Implementation of a Maximin Power Clustering Criterion to Select near Replicates for Regression Lack-of-Fit Tests

Abstract

In earlier work, we presented a maximin power clustering criterion to partition observations into groups of near replicates. Specifically, the criterion selects near replicate clusters for use with Christensen's tests for orthogonal between and within cluster lack of fit. This article further explores implementation of this clustering criterion. In particular, a methodology is developed to determine a collection of candidate groupings to which the maximin power criterion can be applied.