AI evidence extraction

Can GenAI complement supervisor support in shaping postgraduates’ research experiences? A mixed-methods approach

Yating Huang, Sihui Li, Zihan Liu · Studies in Higher Education · 2025

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

4.14

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1080/03075079.2025.2495710

Methodology & findings

Study design

Explanatory sequential mixed methods design: quantitative survey (n=1,515) followed by qualitative interviews (n=20)

Sample

N = 1535, 4 groups

Main result

The findings revealed a complementary alliance between supervisor support and GenAI support in fostering postgraduate students' positive research experiences, and "the synergistic integration of supervisor-GenAI maximized their complementary strengths, enhancing the depth and quality of support while preserving the irreplaceable value of human interactions rather than diminishing the role of human supervisors."

Reports effect sizes.

Research paradigm

Mixed methods (pragmatist)

Author conclusions

"This study contributed to envisioning an educational environment where GenAI and supervisors coexist harmoniously to enrich postgraduate students' research experiences."

Risk of bias

Selection bias (self-selection in survey and interview participants); Confounding variables not controlled for in observational design; Geographic limitation (China-based sample); Selection bias: sample from China only, may limit generalizability; Self-report bias: survey responses on perceived supervisor and GenAI support; Attrition: qualitative sample (20) substantially smaller than quantitative sample (1,515); Confounders: unmeasured factors affecting research experiences not controlled; Selection bias: Sample of 1,515 postgraduate students in China may not be representative of global postgraduate populations; Potential respondent bias: Self-selection into study on GenAI attitudes; Small qualitative sample (n=20) relative to quantitative sample (n=1,515) may limit depth of qualitative exploration; Single country setting (China) limits generalizability; Potential temporal bias: Study timing relative to GenAI adoption trajectories

Open questions raised

The complex interplay of supervisor support and GenAI support on research experiences of postgraduate students remains under-explored
The study addressed the gap that "the complex interplay of supervisor support and GenAI support on research experiences of postgraduate students remains under-explored."

Data: not_statedCode: not_statedExtracted from: pdfAgreement 75%

Explore related topics

Responsible AI Use Human-AI Collaboration Mixed Methods