Examining the Relationships Between Gameplay Behavioural Trace Data and Pupils’ Behavioural Engagement in an Augmented Reality Learning Environment

Learning analytics enables the collection of behavioural trace data from digital learning environments, providing objective records of learners’ interactions. However, the educational interpretation of these traces remains underdeveloped, particularly in augmented reality (AR) learning environments, where little is known about their relationship with pupils’ behavioural engagement. This study examined the relationships between gameplay-derived behavioural trace indicators and five dimensions of pupils’ behavioural engagement in an AR learning environment. A quantitative, non-experimental correlational design involved 63 primary school pupils. Gameplay Duration, Gameplay Score, Total Clicks and Distance Moved were automatically captured across three gameplay sessions using the Falling Leaves AR learning application developed by the research team. Behavioural engagement was measured using the Motivation and Engagement Scale-Junior School (MES-JS), comprising Planning, Task Management, Persistence, Self-Sabotage and Disengagement. Five exploratory multiple linear regression models were estimated to examine the associations between gameplay-derived behavioural trace indicators and the five engagement dimensions. Gameplay Score remained significantly associated with all five engagement dimensions after the Benjamini–Hochberg correction, while Gameplay Duration remained significant only for Planning and Disengagement. Total Clicks and Distance Moved showed no statistically significant associations in the regression models. These findings suggest that selected gameplay-derived behavioural trace indicators may provide complementary information about pupils’ behavioural engagement, although interpretation remains subject to measurement limitations and the exploratory nature of the analysis.