Clustering Facial Recognition Emotions at Start, During, and End of Queer History App Learning

• In Event: Critical Thinking and Technology-Supported Learning

Abstract

Traditional measurements of emotion (e.g., self-reports) often require subjective input whereas the application of facial recognition software to measure emotions has opened possibilities to measure emotions more objectively and in real-time. In this study, we employed a facial recognition software to automatically identify students’ emotions through a video recording during their interaction with an educational app. We utilized two-step cluster analysis to identify distinct emotion groups at three timepoints: beginning, during, and end of learning session. Findings also revealed that students in the neutral emotion group tended to perceive their level of success as higher than other emotion groups (e.g., negative emotion). Implications for examining and measuring emotions over the course of learning will be discussed.

Authors

• Clarissa Hin-Hei Lau, University of Toronto

• Byunghoon Tony Ahn, McGill University

• Jason Matthew Harley, McGill University