"Physics Explains Why University Rankings Won't Change"

astrife · February 19, 2008, 5:19pm

[Physics</a> Explains Why University Rankings Won’t Change](<a href=“http://www.sciencedaily.com/releases/2008/02/080215204726.htm]Physics”>Physics Explains Why University Rankings Won't Change | ScienceDaily)

Physics Explains Why University Rankings Won’t Change
ScienceDaily (Feb. 18, 2008) A Duke University researcher says that his physics theory, which has been applied to everything from global climate to traffic patterns, can also explain another trend: why university rankings tend not to change very much from year to year.

Like branching river channels across the earth’s surface, universities are part of a relatively rigid network that is predictable based on “constructal theory,” which describes the shapes of flows in nature, argues Adrian Bejan, J. A. Jones Professor of Mechanical Engineering at Dukes Pratt School of Engineering.

According to the theory, the hierarchy of university rankings – in which few schools consistently land at the top and many more contend for lesser spots – persists because that structure supports the easiest flow of ideas, Bejan reported.*

“This hierarchy is here to stay,” Bejan said in an interview. “The schools at the top serve everybody well because they serve the flow of ideas. We’re all connected.”

That structure also allows talent to flow and arise naturally in the “right places,” he said.

First conceived by Bejan and published in 1996, the constructal law arises from the natural tendency of flow systems to evolve over time into configurations that make their movements faster and easier.

More recently, Bejan and Gilbert Merkx, also of Duke, co-edited a book entitled “Constructal Theory of Social Dynamics,” including a collection of essays applying the tree-like patterns of constructal theory to business, crowd dynamics, legal systems and written languages, among other human endeavors.

In extending the theory to university rankings, the first step was to define the flow system of the university, Bejan said, “what territory it covers, and what currents flow through it.”

He suspected that a school’s rank might reflect the flow of the ideas its faculty members generate. In support of that notion, he found that the most highly ranked engineering schools are also those with the most people on the Institute of Scientific Information’s most-cited listing, meaning that their work is more often referenced by other researchers.

He also found that university rankings follow a hierarchical pattern that mirrors the distribution of city sizes. The more highly ranked a university or larger a city, the fewer competitors it has. The opposite is also true: the lower the rank, the more numerous are the candidates that compete for that position.

“The similarity is further evidence that the distribution of sources of knowledge is intimately tied to geography,” he said, and to the flow of information across the globe.

So, is there a way to change rankings? In Bejan’s view there is, but he says it takes “cataclysmic” events that encourage the free flow of ideas to alter such deeply ingrained channels. Such shifts have occurred in the past, he noted. For instance, a “brain drain” from post-war Europe after World War II led to significant changes in the academic landscape, catapulting American universities onto the world stage. Similar shifts were also seen after the launching of Sputnik, with the enormous jump in funding for basic science, he added.

“The university is the professors, their disciples, and the disciples’ disciples,” Bejan wrote. “It is the ideas that flow through these human links and into the books of our evolving science and culture. In time, this global vasculature evolves like a river basin during the rainy season: all the streams swell, but their hierarchy remains the same.”

This research was recently published in the International Journal of Design & Nature and Ecodynamics, referenced as Vol. 2, No. 4, (2007) 319-327. For more on constructal theory, see

Adapted from materials provided by Duke University.

tokenadult · February 19, 2008, 6:16pm

Original Duke University Office of News & Communications press release:

[Physics</a> Explains Why University Rankings Won’t Change](<a href=“http://news.duke.edu/2008/02/rankbejan.html]Physics”>http://news.duke.edu/2008/02/rankbejan.html)

debate_addict · February 19, 2008, 7:04pm

I disagree with one thing:

The structure might not change, but the admissions system in my opinion does not accomplish the above. I’m sure everyone knows somebody exceptionally talented who did not flow in his or her “right position” in the hierarchy, but in a “lower” one. Also, I’m sure we all know somebody who is not particularly gifted but managed to “flow in” a higher position in the hierarchy.

But then again I might have misread the quotation
I’m not 100% percent sure that by talent and flows in the right places he is referring to with admissions and students who enroll, but it seems like it…

fumarole · February 19, 2008, 7:38pm

This ‘law’ doesn’t explain anything that we don’t already know: that universities are committed to spending the money and resources to stay atop the rankings, and as long as they have a vested interest in holding on to these spots, not much will change.

It’s absurd to overgeneralise a situation that is about as cut and dry as you can get and apply something that’s only tangentially related to describe what doesn’t need further description.

MidwestMom2Kids_1 · February 19, 2008, 7:56pm

Debate Addict, I haven’t read the paper but I am confident that he doesn’t mean the admissions department and the freshman students. He means the researchers: professors, post docs, grad students. I think he is saying that the best researchers “flow” to the best institutions. To which one might say, Duh.

There best instititutions have the best labs/faculties and the most money and attract the best professors and then those good professors attract other good professors who want to be near them and then they win grants/write books and get more money and become more famous and get better labs and attract more outstanding colleagues. It is circular. (There are exceptions of course!)

I completely agree with your other point: the “assignment” of freshmen to institution has a great deal of variability and randomness to it.

HomeschoolDad · February 19, 2008, 10:27pm

There’s also the concept that what you learned in the past tends to color your judgment if you are not informed of changes. The University of Washington’s USNWR ranking, for example, has remained stuck near the bottom of the first tier for many years.

But this university has 6 current Nobel Laureates on faculty, whose prize-winning research was done on campus. Duke lured Peter Agre from Hopkins, but no Nobel-Prize winning research has ever been done at Duke. Chcago boasts of having more Nobels associated with it than any other American university, but in fact it currently has only 1 Nobel Laureate (prizes established in Alfred Nobel’s will) and 5 Bank of Sweden prize winners in economics, who are colloquially, but incorrectly called “Nobel Prize” winners.

In the past 10 years, the UW has had 13 professors elected to the National Academy of Sciences, which would rank 4th in the Ivy League after Harvard, Princeton and Yale for recent elections.

The UW’s federal research budget in science and engineering is enormous. Its information science and biomedical programs are top-notch–and these are critically important fields in our 21st century economy. The UW has attracted really bright faculty who’ve decided that Washington state, with its beautiful land and seascapes, and world-renowned outdoor recreation, is more attractive than the East Coast or Midwest.

For students who relish challenge, there are abundant honors courses and research opportunities that can make for an extraordinary journey of discovery. If the USNWR rankings gave greater weight to science and technology than social sciences and humanities, the UW might very well rank in the top 10.

sofla951 · February 19, 2008, 11:21pm

not to sound like a prick but aren’t there more important things to be doing research on in physics…

otis · February 19, 2008, 11:38pm

HomeschoolDad:

You are correct that he University of Washington is a fine school. Any student should be proud to attend and graduate from UWashington. My understanding is that Mr. Gates and company have made generous contributions that have allowed UWashington to make strides in bolstering its already solid academics.

The above being stated I must respectfully disagree with your statement that “UW might very well rank in the top 10” among American Universities.

The reality is that the qualifications of the incoming students although very good does not compare favorably with elite American colleges.

Washington’s overall undergrad acceptance rate for the frosh class of 2006 was a less than selective 68%.

The SAT scores of Washington’s frosh class of 2006 ranked 69th among US Universities (as reported by USNews) with Washington’s SAT/ACT 25th-75th percentile ('06) being 1070-1310.

Washington’s SAT range lagged behind University of San Diego, SUNYStony Brook, Clark University, & Clemson University.

Please do not take the above stats as my dissing Washington but rather as a caution that distinguished professors are only one of many factors of a colleges’ worth.

[ps Go Huskies beat Cal]

tomslawsky · February 19, 2008, 11:48pm

What a surprise- a prof from DUKE explaining why his school will always be considered a top ranked school. Consider the source. A better and more clear explanation is that US NEWS rankings have a HUGE “soft” factor and people’s pre-conceptions are hard to change. For instance- ask the average Joe what is the best school in the country, they’ll blirt out “Harvard”. Ask them why, specifically and they’ll have no clue except that’s what has been seared into their head.

ilovebagels · February 20, 2008, 12:26am

tomslawsky, I think the importance and self-perpetuating nature of this “soft” factor is exactly what the professor’s thesis is arguing.

Mr_Payne · February 20, 2008, 12:59am

It’s possible to make a new, good, school. It just takes lots of money.

Free tuition, great faculty. The students will come.

UCBChemEGrad · February 20, 2008, 1:04am

^ Hmmmm… I think it takes a while to build up academic cache, so to speak, and attract top students and faculty. It can’t be as easy as you say…although money makes it a lot easier.

Mr_Payne · February 20, 2008, 1:32am

Olin did just that!

Colodia · February 20, 2008, 1:54am

It’s all gnomes. Look it up.

Calcruzer · February 20, 2008, 3:48am

otis, perhaps you would explain to me how a school’s selectivity ranking has any effect whatsoever on its academic ability.

According to this theory–the better the students, the better the education.

It’s a theory I would agree with if things like top-ranked faculty with knowledge of their fields, available facilities and research centers, alumni advisory committees, great visiting professors, wonderful immersion programs, great internship and research opportunities, a good placement office, good event scheduling, a wonderful athletic and social scene, housing facilities, good dining facilities, fantastic honors programs, proximity to research or innovative businesses, and a good counseling crew didn’t somehow seem a bit more important in determining how good a school actually is.

The difference between the students at the #1 school on USNW’s list and the #50 school on USNW’s list in terms of how good the students are is essentially negligible anyway (0.2 to 0.3 GPA?). Why one thinks that the ability of students (in their high school years)is the determining value in providing the best college education is certainly beyond my comprehension.

Calcruzer · February 20, 2008, 3:54am

I have to question the Duke professor’s argument.

Take Business Week, Princeton Review’s, and the Wall Street Journal’s rankings and compare them to the USNW rankings. You’ll see that once you remove some of the stupid things that USNW considers–such as alumni giving rates (which is naturally higher at private schools than at public schools), return rates of freshmen (which depends not just on ability of the school, but also financial aid provided and the economics of the students accepted), and selectivity rankings (which is based entirely upon the “soft” name recognition from year after year of being repeated)–then the rankings of the schools do change dramatically year-to-year.

Then, we finally begin to see real results based upon actual ability and not on incidentals. Schools that USNW doesn’t rank as highly–such as UVA and BYU–rank in the top 10 undergraduate business schools in Business Week’s ranking. In such rankings (such as Princeton Review’s ranking), Indiana University ranks as the school with the top business faculty in the country, Purdue ranks as the top aerospace engineering school in the country, a school like UCSD ranks first in economics and neuroscience, and a school like Julliard is ranked #1 in music. In these surveys, such as the one from the Wall street Journal, Georgia Tech gets ranked first in electrical engineering, and Dartmouth gets ranked higher than Wharton as the top overall MBA school. Schools like Ohio State rank in the top 5 medical school programs–and even a school like McDaniel College in western Maryland gets ranked as the top school for sending students to medical school–and a school like Emory, with its professors including a former president, Desmond Tutu, the Dalai Lama, and Salman Rushdie (of “The Satanic Verses” fame), as well as top law and medical experts gets ranked as having the most prominent faculty.

The only thing keeping the Duke professor’s theory intact is economic factors–which is what ilovebagels and Mr. Payne have pointed out–with Mr. Payne giving a perfect example of how economics has moved a “brand-new” university to the top in less than 5 years.

DSC · February 20, 2008, 7:22am

Calcruzer -

The best universities are the ‘best’ because they have the best students coming out, attracting the best businesses, having the best research, and so on. The best students coming out are highly correlated with the best students coming in. Therefore, high entrance numbers are a big deal.

otis · February 20, 2008, 8:19am

Calcruzer:

2 points:

<ol>
<li>In response to your question directed to me of “how a school’s selectivity ranking has any effect whatsoever on its academic ability” I direct you to the above response of “DSC”.</li>
<li>The Business Week rankings of Undergrad Business Programs ( see link:[Undergrad</a> B-School Rankings: Interactive Table](<a href=“http://bwnt.businessweek.com/bschools/undergraduate/07rankings/]Undergrad”>http://bwnt.businessweek.com/bschools/undergraduate/07rankings/) ) are flawed because it only lists the schools that offer traditional business programs. Please take a moment to review the listing and you will note that Harvard, Brown, Dartmouth, Yale, Princeton and other elite schools are not included; the reason being that the majority of the elites do not have business departments but rather those schools offer economics majors with classes in finance & entrepreneurial seminars. Don’t kid yourself, there are many, many, many grads from Harvard, Brown, Dartmouth, Yale managing departments at financial heavyweights like Goldman Sachs and the like than the vast majority of the schools listed in the Business Week rankings with the obvious exception of UPenn. Please understand that this observation is not intended to diss the schools included in the Business Week rankings but rather to draw your attention to obvious flaws in the rankings and your apparent on a belief that the rankings are all inclusive.</li>
</ol>

rebelliousduck · February 20, 2008, 9:10am

Sounds like someone has too much time on his hands.

What I find amusing is that schools can make dramatic movement (five or more places) from ranking to ranking. These are large institutions that don’t change that much from year to year. To the degree that an anomaly occurs in an identifiable variable to cause the swing, it should be factored out.

sakky · February 20, 2008, 9:35am

I can immediately think of 2 reasons:

*Students learn from each other. Let’s face it. Most students’ time is not spent interacting with the faculty. It is spent interacting with other students. That is a key component of the educational process. You learn a lot when you’re surrounded by other students who know a lot and who are motivated to learn. On the other hand, when you’re surrounded by students who are not motivated and who don’t know very much, then you tend to learn less. It’s very easy to become lazy yourself when you’re surrounded by other lazy people. {Just like if everybody around you smokes, then you will probably also end up smoking and you will find it extremely difficult to quit smoking.}

*Top faculty tend to be attracted by top students. After all, top faculty would like to have top students working under them. While even the worst school obviously has some top students, it is far easier for a prof to find some top students to work for him if all (or most) of the student body is high quality. Secondly, most profs generally don’t enjoy teaching unmotivated students. Hence, the more motivated the student body, the more desirable the school becomes.

I’ll give you one specific example. I know one Harvard associate professor who is coming up for tenure, which he acknowledges he might not get. Yet he has already said that if he doesn’t get it, he will probably just retire from academia. After all, he doesn’t need to work as his wife is a highly successful hedge fund manager. For the same reason, he isn’t going to move out of the area not just because of his wife’s highly successful career, but because he also has 2 school-age children that he is unwilling to uproot. While he could surely get a tenure-track job at a place like BU or Northeastern, he considers those schools to be undesirable and he specifically cited the quality of the students there as a major reason for why. Hence, if he can’t stay at Harvard, the only local place he would consider working is MIT, but if he gets turned down for tenure at Harvard, it is unlikely he will get a tenure-track offer at MIT. Hence, he will just retire from academia and (I suspect) start his own consulting firm. Either that or join his wife’s hedge fund.

Believe me, there are a lot of people like that. Even now, I hear numerous PhD students saying that if they can’t get placed into a tenure-track position at a top-ranked department, they’ll just go to industry. Granted, these are PhD students who are in fields with industry jobs aplenty (i.e. engineering, business, economics, computer science, finance, molecular biology).

Nevertheless, the point is, academic selectivity does indeed impact academic ability. Note, it is not the only factor involved. But it is a factor.

"Physics Explains Why University Rankings Won't Change"

CONNECT WITH US