This research examines the roles of first-order saturated, supersaturated and minimal level change designs in the problem of minimizing the number of experiments required to differentiate the influential factors from those only contributing noise. Optimal experimental design in an industrial setting may also have as an objective the minimization of the number of experiments. Theoretical solutions have often depended upon unrealistic assumptions about the likely number of influential factors and their interactions. This research focuses on an efficient and effective approach to the design of a minimal series of experiments to identify those factors with the goal of predicting the optimal combination of factors or factor settings for full-scale production.
