The report doesn’t say how they combined popularity with SAT scores, so there’s not much information there. Such lists have to “look right” when they’re published.
The ranking is based on “acceptance rate” & “SAT scores” for each school.
This list can be helpful to those unfamiliar with college admissions–especially so for foreign applicants.
For example, this list helps illustrate that the 8 Ivy League members are not necessarily the ultimate in US colleges and universities as selectivity is an important factor to consider.
I find it interesting to note the prominent schools which are not on the list.
Regardless, some will find this information useful and interesting, while others may not.
P.S. Also unsure as to whether or not you think that this list “looks right”. Stanford University, for example, is ranked at #13 despite an admit rate below 5% which suggests that admission is based on more than just one’s numbers.
Also, does it look right that the top two LACs as ranked by US News are ranked at #18 & at #19 ?
Rice may be a bit of a surprise to some at #6.
Where is Bowdoin College ? Middlebury College ? Bates ?
Regardless, the list offers an interesting perspective as well as an introduction to some schools which may not have been known by some readers.
Julliard and Minerva are notably absent from the list. Perhaps they don’t consider SAT scores, but then again, some that ARE on the list are test optional.
This primarily looks like a list of highest 25th + 75th SAT scores from back in 2017. Acceptance rate seems to have little influence. If you want a list of colleges with highest SAT scores, why not use SAT scores from a more recent year? IPEDS can be used to easily rank for 2018-19, and most of the listed colleges have published 2019-20 in their CDS or elsewhere.
@Data10: I agree. Got the link to this list from another thread concerning “most diverse colleges”. The list claims to have been updated on January 15, 2020 from its original September, 2019 publication date.
Nevertheless, I find this list interesting as it prompts me to re-evaluate the deeply ingrained impression left by decades of exposure to US News rankings.
Differences in acceptance rates were sufficient to separate colleges by at least seven places in relation to their comparative standardized scoring profiles.
Those 7 places likely corresponded to only a small difference in SAT score. It’s still “primarily” based on 2017 SAT score, rather than admit rate, as stated in my earlier post.
For example, the 5 colleges with the highest 25th/75th SAT score in 2017 were Caltech, MIT, Rice, Chicago, and Harvard. This looks very similar to the top 5 “hardest to get into list.” It’s not identical. Rice was ranked #6 hardest to get in to, rather than among top 5. However, it’s a very similar list.
In contrast, if you look at the lowest acceptance rate colleges for 2017, the ranking order is completely different. For example, Stanford had the lowest acceptance rate, yet was only #13. Princeton and Columbia were also among top 4 lowest acceptance rate, yet were only #10 and #16 respectively. SAT scores appear to have far greater weighting than admit rate.
The problem with acceptance rate by itself has always been the lack of information on the underlying applicant pool. A school with higher acceptance rate but a stronger applicant pool may still be more selective and harder to get into. The publisher of this list obviously uses SAT scores as proxies for the strength of the applicant pool, along with the distribution of SAT scores among all test takers to come up this ranking.
Re #10, the ranking linked in the original post weighs two aspects of selectivity (standardized scoring profiles and acceptance rates) to approximate general comparative selectively. Some selectivity rankings use fewer metrics (such as acceptance rates or standardized scoring profiles), others use more. U.S. News, for example, considers the two metrics of the linked analysis plus the HS class standings of enrolled students. The Princeton Review, creatively, adds the percentage of students originating from out-of-state as a metric. A truly sophisticated ranking would adjust for test optional policies, of course.
Extremely selective colleges tend to have near perfect 75th percentile scores, so the 25th/75th score sum gets dominated by 25th percentile score. The colleges with the highest 25th percentile scores relative to selectivity have a lesser degree of influence from ALDC+URM hooks, and a lesser focus on non-holistic stat criteria. This is not HYPSM… type colleges, so they do worse than would be expected based on selectivity.
Not many are interested in hardest to get into colleges list as majority (including me) can’t qualify for or survive at these places. Its easier to discredit the list and dismiss the topic altogether.
It suggests that Stanford is very popular and gets a lot of applications, hence the “Popularity” comment.
Stanford could get someone to solicit tens of thousands of additional apps from unqualified applicants and drive their acceptance rate down to 2%. It doesn’t make it “harder” to get into. The same students get in.
Or, they could publish a guideline of minimum metrics and cut their applications significantly, to drive their acceptance rate to 20%. That doesn’t make it “easier to get into. The same students get in.
Acceptance rate as a metric is not really very valuable, IMHO.
What’s curious about this list is that it was originally date-stamped on September 06, 2019 at 07:02pm, but was supposedly updated on January 15, 2020 at 09:58am.
So the earliest from which the original list and its numbers cold have taken numbers would have been from the cohort entering college in 2018 because the numbers for 2019’s cohort wouldn’t have been generated by any of the colleges at that time. It’s supposedly been updated, and could have replaced the numbers with 2019’s cohort, but it’s still stuck in 2017, which is terribly outdated.
Of the college’s I’m most familiar, Berkeley’s are way low and USC’s are too, and probably just about all colleges’ upper-half median are low along with some colleges’ lower-half. Meanwhile, UCLA has closed the gap with Berkeley, and neither of them superscore which would add a solid 20-30 points to the median on average for both. (Not 40 which I used to believe because in the three-part, with one more variable component, superscoring (“ss”) would have added a good 60 points on average – 20 points per section, and it wasn’t uncommon to have a 100-point addition in ss over non-ss because of the interplay of superscoring with that extra part).
