The world is full of lists that purport to help high school seniors pick a college and help new college graduates pick a graduate institution, but now a German team has come up with a way to rank universities using what they call “spatial scientometrics.” I’m not sure about that word, but their method judges organizations by both excellence and geography, the MIT Technology Review reports.
What they do is fairly simple and easy to understand: They simply count up all the peer-reviewed papers published by people who work at a given university in each discipline and then factor in the number of times those papers were cited in other peer-reviewed papers.
A few problems jump out right off the bat: First, some research generates many, many papers every year, while other research generates a paper a year—maybe even less.
Second, the site makes the following claim:
This web application allows to find [sic] out which institutions (universities or research-focused institutions) are most active in different subject areas of science and which have published the most excellent papers.
Being “active” is one thing, but it’s just a fact that papers are often “cited” by researchers because they’re wrong and being refuted, not always because they’re “excellent.”
So, there are just a few flaws in the methodology—and a small grammatical error in the introduction—but don’t let that stop you in this case. I’ve been known to leave out a word every now and then, even after heavy editing, and the concept behind the site is intriguing, to say the least.
In the long run, I suspect the variations in publication- and citation-counting balance out, as the rankings are mostly predictable. For example, in genetics and molecular biology, the top three institutions are the Broad Institute of Harvard and MIT, the Howard Hughes Medical Institute, and Cold Spring Harbour Laboratory.
The tool, developed by Lutz Bornmann at the Administrative Headquarters of the Max Planck Society in Germany “and a few pals,” can be accessed here, but the application requires a password, which you’ll have to get by email. There’s no automatic way to do it.
The site is visual, presenting a world map that can zoom in on the institutions of interest. It was very interesting to see where the clusters of top-published and top-cited institutions are in the world.