Value-Focused Thinking in the Presence of Weight Ambiguity: A Solution Technique Using Monte Carlo Simulation

摘要

When a Decision Maker is asked to provide his or her preferences, the response represents a snapshot in time. While their preference structure elicited at any given moment may represent their revealed preferences at that point in time, it may change over time. These changing preferences over time represent ambiguity in the decision maker's preferences. Other sources of ambiguity may exist. One weakness of many decision analysis techniques today is the inability to incorporate ambiguity into the basic decision model. The existence of the problem has been known and commented on for many years. This research addresses that problem. It begins with the basic approach and methodology developed by Ralph Keeney, Value-Focused Thinking (VFT). This methodology is then expanded to allow decision makers to specify not just constant weights to demonstrate their preferences, but an entire distribution. These distributions are then incorporated with the value of the attributes and the whole is simulated using Monte Carlo Simulation provided by Crystal Ball. The result of incorporating these weight distributions into the model, is an empirical distribution for the value of an alternative. The alternative distributions can be compared in a number of ways to provide insight to the decision maker.