A Hierarchical Bayesian Approach to Robust Parameter Design

The goal of robust parameter design experiments is to identify signiflcant location and dispersion factors that can be used to set the mean response at the target level and to decrease the sensitivity of the response to uncontrolled noise factors. We present a hierarchical Bayesian model and use empirical Bayes priors to flnd the active factors and to get reliable estimates of the location and dispersion parameters. The approach is particularly useful when the design points are not replicated, a case which is challenging with standard procedures. Keywords: Hierarchical Bayesian model, Empirical Bayes priors, location and