US News 2017 rankings

Agree with rjkofnovi; Stanford exhibits across the board strength. I include it in a separate tier with Harvard only because what Harvard does well it does very well. The only other school in the US with comparable overall strength is UC Berkeley, but that is more apparent at the graduate level.
Also, Prezbucky: I am in general agreement with your groupings.

All the responses saying that metric X makes up only Z% of the USNWR ranking metric (where Z is a small number), and therefore a school doing something to improve its numbers in metric X wouldn’t be effective in moving a school up in the rankings—I don’t think that’s actually correct reasoning.

I’d have to run the numbers to be certain, but I’m pretty sure that the scores for schools (at least those in the upper quartile-ish) are so close that even a small relative improvement or decrement can have an outsize effect on position in the rankings. Maybe I’ll run some simulations in a couple days (in the unlikely event work calms down enough to let me) to check my logic on this.

“Is this not the conventional view of Californians?”

That more or less it. Most of my neighbors here in CA (at least those who think about such things) would rank them roughly like that, although they might sometimes offer a little more granularity. Say something like this:

  1. UCB, UCLA,

  2. UCSD

  3. UCD, UCSB, UCI (in that order)

  4. UCR

  5. UCSC

  6. UCM

Davis sometimes gets a little boost in some people’s minds because it’s bigger than UCI or UCSB, and it has the full range of professional schools (medical, law, business,veterinary) that UCI and UCSB lack one or more of. Plus it has a football team, Berkeley and UCLA are the only other UCs with football. None of which has much to do with the quality of undergrad education but do serve to raise the profile of the school with the public.

@wandlmink I totally agree with your post, especially point #3; this is really a very important point for applicants to keep in mind. But as has been discussed before on this thread, acceptance rate represents just 1.25% of a school’s USNWR score, so in between the top 25 or so I don’t think acceptance rate matters that much. Of course it’s all relative, so if 10 of your peers’ acceptance rates are going down, you may be hurting yourself by not pursuing the same goal.

I do think that schools believe that acceptance rate is a key component of their marketing plan, and agree that it is somewhat deplorable to pursue low acceptance rates at the cost of thousands of unqualified applicants’ confidence.

Also suzyQ7 meant to @mention you in the post above.

@dfbdfb agree but AR is actually 1.25% vs SAT/ACT at 8.125%; 6-yr grad rate is 18% and retention is 4.5%.

You need to focus on everything, but it seems to me that a focused effort on 6-yr grad rate would in some ways be the easiest and most beneficial area to try to maximize your score.

More importantly, if you have a way of removing a subset of students from reporting that you predict will not graduate on time, this is probably the easiest way of “gaming” the rankings.

Getting sick of picking on certain schools, so I’ll just leave that there.

@Scipio, that seems to be the general view here in Southern California, although some would argue that UCSB has switched places with UCD in recent years. The OOS applications to each school indicate that this general view is shared by nonresidents: http://www.ucop.edu/news/factsheets/2014/fall-2014-admissions-table2.pdf

Why is it that Stanford languishes at #5, tied with Columbia, and not in a top-2 or top-3 spot as some claim they merit?

My personal read on HYPSM (and Columbia and Chicago) is that they are all at the very top in terms of overall academic excellence: they are good at (virtually) everything they do. Even Yale STEM and Stanford/MIT Humanities are very good. (Although one has to wonder how valid it is to use US News’ grad school rankings to judge undergrad program strength. My view is that if the school has a well regarded grad program, the undergrad version is likely at least decent.)

They also offer great opportunities and boast high grad rates, freshman retention, test scores, etc. And very high academic rep.

So why do some want to put Stanford above Princeton/Yale/MIT/CC in a dual-king role with Harvard?

I think the popularity and resulting selectivity can be attributed mainly to two things:

  1. Stanford’s rep
  2. California weather

Anyone sick of snow or in the mood to continue avoiding it has two main geographic choices: Deep South/Texas or California. In those areas the Ivy-level elite private universities are Stanford and Rice. That’s really it; there are two choices. (With all due respect to USC, an up-and-comer). And Rice is far more humid than Stanford.

Simply put, Stanford is California’s Ivy. It owes its low acceptance rate and high yield primarily to that fact.

Meanwhile, if you want to go to an elite private U in the Northeast, there are a lot more to choose from (and more yield competition among them).

As for why Stanford is in 5th/6th in the USNews ranking, I think one reason might be this: while average test scores are impressive, they are below those of their peers, and even below those at a handful of schools ranked below them. Stanford is very holistic and I like that, but it does them no favors in the USNews methodology. That’s one area where Stanford may be lacking. There must be others (compared to HYP and Chicago at least).

A question I have is, why does Princeton continue to rule at #1? I get the overall reputation it has for elite academics and the undergraduate emphasis – I guess I’m wondering if someone among us knows what parts of the USNews formula represent those strengths and lead to Princeton winning every year (recently, at least).

You do realize Caltech is vastly underrated by USNWR right? On a pure academic basis, the education at Caltech is one of the most excellent if not THE most excellent.

And again, Princeton can afford to be the top college in the country because it can focus it’s resources on undergrads. When you have one job, you might as well be the best at it. It has the money to pour into financial aid and faculty resources, the peer assessment score that puts it shoulder to shoulder with HYSM, the quality of the student body as judged by SAT and acceptance rate and graduation rates.

In addition, it is the second smallest ivy, which translates into smaller classes size. And finally the alums are extremely happy and loyal despite the grueling academics that tested them while they were there.

How is it still a mystery to you?

^^Just finished book on honors programs, and across 50 programs the 6-yr grad rate for those with new SAT avg of 1440/ACT 31.5–32.0 is 92.6%. For all programs with new SAT 1310–1430, the avg grad rate is 85.2%. There is, in other words, a big difference in multiple metrics (grad rate, ret rate, selectivity, acceptance rates) when the admission scores go up past new SAT 1440 or so. The positive impact for Northeastern, now with an average new SAT of ~1480, is a ripple effect based on high admission stats.

I don’t think you understand, or maybe you choose not to believe. Northeastern likely does not report the SAT scores of at least 25% of it’s incoming students (NUin 17% + Internationals 19%).

So while you are correct that the students who are represented by the “average new SAT of ~1480” have a higher probability of graduating in 6 years, this will not ripple down to the overall 6-year grad rate, unless they are also excluding these same 25%+ in grad rate calculations.

I like what NEU has done recently and think that it’s increase in ranking is largely due to positive changes that improve the student experience and education. And undoubtedly the top 25% of students on campus have a higher academic profile than 10+ years ago. But NUin and a relatively large international population are prime examples of “gaming” reported SAT scores, so I’m not sure I’d be touting them.

From Wikipedia:

Students: 8,088
Undergraduates: 5,391
Postgraduates: 2,697

The just-over-one-third of Princeton’s student population in grad school there (not to mention my family and friends with Princeton graduate degrees) would, I think, be surprised to hear that undergrad education is the school’s “one job”.

Follow the money if you want to truly understand how rankings impact colleges.

Follow the money if you want to learn how different metrics ripple to impact the rankings.

@Jwest22
I’m a fan of Princeton. My “Why are they #1?” question was not imbued with doubt; rather, I’m interested in the relative advantages they enjoy in the USNews methodology that cause them to remain #1. You helped to answer that. But – right on… I am aware of some of its well-known attributes.

But that is the point. Princeton’s grad student program IS comparatively small, 1/3 of it’s undergrad. No one is denying the quality of the grad programs it does have, but that allows it to focus more on undergraduate education programs compared to some of its peers

Harvard has over twice as many grad students as undergrads. There is a big difference between 0.3X, and 2X. So much of Harvard’s focus is away from its undergraduate education.

[quote]
I would like to have a conversation around this concept of gaming, now that we have had some time to process the new ranking… Here are some questions that I have. What do folks mean when they say a university is “gaming” the ranking? Do they mean that a university that should obviously be ranked lower is successfully manipulating some ranking metric to rise in the ranking, but the data they provide on these metrics is nonetheless accurate?

[quote]

Yes.

For US News SAT/ACT is 8.125%; 6-yr grad rate is 18% and retention is 4.5%. Those statistics are a big, significant chunks of the ranking. However, the kicker is they are only counted by first year, first semester students.

Therefore if you have a population of students who you think are weaker, who have lower test scores and are also likely to have lower graduation rates and retention rates. Those are students who would “bring down” your ranking. So what can you do? Hide them. Defer them to spring admission.

Those students are still part of your student body, but all of a sudden they no longer impact those major categories of US News rankings. They are invisible.

I call that gaming the ranking because you did not make your student body perform better, you simply hid the weaker students. What their test scores were, whether or not they actually graduate and get their degree, none it matters now.

I mean I think it is pretty self-explanatory why we would consider practices like that gaming the ranking. Universities that don’t do that are counting all their students, universities who do that are only counting the portion of students they want to count.

@londondad @Scipio @UWfromCA In our SoCal area, the view seems to be close to what Scipio posted, with the exception that UCI is much less desirable than UCD and UCSB. Based on our school’s applications, I would not say UCD is viewed greater than UCSB - basically a toss up which is what their average GPA and SAT scores show. UCSD is not seen as the same tier as Cal and UCLA. It’s probably viewed more the leader of the next tier. UCSC is also becoming more favorable at local schools than UCR. UCSC seems to be on a fast rise.

@Dimnarion: You misread the numbers—Princeton’s grad student population isn’t one-third the size of its undergrad population, it’s slightly more than one-third of the total student population; this means that the grad student population is more than half the number of undergrads.

This is quite in line with many R1 institutions. (And, of course, a much—in many cases, effectively infinitely—larger proportion than most schools, particularly LACs.)

Right you are, but that doesn’t change anything. I was just pointing out compared to some peers (ie Harvard who Princeton keeps beating for the number one spot), Princeton as an institution has a greater focus on undergrads.

Engineering Schools rated lower
The graduation rate performance metric seems to systematically lower the rankings of schools with significant engineering programs. Most schools with significant, rigorous engineering programs have a negative score: MIT, Cal Tech, Stanford, Penn, Columbia, Northwestern, Berkeley, Carnegie Mellon, Case Western, Georgia Tech, RPI, Lehigh etc. A few like schools with significant engineering programs manage to get up to a zero: Cornell and Duke.

I think they are trying to identify schools who do a better job supporting students to graduate, which seems great. Engineering programs are rigorous, and do have somewhat lower graduation rates than most other majors. I do not think it makes sense to penalize schools who offer them. If they had a metric that gave them credit for the higher salaries that those engineering majors earn, then this could make more sense, but the rankings do not include that.

I consider it gaming the rankings when you take actions that have very little to do with the academic experience or quality of students who attend but have a major impact on the rankings. Because the US News formula is well understood, there are as many different ways to do that as the inventive minds of man can come up with.

For example, off the top of my head I can think of numerous ways to improve your reported SAT/ACT scores. You can defer lower scoring accepted students to the spring term, where they won’t count in your stats. USC and Northeastern are examples of doing that.

Alternatively, you can become test optional, so that the higher scoring admitted students will have submitted their scores to be reported and the lower scoring applicants will not. George Washington, Bowdoin and Middlebury do this.

Alternatively, you can cut back on socio-diversity and accept a larger percentage of wealthier students who have higher SAT scores because they can afford to do months or years of private test tutoring. Those students are not actually smarter/more accomplished/better prepared than their peers, but their test scores will be higher on average, because expensive SAT prep absolutely does raise SAT scores. Wash U, Wake Forest and Washington & Lee College all appear to do that if their socio-economic diversity is any indication.

Alternatively, you can pay admitted students to retake the SAT after they already have been accepted, solely to pump up your average SAT numbers. Baylor did that.

Alternatively, you can chuck holistic admissions out the window and just emphasize SAT scores over other admissions criteria that are not as easily compared between schools. You didn’t take difficult courses, your extracurriculars are weak, your recommendations are not enthusiastic, your essay was banal - who cares - you smoked the SAT. Many schools are accused of doing this.

Heck, you can just lie and report false numbers. There is no verification process. Emory, Claremont McKenna, George Washington, Bucknell and others have admitted to doing this, and I’m sure plenty of other schools are doing it right now. US News does not hold it against you if you get caught cheating.

Many aspects of the US News formula can be manipulated in similar fashion.