<p>
</a>
For some years, yes, but other years illustrate my point. For example, for the SAT-V in 1989, the 25th percentile was 600 and the 75th percentile was 710, leading to a split-difference of 655, whereas the actual median was 670. Close, yes, but in a ranking such as put forth in this thread, using the split-difference will lead to inaccurate ranking.</p>
<p>I recently did an analysis similar to what ohmadre did above. For the USNWR Top 30 ranked colleges, here is how they stack up compared to one another for the % of enrolling freshman who scored above 700 on the SAT Critical Reading or Math sections.</p>
<p>Critical Reading</p>
<p>1 Yale 78%
2 Cal Tech 77%
3 Princeton 73%
3 Harvard 73%
5 MIT 68%
6 Columbia 67%
7 Dartmouth 65%
8 Brown 64%
9 Duke 63%
10 U Chicago 61%
11 Wash U StL 60%
12 Stanford 59%
13 Rice 57%
14 Tufts 56%
15 U Penn 54%
16 Northwestern 53%
16 Georgetown 53%
18 Notre Dame 47%
19 Vanderbilt 41%
20 J Hopkins 40%
21 Cornell 38%
22 USC 36%
23 Carnegie Mellon 33%
24 Emory 32%
25 UC Berkeley 31%
25 U Virginia 31%
27 Wake Forest 29%
28 U North Carolina 22%
29 U Michigan 21%
29 UCLA 21%</p>
<p>Math</p>
<p>1 Cal Tech 96%
2 MIT 92%
3 Yale 78%
4 Wash U StL 76%
5 Princeton 74%
5 Harvard 74%
7 Carnegie Mellon 71%
8 Dartmouth 70%
9 U Penn 69%
9 Duke 68%
11 Stanford 67%
12 Brown 66%
13 Northwestern 63%
13 Rice 63%
15 Columbia 60%
15 J Hopkins 60%
17 Cornell 59%
17 Tufts 59%
19 U Chicago 57%
19 Notre Dame 57%
21 Vanderbilt 52%
22 Georgetown 51%
23 USC 50%
24 UC Berkeley 46%
25 Emory 45%
26 U Michigan 43%
27 U Virginia 40%
28 UCLA 39%
29 Wake Forest 38%
30 U North Carolina 29%</p>
<p>
Unfortunately, even these admitted ranges give a rather incomplete picture of a particular student’s chances for admission, since we don’t know the depth of a particular school’s applicant pool in any given score range. Penn is one of the few school’s I’ve seen that actually provides a bit of a window into this parameter, and is therefore instructive. Per Penn’s web site, the middle-50% range for SAT-M scores for students admitted to the Class of 2010 was 680-780, perhaps leading one to believe that a student with a score of 780 (being in the 75th percentile of admitted students) had a better than even chance of being admitted. Unlike most other schools, however, Penn also reveals the percentage of students with math scores in the 750-800 range who were admitted, which was only 22% (i.e., only about 1 out of 5 students with a SAT-M score in the 750-800 range was admitted):</p>
<p><a href=“http://www.admissionsug.upenn.edu/applying/profile.php[/url]”>http://www.admissionsug.upenn.edu/applying/profile.php</a></p>
<p>If other schools revealed this type of data, it would be much more helpful to potential applicants attempting to guage their chances based on SAT scores.</p>
<p>Hawkette, the post # 22 about the % of enrolling freshman who scored above 700 is great. Can you do it for scores above 600 and 650? Thank you.</p>
<p>hawkette, for what it’s worth, some of those numbers appear to be a few years out of date. Not that big of a deal, but it probably changes the ranking a bit (especially if some numbers are old and others are more recent).</p>
<p>45: Interesting. Thanks for the insight.</p>
<p>Kluge: I understand what you are saying. My concern is that a student who is seeking admission to a particular school might look at enrolled stats that make it seem that he/she has a better shot at being admitted than he/she actually has. Wasting an early decision application, for example, can be so detrimental, because we all know what it is like for kids to be thrown into the regular pool, these days. I just want to urge students to call the various admissions offices and get the exact data they are looking for, directly from the source. This goes for whether or not they are basing their decisions on “admitted” stats, or “enrolled” stats. I am not so sure that collegeboard.com is actually posting the most accurate or most recent numbers. It is always best for them to go right to the source.</p>
<p>^ I second what gabriellaah said. There are a lot of numbers flying around this Board, as well as sites like collegeboard.com, Yahoo Education, US News, etc. If applicants really want to compare statistics, they should ensure that they’re comparing accurate statisitics, and that they’re comparing apples to apples (accepted vs. enrolled, most recent data vs. several-year-old data, median vs. mean, etc.). More importantly, they should understand the significance of the data (or lack thereof), and not make additional empirical inferences that are not statistically justified (as, e.g., I illustrated in post #23 above).</p>
<p>Of course, the statistics are only a starting point, and we all know that there’s A LOT more that goes into choosing the right college or university.</p>
<p>I think the error in assuming that an applicant to Penn with SAT scores above the 75th percentile has a “better than even” chance of being admitted isn’t in the correct calculation of the 75th percentile, but in failing to understand that even competitive applicants are never more than 50% assured of being admitted to any of the ultra-competitive schools. A person with a SAT-M score of 800 is certainly above the 75th percentile, but that fact alone does not mean that person has a “better than even” chance of admission based on that fact alone.</p>
<p>I think we all agree that understanding where an applicant falls in relation to benchmarks which are consistent from school to school is a valuable tool when used in conjunction with an understanding of other factors which bear on the issue: overall selectivity, weight of different factors such as GPA, EC’s, etc. Outside of that framework it is of little help, however.</p>
<p>
kluge, you may have misunderstood the point I was trying to make, but we’re actually on the same page. I wasn’t addressing the correct calculation of the 75th percentile, but rather the fact that a kid in that range still has only a 1 in 5 chance of being accepted, which Penn reveals in a table on the page to which I linked, but which most of the other schools would not specifically reveal (instead merely providing the 1st and 3rd quartile scores). Intuitively, we all know that the top schools reject a lot more kids with high scores than they accept but, as I said, Penn is the only school of which I’m aware that gives some specific numbers for this. There may be others, but I’m just not aware of them.</p>
<p>The University of California does this on a per-hundred point score basis for its nine campuses here: <a href=“http://www.universityofcalifornia.edu/admissions/undergrad_adm/selecting/camp_profiles.html[/url]”>http://www.universityofcalifornia.edu/admissions/undergrad_adm/selecting/camp_profiles.html</a>
So you can see, for example, that 40.2% of applicants to UC Berkeley with SAT-M scores over 700, and 20.9% of those between 600 and 690 are offered admission, etc. They break it down in a number of helpful ways; GPA, ACT, SAT, etc.</p>
<p>gabriellah,
Actually, I think the data on collegeboard.com is pretty accurate. From the schools that actually publish their CDS, I think I have seen enough of these that I am comfortable that collegeboard.com’s numbers are taken directly from a school’s CDS. Frankly, I am little less confident about some of the colleges. Marketing being what it is, more than a couple of schools will post 25/75 data on their websites for ADMITTED students without stating that it is for admitted, not enrolled, students. This heightens the public selectivity image of the school, but it is not an accurate picture of who is actually attending.</p>
<p>Re 45 percenter’s comment about yahoo data, I think you are right that this data is older and taken from the 2005-06 CDS. I don’t know when yahoo updates their data set.</p>
<p>t1388,
Here is the data you requested about the percentage of students who scored above 600 on the Critical Reading and Math sections (info drawn from Yahoo Education):</p>
<p>For Critical Reading</p>
<p>1 Wash U StL 98%
2 Yale 97%
3 Princeton 96%
3 Harvard 96%
5 MIT 95%
5 Tufts 95%
7 Cal Tech 94%
7 U Penn 94%
7 Duke 94%
10 Dartmouth 93%
10 Columbia 93%
12 Stanford 92%
12 Northwestern 92%
12 Brown 92%
15 U Chicago 91%
15 Rice 91%
17 Vanderbilt 89%
17 USC 89%
19 J Hopkins 88%
19 Notre Dame 88%
19 Georgetown 88%
22 Emory 87%
23 Wake Forest 85%
24 Cornell 84%
25 Carnegie Mellon 81%
26 U Virginia 79%
27 U North Carolina 73%
28 UC Berkeley 72%
29 U Michigan 70%
30 UCLA 66%</p>
<p>For Math</p>
<p>1 Cal Tech 100%
1 MIT 100%
3 Wash U StL 99%
4 Princeton 98%
4 Harvard 98%
4 Yale 98%
7 U Penn 97%
7 Carnegie Mellon 97%
9 Stanford 96%
9 Tufts 96%
11 Duke 95%
11 Columbia 95%
11 Brown 95%
14 Dartmouth 94%
14 Northwestern 94%
14 Notre Dame 94%
17 Rice 93%
17 Emory 93%
17 Vanderbilt 93%
20 U Chicago 92%
20 J Hopkins 92%
20 USC 92%
23 Cornell 90%
23 Georgetown 90%
25 Wake Forest 88%
26 U Michigan 86%
27 U Virginia 85%
28 UC Berkeley 81%
28 U North Carolina 81%
30 UCLA 77%</p>
<p>Hawkette, thank you.</p>
<p>Re your post #29, Kluge, what you say is absolutely true.</p>