Augmented Reality Enhanced Analytics to Monitor Emotional Engagement in Young Students

Abstract

This paper investigates the use of Augmented Reality Enhanced Analytics (AREA) to monitor emotional engagement in primary students during AR-based learning experiences. A prototype AR application, Falling Leaves, was designed to capture near real-time sensor data such as session duration, movement patterns, and screen interactions while students engaged in three short game play sessions. Emotional states were concurrently evaluated using a structured teacher-observed rubric on seven emotions: happiness, surprise, concentration, frustration, sadness, fear, and anger. Regression analysis revealed that longer AR sessions were significantly associated with fewer frustrations, while increased physical movement predicted higher levels of surprise. These findings support the potential of AREA as a non-intrusive, near-real-time alternative to traditional post-session emotional assessments, providing timely insights for the development of emotion-aware educational technologies.