Many robust parameter design (RPD) studies involve a split-plot randomization structure and to obtain valid inferences in the analysis, it is essential to account for the design induced correlation structure. Bayesian methods are appealing for these studies since they naturally accommodate a general class of models, can account for parameter uncertainty in process optimization, and offer the necessary flexibility when one is interested in non-standard performance criteria. In this article, we present a Bayesian approach to process optimization for a general class of RPD models in the split-plot context using an empirical approximation to the posterior distribution of an objective function of interest. Two examples from the literature are used for illustration.
Citation: Journal of Quality Technology
Pub Type: Journals
Bayesian Predictive Density, Generalized Linear Mixed Models, Hard-to-Change Factor, Markov Chain Monte Carlo, Process Optimization, Response Surface, Restricted Randomization.