Risk-Taking, Mentorship and the Bibliometric Shadow of Scientific Risk: A Comment on Shibayama, Mattsson and Broström (2026)

Shibayama, Mattsson and Broström’s study of risk-taking in early-career academic training makes an important contribution to research-policy debates by shifting attention from funding incentives to doctoral supervision. Their finding that PhD students’ bibliometric risk-taking is associated with that of their supervisors, and that this association may persist after doctoral training, is both provocative and policy-relevant. This comment argues, however, that the paper’s strongest interpretation should be treated with caution. The evidence is consistent with socialization, but it may also reflect selection, research-trajectory inheritance, survivorship bias and the limits of publication-based measures of scientific risk. In particular, bibliometric indicators can observe only risks that survive into published work; failed projects, rejected papers, abandoned experiments and career-ending detours remain largely invisible. The article is therefore valuable less as a definitive demonstration that supervisors transmit risk-taking than as an agenda-setting contribution to the study of how scientific risk is selected, measured, published and reproduced across academic generations.