Quantification of uncertainties is an essential component of probabilistic seismic and volcanic hazard assessments. Formal expert assessment is a structured and documented process for identifying and quantifying uncertainties. Expert judgment is used in any technical assessment, but often is implicit and undocumented. Formal expert assessment explicitly includes judgments of multiple experts to represent the range of views within the scientific community. We use the term “formal expert assessment” rather than the more common term “expert elicitation” to emphasize how this approach differs from the traditional view both in philosophy and in practice.

In a typical expert elicitation, subject-matter experts are asked narrowly defined questions about specific uncertain quantities within their area of expertise, and they provide their judgments in the form of probability estimates or distributions. For example a climate scientist might be asked to provide an estimate of “the equilibrium change in global average surface temperature” given a specific set of circumstances (Morgan and Keith, 1995). In this approach experts are treated as independent point estimators of an uncertain quantity, and the elicitation “problem” is viewed primarily in terms of determining how to ask the right questions as clearly as possible of the most knowledgeable experts. Under this perspective, the elicitors may focus significant effort on ensuring that they have wellcalibrated and informative experts; that is, experts who give both accurate estimates and a narrow range of uncertainty in estimates of quantities similar to those of interest but for which a “true” value can be determined (e.g. Chapter 10 of Bedford and Cooke, 2001).