The science of a career

It’s the envy of nearly every sci­en­tist: a research posi­tion at Har­vard or MIT, Princeton or Stan­ford. But new results from Albert-​​László Barabási’s team at North­eastern sug­gest that moving to one of these “rich club” research insti­tu­tions might not do all that much for a scientist’s career, at least if physics is your game.

The team, lead by Pierre Dev­ille, a grad­uate stu­dent vis­iting from the Uni­ver­site catholique de Lou­vain in Bel­gium, found that moving from a less suc­cessful to a more suc­cessful research insti­tu­tion doesn’t have a mean­ingful impact on one’s career. On the other hand, though, moving in the reverse direc­tion does cause a slight decrease in a researcher’s long-​​term career suc­cess. The results were pub­lished in the journal Nature Sci­en­tific Reports in April.

How did they figure this out? Well, it wasn’t easy–although they didn’t do it by hand, as I naively inquired in our interview…that would have just been crazy talk. Turns out there’s a con­stantly growing body of data inherent in the research com­mu­nity to which they could apply sophis­ti­cated com­pu­ta­tional algo­rithms to ask ques­tions about the mobility pat­terns of sci­en­tists and the impacts those move­ments have on their success.

That data­base is the col­lec­tion of aca­d­emic jour­nals in which sci­en­tists pub­lish their work. Each time they do so, their cur­rent affil­i­a­tion is stamped into his­tory, giving big data researchers like Dev­ille some­thing to play with. For this project, they sifted through more than 400,000 research papers pub­lished in the main corpus of physics jour­nals during the years 1950 to 2012.

After cleaning up the data (turns out more than one “J. Smith” could have pub­lished a physics paper during that time, and she or he could have been asso­ci­ated with either North­eastern Uni­ver­sity in Boston or North­eastern Uni­ver­sity in Shenyang, China…causing a slightly dif­fi­cult dis­am­bigua­tion problem), the team mapped out the way these researchers moved throughout their careers.

They looked at the number of cita­tions that each researcher had to his or her name and used it as a mea­sure of suc­cess. Then they looked at the number of suc­cessful researchers at dif­ferent insti­tu­tions, to give the uni­ver­si­ties, com­pa­nies, and national labs that employed them their own mea­sure of overall “richness”.

A heat map of institutional "richness." The red area in the bottom left corresponds to the richest institutions, while the top right area corresponds to the poorest institutions (in terms of affiliated physics researchers' success rates). The blue green middle area represents those institutions that are neither extremely rich nor extremely poor. Image courtesy of Pierre Deville.

A heat map of insti­tu­tional “rich­ness.” The red area in the bottom left cor­re­sponds to the richest insti­tu­tions, while the top right area cor­re­sponds to the poorest insti­tu­tions (in terms of affil­i­ated physics researchers’ suc­cess rates). The blue green middle area rep­re­sents those insti­tu­tions that are nei­ther extremely rich nor extremely poor. Image cour­tesy of Pierre Deville.

They cre­ated a heat map of insti­tu­tional rich­ness, shown here. The red cor­ners rep­re­sent the most and least suc­cessful insti­tu­tions, while the blue middle ground rep­re­sents exactly that — a middle ground of insti­tu­tions that are nei­ther super suc­cessful nor super limited.

They found that researchers tended to make only a few moves during their careers, most often during their early days. This made sense to Dev­ille and his team, knowing that one usu­ally moves insti­tu­tions when going from grad school to post doc to tenure track employ­ment but then stays put after reaching that latter mark.

They also found that researchers were three times more likely to move within groups rather than between them. “There are move­ments between the two groups, but they’re extremely unlikely,” said Roberta Sinatra, a post-​​doctoral fellow in Barabási’s lab.

When a researcher did move between groups, the team saw some inter­esting things: if the researcher moved to a more suc­cessful insti­tu­tion, his or her pro­duc­tivity didn’t change (they kept pub­lishing roughly the same number of papers) and nei­ther did their overall suc­cess (number of cita­tions). But if they moved to a less suc­cessful insti­tu­tion, they’d still work just as hard (again, pub­lishing the same number of papers), but their suc­cess went down (fewer citations).

All of this strikes a slightly depressing note, as far as I can see it. Basi­cally, if you don’t start out in the rich club, then you’ll never be as suc­cessful as if you did. But if you happen to leave it, well, you can kiss your suc­cess (or at least some of it), goodbye. Of course, as Sinatra pointed out, there’s a bit of a chicken-​​or-​​egg problem here. Is an  insti­tu­tion “rich” because of the suc­cess of the researchers it employs, or are those researchers suc­cessful merely because of the vis­i­bility afforded to them by their affil­i­a­tion with the rich club?

Jury’s still out. The work is part of an over­ar­ching pro­gram in Barabasi’s lab to study the sci­ence of success–to figure out what it is that turns some people into celebri­ties while the majority of us can only dream of such opu­lence. The team believes than under­standing these things, espe­cially in indi­vidual fields like physics, could pro­vide mean­ingful insights for both the cre­ators of knowl­edge and media, as well as the people who reg­u­late it all.

The next step for this par­tic­ular project is already underway: They are expand their exam­i­na­tion beyond the field of physics to look at the entire body of sci­en­tific pub­li­ca­tions over the same time period. They expect it will reveal some nuances of what accounts for suc­cess in the var­ious disciplines.