The Coordination-Rapport Link, Revisited: A Study with Automated Coding Techniques

Rapport signals closeness in relationships. Despite the accumulation of research findings, nonverbal correlates of rapport are not always consistent across studies due to the gestalt nature of rapport experiences. Given that rapport exists only in interaction between individuals, this study examined not specific nonverbal cues, but rather nonrandom and synchronized nonverbal dynamics between interlocutors, i.e., interpersonal coordination. To provide comprehensive coordination measures free from observers’ potential cognitive biases, we employed a video-based automated coding approach. Using the output of pose estimation software, OpenPose, interactional synchrony was quantified in terms of interaction rhythms and simultaneous movement, whereas behavior matching was computed as posture mirroring using body joint angles. Moreover, following the pseudo-synchrony paradigm, a surrogate distribution was generated by repeatedly shuffling data either within individuals or between dyads, which yielded a metric analogous to an effect size. The results showed that synchrony of rhythm and behavior matching were significantly positively associated with rapport. The pseudo-synchrony analysis with within- and between-dyad shuffling yielded mixed resulting pattens, suggesting that the interlocutors exhibited less phase-locked and loosely coupled coordination. This is the first study to show that both interactional synchrony and behavior matching measured by video-based pose estimation software are associated with rapport. Our findings are discussed in terms of the theoretical importance of differentiating synchrony and matching as a coordination construct and the possible contribution of the automated coding approach to illustrating rapport’s behavioral manifestations between interlocutors.