Reputation, Learning, and Coordination in Distributed Decision-Making Contexts

From an organizational perspective, there are many decisions that are not strictly individual. A decision-making process may need to be distributed across multiple participants, each of whom contributes to the final decision by performing one or more tasks. A participant may be a person, a group, a team, or an artifact such as a computerized decision-support system. Computers are routinely used to support individual decision making. However, their potential for supporting distributed decision making is only beginning to be actualized. Further progress in exploring and realizing this potential can benefit greatly from a formal model that accounts for the diverse phenomena that can occur within a distributed decision maker.When considering computer-based support for distributed decision making, the issue of coordinating the multiple participants becomes the central concern. Coordination defines the structural and dynamic patterns of inter-participant relationships in an organization. It has several aspects including planning, control, and review. Planning involves task decomposition, subtask allocation and synthesis. Control indicates mediation, negotiation and execution. And review deals with performance evaluation which will then contribute to organizational learning. Moreover, coordination occurs in a context of concurrent problem-solving tasks where multiple decisions are pending simultaneously. Bits and pieces of these coordination facets have been somewhat supported by existing computer technologies, such as decision-support systems and computer-mediated communication systems. However, systematic study of such support possibilities depends on formal models of distributed decision making as organizing paradigms.In this paper, we present a model of distributed decision making that is particularly concerned with the ongoing coordination among participants in multiple simultaneously active decision processes. Basic outlines of the model are presented as an initial foundation for understanding the possibilities of computer-based support for distributed decision making. The model takes the view that an organization is dynamic in terms of its capacity for improved coordination over time and through experience. It accommodates a bidding perspective as the context for coordination. The use and adjustment of entity reputation offers a means for improved coordination over time, and for capturing the phenomenon of organizational learning.