Networks or Mentors?

Networks or Mentors? How Supportive Professional Networks and Advisors Predict Graduate Student Outcomes


Jeremy Foote
Purdue University


Alejandra Durán Trindiad
Purdue University


Hsuen-Chi (Hazel) Chiu
Purdue University


Bedadyuti (Dyuti) Jha
Purdue University


Melanie Morgan
Oklahoma State University


Joshua Becker
University College London

Graduate School is hard

  • Being a graduate student is stressful and difficult.
  • It is an intense period of acculturation and learning.

Source: Bergvall et al. (2025). “The impact of PhD studies on mental health-a longitudinal population study.” Journal of Health Economics.

Good mentorship helps

  • Effective mentors make a huge difference for their students
    • Productivity, well-being, stress, etc.
  • Good mentorship involved providing:
    • Practical and career support
    • Emotional / social support

Definition of mentorship has broadened a little

  • Peer mentoring
  • Group mentoring
Source: Google Gemini

Our project looks at the role of broader work networks in providing support

  • Interesting context: mix of constrainted and unconstrained ties
  • Very little research on the structure of graduate student networks

Research Questions

  • What is the composition of STEM graduate students’ professional networks (gender, role, etc.)
  • How does perceived support differ?

Hypothesis

  • Measures of social capital (bonding and bridging) will predict positive student outcomes

Survey of over 400 STEM students

  • Ego networks
    • Think of people you work with … get advice from, collaborate with, or regularly have work-related conversations with.
    • Attributes of alters
    • Do alters communicate without ego (Burt, 1992)?
  • Existing scales for well-being, stress, group identity, scientist identity
  • Productivity as number of publications

Results

Networks are small and dominated by PhD students and lab members

Support differs markedly by role and “location”

Support differs by gender

Some support for bonding and bridging capital improving well-being

But mentor quality is clearly more important and attenuates importance of capital

Discussion

Alters outside of university and area most likely to provide support

  • No strong propinquity
  • Possible evidence that people seek support from those with social and physical distance

Mentorship quality is really important

  • Good mentoring can be learned and should be taught
  • Good mentors (apparently) help students build social capital

Limitations

  • Cross-sectional design limits causal inferences
    • We need longitudinal or experimental data
  • Single institution limits generalizability
  • So many possible measures of bonding and bridging capital