A Bayes sensitivity analysis when using the beta distribution as a prior

Abstract

Results that illustrate the determination of the parameters of a beta prior distribution when either the mean and one percentile or two percentiles are given are presented. Further results show how sensitive the parameters of the beta prior are to errors in the given values. The sensitivity of the probability distribution of the reliability of a system to errors in either the given mean and percentile or the given percentiles of the component beta priors is studied. Included in the study are several systems, ranging from simple ones that occur as subsystems of real systems up to a real type of system consisting of 74 components.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>