The Influence of Aesthetic Personalization on Gamified Learning: A Behavioral Analysis of Students’ Interactions

Personalized gamification seeks to address the limitations of the one-size-fits-all approach, mostly by tailoring the selection of game elements to individual preferences. However, there is limited understanding of how aesthetic personalization influences actual student behavior. This paper presents a behavioral analysis of 40 high school students engaged with a Virtual Learning Environment (VLE) over a four-week period. Each participant experienced both the one-size-fits-all and aesthetic personalization conditions for two weeks while submitting homework. Utilizing interaction data, we employed recurrent neural networks and grid search to develop a user model that demonstrated moderate agreement with students’ actual behavior. This model was then utilized to examine student behavior over time. We found that, compared to the one-size-fits-all approach, aesthetic personalization appears to be linked with a higher probability of sustained engagement with the VLE during the initial days of interaction, despite this difference becomes inconsistent thereafter. This discovery suggests that while aesthetic personalization might enhance student learning by optimizing engagement with the VLE, it might suffer from the novelty effect.