Sorry for the awkward title, not enough coffee yet.
This is just a quick study to see which selective schools have the largest and smallest gaps between their 25th/75th percentile test scores. I think it is a reasonable supposition to believe that schools with small gaps care A LOT about test scores, while those with smaller gaps probably weight other factors (academic record) more highly.
The entire list included schools whose admitted students midpoint test scores were above the 90th percentile (102 schools).
(school, average SAT gap, cumulative ACT gap)
The average SAT gap was 94 points, the average ACT gap was 3.6.
Smallest gaps (most “test aware”):
California Institute of Technology 60 2
Washington University in St Louis 67 2
Webb Institute 70 -
Massachusetts Institute of Technology 77 2
University of Chicago 77 3
Washington and Lee University 77 3
Franklin W Olin College of Engineering 77 3
Wake Forest University 80 2
Northwestern University 80 3
Tufts University 80 3
Honorable mention (had significant numbers of students without submitted test scores)
Bowdoin College 77 3
And because someone is sure to ask:
Vanderbilt University 83 2 (in a tie for 14th place with a bunch of other schools)
Largest gaps (least “test aware”):
University of California-Los Angeles 143 7
University of California-Berkeley 133 6
Cooper Union for the Advancement of Science and Art 130 4
Grinnell College 125 4
Bryn Mawr College 120 5
Rose-Hulman Institute of Technology 120 5
Brandeis University 113 4
Williams College 110 4
University of Maryland-College Park 110
Dartmouth College 107 4
My assumption is that the two UC schools at the top of the list indicate how much weight the UC system places on GPA.
Also, some of the data above is from 2013 (including the Wake numbers) and some is from 2014…Wake would likely fall off the list using the current year data.
There are plenty of test optional schools that seem to be pretty “test aware” and plenty of them see large percentages of students submitting tests despite the policy.
Yes, but typically only the high-scoring kids submit scores, since they know the scores will help them. Low and often mid-range scores just aren’t submitted. If only high-scoring kids submit scores, then you have a small gap. What would be interesting to see would be the scores of all accepted students, not just those who submitted test scores. Then you’d likely see a much larger gap.
You see a strong representation of colleges on the top end whose main emphasis is STEM (Caltech, MIT, and Olin, for instance). One theory I have is that these schools have a much higher proportion of students in impacted majors (as many STEM majors are impacted), so the top end of applicants will tend to have pretty high scores, probably arguably above the point where score differences are very meaningful (760 vs 800 etc), so then the admission committees have to turn to other measures to distinguish amongst the students. But there is probably some baseline level that they want to see the students score above. So I think there may be some schools that care about higher scores at every level (for instance, to raise their averages because they want to climb the ranks?), versus others that rank high in part because of structural reasons.
That said, there are some STEM-focused schools on the list with the largest differences (Rose-Hulman, Cooper).
Yeah, I thought it was interesting that there were STEM focused schools on both lists.
Except for the first 3 or 4 schools on both lists, the differences aren’t really all that large, and might not be very significant. I think if you really wanted to learn something concrete about any tendencies of the schools, you would have to study the numbers over a period of years. In any particular year, a college just might over or under shoot its target numbers with a particular incoming freshman class.
This is a very interesting thread, and I’m curious what others make of the data. I think the ranges for schools is a valid way to measure the importance of standardized test scores. After all, a school like Stanford could accept all 36 ACT scorers, but instead its range is 31-34. I disagree with UC being put in the wide range pool. By my understanding, the UCs value GPA/SAT as, essentially, the determinants of your admissions fate, right?
You could also look at the CDS for each school to see how they rank the importance of “Standardized Tests”.
Caltech, WUSTL, Northwestern and Tufts give it the highest rank - “Very Important”, MIT only “important” - but so is everything else, nothing is “most important” there, at U Chicago and Wake it’s only “considered”, not even “important”. W&L and Olin only “important”.