A sequential approach to cross-impact analysis

A knowledge of the likelihoods of future scenarios is needed for planning in industry and government. The approach in this article employs the knowledge and the experience of “experts”, in the form of subjective probabilities, to determine the likelihood of the events. The necessary and sufficient conditions that the elicited information from the experts must satisfy in order to consistently compute the likelihoods of the scenarios are derived. A sequential procedure is developed that uses this information in generating the probabilities of the scenarios. Approximation schemes and sensitivity analysis are recommended to implement the approach with less time, effort, and cost.