“Modern science has been a voyage into the unknown, with a lesson in humility waiting at every stop,” said astronomer Carl Sagan. Staring at the vast cosmos, Sagan rightly explained just why scientists mustn’t be arrogant.
Behind that humility, however, there is a more earthly reason too. Scientists’ career are more scrutable than in most other professions. That’s because their professional currency is the scientific papers they publish, and each paper has attached to it a slew of metrics—such as how often the paper is cited by other papers—that reveal how successful a scientist is.
So can we predict success in a scientist’s career? That’s the question Albert-László Barabási of the University of Notre Dame and his colleagues try to answer in a recent study published in Science.
Barabási and his colleagues used the publication record of more than 2,800 physicists whose career has spanned at least 20 years and who have regularly published studies. The first result they found was that scientists tend to produce their most high-impact work in their younger years, with ”impact” measured as the number of citations a study gets in the years to come.
However, it also turned out that scientists tended to be more productive at younger ages. When that was accounted for, age was no longer a predictor of success. Any given study could become a scientist’s greatest hit.
Unsatisfied with that result, Barabási and his colleagues tried another tack. Any scientific success is the result of two things: factors that a scientist can control (getting the right training, working with the right people, and so on) and factors that a scientist can’t control (weather, natural disasters, serendipity, and so on). They created an equation to represent that:
success (c) = what scientist can control (Q) × what scientist can’t control (p)
Using the data on the same 2,800 physicists, they knew how to measure success (c) based on the number of citations each study received. “Factors that a scientist can’t control” were lumped in as luck (p) and thus were random. With two out of three things known in the equation, they next modeled how Q changes.
Barabási and his colleagues thought that surely a large contributor to Q, the factors a scientist can control, would be the scientist’s ability. And ability, they assumed, would increase as a working scientist learns and gains experience, so Q should increase too. However, when they fed their data into the model, they found that in fact, Q was constant throughout a scientist’s career.
The implication is that scientists are as able at the start of their careers as they will ever get. “I hate to call it innate,” Barabási told Nature, “but [Q] seems to be a combination of your ability and education. Once you start your career, you have it and it stays with you.”
The next job for Barabási is to figure out just exactly what factors contribute to Q. If it is indeed largely made up of ability, the implication that ability doesn’t increase over a scientist’s career could have wide ramifications for how scientists are trained, such as rethinking how doctoral programs (the primary qualification towards becoming a scientist) are run and how further training through post-doctoral appointments is conducted.
If Barabási’s work stands up to scrutiny and can be successfully applied to fields other than physics, Q might become an important measure for judging whether a particular scientist should receive research funding. That’s not going to help the anxieties that many scientists are already facing in a highly competitive environment.