Researchers who incorporate ideas and techniques from multiple mentors while still forging their own paths are the most likely to succeed in academia, according to a study of 18,865 biomedical researchers published in Nature Communications1.
The authors also suggest that mentoring received during postdoctoral training had a bigger impact than mentoring received during graduate school.
The study analysed data from the Academic Family Tree, an online database of academic relationships that launched in January 2005. The authors identified ‘triplets’ — trios comprised of a scientist, their graduate mentor and their postdoctoral mentor — dating back to 1970.
Professional success was gauged in part by the number of trainees a researcher mentored per decade, and an analysis of terms used in abstracts made it possible to track similarity of scientific approaches.
The results give empirical evidence to support some popular career strategies, says study co-author Stephen David, a neuroscientist at the Oregon Health & Science University in Portland. For example, the most successful scientists transferred concepts they learnt in graduate school to their postdoctoral work, suggesting that prospective postdocs should try to join labs that lack their particular skill set.
“You want to be able to offer something new,” David says. That requires stepping beyond the shadow of a graduate mentor without becoming a facsimile of a postdoctoral mentor. “You have to stake out some unique territory, which is always a challenge for postdocs,” he says.
The study found that joining the lab of a prolific mentor — one who has trained many researchers over the years — also increases a scientist’s chance of success. This held true for both graduate and postdoctoral mentors, but a closer look at the data revealed that the qualities of a postdoctoral mentor were especially predictive of success. “You can get a graduate education just about anywhere,” David says. “Postdoc labs are where you establish professional relationships and develop collaborations.”
Researchers should be especially discerning when accepting postdoctoral positions, David says. “You can take a data-driven approach to choosing your mentor.”
Nature 565, 667 (2019)