1782: Leader Emergence in the Digital Realm: Exploring Communication Dynamics via Machine Learning (17615)

Leader emergence in virtual teams has become a pivotal factor in organizational success in today's interconnected business world. Based on expectancy violation theory, this study examines how speaking centrality (the extent of a member's central role in the speaking network) and backchanneling (expressing agreement verbally) influence leader emergence in virtual environments. Diverging from previous research that mainly focused on individual attributes through surveys, our study delves into leader emergence by analyzing verbal communication dynamics within virtual teams. We investigated 51 participants in 12 virtual teams, employing sophisticated machine learning tools to scrutinize group interactions during the idea generation phase of social entrepreneurship projects on Zoom videoconferencing. Additionally, we gathered data on leader votes at a later time using the roster method through surveys. Our results reveal a positive link between speaking centrality and leader emergence. Contrary to expectations, the frequency of a member's backchanneling didn’t moderate this relationship. Instead, it was the rate at which a speaker inspired backchanneling in others (i.e., backchanneling inspiration rate) that significantly positively moderated the relationship between speaking centrality and leader emergence. This thorough mixed-method approach sheds new light on group communication dynamics and makes a substantial contribution to understanding leader emergence in virtual contexts.

view paper (if available)