Use of Performance Shaping Factors and Quantified Expert Judgment in the Evaluation of Human Reliability: An Initial Appraisal.

摘要

The first part of the report considers the nature of human reliability assessment, and the techniques currently employed. It is concluded that most approaches are limited by the availability of data. Approaches to the subjective assessment of error are surveyed. A particular technique which has been developed, the Success Likelihood Index Methodology (SLIM), is described in detail, together with the practical steps for its implementation. The results from a trial application of a questionnaire designed to elicit judges' perceptions of the relative importance of performance shaping factors in determining human reliability are analyzed. A revised form of the questionnaire is presented for future use. A pilot experiment to investigate the relationship between subjectively derived indices of success for six tasks and their objective probability of success is described. The results indicate that the SLIM has potential value as a predictive technique. Some requirements for a program of research to produce a generally applicable methodology are set out.