What college rankings really tell us

http://www.newyorker.com/reporting/2011/02/14/110214fa_fact_gladwell?currentPage=all

by boh • 14 years ago

76 70 14 years ago

I graduated from University of Washington, a school that ranks 41st on that list. My roommate in my freshman year of college was Chinese. His dad, a college professor in Hong Kong, always gave him shit for not attending a university that placed higher in the rankings. This is despite the fact that my roommate was a computer science major, and UW has one of the best Comp-Sci programs in the nation. It always seemed to me that his dad cared more about the prestige factor than the actual quality of education his son received. It wouldn't surprise me to find out he has dreams about bragging to his friends that his son went to Harvard or Princeton.

What a shame.

sliverstorm• 14 years ago

This is despite the fact that my roommate was a computer science major, and UW has one of the best Comp-Sci programs in the nation.

The important question is whether his father would have preferred he attend a higher-ranked school with an inferior CS program, or whether he just wanted his son at a "better" school with a similarly spectacular CS program.

CountHackulus• 14 years ago

Speaking specifically of computer science programs, in eastern Canada (mostly Quebec and Ontario), there's a computer science contest called the CS Games. It's mostly just for fun and networking, but it's interesting to see where different universities ranked in different events. Relatively unknown universities like Carleton (my alma mater) placed excellently in extreme programming and algorithms, but not so well in debugging and shell scripting.

But to draw a parallel to your comment, both Sherbrooke and Harvard attended. Sherbrooke being relatively unknown outside of computer science circles. Well that year, Sherbrooke absolutely dominated every competition placing 4th or better in every event, and 1st overall. Harvard however placed last nearly every event (except AI where they were in the middle of the pack).

Looking at the overall reputations of both of these universities, Harvard is by far the more well known and respected. Just like the article was saying, overall rankings aren't very useful, and that if you're serious about getting a good education, you'll have to dig deeper.

kevhsu5• 14 years ago

Well to be fair Harvard isn't ranked that highly for CS, and this single data point doesn't say a whole lot about it either. You'll find CS undergrads that are willing to enter, but would place last in this competition at MIT, Stanford, Caltech, Carnegie Mellon.... every school.

Just playing devil's advocate here. Rankings are a crapshoot, and I don't love the Ivies at all.

capsule_toy• 14 years ago

A competition isn't necessarily the best metric b/c the competitors are going to be a biased subset of the overall student body, and may not actually represent the quality of the education. I didn't go to a great school but there were still some very brilliant people there.

kevhsu5• 14 years ago

As an Asian American with parents that broke most of the stereotypes regarding "tiger moms", I hate to hear stories like this. My blood boils when I hear Asian parents talking about how they only want their kids to go to either Stanford, MIT or one of the Ivies so they can become a doctor, lawyer, or engineer.

With the crapshoot that is college admissions nowadays, it's completely unreasonable to expect your kid to go to such a small list of schools. Especially when there are a few thousand other 2400 SAT, valedictorian Asian Americans with identical goals...

mng22• 14 years ago

It's just a Chinese Parent thing. I went to Berkeley (EECS) but my mom is always telling me about her friend's son who's doing a PhD at Harvard, or her friend's son who's a director at some chip firm, etc. You get used to it after a while.

SiVal• 14 years ago

If it makes you feel any better: I get my hair cut at a place here in Silicon Valley that's full of Chinese tiger moms (incl. the staff). When I was sitting in the chair a few weeks ago, a tiger mom came in the shop with a book of 8x10 photos. She went from Chinese woman to Chinese woman, holding the book open and announcing, "This is my son. Here he is [pointing at the photos, and slowly turning the pages] at his graduation last week. He graduated from Berkeley in electrical engineering." Then, pausing for effect, turning her head and nodding slowly and meaningfully, added, "...with honors."

She would then move down the line to the next woman and repeat the whole thing verbatim, down to the pause, slow meaningful nod, and "...with honors."

Most women she spoke to tried to look unimpressed, which I thought was a bit rude. Some even sniffed and turned away. It was amazing.

I was the lone white guy in a room full of Chinese women (and some kids) and apparently the only one who was happy for her and willing to look at her pictures. No luck. When she got to me, she sniffed at me and skipped past me to the next Chinese woman.

I wasn't sure whether that was because my opinion was irrelevant to her sense of social status or, more simply, that she knew the Chinese women but didn't know me.

I asked my hair stylist who, like most of the staff, has been a friend of mine for years. She told me that she had seen the woman a couple of times but that she didn't think anybody in the shop knew her by name. I (the only one in the room with a EE background and apparently the only one happy for her) was just not somebody whose opinion mattered to her.

So it seems that your Berkeley EECS is plenty good enough to ruffle the feathers of other Chinese moms and to impress me--if that helps any. ;-)

cupwithyourname• 14 years ago

Doesn't US news have per-degree rankings? If not, taking a more "objective" ranking in computer science of the number of paper awards received at CS conferences [http://jeffhuang.com/best_paper_awards.html#institutions], UW easily places in #3, among Stanford, CMU, and MIT. Maybe the roommate could show his dad that...

duaneb4• 14 years ago

Yes; they publish degree-specific rankings (at least for grad school). However, it remains that people will still consider UW inferior to grad schools even if you went there for CS. When you tell people what school you went to, they're not going to cross reference it with your degree.

santigepigon• 14 years ago

It always seemed to me that his dad cared more about the prestige factor than the actual quality of education his son received.

Sadly, the prestige factor doesn't end when one enters college. My sister and I are Asian American and our parents emigrated from the Philippines. My sister's majoring in Urban Studies at Stanford. Despite this, my father insists on telling family, friends, co-workers, etc. that she's studying law.

Quite a shame indeed.

prezjordan2• 14 years ago

I attend a small tech school outside of NYC, and it's not terribly well-known, and not terribly well-ranked. I was a good student in high school, and got into schools ranked much higher than the one I attend now. Why? I love the location here, and I love the "personal" factor that comes with a small school. I've met lots of interesting people, I live in a beautiful town with beautiful views, and I really enjoy what I'm learning. It just "felt right" and it still does today.

Some days I get a little upset that I don't attend a prestigious university, and I'm not really sure why. There's always grad school, though. Anyway, just a personal experience.

kevhsu5• 14 years ago

RPI? RIT? Those are my guesses...

duaneb4• 14 years ago

IIRC RPI is nowhere near NYC.

daveed• 14 years ago

His profile says: Junior CS student at Stevens Institute in Hoboken, NJ

prezjordan2• 14 years ago

Those were the schools I got into that are ranked higher than where I go now (Stevens Institute - Hoboken). More people know them by name, but even beyond that lies CMU (I was waitlisted, and I'm hoping to maybe pursue a masters there).

ameasure• 14 years ago

I went to a solidly 4th tier university (out of 4 tiers, as I recall). I now work and regularly interact with people from the best universities, Harvard, Princeton, etc, and I can tell you that as far as knowledge and ability are concerned, they have no advantage.

If you're committed to learning and improving yourself, you don't need the ivy league's approval.

Alex39172• 14 years ago

What this article doesn't mention is how much effort the colleges at the top put into gaming the rankings. For example, Harvard spends millions of dollars every year sending direct mail to graduating seniors who they know have no chance of getting in, encouraging them to apply. Why? To increase their rejection rate.

Similarly, Stanford was considered a relatively mediocre school until one year when they rejected every single valedictorian across the country as a marketing stunt, and suddenly jumped in the rankings after guidance counselors started encouraging the kids with the highest grades to apply there.

Similarly, the admissions departments get an update twice a day telling them their projected US News rank based on the grades and SAT scores of the kids who they are admitting. Thus your chances of getting in can vary wildly based on whether your application is read in the morning or afternoon. If the projected rank drops too low then they stop even reading the essays and just admit solely based on grades and SATs until the rankings are back up.

austenallred• 14 years ago

Exactly. Brigham Young University (my university) is an absolute slave to the rankings. They intentionally limit the acceptance rates into certain programs even when there's open space for numbers' sake. If these rankings disappeared and universities operated only for the sake of its students, I think the overall quality of education would actually improve.

gliese1337• 14 years ago

That just further enforces the point that you can't decide on what the best college for you would be by looking at any kind of ranking for the university as a whole. Different colleges and different departments within the university can be run so radically differently that it hardly matters that they're officially part of the same institution.

harryh• 14 years ago

Do you have a citation for the Stanford valedictorian thing? I can't find anything on that topic.

Alex39172• 14 years ago

I heard it somewhere, I can't find a cite either. It's not the kind of thing they would ever admit to doing, so I'm not sure it would even really be possible to find a cite.

ChuckMcM• 14 years ago

Reed College (Portland) goes out of its way to not be ranked. Colin Diver, the outgoing President, suggested as others have, that chasing a rank doesn't do the students any good.

[1] http://www.reed.edu/apply/news_and_articles/college_rankings...

rickdale• 14 years ago

This article brings up some really good points. Ultimately schools are obsessed with rankings because alum and prospective students are also obsessed with rankings. The rankings do a really good job of enthralling the consumer and making it competitive to get into college.

I graduated in 2009. In that same year we got a new president. He pledged to move the school up in the rankings from 22 in a welcome speech. The crowd gave a standing ovation.

bentlegen• 14 years ago

On the subjectiveness of reputation:

I once met a recruiter for the University of Toronto (Canada) who focuses exclusively on visiting US high schools. She told me how U of T was behind other "top" Canadian schools in advertising to potential students in the US. She mentioned how McGill University (in Montreal) started going after US students aggressively beginning in the 70s, and as a result, has a far better reputation there today. Even The Simpsons once joked that McGill is "the harvard of Canada", even though most Canadians would probably say otherwise.

Clip: http://www.youtube.com/watch?v=cc5vN2XReWs

tikhonj• 14 years ago

I suspect that getting US students is important because they pay significantly more than students from Canada itself. I'm not sure exactly how it works in Canada, but at least at my US public university out-of-state and international students pay far more than instate students which certainly doesn't hurt the budget. Since the US is really close, has expensive education domestically and has a large population, it's probably relatively easy to convince high-paying US students to come to universities in Canada.

crusso• 14 years ago

If you're a slave to the school rankings in choosing a school, you get what you deserve.

bane• 14 years ago

I worked for a great guy who also happened to be on the board of a reasonably well respected tech university in the South Western U.S. Every year they have a meeting about doing the kinds of things that boost their rankings in US News and every year they've decided to not do those things because of how it would hurt what they feel to be their academic mission - educating students.

He said that almost everything that you can do to boost your ratings are gimmick events like rejecting more applicants, or increasing spending on getting applicants.

One particularly greusome idea was to hire national survey companies to hang outside of popular teen events and get students to fill out a preliminary application that they would then reject to get boost their rejection percent and appear more "competitive".

They keep rejecting these kinds of meaningless shows of status.

But the lure of prestige keeps this meeting on the agenda anyways.

tkahn62• 14 years ago

I predict the obsession with college and therefore collegiate rankings will dissipate in the next 50 years, maybe even by the close of this decade.

Consider what a college experience offers more readily over what you can accomplish with a high-speed internet connection.

- a community of people around the same age with no liabilities or responsibilities

- guaranteed access to an expert on the subject who can answer direct questions

- verification of knowledge

Anything else?

Why do we need thousands of college professors giving lectures on linear algebra each semester? We have the bandwidth to distribute a complete video lecture series from MIT to all college-going people in the US.

The collegiate model is predicated on the inaccessibility and uneven distribution of information. That's why the top colleges have the biggest libraries. That's how information was primarily stored and accessed and catalogued.

The developed world is a completely different place now.

Locke1689• 14 years ago

The importance of college won't change until hiring practices do. Most employers don't actually test most of their candidates on their skills. Until they do they need a way to distinguish their candidates. Hiring from Harvard is a good way to do it since you can outsource your requirements to the Harvard acceptance department.

kevhsu5• 14 years ago

Well at least in ECE/CS fields, we have the luxury of more liberal hiring practices in which personal projects on resumes and githubs can matter just as much as what school you went to if not more.

It's much harder to measure the quality of candidate in other fields without the context of what school they went to and how well they did there. (Unless, of course, you have time for a full-on interview. I think parent is referring to the screening process though)

aswanson• 14 years ago

CS and engineering jobs must exist in a different universe than the rest of the working world. I have yet to go through an interview process that did not require a verbal/written series of technical questions being answered.

liber82• 14 years ago

You're missing a category because you're focused on the benefits from the students' perspective. Look at it from an employer's perspective:

You have a 100 job applications for one open position. Most of the kids went to mediocre to good colleges. Two or three of the kids went to Harvard/Stanford/other selective college. Who do you offer an interview to?

Keep in mind, you're busy running a business. You don't have time to thoroughly analyze each application. You figure, if they're good enough for Harvard/Stanford/other selective college, they're good enough for me. So you start by interviewing the Harvard kids.

This isn't irrational. Harvard/Stanford/etc. spend a ton of money screening hundreds of thousands of the nations' brightest kids. If you make it into one of those schools, there's a very, very good chance you're at least competent. Elite schools (as defined by things like USNWR) will always serve this function, so kids who want to follow this traditional path to success will always flock to them, or to anything that manages to offer the same signaling/screening effect.

tkahn62• 14 years ago

Sure this is a good point. You're referring to the the brand and stamp of approval a degree represents.

I think that's really an extension of the 'skill verification' bullet point. One can easily imagine a program run by a trusted institution whereby a person pays a fee and they are evaluated over the course of a week or so. If they pass, they are given a stamp of approval, if not they either try to improve their skills or go to a lower-prestige institution and repeat the process.

kenjackson• 14 years ago

The USNWR rankings are especially insidious because they're also not objective.

I wish I could find the citation for this, but I recall reading about how when it first started the person who created the ranking formula showed USNWR the list of top 10 schools and the Ivy League schools didn't do as well as expected. If HYP weren't at the top there was a flaw with the methodology, so they had to go back and get a new methodology that would conform with conventional wisdom.

Also back in '99 Caltech took the top spot. Again this was viewed with some disdain from the East Coast elite. They changed their methodology and added/emphasized a metric they call "value add", which is the difference between actual and expected graduation rates. Because CalTech is difficult to graduate from, they were seriously penalized.

IOW, rankings are often simply a way to perpetuate the status quo, but make it look objective.

_delirium• 14 years ago

Some mathematicians did a reanalysis a few years ago where they investigated how sensitive the rankings were to the particular choice of weightings. They found that the answer is "very sensitive", to the extent that you can produce a wide range of desired outcomes by carefully choosing the weights to assign to each criterion: http://www.sciencenews.org/view/generic/id/37207/title/Math_...

ajross• 14 years ago

I think your broad point about "elite" schools being a matter of reputation rather than objective measurement is basically right. But some of your analysis seems weird.

On the graduation rate: why shouldn't a school be penalized for a low graduation rate? The point of those rankings (beyond selling magazines, of course) is to give prospective students an idea of how each school will help them in life. If I have a choice of going to school A from which I'm 90% likely to graduate and an otherwise equally ranked school B which is 50% likely to kick me out without a diploma, which am I going to choose?

And the bit about CalTech seems a little like sour grapes. I was at one of those "East Coast elite" schools in the 90's and don't remember anyone thinking anything but good things about CalTech (Pasadena, on the other hand...). Basically within its fairly narrow realm of focus CT is absolutely one of the "elite" schools everyone looks to.

gms7777• 14 years ago

> On the graduation rate: why shouldn't a school be penalized for a low graduation rate? The point of those rankings (beyond selling magazines, of course) is to give prospective students an idea of how each school will help them in life.

Its easy to look at this from the other side though.

I think a lot of people look at these rankings as the value of a degree earned from that institution. That is, they are looking at it from the side of the alums as opposed to the prospects. Suppose two schools have the same admissions criteria and everything else, but the graduation rate of one is much higher. I would think the school with the lower graduation rate would be viewed as more rigorous and therefore "better".

majormajor• 14 years ago

"If I have a choice of going to school A from which I'm 90% likely to graduate and an otherwise equally ranked school B which is 50% likely to kick me out without a diploma, which am I going to choose?"

The "otherwise equally ranked school" part shows where a single-dimension ranking system falls on its face. Does it mean they accept the same quality of applicant? If so, graduating from the school with the lower graduation rate might be a much more impressive accomplishment. Or does it mean that the quality of education graduates receive is considered to be basically the same? Does that then mean the second one is just easier to get into?

How would all colleges having high graduation rates avoid pigeon-holing people even further into the box that their high-school performance marked them for? If everybody can graduate, the potential signaling is reduced from "got into and graduated from school x" to just "got in to school x."

duaneb4• 14 years ago

IMHO what weights are the right weights is very subjective - reducing all these colleges to a single comparison is stupid.

For example, if there is 100% graduation, I consider that very suspect and probably a bad sign. It's natural that some people don't finish college. WRT Caltech, they are desirable BECAUSE they offer such a challenging environment.

kafkaesque• 14 years ago

These rankings tend to not be objective, I agree. The sad part is they get away with it by simply defending their methodology of choice (emphasising certain areas).

This is why my alma mater, University of British Columbia (in Canada), opted out of our version of a university ranking, Maclean's. The link is old, but you can read a bit about it here: http://www.cbc.ca/news/canada/ottawa/story/2006/08/14/maclea...

Currently, I am not sure whether or not UBC has opted back in, however.

larrys2• 14 years ago

"If HYP weren't at the top there was a flaw with the methodology, so they had to go back and get a new methodology that would conform with conventional wisdom."

Most importantly, I think it's important to understand that if HYP (and others in the Ivy League) weren't in the ranking nobody would take the ranking seriously.

So you have to have a smattering at least of what people are expecting to see or you don't have a list worth considering and reading.

If you are going to have a list of anything, in order to elevate lesser known members of the list, you have to have credible proof of known entities that everyone can relate to.

For example take the Rolling Stone list of the 100 greatest acts of all time. Particularly the top 10:

http://www.rollingstone.com/music/lists/100-greatest-artists...

The top four positions are taken by the Beatles, Bob Dylan, Elvis, and the Rolling Stones.

Can you imagine in this list if those four didn't appear until position 71 85 93 99?

Nobody would believe the rest of the list.

api• 14 years ago

I love how everyone responsible for ranking colleges lives along the North-Eastern seaboard. That area is such a bubble of regional self-importance.

jmduke2• 14 years ago

True, but yesterday a post titled 'How Coffee Meetings Power Silicon Valley' hit the front page.

Every region has people who are self-important and people who are not.

mullr• 14 years ago

I wonder where the data for these rankings comes from. Is it publicly available? It would be nice to have a dynamic ranker which prioritizes the things you care about.

jmduke2• 14 years ago

http://en.wikipedia.org/wiki/List_of_colleges_and_universiti...

Sort by total endowment per student. Cross-reference this list with the USNWR rankings. The overlaps are striking.

justaddwater57• 14 years ago

Interesting that Tulane is called out here as an example:

"One common statistic used to evaluate colleges, for example, is called “graduation rate performance,” which compares a school’s actual graduation rate with its predicted graduation rate given the socioeconomic status and the test scores of its incoming freshman class.... Tulane, given the qualifications of the students that it admits, ought to have a graduation rate of eighty-seven per cent; its actual 2009 graduation rate was seventy-three per cent. That shortfall suggests that something is amiss at Tulane."

Anyone else think that those numbers for the 2009 graduating class has anything to do with the fact that Hurricane Katrina hit during the first few weeks of '09's freshman year, forcing students at Tulane (which is located in New Orleans) to take leaves of absences or transfers?

Just goes to show, in addition to inherent biases, rankings also don't capture extenuating circumstances that can have drastic short term effects that go unnoticed.

sfny• 14 years ago

Have you read Gladwell's book "Outliers"? Basically deals with the situational factors that create outliers like Bill Gates. Circumstances have incredible life-determining effect that go totally unnoticed too.

RedwoodCity2• 14 years ago

Hurricane, or perhaps Tulane is a party school.

geebee• 14 years ago

The Washington Monthly had an interesting angle on college rankings, and how different their results are from US News:

http://www.washingtonmonthly.com/magazine/septemberoctober_2...

mnemonicsloth• 14 years ago

    (sort-by  
       (fn [u]
         (* (:age u)
            (:tuition u)
            (if (rhymes-with (:name u) 'Warvard)
               1
               (+ 0.9 (rand 0.1))))
       univs)

appleflaxen• 14 years ago

I liked malcolm gladwell a lot more before I found out he is a big tobacco shill.

http://exiledonline.com/malcolm-gladwell-tobacco-industry-sh...

sigil• 14 years ago

Watch for recommender systems to disrupt and replace many of these rating systems.

As Gladwell pointed out in the Car & Driver example, the "scores" assigned in these rating systems are arbitrary linear combinations of factors chosen by "experts." You the car buyer will personally care more or less about certain factors than they do, but they've already chosen the coefficients for you, so too bad. Same for the college ratings assigned by the USNWR.

I think we're asking the wrong questions with college rankings. Prospective students should be asking, what education will maximize the expected value to me personally in the future? It's not a heterogenous and inexplicable "score" they're looking for, it's the probability that the match between student and college will be a productive one. Given data about individual students, colleges, and historical outcomes, surely a better predictive model is possible.

RedwoodCity2• 14 years ago

Thank You to who ever posted this article. It points out how stupid college/car/etc... rankings are. I think people have a natural tendency to want to rank things, in the end it just makes us unhappy.

What college rankings really tell us

Related Stories