Affective computing is an interdisciplinary field that aims to automatically recognize and interpret emotions. Recent research has focused on using physiological signals (e.g., electrodermal activity) to improve emotion recognition. However, the theoretical emotion models that underlie these systems have received comparatively little attention. We conducted a systematic review and meta-analysis on electrodermal-activity-based emotionrecognition systems. Our findings suggest that arousal prediction models outperform valence prediction models, supporting our preregistered hypothesis. This correlates with arousal’s association with autonomic nervous system activity and its direct link to electrodermal activity. We also observed a mismatch between the machinelearning approaches most often used—chiefly classification models—and the predominantly dimensional emotion frameworks adopted in the literature. Specifically, although dimensional affective models are increasingly popular, there has not been a parallel rise in regression models that would better reflect the continuous nature of the underlying data. We conclude that a comprehensive understanding of affective states requires consideration of both psychological and computational perspectives in affective computing research.
Tomás Ariel D'Amelio (Sat,) studied this question.