Title: Combining Probability Distributions from Experts in Risk Analysis
Authors: Robert T. Clemen  Duke University
Date: April 15, 1996
Status: working paper

How effective are decision-analysis techniques? Certainly decision analysis can show an individual how to be coherent in making inferences and choices. On the other hand, behavioral decision theory shows that people are not always coherent or internally consistent. And many other techniques have been developed to help individuals and groups understand their preferences, construct models, and make decisions. But are the techniques really effective in helping people to achieve their objectives? In this paper, we distinguish between two types of effectiveness. First is the notion that individuals must typically make choices before all uncertainties are resolved; such decisions must be made on the basis of preferences over those uncertain prospects, and the question is whether decision analysis leads to options that are broadly viewed as preferable. Second, after all uncertainties are resolved, the decision maker experiences the consequences, and the question here is whether decision analysis really leads to better outcomes: Are users of decision analysis happier, healthier, wealthier, or wiser in the long run? We use the terms weak and strong effectiveness, respectively, to refer to these two ways to think about the effectiveness of decision-making techniques.

Unfortunately, no body of literature exists that demonstrates unequivocally that decision analysis is either weakly or strongly effective. In this essay, we discuss how to determine the answers to these two questions. We review some existing related work, but much of the discussion is prospective, describing a number of possible research paradigms and experimental procedures. The essay concludes with suggestions for specific research to be done and some dilemmas that researchers will face.

Click here to access pdf version of the paper
Word V6.0 version of the paper available from Robert T. Clemen's personal web site

Back to the Decision Analysis Working Paper Index