Task sequencing within a distributed human team: experimental paradigm and normative modeling

An experimental paradigm and a preliminary normative model for studying task sequencing within a distributed human team are presented. In the problem, a team consisting of one leader and and three subordinates is to maximise the team reward by processing two of randomly arriving tasks: individual tasks and multioperation tasks (requiring sequential processing by two or three team members). The team decision-making and coordination mechanisms are investigated with respect to leader's involvement, communication structure, and workload. The experimental paradigm is designed to provide a testbed for conducting human experiments. The preliminary normative model is developed to provide a baseline for evaluating team performance and for follow-on normative-descriptive model development. By simplifying the dynamic and stochastic aspects of the problem the model is formulated as a centralized static scheduling problem, to be solved in a moving window fashion. Using Lagrange relaxation and duality theorem, the model is converted to a two-level framework representing the decision-making process of the distributed team.<>