Using common random numbers and control variates in multiple-comparison procedures

Using common random numbers and control variates in multiple-comparison procedures

0.00 Avg rating0 Votes
Article ID: iaor19931221
Country: United States
Volume: 39
Issue: 4
Start Page Number: 583
End Page Number: 591
Publication Date: Jul 1991
Journal: Operations Research
Authors: ,
Keywords: statistics: experiment
Abstract:

This paper considers the determination of the relative merits of two or more system designs via stochastic simulation experiments by constructing simultaneous interval estimates of certain differences in expected performance. Tukey’s all-pairwise-comparisons procedure, Hsu’s multiple-comparisons-with-the-best procedure, and Dunnett’s multiple-comparisons-with-a-control procedure are standard methods for making such comparisons. The authors propose refinements for all three procedures through the use of two variance reduction techniques: common random numbers and control variates. They show that the proposed procedures are better than the standard multiple-comparison procedures in the sense that they have a larger probability of containing the true difference and, at the same time, not containing zero when a difference exists.

Reviews

Required fields are marked *. Your email address will not be published.