De-aliasing in two-level factorial designs: A Bayesian approach

Given limited resources for conducting follow-up trials, the inability to separate aliased factorial effects hinders the ubiquitous practicality of regular fractional factorial designs in the analysis of experiments. Wu (2015) proposed a frequentist remedy for "de-aliasing" aliased effects by using conditional main effects. Although Su and Wu (2017) systematized the remedy to make it easy to implement, it might miss truly active effects. Missing active effects can be a severe drawback if the purpose of experimentation is to determine the mechanism of a process rather than to make predictions. In this paper, we propose a Bayesian remedy for de-aliasing in two-level regular factorial designs. Through numerical studies, we show that our method can yield desirable model fittings and reliable de-aliasing results. (C) 2019 Elsevier B.V. All rights reserved.