Model of Coordination Flow in Remote Collaborative Interaction

We present an information-theoretic approach for modelling coordination in human-human interaction and measuring coordination flows in a remote collaborative tracking task. Building on Shannon's mutual information, coordination flow measures, for stochastic collaborative systems, how much influence, the environment has on the joint control of collaborating parties. We demonstrate the application of the approach on interactive human data recorded in a user study and reveal the amount of effort required for creating rigorous models. Our initial results suggest the potential coordination flow has - as an objective, task-independent measure - in supporting designers of human collaborative systems and in providing better theoretical foundations for the science of Human-Computer Interaction.