Combining probability distributions from experts in risk analysis

This paper concerns the combination of experts' probability distributions i n risk analysis, discussing a variety of combination methods and attempting to highlight the important conceptual and practical issues to be considere d in designing a combination process in practice. The role of experts is im portant because their judgments can provide valuable information, particula rly in view of the limited availability of "hard data" regarding many impor tant uncertainties in risk analysis. Because uncertainties are represented in terms of probability distributions in probabilistic risk analysis (PRA), we consider expert information in terms of probability distributions. The motivation for the use of multiple experts is simply the desire to obtain a s much information as possible. Combining experts' probability distribution s summarizes the accumulated information for risk analysts and decision-mak ers. Procedures for combining probability distributions are often compartme ntalized as mathematical aggregation methods or behavioral approaches, and we discuss both categories. However, an overall aggregation process could i nvolve both mathematical and behavioral aspects, and no single process is b est in all circumstances. An understanding of the pros and cons of differen t methods and the key issues to consider is valuable in the design of a com bination process for a specific PRA. The output, a "combined probability di stribution," can ideally be viewed as representing a summary of the current state of expert opinion regarding the uncertainty of interest.