STOPPING RULES FOR SEQUENCES OF FACTORIAL-DESIGNS

When the sequential design of response surface and fractional factoria l experiments was introduced in the 1950s, the question of deciding wh en to stop was raised but not answered. In this paper the decision is made using a Bayesian analysis, by obtaining the posterior distributio n of the difference between the optimal response and the response from the predicted optimum. The experimentation is stopped, for example, w hen the mean of this posterior distribution falls below a specified va lue. The method works best if the variances of the prior distribution are not too low.