The Science of
Effective Mentorship in STEMM

Online Guide v1.0

Mentorship is essential in developing science, technology, engineering, mathematics, and medicine (STEMM) professionals. It is a set of skills that can and should be learned, practiced, and improved upon with self-reflection and feedback. If you are a mentor or mentee yourself, or if you are a leader in your organization responsible for ensuring that your faculty and their mentees have the skills to engage in the most effective mentoring relationships, this website is for you.

Throughout this website, you can find:


Help develop and maintain strong and effective mentorship education initiatives in STEMM

Mentoring Tools

Successful mentorship programs and tools for establishing strong mentoring relationships, particularly between faculty and students who have been traditionally underrepresented in STEMM fields


Convince STEMM faculty and institutional leadership of the value of effective mentorship and the harms that can result from negative mentoring experiences.


Mentoring has an important role in the science, technology, engineering, mathematics, and medicine (STEMM) fields. Mentoring experiences can be transformative for both mentors and mentees. Despite its essential place in the academic STEMM culture, mentorship rarely receives the focused attention, evaluation, and recognition of other aspects of the professional development process such as teaching and research.

Undergraduate faculty

My institution takes mentoring into consideration during performance reviews

48% agree

I have participated in training to be a mentor

6.9% agree

Science and engineering majors

I had a mentor at my undergraduate institution

22% agree

For all the effort academic institutions have spent formalizing the education and training of budding STEMM professionals, mentorship has largely been left to happen organically or on an ad hoc basis. There is a well-developed science of mentorship and a literature that shows the importance of mentorship and how to do it better.

Why Mentorship Matters

Effective mentoring relationships can engage and develop the talent of a broader group of students interested in STEMM careers, and as a result, help with the development of STEMM professionals by increasing access, equity, and inclusion in STEMM

Talent is equally distributed across all sociocultural groups, but access and opportunity are not. The underrepresentation of women of all racial/ethnic groups and individuals specifically identifying as Black, Latina/o, and American Indians/Alaska Natives in STEMM contexts is pervasive. Effective mentoring relationships can engage and develop the talent of a broader group of students interested in STEMM careers, and as a result, help with the development of STEMM professionals by increasing access, equity, and inclusion in STEMM.

More diverse and inclusive STEMM workplaces are more responsive to current and emerging problems because teams comprising individuals with diverse experiences and areas of expertise often ask different questions and tend to be more creative and innovative in how they answer those questions. More diverse research teams also, on average, result in higher impact research and make better decisions than less diverse teams. If access to the talent development and support is not provided equitably, the diversity of the STEMM workforce will continue to be insufficient.

Mentoring relationships can catalyze:

individuals’ potential for discovery, curiosity, and participation in STEMM—and subsequently improve the STEMM educational environment.

developmental spaces in which students can hone their STEMM skills and discover pathways into STEMM fields.

nascent STEMM professionals seeing themselves through the eyes of an influential guide, finding their place in STEMM education and careers, and receiving support to realize their next stages in professional development.

Some outcomes of mentorship include:

  • Graduate students are more likely persist in their academic decisions if engaged in positive mentoring experiences (McGee and Keller, 2007; Williams et al., 2016)
  • Graduate students cite positive mentoring experiences as the most important factor in completing a STEM degree (Ashtiani and Feliciano, 2012; Solorzano, 2000)
  • Women and underrepresented students are better integrated into the STEMM academic community if engaged in positive mentoring experiences (Anderson and Kim, 2006; Byars-Winston et al., 2015; Estrada et al., 2018; Felder, 2010; Good et al., 2000; Griffith, 2010; Huang et al., 2000; Lewis et al., 2016; Lisberg and Woods, 2018)
  • Positive mentoring experiences increase recruitment of underrepresented mentees into graduate school and research-related career paths (Hathaway et al., 2002; Junge et al., 2010; Nagda et al., 1998; Thiry and Laursen, 2011).
  • A focus on psychosocial needs is associated with increases how mentees perceive the quality of the mentoring relationship and how satisfied they are with that relationship, which in turn enables them to see themselves as more competent as STEMM researchers (Tenenbaum et al., 2001; Waldeck et al., 1997).
  • Mentored graduate students and medical trainees are more likely to publish their research than those who are not mentored (Steiner et al., 2004; Steiner et al., 2002; Wingard et al., 2004)

About the Report

In February 2017, the Board on Higher Education and Workforce (BHEW) convened a workshop to explore some of the major challenges for ensuring high-quality mentorship for undergraduate and graduate STEMM students (National Academies of Sciences, 2017a). One clear outcome of the workshop was the need for a deeper investigation into the science of effective mentorship.

Over the past decade the National Academies have convened several consensus study committees and conferences that assembled experts across disciplines to examine the research behind mentorship and related issues or that highlighted the importance of mentorship in building and maintaining the STEMM workforce (NAS-NAE-IOM,, 1997; NASEM, 2017b, 2018a, 2018b; NASEM., 2007, 2009, 2011; National Research Council, 2010, 2013, 2015). In addition, there are several previous and on-going studies, workshops, and programs conducted by the National Science Foundation, Alfred P. Sloan Foundation, Howard Hughes Medical Institute, National Institutes of Health, among others. Although this prior work contributed to the body of literature on mentorship, these activities were not aimed at compiling, reviewing, analyzing, and presenting research systematically and in an accessible format that can inform and drive practice.

In December 2017, the National Academies assembled the Committee on the Science of Effective Mentoring in Science, Technology, Engineering, Mathematics, and Medicine to produce a new consensus study on evidence-based approaches to mentorship in STEMM for students at the undergraduate and graduate levels. To inform the committee’s deliberations and supplement an intensive literature review, the committee convened 3 public workshops and held 18 listening sessions. The committee incorporated the output of these evidence-gathering activities throughout the report and in this online guide. This website highlights the key lessons from the committee’s 22 months of work.


The committee presents nine sets of recommendations to encourage a shift away from a culture of ad hoc mentorship and toward one of intentional, inclusive, and effective mentorship in all institutional contexts (e.g., minority-serving institutions, undergraduate-only institutions, research-intensive institutions, academic medical centers). For the first seven sets of recommendations, the committee lays out specific roles for various participants in the mentorship ecosystem—including institutional leadership (e.g., presidents, provosts, deans), department chairs, program leaders (e.g., research, training, and graduate program directors), mentors (faculty members, staff, and others who have extensive contact with graduate and undergraduate students), and mentees (undergraduate and graduate students participating in mentoring programs and other mentoring relationships), and professional associations. The last two sets of recommendations are directed at agencies that fund mentorship programs and scholars of mentorship.

The committee’s recommendations are best understood in the context of a common understanding of mentorship. Therefore, the first recommendation is directed toward all participants in the mentorship ecosystem.

Institutions and programs should adopt an evidence-based, operational definition of mentorship, such as the one used by the committee in its work:

Mentorship is a professional, working alliance in which individuals work together over time to support the personal and professional growth, development, and success of the relational partners through the provision of career and psychosocial support.

Mentorship is operationalized for STEMM contexts through the career support functions (e.g., career guidance, skill development, sponsorship) and psychosocial support functions (e.g., psychological and emotional support, role modeling) aimed at mentee talent development. Mentorship complements other developmental processes like teaching or coaching to support mentees in developing knowledge and skills, and is essential to the holistic development of scientists, technologists, engineers, mathematicians, and physicians, including but not limited to developing a strong identity as a STEMM professional, developing confidence in one’s ability to work as a STEMM professional, and successfully navigating the culture of STEMM.

  • 2.1: Institutional and departmental leadership should support the use of evidence-based mentoring practices by both mentors and mentees, starting with new faculty and student orientation. Support should include tested mentorship education curricula, resources, and tools (guided discussions, mentoring compacts, individual development plans, and mentor maps) as well as time for professional development and mechanisms for feedback, improvement, and accountability.
  • 2.2: Program leaders should support mentorship by ensuring there are evidence-based guidelines, tools, and processes for mentors and mentees to set clear expectations, engage in regular assessments, and participate in mentorship education. Program design should take into account the stages of mentoring relationships and ensure that the evolving needs of undergraduate and graduate students are met as they shift to career stage–appropriate independence.
  • 2.3: Department chairs should deliver professional development on effective mentorship to support mentors and mentees in understanding how successful mentoring relationships can be created, cultivated, and nurtured; addressing challenges such as those caused by biases and micro- and macro-aggressions; encouraging self-reflection; and mastering critical skills over time.
  • 2.4: Mentors should learn about and employ evidence-based mentorship tools and strategies through a process that includes exploring evidence-based mentorship resources, dedicating time for mentorship education, and participating in relationship-level, departmental-level, and institutional-level mentoring accountability mechanisms.
  • 2.5: Mentees should acquaint themselves with evidence-based mentorship tools and strategies, including compacts, individual development plans, mentor maps, and mentoring accountability mechanisms. When possible, mentees should take advantage of any mentee-focused mentorship education and resources and be aware of which faculty members in their program, department, or institution have participated in mentorship education and which faculty use evidence-based mentorship tools.

Assessment and evaluation of mentorship are necessary to identify areas of strength and opportunities for improvement. Evaluation through structured systems may reduce unintentional bias and protect mentees who are in inherently more vulnerable positions as students and trainees.

  • 3.1: Institutional and departmental leadership should regularly and systematically review formal mentorship activities and programs to support development of mentorship skills and student success and well-being. Such reviews should involve different stakeholders groups, check for alignment with stated program goals and missions, ensure that practices for effective mentorship are incorporated throughout activities and programs, and work to create a culture of accountability.
  • 3.2: Program leaders should establish and systematically review formal mentoring activities and programs and other structured feedback systems to make programmatic decisions such as who is allowed to serve as a mentor, when to intervene if relationships are not effective, and how to help mentors improve their skills over time using established methods and instruments for measuring mentorship effectiveness. Program leaders should regularly provide deans, department heads, and other program leaders with program metrics, including data on mentorship processes and outcomes.
  • 3.3: Mentors and mentees should work with each other and their institutions to develop feedback systems to document, evaluate, and advance mentorship competencies over time using established methods and instruments for measuring mentorship effectiveness. They should also participate in institutional reviews of formal mentorship activities and programs to enhance mentor and mentee outcomes and inform periodic self-reflection.
  • 3.4: Professional associations should regularly review and gather evidence on formal mentorship activities and programs that are designed to enhance students’ success outside of their home institution. Such reviews should also check for alignment with stated program and association goals, missions, and accountability mechanisms and for widespread use of effective mentorship practices.

All participants in the mentorship ecosystem should recognize that identities influence academic and career development and thus are relevant and significant for effective mentorship.

  • 4.1: Institutional leadership should intentionally support mentorship initiatives that recognize, respond to, value, and build upon the power of diversity. Leaders should intentionally create cultures of inclusive excellence to improve the quality and relevance of the STEMM enterprise.
  • 4.2: Mentors should learn about and make use of inclusive approaches to mentorship such as listening actively, working toward cultural responsiveness, moving beyond "colorblindness," intentionally considering how culture-based dynamics like imposter syndrome can negatively influence mentoring relationships, and reflecting on how their biases and prejudices may affect mentees and mentoring relationships, specifically for mentorship of underrepresented mentees.
  • 4.3: Mentees should reflect on and acknowledge the influence of their identities on their academic and career trajectory, including the potential for imposter syndrome to disrupt mentorship. Mentees should seek mentorship that is intentional in considering their individual lived experiences.
  • 4.4: Professional associations should intentionally address sociodemographic factors in mentoring relationships, specifically for mentorship of underrepresented mentees. Professional associations should also intentionally create cultures of inclusive excellence to improve the quality and relevance of the STEMM enterprise.

All participants in the mentorship ecosystem should recognize that identities influence academic and career development and thus are relevant and significant for effective mentorship.

  • 5.1: Institutional leadership should support policies, procedures, and other infrastructure that allow mentees to engage in mentoring relationships with multiple individuals within and outside of their home department, program, or institution, such as professional societies, external conferences, learning communities, and online networks, with the ultimate goal of providing more comprehensive mentorship support.
  • 5.2: Mentors should provide opportunities and support for mentees in mentoring relationships with other individuals within and outside of their home department, program, or institution (such as professional societies, external conferences, learning communities, online networks) who can provide complementary or supplementary functions that enable mentees to progress and succeed.
  • 5.3: Mentees should consider developing, as needed, a constellation of mentoring relationships with multiple individuals within and outside of their home department, program, or institution using tools designed for this purpose such as mentoring maps and individual development plans.
  • 5.4: Professional associations should proactively facilitate the development of mentoring relationships among individuals from different programs or institutions, as needed, who can provide complementary or supplementary mentorship functions. This could include activities such as pairing first-time conference attendees (mentees) with returning conference attendees (mentors) to orient them to conference events and support their networking or establishing and supporting online communities for mentees to find and make supportive connections outside their own institutions and environments (e.g., academia).

All participants in the mentorship ecosystem should recognize that identities influence academic and career development and thus are relevant and significant for effective mentorship.

  • 6.1: Institutional leadership should reward and visibly recognize mentors for documented, effective, and inclusive mentorship in the same manner as effective teaching is recognized, including through annual awards. Consideration should be given to all forms of mentorship, including informal and formal relationships that occur beyond the research advisor or other academic advisor and the student. Leaders should also structure job recruitment, application, and selection procedures to make evident an applicant’s commitment to and success with mentorship and ensure mentorship quality and potential are weighed in hiring decisions, possibly through the inclusion of mentoring statements in applications.
  • 6.2: Department chairs, in consultation with institutional leadership, should use promotion, tenure, and performance appraisal practices to reward effective mentorship. Elements of a promotion or tenure package could include descriptions of approaches and resources used in mentoring, reflective statements of ways the candidate has worked to improve their mentoring over time, evidence of mentored scientists as coauthors on manuscripts and grants and their placement into positions, letters from program leaders and testimonies from students, institutional and national award for mentorship, and process measures that assess mentoring relationship quality from the perspective of the mentee and the mentor.
  • 6.3: Professional associations should provide visible recognition of effective mentorship through prominent rewards for documented, effective, and inclusive mentorship, such as certifications for completing substantive mentorship education, named awards for sustained contributions to mentorship, and noteworthy track records of effective mentorship supported with assessment data.

Mentorship education for both mentors and mentees can help to reduce or prevent negative mentoring experiences. However, negative mentoring experiences do and will occur, and direct steps should be taken to mitigate harm from such occurrences.

  • 7.1: Institutional leadership should appoint and make visible one or more neutral third parties (e.g., ombudspersons, research integrity office) to serve as a point of contact to identify, investigate, and address negative mentoring experiences. These individuals, offices, or committees should be selected based on their potential to engender a sense of trust and approachability among mentees and mentors. The appointed neutral third parties should also be prepared to carry out their role effectively by participating in professional development on mentorship, conflict management, and workplace laws and ethics.
  • 7.2: Program leaders and department chairs should periodically review mentorship assessment results to identify and mitigate negative experiences. They should be open to the possibility of having to serve as a neutral third party to improve ineffective or negative mentoring experiences, and they should also be prepared to carry out their role effectively by participating in professional development on mentorship, conflict management, and workplace laws and ethics.
  • 7.3: Mentors should recognize that negative mentoring experiences can occur even with well-intentioned mentors and mentorship practices and be open to addressing unintended negative mentoring experiences with a neutral third party. In addition, mentors should become familiar with and recommend resources, such as ombudspersons, who can help identify, investigate, and address negative mentoring experiences.
  • 7.4: Mentees should maintain relationships with a network of faculty outside of their primary advisor, research supervisor, or mentor, and when necessary, seek out an ombudsperson or other neutral third party who can serve as a resource to address negative mentoring experiences.

Funding agencies play a key role in shaping the values of institutions and the projects that scholars pursue. As such, funding agencies’ role in encouraging and supporting effective mentorship practices is essential.

  • 8.1: Funding agencies should encourage the integration of evidence-based mentorship education for mentors and mentees and assessments of mentorship into grant activities that involve undergraduate and graduate student research, education, and professional development to support the development of the next generation of talent in STEMM.
  • 8.2: Funding agencies, when supporting STEMM student development, should require tools such as mentoring compacts and individual development plans to operationalize intentionality and promote shared understanding of the goals of mentoring relationships on sponsored projects.
  • 8.3: Funding agencies should support the study of the process and impacts of mentorship and the development and validation of new or adapted measures for use in STEMM mentorship to comprehensively understand the relationship between mentorship processes and outcomes, as well as demographic disparities in student outcomes.
  • 8.4: Funding agencies should support in-depth, cross-program evaluation and research to better understand the processes and outcomes of mentorship, particularly on the outcomes of diverse student populations.

When the committee reviewed the literature on mentorship and mentoring relationships, it became apparent that more scholarship is needed on specific aspects of mentorship and mentoring relationships. Items 9.1–9.5 represent some of the areas that would benefit from additional scholarship and make contributions to advance the science of mentorship.

  • 9.1: Scholars should conduct multidisciplinary research on mentorship in STEMM, including employing advanced multimethod approaches, using current technologies, and establishing standards for measurement to uncover the relational processes that drive effective mentorship. Scholars should particularly attend to the reciprocal and dynamic nature of mentoring patterns, processes, and outcomes in STEMM to advance theories of mentorship in STEMM.
  • 9.2: Scholars should make greater use of study designs that allow for causal and longitudinal inferences, paying particular attention to the antecedents, processes, correlates, and outcomes within effective mentoring relationships in STEMM to determine the effects of mentorship on persistence and success in STEMM as well as on the STEMM enterprise.
  • 9.3: Scholars should define and characterize negative mentoring experiences or ineffective mentorship in STEMM and investigate their prevalence and impacts, specifically addressing the possibility that negative mentoring experiences may disproportionately harm underrepresented students and compromise science and research itself.
  • 9.4: Scholars should intentionally expand the knowledge base for populations that remain little-studied in STEMM and account for how differing conditions and contexts of mentorship may differentially affect individuals with diverse sociocultural identities. Scholars should examine mentorship assets at the individual, department, and institutional levels to assist STEMM researchers and universities in creating targeted recruitment and retention programs for underrepresented and underserved populations.
  • 9.5: Scholars should investigate how different aspects of mentor-mentee sociocultural similarity may help shape mentorship outcomes to elucidate the effectiveness of matching practices and processes in formal mentorship programs and provide greater access to quality mentoring.
Comments or Suggestions?