Automated Real-Time Detection of Emotions in Writing With Reflective Questions and Events

• In Event: Poster Session 14
  In Poster Session: Advances in Reading and Writing Research

Abstract

A previous study indicated that facial behavior is the function of the evoking events (Ekman, 1999). In this paper, we applied an automatic facial expression monitoring technique to detect middle school and university students’ real-time facial behavior in an attempt to examine the relationship between emotion, reflective thinking, and writing performance. Results from correlation and ANOVA analyses suggested an inhibitory effect of negative emotion; students with negative emotion (sadness) tended to make more spelling errors. The association between prior knowledge, analytical thinking, and confusion observed in this study is consistent with previous studies, suggesting that confusion is both an identifiable facial marker and a cognitive state. We also observed group and gender effects on emotion and writing.

Authors

• Zhenhua Xu, University of Toronto

• Earl Woodruff, OISE/University of Toronto