Making Sense of Output and Increasing Efficiency

  • George S. Fishman
Part of the Springer Series in Operations Research book series (ORFE)


A simulation run customarily generates sample-path data on user-specified dynamic processes that evolve as simulated time elapses. These often include delay time, queue length, and resource utilization. Sample averages summarize these data and usually become the focal point of reports and presentations to management. Since they merely approximate corresponding long-run averages that would be observed if the simulation were run for an infinite rather than a finite amount of time, some measure of error should accompany each sample average. Confidence intervals, standard errors, and relative errors provide three alternative means of assessing how well sample averages approximate true long-run averages. As illustration, the Final Tableau in Figure 6.12a displays 99 percent confidence intervals under the headings Lower and Upper, standard error under sqrt[B*W(L, B)/t], and relative error under (Upper — Lower)/\(\left| {\bar X} \right|\). A report that omits an error assessment can lead to serious misconceptions of how well sample averages characterize unknown long-run averages.


Sample Average Variance Reduction Interval Length Exceedance Probability Percent Confidence Interval 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Banks, J., J.S. Carson, III, and B. L. Nelson (1996). Discrete-Event System Simulation, second edition, Prentice-Hall, Upper Saddle River, New Jersey.Google Scholar
  2. Bechhofer, R.E., T.J. Santner, and D. Goldman (1995). Design and Analysis for Statistical Selection, Screening and Multiple Comparisons, Wiley, New York.Google Scholar
  3. Fishman, G.S. (1996). Monte Carlo: Concepts, Algorithms, and Applications, Springer-Verlag, New York.zbMATHGoogle Scholar
  4. Kendall, M.G., and A. Stuart (1961). The Advanced Theory of Statistics, 2. Hafner, New York.Google Scholar
  5. Law, A.M., and W. D. Kelton (1991). Simulation Modeling and Analysis, second edition, McGraw-Hill, New York.Google Scholar
  6. Matejcik, F.J., and B.L. Nelson (1995a). Two-stage multiple comparisons with the best for computer simulation, Operations Research, 43, 633–640.MathSciNetzbMATHCrossRefGoogle Scholar
  7. Matejcik, F.J., and B.L. Nelson (1995b). Using common random numbers for indifference-zone selection and multiple comparisons in simulation, Management Science, 41, 1935–1945.zbMATHCrossRefGoogle Scholar
  8. Scheffé, H. (1970). Practical solutions of the Behrens—Fisher problem, J. Amer. Statist. Assoc., 65, 1501–1508.MathSciNetzbMATHCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2001

Authors and Affiliations

  • George S. Fishman
    • 1
  1. 1.Department of Operations ResearchUniversity of North Carolina at Chapel HillChapel HillUSA

Personalised recommendations