AR for Science Education: Students’ Behaviour Patterns and the Relationship between Cognitive Load, Knowledge Acquisition and Performance

Abstract

Augmented Reality (AR) has been gaining increased attention in science education, as research has demonstrated its ability to pro-
vide students with an engaging and interactive way to learn. Students are able to visualize and experience phenomena that would otherwise remain hidden, practice in simulated hazardous conditions, and develop their spatial skills. While the potential benefits of this technology are clear, more research is needed to understand how students actually interact with and experience AR. Due to its novelty, using AR is challenging for some users. This sometimes leads to a high cognitive load during the learning task, which can ultimately impact performance. To investigate how students interact with an AR application, the present study examined students’ in-app performance, knowledge acquisition and perceived cognitive load through a Bayesian hypothesis testing approach. More-
over, process mining was used to qualitatively assess students’ behaviour patterns. The contribution of this work is twofold: 1) the observed interactions during an AR intervention provide insights into the relationship between students’ cognitive load, in-app performance and knowledge acquisition, and 2) the process analysis helps to derive general recommendations for practitioners who are designing or using a similar AR application in their practice.