<p>Some posters claim that SAT scores don't always give an accurate picture of the standardized test achievement level of the students at certain schools. Are there great discrepancies in the SAT and ACT results? I looked at 87 colleges that provided data for SAT and 72 that provided data for ACT. My conclusion is that there are not great differences when you compare the numbers and national standings of SAT and ACT scores. </p>
<p>Here is the data and the ranks that I was able to find on collegeboard.com:</p>
<p>SAT rank, school, SAT level</p>
<p>1 Cal Tech 1520
2 Harvard 1490
3 Yale 1485
4 Princeton 1480
5 MIT 1470
6 Duke 1465
7 Dartmouth 1450
7 Wash U StL 1450
9 Stanford 1440
9 Brown 1440
11 Rice 1435
12 U Chicago 1425
13 U Penn 1420
13 Columbia 1420
15 Northwestern 1410
15 Tufts 1410
17 Carnegie Mellon 1395
18 J Hopkins 1390
18 Georgetown 1390
20 Cornell 1385
21 Vanderbilt 1370
21 USC 1370
23 Brandeis 1360
24 Emory 1350
24 W & M 1350
26 Notre Dame 1345
27 Wake Forest 1340
28 Boston College 1335
29 Case Western 1330
30 UC Berkeley 1325
30 U Rochester 1325
30 U Virginia 1325
33 Tulane 1323
34 Rensselaer 1320
35 U Michigan 1315
35 Georgia Tech 1315
37 NYU 1310
38 Lehigh 1305
39 UCLA 1295
39 U North Carolina 1295
39 Worcester 1295
42 U Illinois UC 1285
42 G. Washington 1285
44 U Wisconsin 1275
44 Boston Univ 1275
46 U Miami FL 1270
46 American U 1270
48 SUNY-Bing 1260
48 Stevens Inst 1260
50 UCSD 1250
50 U Florida 1250
52 Pepperdine 1240
52 U Minnesota 1240
54 Yeshiva 1235
54 BYU 1235
56 U Maryland 1230
56 SMU 1230
58 U Texas 1225
58 Syracuse 1225
58 Clemson 1225
58 U Georgia 1225
62 U Pittsburgh 1220
63 Rutgers 1210
64 Miami U OH 1205
64 Baylor 1205
66 Ohio State 1200
66 U Delaware 1200
66 Clark Univ 1200
69 Va Tech 1195
69 Fordham 1195
69 U Washington 1190
69 UC S Barbara 1190
69 U Iowa 1190
69 U Connecticut 1190
69 St. Louis U 1190
69 Iowa State 1190
77 UC Irvine 1185
77 Texas A&M 1185
79 Penn State 1180
79 NC State 1180
79 Marquette 1180
82 U Colorado 1170
83 UC Davis 1155
84 UC S Cruz 1145
85 Purdue 1135
86 Michigan St 1130
87 Indiana U 1120</p>
<p>ACT rank, school, ACT level</p>
<p>1 Harvard 32.5
2 Princeton 32.0
2 MIT 32.0
2 Rice 32.0
5 Duke 31.5
5 Wash U StL 31.5
5 Notre Dame 31.5
8 U Penn 31.0
8 Dartmouth 31.0
8 Northwestern 31.0
11 Stanford 30.5
11 U Chicago 30.5
11 Columbia 30.5
11 Tufts 30.5
15 Cornell 30.0
15 Brown 30.0
15 J Hopkins 30.0
15 Vanderbilt 30.0
15 Carnegie Mellon 30.0
15 USC 30.0
21 Georgetown 29.5
21 Brandeis 29.5
21 W & M 29.5
24 Emory 29.0
24 U Michigan 29.0
24 NYU 29.0
24 Tulane 29.0
28 U Virginia 28.5
28 U Rochester 28.5
28 Case Western 28.5
28 U Miami FL 28.5
32 U Wisconsin 28.0
32 Georgia Tech 28.0
32 G. Washington 28.0
35 U North Carolina 27.5
35 U Illinois UC 27.5
35 Worcester 27.5
35 BYU 27.5
35 American U 27.5
40 UCLA 27.0
40 Rensselaer 27.0
40 SUNY-Bing 27.0
40 Boston Univ 27.0
44 U Florida 26.5
44 Syracuse 26.5
44 Pepperdine 26.5
44 Ohio State 26.5
44 U Georgia 26.5
44 Miami U OH 26.5
44 Clemson 26.5
44 SMU 26.5
44 St. Louis U 26.5
44 Marquette 26.5
54 UC SD 26.0
54 U Texas 26.0
54 Fordham 26.0
54 Clark Univ 26.0
58 U Washington 25.5
58 Texas A&M 25.5
58 U Delaware 25.5
58 U Minnesota 25.5
58 U Colorado 25.5
58 Baylor 25.5
64 UC S Barbara 25.0
64 Purdue 25.0
64 U Iowa 25.0
64 U Connecticut 25.0
64 UC S Cruz 25.0
69 Iowa State 24.5
69 Indiana U 24.5
69 Michigan St 24.5
72 NC State 23.5</p>
<p>Schools without ACT data</p>
<p>Yale
Cal Tech
UC Berkeley
Wake Forest
Lehigh
Boston College
UC Irvine
Yeshiva
UC Davis
Penn State
U Maryland
U Pittsburgh
Rutgers
Stevens Inst
Va Tech</p>
<p>You can't determine the median (or mean, or course) by simply splitting the difference between the 1st and 3rd quartiles. So if that's how these numbers were derived, statistically they're sort of meaningless.</p>
<p>Please check with the websites or the college admission office of the school of your choice concerning SAT scores. Also, the scores that students who will be applying to various schools need to be most interested in are those of the admitted students. The average SAT/ACT score for the students who accept admission may vary from year to year. One year they may be several points higher, and the next, several points lower. You need to concern yourself with the numbers you need to get admitted. In terms of accuracy, even the college board, as in the numbers above, is a secondary source to the actual college/university. Go there for the most accurate information.
Example: Hopkins, '06 scores for admitted students: 1440 SAT; 32 ACT.
Please see: <a href="http://apply.jhu.edu/facts/facts.html%5B/url%5D">http://apply.jhu.edu/facts/facts.html</a>
Once again, please check with the individual schools.</p>
<p>"You can't determine the median (or mean, or course) by simply splitting the difference between the 1st and 3rd quartiles. So if that's how these numbers were derived, statistically they're sort of meaningless."</p>
<p>No, but you can get close enough for most purposes.</p>
<p>
[quote]
My conclusion is that there are not great differences when you compare the numbers and national standings of SAT and ACT scores.
[/quote]
</p>
<p>Except for the fact that Midwestern schools, overall, tend to do better when looking at ACT scores as opposed to the SATs (which makes total sense).</p>
<p>Both the admitted and the enrolled student numbers are useful. A college that admits lots of high scoring kids who do not attend will have a low yield and a large difference between admitted and enrolled scores. For a prospective student, knowing the academic background of their potential classmates (enrolled stats) is quite useful. Since all the enrolled students were admitted, the enrolled scores is a representation of what it takes to get in.</p>
<p>Like Barrons says, midpoint is a crude estimate, but good enough for government work.</p>
<p>SAT ACT comparison may be misleading, since at many colleges only a small proportion of students present the ACT. The ACT numbers may therefore be more volatile at such places.</p>
<p>
[quote]
hawkette, I realize that you did it just to come up with some rough numbers, but be careful about splitting the difference between the first and third quartiles (25th percentile and 75th percentile) to derive some sort of median (or "average", as you called it). The split difference isn't really statistically meaningful since, among other things, it doesn't account for the actual distribution of students within the middle 50%, or for the total "spread" of that group. Also, in most--if not all--cases, it won't give you either the true median or the true mean SAT scores for that school.
And no, barrons, it may not be "close enough for most purposes." The distribution of students within that middle-50% range can vary widely from school to school, greatly undercutting the usefulness of the split-the-difference approach. For example, 2 schools could both have a middle-50% range of 1300-1400, but at school "A" most of the students in that range are clustered nearer to 1300, with most of its students in the upper quartile (above the 75th percentile) clustered near 1400. Whereas at school "B" most of the students in the middle-50% are clustered closer to 1400, with the upper quartile distributed more evenly from 1400 to 1600. If we just split the difference of the middle-50% range, the number for both schools would be 1350. However, that would be very misleading, since in fact the median scores of the 2 schools would be very different, as would their mean scores.</p>
<p>Sorry, in the real world it does not work that way for any large school. I have looked at many that have both scores available and it is close enough.</p>
<p>This is perhaps accurate. Nevertheless, if one is applying, that person needs to know the stats for admitted students. In any particular year, those who choose to attend may or may not have lower stats than those for the admitted students. It depends upon the year. In terms of the figure for enrolled students, college board.com may very well be correct. To be safe, though, if one is going to draw conclusions based on that number, call the school or check the website, just to be sure.</p>
But it DOES work that way for the Ivies and other elites. Unless you can demonstrate to the contrary with actual SAT medians and means for those schools (which they generally don't release).</p>
<p>which gives SAT data in the form of what % of students have scores above the 500, 600 and 700 centiles. It appears to come from common data set information. For OP's list of top 20 schools based on average SAT, the % of students scoring above 700 mark in CR and Math, respectively, is as follows:</p>
<p>I just finished reading the very comprehensive article about Hopkins on the Wikipedia site. For those interested, it is very detailed and informational. Also, the stats for the class of 2011 are included in the info. I would imagine that would be the case for all colleges and universities, so if you want a heads up on the new stats that will be published in August, check out the articles on your schools of interest. Again, any stat should be checked out by contacting the school, to double check if contemplating applying.</p>
<p>Well, again, that's the data on the admitted students. The average of those who then choose to* enroll* will be different. For last year's JHU freshmen, the ACT midrange was reported to be 28-32; Verbal 630-730, Math 660-760, yielding the numbers Hawkette posted. It's always important to recognize which set of numbers you're dealing with.</p>
<p>Oh, gosh. We are talking about the stats one might need in order to be admitted in subesequent years. That is what is relevant to students who want to apply. Hopkins, nor any other school, has pre-knowledge concerning which students will accept admission. And again, I would encourage anyone interested in the enrolled stat to check with the school. These numbers are apparently Hopkins' current mid standard for admitted students.</p>
<p>You're right, of course, Gabriellaah - but the most widely and consistently reported data (via the common data set) is the stats of the most recent freshman class. There is no comparable set of similarly calculated numbers for total admittees at a large number of schools. If you go by the average numbers reported for the admittees at one school and compare it to those of the enrollees at another you'd be comparing apples to oranges. As long as the data is consistent, you can use it for guidance.</p>
<p>And actually Hopkins - like all Universities - has a pretty good idea how many students in each band of admittees will enroll and how many will go elsewhere. That's how they admit (approximately) the right number of excess applicants so as to arrive at the (approximately) right number of actual matriculants. Of course, every once in a while they screw up.</p>