AN ALTERNATIVE VARIANCE-REDUCTION STRATEGY FOR RSM SIMULATION STUDIES.

In the previous paper, Cooley and Houck examined the simultaneous use of common and antithetic random number streams as a variance-reduction strategy for simulation studies employing response surface methodology (RSM). Our paper supplements their work and further explores pseudorandom number assignments in response surface designs. Specifically, an alternative strategy for assigning pseudorandom numbers is proposed; this strategy is more efficient than that given by Cooley and Houck, especially when more than two factors are involved. [ABSTRACT FROM AUTHOR]