@Alexandre your second and third tiers are way too big and include too many schools that should be higher or lower. But enough of that.
@alexandre-
Been there, done that. (over 30 years ago!)
As my elementary school History teacher used to say “Those that forget the past are condemned to relive it.” 
I can’t help but think that this rating stuff always ends up in a morass. I think this usually the result of lack of clarity on what it is trying to accomplish. (I thought this thread was intended to represent option 1 below, but perhaps I misread.) Institutions can be rated in so a number of ways:
- Undergraduate perception - if you encounter a graduate (may need to define time period) and don’t know anything else about them other than where they attended school, what is your relative perception of their capabilities? (this is what I attempted to answer subjectively since it requires some synthesis)
- Value add – given that smart, driven people may succeed if given a reasonable opportunity regardless of where they go to school, how much value does the institution really add relative to alternatives in making the person more productive, fulfilled etc.? (I have my own views here but have not shared.)
- Research - how good is the institution at producing research that is valuable to society and the advancement of knowledge? There are ratings out that do this looking at publications, cited publications, etc.
- Selectivity – How objectively selective is the institution? USNews and other rankings can be used, but keep in mind that they can be subject to some manipulation.
- Overall reputation – this can be a combination of a number of things – undergraduate, graduate, research, value, etc. etc. I question how valuable this would be specifically for assessing undergraduate study. The USNews reputation rating may be in this category.
- Value for money – how efficient and effective is the institution in providing education given the required investment?
- Graduate and professional programs . . .
I’m sure this list could go on. . .
From (mostly…) 2016-2017 CDS forms, this is the percentage of undergraduates at each of these schools (undergrads/total enrollment):
Wisconsin 73.17% (31710/43336)
Cal-Berkeley 71.97% (27496/38204)
Brown 70.81% (6926/9781)
Notre Dame 68.83% (8530/12393)
UCLA 68.69% (30873/44947)
Virginia 68.34% (16331/23898)
Dartmouth 67.25% (4310/6409)
Princeton 66.01% (5400/8181)
Cornell 65.26% (14566/22319)
Michigan 64.81% (28983/44718)
UNC 62.86% (18523/29469)
*Rice 58.19% (3910/6719)
Vanderbilt 54.59% (6871/12587)
*NYU 51.42% (25722/50027)
*Washington U 51.09% (7504/14688)
Emory 50.9% (7591/14913)
Tufts 47.94% (5508/11489)
CMU 47.08% (6574/13961)
*Penn 45.46% (9726/21395)
Yale 43.92% (5472/12458)
Caltech 43.7% (979/2240)
*USC ~43.18% (~19000/~44000)
Stanford 41.59% (7034/16914)
Duke 41.49% (6609/15928)
*Georgetown 40.97% (7562/18459)
MIT 39.77% (4524/11376)
Northwestern 39.39% (8353/21208)
Chicago 35.99% (5547/15413)
Harvard ~30.52% (~6700/~21950)
*Johns Hopkins 28.64% (6524/22783)
Columbia 27.82% (8712/31317)
*2015-16 CDS numbers (school has not posted 2016-17)
**JHU was awful. I had to get their figures from their USNews profile, which I think are 2015-16 stats
***USC gives 2016-17 enrollment stats to the nearest 500. Way to be accurate…
General gripe: Please post your most recent CDS, schools that haven't. It makes researching so much easier.
Any surprises?
You can look at conditional probabilities based on the parents’ income quartile. It doesn’t change the rankings that much. You still end up with the Ivies+MIT+Stanford+Duke+Georgetown+Rice in 13 of the top 15 slots.
Top 10 alma maters of employees at Silicon Valley’s top 25 companies:
- University of California, Berkeley
- Stanford University
- Carnegie Mellon University
- University of Southern California
- The University of Texas at Austin
- Georgia Institute of Technology
- University of Illinois at Urbana-Champaign
- San Jose State University
- University of California, San Diego
- Arizona State University
Silicon Valley Most Hired in 2017
- University of California, Berkeley
- Stanford University
- Carnegie Mellon University
- Georgia Institute of Technology
- University of Southern California
- San Jose
- U Waterloo (Ontario)
- The University of Texas at Austin
- University of Illinois at Urbana-Champaign
- UCSD
- Cornell
- ASU
- NCSU
- California Poly
- MIT
- Purdue
- Michigan
- Texas am
- UCLA
- University of Washington
@nostalgicwisdom post #53
“I really just don’t see what makes a current Dartmouth student (admit rate 10.4%) or recent graduate two tiers more distinctive than a Vanderbilt (admit rate 10.2%) or Swarthmore student (admit rate 10.2%). I don’t find it fair to judge current students on the history of their institution”
There is more than one way to look at this. You are going solely by admit rate. I was looking at it more broadly. For instance, if you look at value add based on ratings that were done by the Economist ( http://www.economist.com/blogs/graphicdetail/2015/10/value-university ) and the Center for Education and the Workforce at Georgetown ( https://cew.georgetown.edu/cew-reports/college-rankings/#interactive), Dartmouth comes out well ahead of either Vanderbilt or Swarthmore:
Economist: Dartmouth – Expected Earnings $65,950, Median Earnings $67,100, Value Add +$1,150; Swarthmore – Expected Earnings $58,115, Median Earnings $49,400, Value Add -$8,715; Vanderbilt – Expected Earnings $61,344, Median Earnings $60,900, Value Add -$444
CEW: Dartmouth – Expected Earnings $62,300, Median Earnings $67,100, Value Add +$4,800; Swarthmore – Expected Earnings $57,900, Median Earnings $49,400, Value Add -$8,500; Vanderbilt – Expected Earnings $62,000, Median Earnings $60,900, Value Add -$1,100
So which is more important, higher earnings, or the satisfaction you go to a school with a .2% lower admit rate?
Admit rate is just one variable and its importance is limited, IMO. Stats are at least as important if you’re looking to define class/student quality at a school.
Neither is much of a factor in the quality of teaching/academics, IMO.
@prezbucky As most of the schools in your list of % undergrads are from the USNews top 30, if you go down another 10 you get UCSB with 87.7% (20,607/23,497).
UCSB: the big LAC masquerading as a state school.
Obviously other things make a school undergrad-focused, like access to faculty, spending on undergrads, low S/F ratio, and undergraduate support and opportunities. But I think among peer groups (elite privates, publics, etc.), that undergrad proportion (%) can at least tell us which schools must, out of necessity, pay relatively more attention to their undergrads. The fewer the grad programs and grad students, the more time and funds the school has to focus on its undergrads. And… the more time the faculty members have to spend with undergrads.
Brown, for instance, claims that all faculty teach undergraduates. That’s the sort of thing I would hope to hear from a school that is undergrad-focused.
@IzzoOne This will be my last post on here because I think we’re starting to stray from topic. Feel free to PM me if you want to continue the discussion.
Starting salaries are one of the least useful ways to consider the strength of an institution, in my opinion, because they give no indication of what students are studying and pursuing. That’s what salaries are based off, far more than anything else.
For instance, Harvey Mudd has a higher salary than most colleges in the country. Is it because of prestige? Slightly, but it’s more because some 70-80% of their students major in computer science or engineering.
The starting salary data for the Economist is based of students who take out loans. To my understanding, that’s under 35% of the student body at these three schools. So not necessarily an indicative group. Furthermore, in Dartmouth’s case, it is one of the most pre-professional schools out there- very business/economics/finance oriented, which pay well. It doesn’t show up until #39 on the PhD per capita list; Swarthmore is 3. Some 85% of Swarthmore students go onto graduate school because that’s the type of students it attracts; the route to take a PhD can mean a meager salary for many years. Swarthmore has also been known to be a big community service/non-profit/education oriented school especially relative to the more pre-professional Williams and Amherst. I’m not sure about Vanderbilt’s history (which is important, as these salaries are 10 year historical data), so I can’t comment on it.
Here’s the question I pose. 91% of Swarthmore kids rank in the top 10%. Their median SAT is 2155. 93% of Dartmouth kids rank in the top 10%. Their median SAT is 2170. Many Swarthmore students turned down Dartmouth to go there (and vice versa too, of course). When you meet a student from either school, who may be virtually indistinguishable by academic profile as the above points out (no, I don’t consider that a big difference, especially since Swarthmore takes more low-income students/URMs who statistically don’t test as highly than Dartmouth), are you going to assume that the Dartmouth student is two tiers more capable than a Swarthmore student?
I can’t. In my opinion, Swarthmore and Dartmouth have always been and still do attract comparably talented student bodies with different perspectives and interests. Vanderbilt has started doing so recently as well; their yields and testing profiles have shot up over the last five years. Not really a trend when it has consistently happened and the new admitted class shows a continuation of it.
I’m in your corner regarding the flaws of including outcomes in a college ranking – it’s mostly reflective of what types of students the schools attract, what their majors/interests are, and where (geographically) they become employed – differences in cost of living mean differences in salaries for the same jobs.
Now, if we could compare results by major, that might be worth more. But we’d still have the “work force vs. grad school/PhD path” and “cost of living” issues to overcome.
^Starting salaries may not be the best measure, but by 32-34 most people’s career paths are fairly well set.
@nostalgicwisdom and @prezbucky , I understand your points, but if you look at the CEW Georgetown rating for instance, they are adjusting (or at least attempting to adjust) for mix of majors. (It also adjusts for relative student preparation.) So if a school is engineering or business heavy as you say, the expected salary is adjusted, and the relative performance is indicated. In other words, they are trying to make it an apples to apples comparison. They are also not using starting salaries. It is 10 years in.
Regarding Dartmouth vs Swarthmore, I’ve known some grads from both. The Dartmouth ones were real “go getters”. I’m not saying the Swarthmore ones were bad in any way, but that shaped my impression of Dartmouth. As I said several times, the original post was based on my perception based on a number of inputs (selectivity, information on outcomes, perception of graduates). I recognize that perceptions can be skewed by personal experience, but this is really what everyone does. I didn’t even look at USNews as you had suggested (which probably explains some oversights on my part). If we just wanted to do a ranking by SAT and admit rates, we could just default to USNews (with all its manipulated warts).
If I did V2 of my rating, I’d probably condense to 3 tiers, with 1 remaining the same, collapsing 2 and 3, and 4 and 5.
By “academe” writ large? Does academe ever speak with one voice about anything?
The US News peer assessment survey goes to presidents, provosts, and deans of admission.
If only the ranked colleges are represented, then it’s probably going to many hundreds of peers (but not thousands) each time. After that, what’s the completion rate? Some college presidents earn more than $1M per year (from salary alone); I doubt nearly all of them personally spend the many uncompensated hours it could take to complete an entire US News PA survey conscientiously.
AFAIK, the PA produces scores, not tiers. What bright line separates a 4.7 from a 4.8?
Indeed, it’s not even clear how much difference a full point makes in, say, student satisfaction or alumni outcomes.
In 2009, Inside Higher Ed reviewed the annual US News assessment and found it “subject to apathy and glaring disparities”. One university’s president listed “don’t know” for ~half the universities surveyed. That sounds like an honest answer. How many colleges can the average president, provost, or dean know well? Surely, they aren’t immune to bias that favors their own employers, their alma maters, or research universities that publish heavily in their own fields of interest.
Nevertheless, for the most part, the PA scores do tend to corroborate the rest of the US News ranking factors (at least in the set of top colleges identified if not in the precise order). Rice, Vanderbilt, and WUSTL are a bit lower, while Berkeley and Michigan are higher, in the PA than in the overall rankings.
The 2014-2015 Common Data Set for Johns Hopkins is available online. The values reported under section B are as follows:
Johns Hopkins 73.43% (5365/7306)
These numbers are for the Homewood Campus only, so they exclude A&S Advanced Academic Programs, Engineering for Professionals, Business School, School of Education, School of Medicine, School of Public Health, Nursing School, SAIS, and Peabody Music, which are located at different sites and have different faculties. I don’t know if this is consistent with the Common Data Sets from other schools, but if you are going to use a standard methodology, these are their numbers. The student-to-faculty ratio in the CDS is also reported using only the Homewood Campus, so this seems like a fair indicator of the undergraduate experience there.
If you look at the general quality of the undergraduate experience and add to that the quality/impact of the faculty the University attracts and the quality/impact of the students it produces, I would generally classify Research universities into the following Tiers
Tier 1: Harvard, MIT, Stanford
Tier 2: Chicago, Columbia, Princeton, Yale
Tier 3: Berkeley, Caltech, Cornell, Johns Hopkins, Penn
Tier 4: Dartmouth, Duke, Michigan, Northwestern, UCLA
Tier 5: Brown, Emory, Notre Dame, NYU, Rice, UCSD, UVa, USC, Vanderbilt, WashU
Tier 6: CMU, Georgetown, Tufts, UNC-Chapel Hill, Univ of Washington, UIUC, Univ or Rochester, Univ of Wisconsin, UT Austin, Wake Forest
If I had the good fortune of being admitted to several schools that spanned these tiers and the difference were just one Tier level, I would be inclined to pick on personal fit preferences.
The 2014-2015 Common Data Set for Johns Hopkins is available online. If you are going to employ a uniform methodology, you should use those numbers, which report 5365 undergraduate and 1941 graduate students. The entry should therefore read
Johns Hopkins 73.43% (5365/7306)
The Common Data Set reports values for the Homewood campus only, so it excludes schools at different sites and having separate faculties, such as the Business School, School of Nursing, and Peabody Music. But the undergraduates there spend their time on the Homewood campus, so this may be a fair indicator of their experience.
mdphd92 I do not see how we can ignore the more than 10,000 graduate students enrolled at JHU. Just its engineering college enrolls 1,500 full time graduate students and an additional 2,200 part time graduate students. The medical school enrolls another 500 graduate students. And how about all the PhD programs? Biology? Chemistry? Physics? Economics? History? English? International Relations?
Then you would have to do the same for Tufts, because its medical, dental, vet schools are in downtown Boston, not on the Medford campus, which is why I think listing schools by undergrad vs grad school populations isn’t helpful. Tufts is hugely focused on its undergrads despite having lots of grad students. Brown, too, has its med school in downtown Providence.