AI-Driven Adaptive Gamification: Student Engagement and Personalised Academic Achievement

Abstract

Gamification is now widely used in tertiary instruction, yet evidence that generic points, badges and leaderboards improve learning remains inconsistent, partly because uniform game elements ignore differences among learners. Artificial intelligence (AI) makes it possible to adapt game mechanics to each student in real time. This study examined whether students’ perceived exposure to AI-driven adaptive gamification (AAG) predicts student engagement and personalised academic achievement, and whether engagement mediates that relationship, among undergraduates in Enugu State, Nigeria. Grounded in self-determination theory, flow theory and the theory of gamified learning, a correlational cross-sectional survey was conducted with 362 undergraduates from four tertiary institutions selected through proportionate stratified random sampling. Sample size was justified with the Krejcie–Morgan formula (minimum = 346) and an a priori power analysis (minimum = 262). Validated scales measured AAG and behavioural, emotional and cognitive engagement (Cronbach’s α = .861–.888; composite reliability = .862–.888; AVE = .524–.614), while achievement was indexed by end-of-semester course scores with prior cumulative grade point average (CGPA) controlled. Hierarchical regression showed that AAG explained an additional 15.7% of the variance in engagement (β = .40, p < .001, f² = 0.19) and 9.5% of the variance in course achievement (β = .31, p < .001, f² = 0.13). Bootstrapped mediation analysis (5,000 resamples) showed a significant indirect effect of AAG on achievement through engagement (ab = 1.85, 95% CI [1.26, 2.49]; completely standardised effect = .14), accounting for 46.6% of the total effect. The findings indicate that adaptive, AI-personalised gamification improves achievement chiefly by deepening engagement, with implications for instructional design and digital-learning policy in Nigerian tertiary education.