Wen Dong, Bruno Lepri, Taemie Kim, Fabio Pianesi, Alex 'Sandy' Pentland

Abstract

In this paper, we describe a novel approach, based on Markov jump processes, to model group interaction dynamics and group performance. In particular, we estimate conversational events such as turn taking, backchannels, turntransitions at the micro-level (1 minute windows) and then we bridge the micro-level behavior and the macro-level performance. We test this approach with a cooperative game dataset and we verified the relevance of micro-level interaction dynamics in determining a good group performance (e.g. higher speaking turns rate and backchannels rate and lower turns competition rate).