New WSJ-Times Higher Education Rankings

@prezbucky, I’m pretty certain that alum salaries isn’t what’s keeping Harvard and Princeton out of the top 5.

Well what on earth is it then? Princeton is the most undergrad-focused of the HYPSM crew (by % of total student body) and Harvard is… Harvard. hehe

@prezbucky -

Note that the top eight schools (Princeton is # 8) are only separated by 2.4 points on a scale of 100, so the differences in the scores are well within the margin of error of this type of measurement/calculation.

Harvard and Princeton are “hurt” by their relatively low scores in the “engagement” category and the “environment” category.

The score in the “environment” category attempts to measure diversity and the spread among the 8 schools in this category is 1.2 points.

The score in the “engagement” category includes the results of survey of students taken by the Times. One question asked students about the teaching quality of their professors. Neither Harvard or Princeton made the top 150 in the teaching area because the students said they were “aloof”.

The spread among the 8 schools in the “environment” category is 2.4 points

Harvard and Princeton do well in the “outcomes” category which includes salary. The spread among the 8 schools in this category is 1.1 points.

Harvard and Princeton both had high raw “Scorecard” salaries and the methodology employed by Times/WSJ creates a “value added” salary that attempts to correct for geography, wealth and test scores of incoming students, as well as majors. This methodology has issues that I can go into if you are interested, but in comparing among the schools you are talking about it is pretty fair, if not favorable to both Harvard and Princeton. Note that the Times/WSJ is vague on their exact methodology, but they say it is based on the “Brookings methodology” which is what I used as a reference.

Within the “outcomes” category, Harvard does get “hurt” by its relatively low percentage of students paying back their loans three years after graduation. This may be due to the small percentage of students at Harvard who take out loans (4%). But this small percentage of students taking out loans also makes the sample size for Harvard students rather small and that appears to artificially inflate Harvard’s raw scorecard salary (based on a quick cross-correlation analysis of Scorecard and Payscale salaries for the schools).

Note that the Scorecard salary calculation does not include students whose loans are in deferment, so Harvard and Princeton cannot be “hurt” by the incomes of students in grad school.

Also note that the “outcomes” category includes the results of survey of academics of the schools teaching quality (how this qualifies as an “outcome” is rather questionable). To assess the impact of this category, I compared the “outcomes” scores of Harvard and Princeton to those of Babson, which has equivalent or higher “value added” salary and higher loan repayment rates, but a lower graduation rate (90% vs 97%). The difference in the outcome score was a full 10 points, which is just a little less than the difference separating the top 30 schools.

Based on the data, I would say that Harvard and Princeton are actually helped by their salary data relative to most other schools, and that the difference between the top 8 schools is noise.

I would suggest that given the large boost that Harvard and Princeton get from the inclusion of the results of a survey of “teaching quality” as seen by academics that has a strong inverse correlation to the results of a survey of “teaching quality” as seen by students, that a case could be made that the ranking of both Harvard and Princeton are grossly inflated.

Interesting. Brookings has a solid critique of the Scorecard salary data: https://www.brookings.edu/research/deconstructing-and-reconstructing-the-college-scorecard/ (see esp. graf 3).

@Mastadon, thank you, that was interesting. Maybe the Princeton and Harvard students who responded to the surveys were being smart-alecks or were in the “sour grapes” crowd. Certainly both schools are known for having among the very best profs. (and Princeton has relatively fewer grad students taking up their time and attention…)

The slimness of the point differential is reassuring. Given that it’s typical of many rankings, I think ranking by tiers is more accurate and helpful for prospective students, if rankings are helpful at all.

Regarding the salary gripe, my concern was mainly that schools not near the coasts – UChicago, Northwestern, Rice, Vandy, Notre Dame, WUSTL, Michigan, Texas, Wisconsin, etc. – would be hurt due to relatively lower cost of living. And that all schools without a lot of Engineering, CS or Finance majors would be affected negatively.

@prezbucky Princeton, yes, but Harvard actually is known for having renowned researchers and not caring if they are good teachers. At least according to a report prepared by Harvard itself a few years back.

"CAMBRIDGE, Mass., May 8 — Joshua Billings, 22, says he did not come to Harvard for the teaching.

“You’d be stupid if you came to Harvard for the teaching,” said Mr. Billings, who will graduate this spring and then go to Oxford as a Rhodes scholar. “You go to a liberal arts college for the teaching. You come to Harvard to be around some of the greatest minds on earth.”

http://www.nytimes.com/2007/05/10/education/10harvard.html?_r=0

Let’s hope Harvard is keeping its eye on the ball regarding the quality of undergrad teaching. They certainly have the resources to inspire and reward it.

Your summary of my post history is appreciated, and paints a consistent picture of an anti-PC crusader. Or do you need a trigger warning with that?

“Can someone please explain what mathematical theorem they are using when they conclude that adding more inaccurate data improves the accuracy of a calculation?”

https://en.wikipedia.org/wiki/Wisdom_of_the_crowd

I’m thinking the law of large numbers would show the more data that’s collected, even if it’s inaccurate, results in a better overall result. The condition is that the error is uniformly distributed around the actual value(s).

Why not include the Forbes ranking as well? WSJ, US News and Forbes (1/3 each)

…because of the flaws inherent to using ROI to judge the quality of undergraduate education.

…then why include WSJ?

USC places its students really well. Rice and Brown don’t place like USC does.

It is a pity that Suffolk University is not ranking that high
This is for John: Nami Farkhondeh https://www.f6s.com/namifarkhondeh had recommended it.

The attend, but didn’t graduate bias is highest for the lower tier schools. At the top tier schools, the 4-year graduation rates are over 80% and the 6-year graduation rates are over 90%.

Any ranking that places UCLA more than 10 places above UC Berkeley is seriously flawed IMO. Without any doubt, for most CA residents students who get into both UCLA and UC Berkeley would choose Berkeley over UCLA. I would say around 70% would choose Berkeley over UCLA.

@websensation As a Californian, I agree. There are of course some who would choose UCLA because they wanted to stay closer to home (or maybe get further away from home) or really, really cared about sports success, but absent that, Berkeley would be the choice every time. UCLA is a great school, but Berkeley is the flagship, the place with the history and the Nobel Prizes and the best students and the superior programs in virtually every field.

@marvin100 [snip]

I read that argument in the Brookings report and it presents an interesting problem. Basically, it’s saying that a school’s overall ranking is hurt to the extent that its best future earners are students coming from wealthy families, and didn’t need financial aid. I guess my question is, knowing full well that wealthy students are likely going to skew the results, isn’t excluding them actually a good thing?

Knowing that something happens is quite a different thing from knowing how it happens or how much it happens. It’s also problematic that it only includes federal FA–there are many other avenues of FA (I know, I had and repaid a good deal of it haha). @circuitrider