Microaspects of Leadership: Evidence from Audio and Tracking Data

Abstract

Previous results have shown that encouraging teams to choose a leader improves their performance in non-routine analytical team tasks. Using real-life escape challenges as a lab-in-the-field setting, we study more than 270 student teams solving such tasks. Before teams start working on their tasks, we ask treatment groups to jointly decide on a team member, who takes on the role of the leader (Endogenous) or we randomly select a team member, who fulfills the role of the leader (Exogenous). Using an objective measure of team performance (i.e. the time teams need to solve the non-routine task), two main observations stand out: Performance is significantly lower for teams with exogenously appointed leaders. But when controlling for random selection of the most-likely endogenous leader, the result of performance-mediating leadership can be restored. To uncover microaspects of leadership, we leverage unique rich data on team communication and team organization and study differences in how teams talk and move. The analysis of audio recordings from the task reveals differences in speaking time shares across and within teams, as well as differences in the use of collective versus individual phrasing and in overall sentiment. Motion measures from tracking data reveal heterogeneous, but noisy, patterns: more time spent in meetings can often signal stagnation, while greater search effort correlates positively with performance.