Aggregated College Ranking

Applying to College
1

REVISED 9/24: 2 of the 8 data sources had some missing data. I corrected and rescrubbed the entire list. A few changes reflected in the updated list below.

So for kicks and giggles with all the discussion of college rankings, I did an aggregation of 8 rankings into one data set, which I both sorted by average and median (which tended to reduce the impact if there was 1-2 aberrant rankings for a given school). The rankings include US News, WSJ, Niche, Forbes, Washington Monthly, Times Higher Education, College Raptor and the Shanghai Ranking (sorted just for US colleges). A few of these lists only included a limited number of top colleges or started aggregating a range above a certain top threshold, in which cases excluded them from the average and median calculations for those colleges (in other words, some represent 8 rankings others 6-7 ranks). I only include the top 35 here which tended to be on most lists.

To be clear, I fully understand this list is meaningless as a valid rank of colleges. By using a bunch of rankings with different methodologies, many of which have been demonstrated to have compromised data in some cases, or questionable allocations of weighing, this list is kind of a Frankenstein’s monster of dead parts stitched together. So I do not present it as a valid ranking, nor to I endorse any of the rankings that make it up. So no need to debate why its a poor way to rank.

I simply present it for the curiosity factor, and to observe patterns about the impact of which schools tend to be doing well across a broad rang of ranking systems.

I show the top 35 national universities here, and may do an equivalent reply with liberal arts colleges another time.

NU Top 35
NU Top 35900Ă—1546 217 KB

A few observations:

  • HYPMS is very secure as the T5 across almost every list and in aggregation. Though no one school consistently dominated between them. Princeton and Harvard tied for the most 1st ranks (3 each), yet Princeton ended up 4th in both average and median. MIT is the only college that was in the top 5 on all 8 rankings. The averages for Harvard, MIT, Stanford and Princeton were extremely close to each other (all 3.1x-3.3x) with Yale meaningfully behind (5.13).

  • Despite its #12 in the USN ranking, Columbia is #6 by both aggregated ranking measures, with Penn ~1 point behind it. There is a major gap after them before #8 in average. Penn’s average is 8.38 while #8 Caltech and Berkeley are 17.75. (Caltech’s average was hurt by a 47 rank with Forbes.)

  • In general the top publics seem to do better in this aggregation of rankings than they do even after the changes this year to the US News ranking. In particular, UC Berkeley appears to be somewhat polarizing in the rankings. It is a top 10 school in 4 of 8 rankings, including #4 with Shanghai and #5 with Forbes, but is pulled down by a #47 with Niche and #51 with WSJ. Overall it ties for #8 in the median rank.

  • University of Washington made a surprising overall showing on the list, particularly the median where it ranked #29. This was despite a ranking of #134 in WSJ, #40 in USN and #60 in Niche. But it made a comeback with strong showings in the Shanghai and Washington Monthly rankings both of which put it at #14.

  • Rice was surprising at #19 (median) and #34 (average), as was Carnegie Mellon at #31 (median) and #25 (average).

  • U of Virginia was #27 (median) and #25 (average).

4 Likes
2

I’ve always looked at all these rankings in two distinct separate categories: college rankings and global rankings (THL, Shanghai), the same way we look separately at the national universities and LACs and for this reason you left the LACs out of this aggregate rankings. The global rankings are heavily based on faculty research output, so I never found them as meaningful for “college” rankings. So, even for kicks and giggles, it’d have been more consistent and interesting to see the aggregate rankings in three separate categories: college, LAC and global.

3

I debated inclusion of the Shanghai ranking and ran the averages and medians with and without it and it did not have too large an impact on this top list (perhaps because it was only 1 of 8 rankings and I only included the top 39 as beyond that it started bucketing colleges in ranges rather than specific ranks). I also wanted to see what happens when you are methodologically agnostic, which in this case included a counter balance to most of the US ranks which consider cost of attendance and social mobility in some form or other – criteria not of equal importance to everyone. Also in its favor is its consistency – it does not fluctuate methodology as often as the US rankings do, which often comes across as a subjective realignment to conform results to desired outcomes.

1 Like
4

So the global rankings are basically postgraduate rankings as applied to US institutions. I say this because the dominant type of US “elite” college–with general admissions, holistic review, core/distributional requirements, exploration periods before deciding on majors, and so on–is virtually non-existent outside the US.

Global rankings therefore basically have to ignore this aberrant college issue with US institutions (including by, say, entirely ignoring LACs), in order to have something comparable to non-US institutions. And the common denominator is in effect grad/professional programs in the US, because those work much more like how the rest of the world works starting with undergrad programs.

At a high level, HYPSM is such a stable concept because it really doesn’t matter for them whether you are talking about undergrad or postgraduate–they are among the consensus “top” US institutions, indeed “top” global institutions, either way.

Where it definitely gets tricky is when you start getting to institutions like Brown or Dartmouth. These are still research universities too, but they have such good colleges in comparison that realistically their colleges should be “ranked” higher than their postgraduate programs.

And then institutions like Berkeley, Michigan, UCLA, Texas, Washington, and so on are sort of on the other end of that problem. Not that they don’t have good undergrad programs, but again their postgraduate programs are so good, both in the US and globally, that their postgraduate programs should realistically be “ranked” higher than their colleges.

So, uh, what to do about this?

Well, the US News separates LACs from research universities, so that’s a start. But then people can and do cross-shop some of the colleges at research universities with LACs when making college choices. So how do you handle that with two separate lists?

And then the US News does not separate out research universities with realistically a better undergrad ranking like Dartmouth from more of the opposite like the top research publics. And that also does not really capture the reality of how people looking for a college are likely to cross-shop.

So . . . again it is easy to pick holes in rankings in general, because rankings don’t make sense when there is such a variety of institutions with different pros and cons, strengths and weaknesses, and so on.

But I think exercises like this are interesting in that they really expose some of the specific issues like that, including explaining why HYPSM is actually kinda a thing, and yet Dartmouth versus Michigan leads to such inconsistent answers.

1 Like
5

REVISED 9/24: 2 of the 8 data sources had some missing data. I corrected and rescrubbed the entire list. A few changes reflected in the updated list below.

Here’s the top 25 LAC’s by the same process.

A couple important caveats:

  1. This list excludes the military academies and all women colleges. Not because they are not worthy and they certainly would have taken some of these top spots. But the military academies are a very unique category and the women’s colleges are not available to everyone. So, again, exclusion is not a judgment on them – simply not this list.

  2. None of the LAC’s were in the international rankings so that is not a factor here.

LAC Top 25
LAC Top 251050Ă—1328 181 KB

  • Despite the historical dominance of the NE and the NESCAC in LAC’s, Pomona, CA is a powerhouse of LAC’s with 3 in the top 10 by either measure on one larger campus. Acre for acre it is top LAC ground zero.

  • Carleton did surprisingly poorly – barely making the median list and not making the average cut. It’s #9 in USN, but scores very low in all the other ranks that include LACs.

  • Reed didn’t come close to making the top 25 by either metric, so it’s not just USN that doesn’t love them. They ranked well into triple-digits on half the lists.

  • This could be my misperception, but I feel like historically the informal hierarchy in Maine was Bowdoin, Bates, Colby. If so, Colby seems to have surpassed Bates by these rankings (which, again, mean little in the real world). I know Colby plays games in admissions, but that wouldn’t account for their position in most of these rankings.

2 Likes
6

I would say as a general rule it seems like Great Lakes LACs are not doing as well as coastal LACs, which means mostly East Coast LACs plus the Claremont Colleges.

At a high level, I think when it comes to things like peer reputation and observed placement adjusted for self-selection, Carleton and Grinnell really belong in that same group immediately after “WASP”+HM, so in the same range as like Claremont McKenna, Bowdoin, and Davidson. For that matter, I also think Middlebury, Vassar, and Wesleyan should all be in there too, along with Wellesley and Smith (I understand why they were not included here, just pointing out how I see all this).

And arguably more, but my point is even if you buy WASP+HM as a thing, that next group is really a pick your flavor of ice cream tier, there is no real hierarchy in my view.

But something about these rankings is not seeing things quite that way, and I have some suspicions. Like, I know there are big differences sometimes in averaged graduate salary studies. Which I think is almost entirely self-selection.

Like, I don’t think kids coming out of HS with, say, the ambition to work in highly selective finance positions in NY or LA choose these LACs at random. Rather, if you are that sort of kid, you know which LACs make sense for you.

But other kids are interested in becoming PhD students and professors, say, and then they know which LACs make sense for them.

WASP is a sort of analog to HYPS in the sense those LACs are really good at all that. HM is an MIT analog in the sense it is really good for certain fields. But then all these other LACs I am mentioning are sort of like the rest of the Ivy+, in that any could be great for a given kid, and ranking them is kinda silly.

But anyway, as a bit of a general rule, the Great Lakes LACs tend to be more on the academicky side. Not completely, but again people know if you want a path to the top business positions in NYC or LA, Carleton should not necessarily be your top choice of highly selective LACs.

But if you want to go to a PhD program? Well . . . .

Per capita, the total PhD list starts with Caltech, Harvey Mudd, Swarthmore, and MIT, and then next is Carleton, Reed, and Grinnell. Reed is a whole other subject, by the way. Then after that is Williams, Haverford, and Pomona, followed by St John’s (another special case), the University of Chicago, Vassar, Bryn Mawr, Princeton, Oberlin (also not doing great in these LAC rankings), Yale, Olin, Wesleyan, and Amherst at #20.

OK, so we finally got all of WASP + HM within the top 20 (and this includes universities), which I think is quite plausible. But for this particular purpose, Carleton, Grinnell, Haverford, Bryn Mawr, Vassar, Oberlin, and Wesleyan are also mixed in with WASP, which I personally think is exactly right in the sense those are all great LAC choices for an academicky LAC student.

Which is then going to show up, I suspect, in things like graduate salaries.

I guess the other thing that is going on is social mobility measures of various sorts. How these intersect with LACs as a group is very complicated, but I actually would kinda agree such measures are not necessarily great for academicky LACs, particularly in the Great Lakes. Like, the great engines of social mobility in the Great Lakes have for like 200 years been colleges like Michigan, Indiana, Illinois, Wisconsin, Minnesota, Ohio State, Purdue, and so on. That was their purpose, and they do it well. And part of how they do it well is with a particular strength in more practical fields, including these days business and engineering.

The LACs have long been more of a niche thing for the academicky sorts who do not necessarily need to be deeply worried about things like average graduate salaries. And I personally think that is fine, but if you are looking for engines of social mobility, well, not exactly the first thing that comes to mind when I think about Carleton.

But then things are a little different on the coasts, particularly the East Coast, because public universities have never been quite as prominent there as they are in the Great Lakes. So, the Ivies and top LACs have always had a bit more of a dual mandate as not just a niche for certain sorts of students, but also as just the generally top colleges and universities overall.

And personally, I think that explains a lot of why social mobility measures can be a bit kinder to some of these coastal LACs, at least the ones who have been investing in trying to get and support a lot more lower-income students, FGLI students, and so on, as part of their perceived dual mandate.

And it is not like the Great Lakes LACs don’t do that at all, but I think they are just not positioned to really do that to the same extent given the competition they face in those areas from these great public universities.

Anyway, that is my two cents on what is going on with some of these more academicky LACs, particularly in the Great Lakes.

For that matter, I think something similar happened to Chicago and WUSTL among universities. Some would argue those also seem a bit underrated, and I think a similar big picture issue applies. In the Great Lakes/Midwest, these colleges have faced competition from the likes of Michigan et al that sort of directed them into a particular positioning. And while I personally don’t see that as hugely problematic, it again is not doing them favors when social mobility becomes a criteria.

And indeed long before this phase of rankings, Chicago and WUSTL were working on that sort of thing, with some success actually. But their coastal competitors are too, and I think it remains an issue for them that the prominent public universities are tough competitors on these issues in their region.

3 Likes
7

By the way, I feel like I am sort of an unintentional lay expert in this subject because my S24 is looking at colleges that mostly are doing poorly in the new rankings phase! That is precisely because he is the sort of well-rounded but academicky upper-middle-class kid who is their bread and butter, but that is not doing them any favors in the new era.

So, in addition to LACs, he is looking at universities like WUSTL (15 to 24), Wake (29 to 47), Rochester (36 to 47), and William & Mary (41 to 53). Basically, if we scheduled a campus visit, your college was doomed to drop . . . .

Which I find amusing, and he seems OK with it too.

6 Likes
8

That’s actually a surprisingly high rank considering the amount of venting that happens on the Reddit forum. I think of Niche mostly as a customer satisfaction survey and those usually never reflect the median college experience and by nature attracts the polarities on the extreme in terms of survey respondents.

3 Likes
9

Updating both lists here as I found some missing data from 2 of the 8 sources (then rechecked all of the sources). I also corrected the lists in the original posts to avoid confusion.

Most of the impacts are minor in NU but there’s a couple new additions to the LAC list and a couple displacements.

I looked into adding the military academies and women’s colleges to the LAC rank but it didn’t work. Only 2 of 8 sources ranked the military academies and they did so wildly differently making the averages and medians pretty meaningless. Similarly, the women’s colleges didn’t have enough rankings to make up for the WSJ’s methodology negatively skewing them.

NU Top 35
NU Top 35900Ă—1546 217 KB

LAC Top 25
LAC Top 251050Ă—1328 181 KB

6 Likes
10

If you have any questions about William and Mary, I’m happy to help.

11

I’m really digging this aggregated LAC list. It’s helpful to see the classic, coed LACs listed and ranked as they’re sometimes a hard bunch to categorize due to differences in size, vibe and mission/focus etc.

1 Like
12

To your last point – I’m also wondering why Bates falls further down on the list. I have a sophomore at Bates, and it’s a very academically demanding school. If you compare requirements for similar majors, Bates has higher standards than some of the other Top 20 SLACs. For example, 11 courses in the major vs 9-10 at other colleges. A thesis is required senior year, which is not the case at many other SLACs. So, it’s an opportunity to work closely with a professor for 1-2 semesters and produce a significant work. A secondary focus area (minor or concentration) is also required to graduate. I think this puts Bates in the same league as quite a few of the top 15 on the two lists above, at least from an academic standpoint. It’s really a fantastic school - and known for great teaching, too! Thanks for raising the question.

1 Like
13

Bates scores lower (sometimes only slight, sometimes more) than Colby and Bowdoin in every ranking . But what really pulls Bates down the list overall is a very poor ranking in the WSJ. They aggregate NU and LAC together so in general all LAC’s will have higher numbers in WSJ, but Bates was particularly hard hit at #127. Given the WSJ’s methodological focus on ROI versus net cost for students with federal loans, Bates probably performed poorly in this category versus its peers.

More generally, this exposes one of the many weaknesses of aggregated rankings. It’s often possible for one ranks really bad score for a college to drag down its average and median versus colleges that didn’t have an aberrant rank.

1 Like
14

My observation is that rigor hurts ratings.

It doesn’t hurt MIT, Princeton et al because the rigor is part of their brand. But drop down- and rigor becomes significantly less popular as a school’s characteristic, and often, the “rankings” reflect that.

How many kids post here looking for the “easiest finance degree” with the best reputation, or how many parents post wondering if XYZ college is “worth it” because its grade deflation means getting a job on Wall Street or admission to med school will be “much harder”? (I’d argue that if you are just looking to get your ticket stamped in organic chemistry, you’re likely not a candidate for med school anyway… but that sorting is down the road).

And how many people post wondering if an Honors College is “worth it” since their kid ONLY wants the good dorm and early registration, but has zero interest in the seminars, research opportunities, etc?

Rigor is out of fashion- sadly. Bates has ALWAYS been considered a top notch college among Northeast employers (its appeal is regional-- have to admit- given its size, and that “back in the day” when companies sent recruiting teams, it’s not always the easiest place to get to). So rankings shmankings…

4 Likes
15

Carleton appears weirdly low, too, and also requires a senior comprehensive exercise (thesis or similar).

Carleton is a powerhouse for producing PhDs and also known for great teaching.

But this aggregate list does not appear to reward those things. Not sure what it does reward exactly: number of mentions in google searches? name familiarity with corporate board members? Or maybe it just comes down to a fluke.

2 Likes
16

Thank you! We had a visit that went really well, so I think for now it has a firm place on my S24’s application list. But if anything specific comes up and we can’t figure it out easily, we will make sure to reach out.

1 Like
17

A cynic might suggest that at least among private universities and colleges, it rewards the cycle of sending a lot of people to Wall Street-type careers, getting big donations, and then using those donations for a combination of attracting more people interested in Wall Street-type careers, plus somewhat more generous need-based aid.

In a way the WSJ was admirably transparent about this. But I think the US News changes reflect something similar, as moderated to make sure they stick within a comfortable range of pre-existing conventional wisdom.

1 Like
18

Carleton is substantially hurt by WSJ (#218) and Forbes (#172) – both of which aggregate LAC and NU in the same list so all LAC’s have higher ranks, but not nearly this high. Clearly those methodologies are incompatible with Carleton’s institutional priorities. If you only looked at the 2 ranks which isolate LAC’s, it’s #9 in USN and #21 in Niche. It’s scores in the ~20th position for College Raptor (which also aggregates LAC and NU) once you isolate LAC-only scores.

I believe both WSJ and Forbes are heavily economic outcome weighted.

1 Like
19

I’m not disagreeing, but curious what specifically in the methodologies would negatively correlate with rigor? I could see it with Niche which I believe is partly based on student reviews. Though Bates does fine in Niche. Perhaps there are other measures. Would be curious to see, particularly among the those more expert.

What is likely is that several of the rankings are biased toward ROI outcomes, which preferences either big universities where at least a healthy portion of their students go onto either professional schools or high paying jobs, or LAC’s that similarly tend to send more people to high paying jobs and less to humanities-based careers or non-professional grad degrees and academic or research work. Perhaps it could be argued that those who seek out known high rigor schools are more likely headed for lower paid academia.

1 Like
20

I think this is basically it. It is not a simple correlation but I do think a loose correlation exists.

I think at least three things are going on.

One is just that if you love academic activities enough to want to make them a career, you may actually see academic rigor as a positive.

Another is if you are interested in competitive but lucrative pre-professional careers, with some reason you might be more worried about things like grade deflation, robust core/distributional requirements outside your areas of strength, and so on.

And finally, I do think there is something to the fact that pursuing certain sorts of lower-pay/high-status careers–including many academic careers, but also certain sorts of government and NPO types of careers–is perceived as a risk and/or luxury by certain families, not least families without a long multigenerational relationship with these colleges. Like, these are the sorts of families who will say they are not going to pay private college costs just so their kid can study something like Art History.

So for some families, the whole concept of rigorous Art History is nonsensical.

And even if they bought it could exist at some colleges, it would be a negative if their kid was just trying to get the grades for a lucrative tech, or medical, or business, or so on career.

2 Likes