Error Reduction in Distributed DSS Through Coordination of Modeling Activities: Simulation Study

As end-user computing in general moves toward maturity, the integration of departmentally developed computer models has been a major concern for information systems (IS) researchers. However, decision support systems (DSS) have placed emphasis on performance of individual modeling activities, overlooking the need for coordination mechanisms to integrate the models into a reliable whole as a distributed DSS. In this article, we investigate coordination issues associated with end-user models with a focus on how such coordination affects decision quality through reduction of errors. The article first enumerates the various types of model interdependence and then proposes a coordination strategy for accommodating each interdependence type. Using simulation methodology, we demonstrate that quality, as measured by the number of errors, can be improved through coordination of individual modeling activities. The simulation results help both practitioners and researchers define a new role of the IS function around the coordination of end-user modeling activities, in addition to the current role of transferring technological skills to end users.