John Grego

Department of Statistics

University of South Carolina


Empirical Bayes Models for Process Mean and Variance

Several different approaches have recently been proposed for analyzing process mean and variance in industrial experiments. I will review some of these methodologies and discuss in particular an Empirical Bayes approach due to Sohn and Park (1998). Wenkai Hu (a former Master's students) and I have made extensions to Sohn and Park's random intercept models to include random slopes for the mean effects. Progress on random slope models for the variance effects and analysis of unreplicated experiments will be discussed.


Back to Colloquium Series