Statistical tests have always played an important role in the analy sis of comparative simulation experiments. This paper argues for a shift of emphasis toward a more exploratory approach to the analysis of such data and, in particular, the wider use of Tukey's median polish.
Balmer, D.W.1986. "Statistical analysis of simulation output." CASM Research Report DWB/86/1 (October).
2.
Crookes, J.G.; D.W. Balmer; S.T. Chew; and R.J. Paul.1986."A three-phase simulation system written in Pascal." J. Opl. Res. Soc. 37, no. 6: 603-618.
3.
Emerson, J.D. and D.C. Hoaglin.1983. 'Analysis ofTwo-Way Tables by Medians ." In Understanding Robust and Exploratory Data Analysis (David C Hoaglin, Frederick Mosteller, and John W. Tukey , eds.). John Wiley and Sons Inc., New York, N.Y., 166-210.
4.
Friedman, L.W. and H.H. Friedman.1985. "The probability distribution as a performance criterion when comparing alternative systems." SIMULATION45, no. 5 (Nov.): 262-264.
5.
Friedman, L.W. and H.H. Friedman.1986. "Comparing simulated alternatives using a distribution-free statistic with blocking by random number stream." SIMULATION47, no. 2 (Aug.): 68-70.
6.
Friedman, M.1937. "The use of ranks to avoid the assumption of normality implicit in the analysis of variance." Journal of the American Statistical Association32: 675-701.
