AI evidence extraction

How postgraduates navigate GenAI’s dual pathways between research creativity and academic plagiarism: research engagement and self-regulated learning matter

Yating Huang, Keying Zhang, Sihui Li · 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

0/4

Quality (LMQS)

Evidence

Citations

6.35

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1080/03075079.2025.2543438

Methodology & findings

Study design

Sequential explanatory mixed methods approach combining quantitative questionnaire survey (n=1,515) with semi-structured interviews (n=20).

Sample

N = 1535, 4 groups

Primary method

Moderated mediation model analysis using path analysis. Quantitative data analysis methods are referenced but specific software and detailed procedures are not provided in the abstract.

Main result

The study found that "GenAI usage's effect on research creativity (β = 0.171, p < 0.01) and GenAI plagiarism (β = 0.184, p < 0.01) were significantly mediated by research engagement with GenAI, while self-regulated learning moderated these paths differentially – positively for research creativity (β = 0.077, p < 0.01) and negatively for GenAI plagiarism (β = −0.104, p < 0.05)." Qualitative findings demonstrated how institutional and individual factors shape postgraduate students' engagement with GenAI and its influence on research creativity and plagiarism.

Reports effect sizes and confidence intervals.

Research paradigm

Mixed methods (quantitative-qualitative integration); grounded in self-regulation theory

Author conclusions

The authors conclude that "This study highlighted how individual and institutional factors influence GenAI plagiarism, emphasizing higher education institutions' responsibility to guide postgraduate students in using GenAI productively while upholding academic integrity."

Risk of bias

Selection bias: Study conducted in China only; may not generalize to other cultural or educational contexts; Self-report bias: Questionnaire-based data on plagiarism and creativity; Attrition: Potential differential dropout between survey (n=1,515) and interview (n=20) participants; Confounding: Performance-driven academic environment mentioned as institutional factor but not fully controlled; Single-country study (China) - limited generalizability; Self-report bias inherent in survey methodology; Potential selection bias if participation was voluntary; Cross-sectional survey design limits causal inference; Interview sample size (n=20) may not adequately represent diversity of 1,515 survey participants; Selection bias - participants self-selected into survey; Potential social desirability bias in self-reported plagiarism behaviors; Geographic limitation - data collected in China only; Temporal limitation - cross-sectional survey design limits causal inference; Common method variance - reliance on self-report questionnaires

Open questions raised

The study emphasizes the need for institutions to provide guidance on productive GenAI use while maintaining academic integrity standards; qualitative findings suggest further investigation into how performance-driven academic environments shape student engagement with GenAI.
The study emphasizes the need for higher education institutions to develop better guidance mechanisms for postgraduate students using GenAI, and highlights the importance of understanding how institutional factors and self-regulated learning interact in shaping student outcomes.
Future research should examine how performance-driven academic environments influence GenAI use patterns; investigate differential impacts across discipline areas; explore longitudinal effects of sustained GenAI engagement on research creativity and plagiarism; examine institutional policy implementation effectiveness

Data: not_statedCode: not_statedExtracted from: pdfAgreement 57%

Explore related topics

AI Manuscript Writing Academic Writing Mixed Methods