Emotion Theory and Learning Analytics: A Theoretical Framework for Capturing Emotion Regulation Using Process Data

Emotion regulation (ER) refers to the constant monitoring and modulating one’s dysfunctional emotion states and is essential for increasing learning outcomes with advanced learning technologies. Prior studies have typically captured ER using static instruments (e.g., self-reports) while little has focused on using various types of process data, including facial expressions of emotions, log files, screen recordings, physiology, verbalizations, and eye tracking. This possibly stems from the general lack of guidance for researchers on how to capture ER processes as individuals learn with advanced learning technologies. With an enormous increase in the access to process data for researchers, this chapter contributes to the field of learning analytics by reflecting on state-of-the-art methods and describing a theoretically-driven framework to study each phase of ER using process data. Specifically, this chapter extends McRae and Gross’ (2020) ER model by identifying modalities of data that represent different phases of ER. This framework lays the theoretical groundwork for how future studies should incorporate multimodal data to capture the ER process, increasing the accessibility and utility of leveraging process data to a range of stakeholders for capturing and identifying ER using learning analytics.