Mentoring Changes the Odds: What the Research Suggests

Part 1 of a series on mentoring, sponsorship, and the systems that shape careers in STEMM

Mentoring is often talked about as a “nice to have.” A generous act. Something people do when they have extra time.

Research tells a different story.

Across education, STEM, medicine, and professional careers, mentoring is consistently associated with who stays, who advances, and who leaves and while most studies are observational, these associations are reproducible across cohorts, disciplines, and career stages.

This newsletter isn’t about inspirational anecdotes. It’s about what the data actually show.

1. Mentoring Improves Academic Performance and Retention

Let’s start early: college and graduate school.

Multiple large reviews show that students who receive mentoring are more likely to persist, graduate, and perform well academically, especially in demanding fields like STEM.

A critical review by Crisp and Cruz  looked across mentoring programs in higher education and found that mentoring is most strongly linked to higher GPAs, better retention, and greater academic self‑efficacy for students.

But here’s the important part: Mentoring worked best when it included concrete guidance (e.g. career planning, goal setting, navigation of systems), not just encouragement.

A meta-analysis by Eby and colleagues confirmed this across education and early career stages including higher achievement and lower attrition.

For students from underrepresented backgrounds, the impact is even clearer.

In a study of a longitudinal national panel of underrepresented minority (URM) STEM undergraduates, Estrada et al. found that strong mentorship and research experiences predict higher science identity among URM undergraduates. Science identity and values predict persistence in STEM careers even years after graduation.

Mentoring doesn’t just help students feel supported. It changes who stays in the pipeline.

2. Mentoring Accelerates Career Advancement and Productivity

The effects don’t stop at graduation.

In the workplace, and especially in science and medicine, mentoring predicts objective career success.

A classic study by Allen et al. showed that mentored professionals have higher compensation, faster promotion rates, and greater career satisfaction. This wasn’t about confidence alone. It was about outcomes.

A meta-analysis by Eby et al. reinforced this, showing that mentoring is associated with promotions and performance, not just perceived success.

In biomedical research specifically, mentored scientists were more likely to secure grant funding, publish more, and persist in academic medicine.

In other words: mentoring increases productivity and retention in environments where losing talent is expensive.

3. Mentoring Reduces Attrition in STEM and Medicine

If you care about the “leaky pipeline,” mentoring is one of the few interventions with consistent evidence behind it, though much evidence is observational.

In a recent scoping review on mentorship interventions in postgraduate medical and STEM settings, Gangrade et al. found that formal mentoring programs in academic medicine were associated with higher promotion rates, greater leadership attainment, and reduced burnout.

And Beech et al. showed that mentoring was linked to is linked to lower attrition of minority faculty and higher academic rank over time.

Mentoring doesn’t fix every structural problem. But it measurably reduces the rate at which talent is lost.

4. How Mentoring Actually Works (The Mechanisms)

This is where mentoring stops being vague and starts being scientific.

The literature points to a few consistent mechanisms:

• Mentoring transfers social capital: it gives people access to networks, sponsors, and decision‑makers.

• Mentoring is cognitive apprenticeship: it teaches how experts think, decide, and prioritize, not just what they know.

• Mentoring shapes identity: it helps people see themselves as someone who belongs in the field.

• Mentoring transmits tacit knowledge: it reveals the unwritten rules for how promotion works, how grants are really reviewed, and how leadership decisions are made.

The Ragins & Kram framework divides mentoring into:

• Career functions (sponsorship, exposure, coaching)

• Psychosocial functions (confidence, belonging)

Both matter. But career functions drive advancement. And quality matters more than presence.

A study by Pfund et al. showed that training mentors—not just assigning them— improves mentee satisfaction, research outcomes, and alignment of expectations.

In a randomized controlled trial, Byars‑Winston et al. showed that a mentor training intervention improved mentors’ attitudes, confidence, and mentees’ ratings of mentor effectiveness, particularly in respectfully broaching race and ethnicity topics. Faculty‑level studies tell the same story.

Poor mentoring can be neutral. Sometimes, it’s harmful.

5. The Strongest Evidence: Meta-Analyses

A meta-analysis of 69 studies by Eby et al. found mentoring was associated with:

• Increased performance

• Increased career commitment

• Reduced turnover intentions

Another analysis by Allen et al. showed that mentoring has an effect size comparable to formal training programs, but at a much lower cost.

Mentoring works. But only when it’s done intentionally.

6. What the Science Does Not Say

The evidence does not support the idea that:

• Mentoring alone eliminates structural inequities

• One-off or symbolic mentoring changes outcomes

• Good intentions are enough

Mentoring is not a substitute for fair systems. But it is one of the most reliable tools we have to help people navigate them.

One final note.

The evidence is clear: mentoring is one of the best bets we have for boosting performance, retention, and advancement.

But mentoring alone does not explain who ultimately reaches leadership.

In my next newsletter, I’ll examine the distinction between mentorship and sponsorship—and why many women and underrepresented professionals receive plenty of advice, but far less advocacy.

Mentoring helps people grow. Sponsorship decides who moves.

The Bottom Line

Mentoring isn’t a favor, and it’s not just about being nice.

It is an evidence-based intervention that shapes who stays, who advances, and who leads.

The question then isn’t whether mentoring should be a core part of programs, but how to design it so it actually works for everyone.

If we care about outcomes, we should treat mentoring the same way we treat any other intervention that works:

• Design it deliberately

• Measure it honestly

• Invest in quality

Because talent doesn’t leak out by accident. And it doesn’t stay by luck.

How You Can Help

If you’re in a position to mentor, focus on the mechanisms that matter: transfer social capital, coach decision-making, and help others see themselves as belonging in their field. 

If you’re designing a program, treat mentoring as an evidence-based intervention, not a ‘nice-to-have.’ Invest in training, measure outcomes, and focus on equity.

---

In the next issue, we’ll dive into what data says about sponsorship and why it often remains the missing piece for underrepresented and women in STEMM.

---

If you find this newsletter informative, please don't forget to share it with others.

This newsletter is also published on LinkedIn