A Multimodal Methodology for the Detection of Emotion Regulation

• In Event: Roundtable Session 6
  In Roundtable Session: Conceptualizing Multiple and Multimodal Methodologies

Abstract

Methods for capturing emotion regulation (ER) have relied primarily on self-report data with limitations such as inability to capture unconscious ER. One way to address these limitations is to use multiple observational channels (i.e., vocal, facial and postural expressions) to detect ER. We describe this multimodal methodology in a multiple case study of four medical students learning to regulate their emotions when presenting bad news to patients. Through qualitative analysis of the data sources we identified: (a) misalignment between emotion channels (e.g., neutral facial expression but agitated voice); (b) sudden changes in emotion expression; and (c) occurrence of unexpected emotions. We also document co-occurrences of the data channels with regulatory strategies as a model framework to determine instances of ER.

Authors

• Maedeh Assadat Kazemitabar, McGill University

• Susanne P. Lajoie, McGill University