Assessing the impact of information on decision-making in contexts lacking well-defined utility functions

This dissertation provides a methodology that enables analysts to investigate and measure the impact of information on decision-making in distributed supervisory control systems. These systems are characterized by remote supervisors that provide direction to subordinates in a changing environment. In many cases, the local subordinates must dynamically adapt their plans and decisions in a manner consistent with the intent of the supervisor. Situations exist where decision-making is required in difficult conditions that require urgency in a dynamic environment. In these cases, communications are critical to a successful operation.;For example, wild fire fighting is a profession where rapid response is required with significant consequences at stake. The decision-making process made to combat forest fires in an epistemically uncertain environment are very closely related to the decision-making process of military leaders. This research builds upon advancements made in the modeling and simulation of information systems in the military to the profession of fighting forest fires.;This dissertation extends the field of industrial engineering by creating a framework for analyzing the affect of information on decision-making in distributed supervisory control systems. New measures and techniques have been implemented and instantiated within an agent-based model to quantitatively examine the impact of technological systems on the perceptual and cognitive systems described in the DMSC. This research extends the previous work of modeling the affect of information beyond the technological systems into the cognitive domain.;The foundation of the analysis is the development of relevance filters to segment the data into discrete time intervals based on events within the simulation. The end result of this research is a mathematical model and response surface that can be used to assess the contribution of the factors (sensors, network) to the comprehension and projection of the decision-maker. The resultant model provides correlations between the factors and the accuracy of decisions made at the platoon level. This research and methodology can be applied to similar industries such as fire fighting.