Ivy-equivalents (ranking based on alumni outcomes) take 2.1

College Search & Lists
1

A couple years ago, I decided to do an outcomes-based ranking to determine tiers of schools (again, mostly for bragging rights; for different career paths, like Wall Street vs. engineering vs. art vs. design, etc., I’d actually recommend different combinations of school options that don’t neatly fit in to these tiers).
(Original thread here: http://talk.collegeconfidential.com/college-search-selection/1682986-ivy-equivalents-p1.html)

At the time, I used 4 metrics:

  1. Percentage entering elite professional schools (for MBA/JD/MD).
  2. Production of “American Leaders”, who are mostly leaders in business, government, and the arts.
  3. Percentage winning prestigious national student awards.
  4. Percentage getting PhDs.

For the second try, I had added in a ranking based on “Who’s Who’s appearance frequency”, but after discovering some problems with the methodology there, I’ve decided to discard that and just use the original 4 criteria (but using 2015 instead of 2014 data). Also, I’m using a point-based system this time.

Since I value professional success more than academic success (since most graduates will join the workforce rather than academia, I again value those higher). Though I assign different points to different tiers.
So for “American Leaders”, the top 25 get 3 points, the next 25 2 points, and 51-75 1 point each.
For Elite Professional Schools and Prestigious National Awards, the top 25 get 2 points each and 26-50 1 point each
For PhDs (since a PhD may not actually be all that useful), only the top 25 get 1 point.

So here’s the list (and keep in mind that schools within 2 points of each other are essentially the same level):

Ivies & equivalents (16 RU’s and 14 LAC’s):
8: HYPSM + WAS LACs and Brown
7: Chicago, Cornell, Rice, Pomona, Haverford
6: Columbia, Dartmouth, Duke, Northwestern, Bowdoin, Wesleyan
5: Caltech, Georgetown, UPenn, Bryn Mawr, Carleton, CMC, Oberlin, Reed, Smith, Wellesley

Near-Ivies (8 RU’s, 2 service academies and 7 LACs):
4: Cal, UMich, UVa, ND, West Point, Naval Academy, Barnard, Grinnell, Middlebury, Vassar
3: NYU, Tufts, UCLA, Wisconsin-Madison, Harvey Mudd, Macalester, New College of Florida

Other good schools:
2: Johns Hopkins, UNC, USC, UT-Austin, Rochester, UIUC, Indiana, Colorado, Bates, Bennington, Davidson, Holy Cross, Lehigh, Occidental, Pitzer, Scripps, Trinity, W&M, W&L

2

What is a “WAS” LAC? Williams, Amherst, Swarthmore?

3

You should add students getting MD’s. The process of becoming a medical doctor is far more difficult, physically and mentally demanding and a greater test of character than getting a PhD.

New College of Florida is a near Ivy?

4

I don’t think Indiana(Isn’t Purdue better?) and Colorado should be in the same category as JHU and USC. They have really high acceptance rates and are really easy to get into.

5

WashU didn’t make the cut, huh?

6

@ShrimpBurrito: Nope. WashU just misses the cut in a few of these categories, though (like CMU and Kenyon).

@Ballislife23mjs: Arguably. I just tallied up the numbers, and while I would put JHU a cut above CU, IU-Kelley is about the same level as USC-Marshall and while the USC film school is at the top in it’s field, so is IU Music. USC is definitely harder to get in to as a freshman than IU, but note that USC takes a ton of transfers: less than 3000 freshmen entering in the fall but almost 5000 graduate with USC bachelors degrees each year. (You may even suspect that they are doing so to game the rankings).

@OnTheBubble: NCF puts an insanely high percentage of its grads in to grad schools and professional schools. Just because it isn’t well-known doesn’t mean it isn’t good.
Also, MD’s are included in category 1. I understand that entering any med school is impressive, but nobody has rates for that by school (that I know of), so I don’t have that data. Otherwise, I’d use it (and it would likely bump up JHU and WashU).

@merelyaboveavg: Yes, WAS stands for Williams, Amherst, and Swarthmore.

7

Re #3

OP was trying to make an output based ranking, not an input based ranking.

However, output and input are related, so if one wanted to try to isolate the school’s treatment effects, one would have to measure output compared to that expected from its input.

8

OP preference and bias toward schools in the east is very obvious!

9

You just count the absolute numbers?

10

@hzhao2004: Nope. These are almost all by percentages of the graduating class (“American Leaders” has a small weight for absolute numbers).

If you go by absolute numbers, Cal and UMich would be up there with HYPSM
(look here: http://www.nsf.gov/statistics/infbrief/nsf13323/ and here: http://inpathways.net/ipcnlibrary/ViewBiblio.aspx?aid=1577). Actually, Cal destroys everybody by the absolute number of its undergrads who get STEM PhDs.

11

@uclaparent9:

Not my preference (trust me; I have no allegiance to any East Coast school). The numbers are what the numbers are. Granted, the WSJ ranking that is used for prestigious professional schools is biased towards the East Coast. As I noted in my original post a few years back:
“Though notice that there is a heavy East Coast bias in the grad schools deemed elite by WSJ, with 12 on the East Coast, 2 in the Midwest, and 1 on the West Coast. Compare with the composition of M7 b-schools + T6 law schools, which is 8 on the East Coast, 3 in the Midwest, and 2 on the West Coast. If you add the top 2 med schools according to USN to get to 15, it becomes 9-3-3 (2 med schools is probably enough because unlike b-school and law school, what med school you go to likely won’t have a big effect on earnings). Still, this is the best ranking we have for admittance to elite professional schools.”
If it makes you feel better, bump all the non-Eastern school up a point to compensate.

12

@PurpleTitan A school with a 55% four year graduation rate where 19% go to graduate school is not even remotely a near Ivy.

13

@OnTheBubble, the NCF graduation rate is low, but then, its inputs are also much lower as well (BTW, once upon a time, the UChicago graduation rate was also atrociously low; around 50%; but it was seen as an Ivy-equivalent even back then because it had stellar outputs).

The percentage of NCF grads who go on to grad school is quite high. Have you seen what the grad school percentages are for the other near-Ivies or even Ivies?

BTW, http://www.previous.ncf.edu/graduate-school
Those numbers compare favorably to a lot of elite schools (and are higher than 19%). Where did you get that number from?

14

@PurpleTitan Where does University of Scranton rank in your model. Over 40% of students go directly to grad school.

15

@OnTheBubble, not PhD programs or elite professional schools, though. They also don’t win prestigious undergraduate awards at the rate that NCF students do.

16

Very interesting exercise, @PurpleTitan . Philosophically, I agree with your goal of an outcome-based ranking.

Can you please repost the sources of data? Some links in your old thread are no longer active.

I would say the surprise omissions for me are: Vanderbilt, CMU, Washington Univ S.L., and Emory (all top 25 in USNWR)

The biggest surprise inclusions for me are : Reed, New College FLA, Bennington, Colorado (I assume U. Colo Boulder?) These are all #82 or lower in USNWR, either National U’s or National LAC’s.

I’ll be the first to admit that USNWR’s methodology is suspect, however I do think their ranking (+/- 10 or so places) is roughly representative of the “quality” of the institution.

I’d be interested in your analysis of my ‘surprise omissions/inclusions.’

17

@pickpocket, I saved the PDF’s, but I can’t find online versions of the 2015 Forbes components any more. If there is a way to send over PDF’s over, I could do that.

In any case, the sum is so low for the last group (“Other good schools”) that you probably could come up with another group of “Other good schools” that are, in aggregate, about or almost as good (and as you would expect in any normal distribution, there are more in each tier the farther away from the top you are).

In terms of the surprises, note that the USNews rankings are heavily input-based and can be gamed. Schools that heavily value high test scores like Vandy and WashU may be higher up in the rankings than their outputs would indicate. Also, until recently, WashU, Vandy, and Emory were only regionally acclaimed, so they would draw the best talent from only their immediate area (and none were in big metros; Atlanta now is one, but that’s because it has been growing rapidly; it certainly wasn’t in the past). Also, schools where a lot of their top students go to med school (WashU and JHU & arguably Emory) may be dinged by my methodology as those schools would be less represented at the top in other areas (like the arts, politics, business, law, etc.).

CMU is a lot like NYU, USC, UIUC (and IU) in being unbalanced: very strong in some fields/schools and not so much in others. They were a bit unlucky to finish just a little on the wrong side of the line in a few of the criteria. Likewise, you could argue that CU was a bit lucky to finish in the second lot of 25 in “American Leaders”. Though note that USNews tends to underrate giant publics with its methodology with criteria such as median test scores & acceptance rate (tough for a giant public to bring very high and very low, respectively, even if a large absolute number of their alums become successful by various criteria) as well as per capita endowment. Bennington does do well in the arts.
In any case, they’re both in that lowest group and could be easily in or out.

Reed and NCF, however, are for real. They both put a high percentage of grads in to PhD programs and elite professional schools and also win prestigious awards at an good rate. Reed, for the longest time, disdained playing the US News rankings game (and got punished for it). In that sense, they were like UChicago a decade ago. If you don’t market like crazy to drum up a ton of applicants and thus your applicant & accepted pool is weaker and thus fewer of them graduate, you will be dinged by USNews even if the percentage of your grads with notable achievements is similar to other near-Ivies or Ivies.

I personally don’t think much of the USNews rankings. If you want a ranking that is less-easily gamed and reflects the real-world better, look at my tiers or the Forbes rankings (which uses a lot of the same inputs as mine).

18

PurpleTitan, while percentages are certainly worth noting, absolute numbers should also be factored in.

19

@Alexandre, they are to a degree in the “American Leaders” category.

20

Suppose we want to remove alleged “East Coast bias” by including in outputs the number of alumni surfing and rodeo champions. Then the schools can adjust their admission process and add classes, or build training facilities, to encourage the production of more champion surfers and rodeo riders. UChicago’s Midway would make a wonderful bull-riding arena.

Of course, the kinds of colleges we’re talking about do not exist to train surfers and rodeo riders … but they also don’t exist, really, to crank out MBA-earning, JDA-earning, or big city mayoral election-winning alumni. They exist primarily to discover and share knowledge. That really is their purpose in life, regardless of what parents and students think they want to get from these places. So I’d weigh per capita PhD production more heavily.

The weights don’t really matter all that much, though. Whether you use the US News criteria, the Forbes criteria, or PT’s output criteria, you’re still likely to wind up with roughly the same set of top schools. That may because those schools really are “better”. Or it may be because all these measurements (including output measurements) are telegraphing the effect of admission selectivity and institutional wealth. The richest schools can afford to cherry-pick the most talented and motivated students, regardless of ability to pay. Whether we’re talking about academia or investment banking, talented and motivated people tend to be successful in life. Surprise!

I’ve found that the highest per capita alumni PhD producers, *after adjusting for “inputs”/i, do include some very high ranking schools such as Caltech, Harvey Mudd, and Swarthmore. However, they also include schools like Allegheny College, NM Tech, and Earlham.
What do the over-achieving colleges (w.r.t. PhD production) seem to have in common? They tend to be small. They tend to have consistently small classes. They have an understated (or no) Greek and D1 sports presence. They offer high levels of student-faculty engagement and relatively few distractions from academics. They’re Nerd Schools!

Without a doubt, in a knowledge-driven economy, building more Nerd Schools would Make America Great Again, and we should make China and India pay for them.