Differential engagement with generative artificial intelligence in higher education: Gender, motivation, and achievement trajectories

Synthesis: This study of 97 graduate students and 2,819 queries to a RAG-based GenAI chatbot (StatBot) for statistics learning reveals that engagement with GenAI is strongly differentiated across learner subpopulations. More than one-quarter of students never used the chatbot; female students, those with stronger autonomous Motivation, lower prior knowledge, and higher course performance interacted more frequently. Using zero-inflated negative binomial modeling and k-means achievement profiling, the authors show that "Growing Achievers" posed more diverse, conceptually oriented questions, while "Declining Performers" engaged minimally with narrow procedural inquiries. The authors argue that effective AI integration requires moving beyond uniform access toward Scaffolding concept-focused support that prioritizes autonomous motivation and shared human-AI agency.

Core Finding

GenAI efficacy in higher education depends less on universal access and more on learner-specific engagement: autonomous motivation and achievement trajectory shape both the frequency and the quality of AI interaction. GenAI tools can amplify growth for some learners while remaining underutilized by others, so the benefits are not distributed equally across student subpopulations.

Significant Non-Participation

More than one-quarter of students did not use the chatbot at all. This meaningful non-participation challenges the assumption of universal AI adoption and highlights that engagement is not guaranteed by providing a tool. Higher-performing students were less likely to abstain from use, suggesting that prior success correlates with readiness to engage AI support.

Determinants of Engagement Frequency

The zero-inflated negative binomial model showed that engagement frequency was shaped by learner differences: female students, those with stronger autonomous motivation, lower prior knowledge, and higher course performance interacted more frequently. Notably, the gender difference (female students engaging more) and the role of autonomous motivation suggest that motivational quality — not just access — drives who benefits from AI-supported learning.

Inquiry Quality and Achievement Trajectories

Engagement was not merely a matter of frequency but of inquiry quality and diversity. k-means clustering identified distinct achievement profiles. "Growing Achievers" — students who began with weaker achievement but demonstrated substantial performance gains — posed more diverse and conceptually oriented questions, leveraging the chatbot for genuine learning gains. In contrast, "Declining Performers" engaged minimally and focused on narrower procedural inquiries despite stronger initial knowledge. The predominance of Initial Inquiry behaviors (70.3%, especially Definition and Copy queries) indicates that many students used the chatbot primarily for cognitive offloading of basic concepts rather than higher-order reasoning.

Relevance to the knowledge base

This paper contributes directly to the knowledge base's understanding of Student-AI Interaction, Motivation, and the equity of GenAI in higher education. It shows that engagement is patterned by gender, motivation, prior knowledge, and achievement trajectory — a differentiated, non-uniform picture that refines simplistic "AI improves learning" claims. It connects to achievement research, chatbot design, personalization, and the goal of shared human-AI agency. Its finding that concept-focused scaffolding and autonomous motivation matter most informs instructional design and aligns with Learner Agency-oriented frameworks on the knowledge base.

Connected Concepts

• Generative AI
• Conversational AI
• Motivation
• Student-AI Interaction
• Personalized Learning
• Learning Gains
• Self-Regulated Learning
• Higher Education

Connected Articles

• Designing a mobile chatbot-based learning journaling system for intrinsic motivation and engagement
• Not All Students Engage Alike: Multi-Institution Patterns in GenAI Tutor Use
• Using AI-based Learning Assistants in Higher Education: A Large-Scale Descriptive Analysis
• Kutti AI: A Voice-First, Offline-Capable Learning Companion with Real-Time Struggle Detection for Visually-Impaired Children
• "It is a temptation to get it to do the work…" Student Experiences of Navigating the Generative AI Landscape in UK Higher Education: A Cross-Institutional Survey with International Comparison

Citation

Lee, Y.-H., & Wu, J.-Y. (2026). Differential engagement with generative artificial intelligence in higher education: Gender, motivation, and achievement trajectories. International Journal of Educational Technology in Higher Education.