GenAI dependence and graduate students’ research creativity: the roles of critical thinking and human–AI collaboration quality

Abstract

Introduction As generative artificial intelligence (GenAI) becomes increasingly integrated into graduate students’ academic research, the relationships between GenAI dependence and research creativity have attracted growing scholarly attention. This study distinguishes task-oriented functional dependence from deeper psychological existential dependence and examines their associations with research creativity to examine their associations with research creativity, as well as the potential mediating roles of critical thinking and human–AI collaboration quality. Methods Questionnaire data were collected from 1,157 graduate students recruited through online convenience sampling. Structural equation modeling (SEM) was used to examine the overall structural framework linking GenAI dependence, critical thinking, human–AI collaboration quality, and research creativity. PROCESS Model 6 with 5,000 bootstrap resamples was subsequently used to examine and compare the indirect pathways associated with functional and existential dependence while controlling for demographic variables. Results The results indicated that: (1) Functional dependence and existential dependence were both positively associated with research creativity, with the total association for functional dependence (β = 0.457) being substantially stronger than that for existential dependence (β = 0.147); (2) critical thinking was significantly associated with the relationship between GenAI dependence and research creativity; (3) human-AI collaboration quality was also significantly associated with research creativity; (4) critical thinking and human-AI collaboration quality jointly constituted a significant serial mediation pathway linking GenAI dependence and research creativity. Discussion These findings contribute to the literature on GenAI use and graduate education by highlighting the differentiated associations of functional and existential dependence with research creativity. The findings also suggest that critical thinking and high-quality human–AI collaboration may be important factors associated with graduate students’ self-reported research creativity. Given the cross-sectional design of the study, the findings should be interpreted as associative rather than causal.