Behaviorally-Anchored Team Member Feedback in Undergraduate Learning: Linking Verbal Communication to AI-Detected Emotional Expressions

This study examines how undergraduate team members’ verbal communication behaviors are associated with the emotional expressions of their teammates during student project team interactions. Drawing on data from a semester-long, team-based undergraduate course, three key team meetings were video-recorded across the team life cycle. Verbal communication was analyzed using a systematic content analysis with an a priori codebook capturing emotionally relevant behaviors such as optimism, empathy, building bonds, developing others, service orientation, and task-focused communication. Nonverbal emotional expressions were assessed using AI-based facial expression analysis grounded in Facial Action Coding System research, generating average emotion scores for each team member. Emotion indices were calculated to examine how one member’s verbal communication corresponded with the aggregated emotional expressions of their teammates. Correlations revealed consistent and meaningful associations between communication behaviors and teammates’ emotional responses. These findings conceptualize emotional intelligence in teams as an observable, interpersonal phenomenon and demonstrate the value of integrating content analysis with AI-based emotion detection to provide behaviorally anchored feedback in team-based learning environments.