The most successful early-career researchers often rely on a network of mentors who each have different roles — from advising on research and grants to making professional introductions and providing emotional support, finds a study of nearly 900 scientists just starting out¹.

The findings are consistent with what mentorship scholars and practitioners have long argued: no single mentor can meet all of a trainee’s scientific, professional and psychosocial needs, says Keivan Stassun, an astrophysicist at Vanderbilt University in Nashville, Tennessee. “It is encouraging to see these patterns emerge so clearly in a large data set of successful early-career investigators,” says Stassun, who has won awards for his mentorship.

For the study, which was posted on the bioRxiv preprint server and hasn’t been peer-reviewed, researchers analysed responses from recipients of K awards, issued by the US National Institutes of Health for career development. These prestigious grants support early-career researchers in strengthening their research skills and provide time for training activities that prepare them to become independent investigators.

The team found that recipients had a median of 3 mentors when they applied for their awards, although some reported as many as 15. More than 90% of the awardees said that mentoring relationships resulted in publications and grants. Other common mentoring outcomes were new professional connections (88.5%) and enhanced skills (81.8%).

The authors acknowledge that although recruitment for the study targeted successful and unsuccessful award applicants, most respondents were recipients. This prevented comparisons of characteristics between those who won awards and those who didn’t, the authors write.

Different outcomes

Participants were asked to categorize their mentors into six predefined roles. All of them had a research mentor, but many others served as coaches, sponsors, connectors or peers. Connectors used their professional network to introduce the early-career researcher to useful contacts, and sponsors used their seniority and influence to advocate for their trainees and help to create opportunities. Some awardees also had an ‘identity mentor’, with whom they shared certain traits, such as gender, race, ethnicity, nationality or sexual orientation.

Different mentor roles were associated with different types of success. For instance, research mentors were most strongly associated with achievements such as writing publications, winning grants and gaining research skills. Psychosocial benefits — such as building confidence, personal growth and improved work–life balance — came from relationships with coaches, peer mentors and identity mentors. And early-career researchers were more likely to land a job because of their relationship with a connector or sponsor than through their research mentor.

This is an important takeaway, says co-author Felicia Setiono, a health-equity researcher at SUNY Downstate Health Sciences University in Brooklyn, New York. Most postdocs aspire to continue in academia, and these positions are competitive, she says. “Yet postdocs still tend to primarily rely on their main research adviser to navigate the process for pursuing their next steps,” she adds.

Under-represented scientists

Roughly 15% of awardees were from groups under-represented in science, and they in particular seem to benefit from larger mentoring networks compared with researchers outside those groups.

On average, these researchers had worse K-award scores than did their counterparts. But each mentor beyond the primary supervisor was associated with roughly a one-point improvement in award score. By having more mentors, researchers from under-represented groups were able to catch up to the scores of their peers, the authors write in the study.

The researchers found that having a peer mentor or an identity mentor was more strongly associated with better K-award scores among researchers in under-represented groups than among their peers without such identities.

Monica Cox, an engineering-education researcher at the Ohio State University in Columbus, says that mentors or peers from similar identity backgrounds often offer safe relationships. They consider the whole person and take a holistic approach when determining what success looks like for their trainees, accounting for factors such as family care and maintaining healthy relationships at work.

Stassun says the findings apply beyond biomedical research and grant-funded academic positions. Modern scientific roles are increasingly complex and interdisciplinary, and researchers need scientific guidance, career advice, sponsorship, networking opportunities, psychosocial support and often identity affirmation, he says. “It is unrealistic to expect any one mentor to fulfil all of those functions,” he adds.