Affect Dynamics and Behavioral Patterns in Intelligent Learning Environments

This research examines the interplay between emotions and learning behaviors in Intelligent Learning Environments (ILEs) using data from Brazilian students. We analyze associations between emotions—including confusion, frustration, boredom, and engagement—and learning behaviors, focusing on the influence of gender and the duration of these emotions and behaviors. Data from 30 students in a math step-based tutoring system were annotated by human coders, with transition probabilities assessed using the L metric. Key findings indicate that engagement and confusion correlate with active task involvement, while boredom and frustration are linked to task withdrawal. Additionally, gender-specific dynamics between emotions and behaviors were observed. These results highlight the potential to use the observed relationships between emotions and behaviors to enhance personalized learning experiences and inform the design of adaptive learning environments.