Abstract
Primary purpose of the article is to develop outlier robust designs. As a matter of fact, negative effect of outliers in any experimental settings is established where the outliers at any specific design point can destroy the features of the design for which it is being developed. It is attempted here in this article to develop a version of robustness for central composite designs which may protect it for outliers by introducing the idea of minimax outlying effect. This involves the calculation of the degree of outlying effect(s) outlier(s) may produce and then minimize the maximum of these outlying effects in an attempt to equalize the influence of all design points. On comparison, these outlier robust designs are proved to be more optimal, on the scales of A, D, and E optimalities, against existing conventional rotatable, orthogonal, and other such designs. The outlier robust designs, developed here, are suitable for settings prone to outliers where conventional designs fail to represent and analyze the processes and systems.