Controlled sequential factorial design for simulation factor screening

Screening experiments are performed to eliminate unimportant factors so that the remaining important factors can be more thoroughly studied in later experiments. Sequential screening methods are specifically fit for simulation experiments. They are usually more efficient than one-step procedures. The challenge is to prove the "correctness" of the result. This paper proposes controlled sequential factorial design (CSFD) for discrete-event simulations. It combines sequential hypothesis-testing procedures with the traditional factorial design to control the type I error and power for each factor under heterogeneous variances conditions. CSFD requires minimum assumptions and demonstrates robust performance with different system conditions.