A normative-descriptive approach to hierarchical team resource allocation

Abstract

Dynamic, distributed, human team resource allocation and task processing is considered in an abstracted Navy-like command and control environment. A hierarchical team of one leader and three subordinates is to process multiple types of randomly arriving tasks (threats) that have different processing resource requirements, time requirements, values, and deadlines. Each subordinate is responsible for processing a subset of these tasks. Two kinds of team leader (resource coordinator) are considered: an active leader who transfers resources among the subordinates, and a passive leader who provides offline guidance only. The individual decision-making and coordination processes of both types of leader are described and analytically modeled. A team-in-the-loop experiment provides data to compare with model predictions. The self-centered bias (wherein human decision-makers overvalue their own responsibilities) is identified as a major contributor to model-data mismatch. Incorporating such human cognitive limitations and biases into the normative models successfully replicates the experimental results. >