Your most impactful publication can come at any time

An international research team led by Albert-László Barabási (Northeastern University) has just published a wide-ranging study of how the impact of a scientist’s oeuvre varies over the course of a career. To gauge a publication’s impact, the researchers considered the cumulative number of citations it received in the first 10 years after publication, a metric called c₁₀. One unexpected discovery came when they analyzed the probability that a scientist’s nth paper was their most impactful one and found that all papers were equally likely to have the largest value of c₁₀.

The flat probability distribution is in accord with the disheartening hypothesis that a scientist’s ability is immaterial to a publication’s success. But Barabási and colleagues found that possibility inconsistent with other trends in their data—for example, that scientists who pen specific papers with truly high impact are likely to produce other important work. Instead, the researchers found, the impact of a scientific publication is well modeled as a product of a scientist-independent luck factor and a quantifiable scientific ability, Q.

The figure below shows the significance of Q according to the researchers’ detailed numerical work. The y-axis indicates the probability that a paper will have a c₁₀ value at least as great as that given on the x-axis. Ability, as measured by Q, increases from violet to red. The colored shapes fitted by bold lines give real-world data for three scientists.

The ability that helps determine the impact of a scientist’s career may be influenced by education and other factors. For now, the dependence of Q on such factors remains unknown. (R. Sinatra et al., Science 354, 596, 2016, doi:10.1126/science.aaf5239 .)