ON ROBUSTNESS OF DESIGNS AGAINST INCOMPLETE DATA

SUMMABY. In this paper, we characterize the robustness property of designs against incomplete data in the sense that, when any t (a positive integer) observations are missing, all parameters are still estimable in the model assumed. We also present some examples of Srivastava-Chopra Optimum balanced resolution V plans for 2m factorials which are robust against missing of any two observations.