Combining Expert Judgments: the Classical and Copula Methods

Introduction Simulation models are frequently used to evaluate the effects of the energy and other economic sectors on the environment, safety, and other areas of concern. In many important contexts, such as risks of global climate change and risks of accident at nuclear power plants, key model parameters cannot be empirically estimated, and so values are obtained by expert judgment. While it seems clear that using judgments from multiple experts is superior to using judgments of a single expert, the best method for combining information from multiple experts is not obvious. Two alternative methods show promise: the classical (Cooke, 1991) and copula (Jouini and Clemen, 1996) methods. This paper describes the initial phase of a project to evaluate and compare the performance of these combination methods.