inversely correlated to GPA). Really, a school is only as strong as its weakest students.
[/quote]
</p>
<p>The point of my first post wasn't to diminish the accomplishments of those with lesser SAT scores. Rather, I wanted to show that there is an elite community of intellectuals that is significant in number at nearly all of the top 20 universities. </p>
<p>Too often around CC I see people making tacit assumptions that Ivy League students are inherently smarter than Rice or WashU students, for example. I simply wanted to show that the Top 25% of many Top 20 schools is comparable in raw intelligence to the Top 25% at many Ivy League schools, making ANY ONE OF THESE TOP SCHOOLS an excellent place to pursue an undergraduate education for top students.</p>