Bayesian design for life tests and accelerated life tests

This dissertation, consisting of three separate papers, describes Bayesian methods for life test planning and accelerated life test planning with censored data from a log-location-scale distribution, when prior information of the model parameters is available. The first paper studies Bayesian life testing planning with Type II censored data from a Weibull distribution with given shape parameter, where closed form solutions are available. The second paper presents Bayesian methods for life test planning with a general distribution in a log location-scale-family, in which a large sample approximation approach and a simulation approach are developed to evaluate the criterion and provide plan solutions. The third paper describes Bayesian optimum design methods for accelerated life tests with one accelerating variable and a linear acceleration model, where a large sample approximation is used for the test solutions and simulations are used to evaluate the resulting designs. Appropriate Bayes criteria are developed for each of the situations discussed, and numerical examples are used to illustrate the practical use of the Bayesian design methods