In a different thread@merc81 shared a link to this listing of colleges in America, which is essentially a ranking of 610 colleges based on their average SAT score from 2015 (i.e., before test optional became a major thing).
It combined the categories of national universities & liberal arts colleges and regional universities and college all into one listing (and used correspondence charts to convert ACT scores to SAT scores).
Although there are certainly issues with standardized test scores (and with the naming of the schools as the “smartest”), I think there are many that view the scores as a way of determining the quality of a school. And although there are some schools that may have had some strong surges up in scores over the last 10 years (perhaps Tulane and Northeastern?) I suspect that the vast majority of these schools are as similarly-situated to one another now as they were then.
I find this “ranking” interesting for several reasons:
It shows how minute the differences are between “ranked” schools. The way some people talk about schools is as though #20 is so different from a #3 school, when in reality, the differences are very small (i.e. all of which put someone in “range” of a school and will not eliminate someone from consideration). Similarly, there are 15 schools ranked at #139.
Sidebar: I wish the list was sorted as a table of SAT scores rather than a ranking, because a regular person might think that a school ranked #135 has some marked differences over one ranked #154…except that it’s the difference between a 1250 and a 1240 (i.e. pretty much no difference). If people would say, I’m looking for schools where the SAT range is between 1200-1300, then they’d have a huge list of schools vs. if they go in saying, I want a school that ranks in the top 100
The other part of the high end that was surprising to me was Cal Tech whose SAT score was 40 points higher than the #2 school, the biggest difference I saw anywhere on the rankings.
But as people who may have seen my posts elsewhere on the forum know, I like to see which schools don’t seem to get as much attention on CC as some of their similarly-ranked brethren. (Note: When I speak of “rankings” it just happens to be the number on this list. This is simply a listing of schools’ average SAT scores and schools are more than just their score. This is just one data point.)
So, with all that prefacing, on to some of the schools that caught my eye:
Reed (OR) & Cooper Union (NY): same SAT as Colgate, 5-points below USC and Michigan, and 5-points above Emory, Case Western, and William & Mary. Not mentioned nearly as often on CC as any of those.
Brandeis (MA): Same score as Boston College & Middlebury, within 5-points of Macalester, Case, Emory, & William & Mary and higher than Tulane and UVA.
Wheaton (IL): Same score as Lehigh, SMU, U. of Miami and within 5 points of Kenyon and U. of Richmond.
Whitman (WA): Same score as Wake Forest and within 5 points of Lehigh & Villanova.
U. of Tulsa (OK): Except when talking about National Merit Finalists, Tulsa gets little recognition here. It has the same score as Bucknell, College of the Holy Cross, Union, Franklin & Marshall, and St. Olaf.
Brigham Young (UT) and Hillsdale (MI): Same score as UC-San Diego, Boston U, Lafayette, Occidental, Rhodes, and Santa Clara.
Centre (KY): Same score as Gettysburg, Denison, and Lewis & Clark and 5-points above Dickinson.
New College of Florida: This school is undergoing some upheaval, but its score was same as Dickinson and within 5 points of Lewis & Clark, Denison, and Gettysburg.
Hendrix (AR) & Thomas Aquinas (CA): Same score as Trinity U. and higher than Fordham & U. of Denver.
UT-Dallas: This one is often only mentioned for National Merit Finalists, but has the same score as Clemson and is higher than U. Conn, Stony Brook, and Miami-Oxford.
#139: American, Miami (OH), U. of Denver, NC State, Florida State, Saint Louis, Michigan Tech, Trinity (CT), College of Wooster, Kalamazoo, Illinois Wesleyan, Transylvania, Butler, Drake, Milwaukee School of Engineering
Yeshiva (NY): Same score as Skidmore and Bentley and within 5-points of Butler, U. of Denver, U. of Georgia, and Furman.
There is so much more on the list, but I figure I should let some other folks have some fun with it. Looking forward to hearing what others found interesting on this list!
Fixation on 5’s in standardized test scores is just laughable. Any college above about #50 or so would have absolutely no difference in the overall academic ability of the students in the classes at the various schools.
I love data, support the use of standardized tests in admissions, and find it helpful for context, but I have enough training with both quantitative and qualitative research to know that single metric scrutinized to the 5th point is simply meaningless above a certain threshold. One would need to account for other factors to determine exactly what that threshold is. I’m not in a position to do that analysis, but I live surrounded by anecdotal data to support this hypothesis.
If I am understanding your message correctly, I agree with you. I don’t think that the 5-point difference between schools is meaningful, and I don’t think anyone else would, either. That’s why I used a 5-point measurement as an example of comparable schools, because pretty much nobody would think that there was a difference between those schools. And yes, a school with a 1350 vs a 1345 vs a 1340 should essentially all be considered the same. I don’t think that dilutes the worth of looking at what schools in those ranges get far less attention than others in a similar score range.
As to the reason I’ve posted this analysis occasionally, it allows students to compare colleges with which they are already familiar to those that they may not have previously known. If the list is considered not as a ranking, as such, but as a tool for exploration, it has the potential to help students match themselves to additional suitable colleges during the search and selection process. The drawback may be that the concept of the analysis suggests several attendant caveats. However, @AustenNut did a nice job of including some of these caveats in the original post.
I think we are getting at the same thing from different perspectives. To illustrate, what I mean, I will pull from your highlights:
Hillsdale vs. BYU vs. Occidental. Same average SAT score, but vastly different schools with dramatically different class constituencies. I could see myself as a high school student choosing Occidental, but not the other two, and my SAT score was much higher than the average listed here, but I still would have preferred Occidental over, say, Columbia for undergrad. Not knocking Columbia- it’s a great school. It’s a question of fit, and there is much more to fit than score deciles, let alone even more dissected score intervals.
I completely agree with you and I definitely think that fit needs to come into account, as several of those schools mentioned are very much fit schools. But how many times does Yeshiva get mentioned for a family where it’s important to have a sizable Jewish population? Or if we take a look at all the schools at #139, people interested in NC State and Florida State would be very unlikely to be interested in Illinois Wesleyan. But as @merc81 mentioned, it’s a tool to explore what other schools might be of interest. For instance, if I’m interested in College of Wooster and Kalamazoo, then maybe I’d also want to look at Illinois Wesleyan and Transylvania.
First off, the data was for 2015, prior to the time that the majority of colleges went test optional. But, even back then, there were schools that were TO, as demarcated by a number next to its ranking (whether it was TO, or didn’t ask all students to report data, or itself didn’t report all scores, etc). Thus, I didn’t mention schools with a footnoted number as any of the “surprise” ones, since one could argue that the test reporting policies helped with that (though I did include some schools as comparison points, if those schools were more popular ones on CC).
I agree that looking at these numbers is an interesting exercise. Some things that stand out to me:
I agree that it is striking how much higher Caltech is than the rest, even than MIT which seems like a similar school in so many ways. I suspect it comes down to Caltech not recognizing hooks essentially at all. I also noticed that MIT and Harvey Mudd are neck and neck (but MIT is seen as the “dream school”, while HMC is unknown except to those “in the know.”)
Surprised that the averages of many “elite” schools are as low as they are. I think there are plenty of posters on CC or Reddit a2c who would feel that any score under a 1500 would be a liability. But only 5 schools have an average at or above 1500. Again I think this shows that most elite schools really do practice holistic admissions. The SAT score was not as big of a piece of the puzzle as people imagined, even back before TO.
Interesting that certain religious schools and Midwest schools seemed to punch above their weight.
It makes sense to me that schools that are mostly STEM focused, like Cal Tech, MIT, Mudd, etc… skew much higher. If you take the public flagships and look at their stats for CoE, the numbers are much higher than the university as a whole.
Please note that the SATs have been recalibrated (as well as changed in some fundamental ways) since the time of the BI analysis linked in the original post. I believe an average score now is inflated by about 60 points in relation to that at the time of the BI analysis.
The interesting question is why that’s the case. The math portion of SAT is commonly regarded as “easy”, or even “trivial”, and the verbal portion more challenging. One would think since SAT math scores matter more than verbal scores to STEM schools (and engineering colleges in other schools), the average scores for these colleges would be lowered by verbal scores which they, as well as their STEM students, paid less attention to.
Caltech is tiny – only 1k undergrads – and they don’t care about hooks, so its easy for them to select the highest scoring Math and Verbal students.
Percentiles have shifted. As an example, a score of 1300 currently represents the 86th percentile. Circa 2015, the 86th percentile was reached with a score (verbal + math) of about 1230.
For example, USC and UCLA are widely viewed as being comparably selective for admission overall. However, USC weights test scores more heavily than grades compared to UCLA, which is reflected by its higher average SAT score in the linked page (1380 versus 1325), and also by its lower unweighted HS GPA shown in the frosh profiles of both schools.
Colleges that lure lots of NMFs with scholarship money are (by definition) luring those who do well at standardized tests similar to the PSAT (like the SAT). So it is no surprise that such colleges tend to have a higher SAT average than their general admission selectivity may suggest.
Another example is University of Alabama, whose 1205 SAT average is the same as that of Purdue University - West Lafayette, even though no one around here thinks that Alabama is anywhere near as selective in admissions as Purdue. But Alabama lures high test scorers and NMFs with large scholarships, which boosts its test score profile.
Caltech has a reputation of being hard, so it likely has a self-selected applicant pool. In addition, its admissions is (compared to other highly selective colleges where “hooks” are weighted more heavily while admitting larger classes) heavily biased toward looking for academic strength through indicators beyond the usual ones of SAT/ACT and HS GPA. It is likely that those who have that level of academic strength indicators to get admitted to Caltech find the SAT to be an easy 800 math and 700+ V/CR/EBRW (probably with no or minimal test prep).
However, most posters are unhooked, so they may be assuming that SAT/ACT scores and other academic indicators much be significantly above average to gain admission out of the unhooked applicant pool – especially at smaller colleges where the recruited athletes consume a larger percentage of the class than at larger colleges.