Generative AI Competence and Student Engagement in Higher Education

Abstract

This study investigates university students' competence and engagement with Generative AI applications, analyzing usage, perceived autonomy, and AI formal learning as predictive mediators. The study used a cross-sectional survey design to collect data from 262 students across various disciplines. Analysis using PLS-SEM revealed that GenAI competence is a powerful, direct driver of students’ engagement (β = 0.255, p < 0.001), accounting for 64.5% of its variance (R² = 0.645), with the direct path showing a small to medium effect size (f² = 0.117). This relationship is significantly mediated through three distinct pathways: GenAI utilization (β = 0.155, p < 0.001), Perceived Autonomy (β = 0.121, p = 0.005), and AI Formal Learning (β = 0.092, p = 0.005).