Uygun Simulasyon Sayısının Belirlenmesi: Monte Carlo Simülasyon Çalışması

Biyolojik, sosyal, ziraat, tıp ve ekonomik olayların modellenmesinde simülasyon teknikleri oldukça önemlidir. Dolayısıyla, denemenin planlanması aşamasında kaç simülasyon denemesinin yapılacağı sorusu oldukça önemli bir sorudur. Çünkü, yapılacak parametre tahminlerinin kararlı olup olmaması, simülasyon sayısı ile oldukça ilişkilidir. Bu çalışmada, simulasyon sayısının gerçekleşen 1.Tip hata olasılıklarının kararlılığı üzerine etkisi araştırılmıştır. Çalışmada, 16 farklı simülasyon sayısı bakımından gerçekleşen 1.Tip hata olasılıkları tahmin edilmiştir. Çalışma sonuçları, simülasyon sayısının az olduğu durumlarda aynı deneme koşullarında gerçekleşen 1.Tip hata olasılıkları arasındaki farkın büyüdüğünü göstermiştir. Diğer yandan, çalışma sonuçları örnek hacmi ne olursa olsun genel olarak 50000-70000 simülasyon sayısının uygun bir simülasyon sayısı olduğunu göstermiştir

Determining of Suitable Simulation Number: A Monte Carlo Simulation Study

Simulation is an important method for modeling biological, social, agricultural, medicine and economic processes. Therefore, at the planning stage of a simulation modeling the question of number of simulation runs N is very critical. Because, stabilization of parameter estimation changes depend on number of simulation. This study aimed to investigate effect of number of simulation on stabilize of actual Type I error rate. For this aim, Type I error rates were estimated for 16 different number of simulation conditions. Results of this simulation study suggested that while number of simulation was small, differences among the actual Type I error rates increased. On the other hand, it could be said that 50000-700000 simulation numbers were optimum for all sample sizes

___

  • Anonymous, 1999. SAS Institute Inc., SAS OnlineDoc®, Version 8, Cary, NC: SAS Institute Inc., Sons.
  • Bradley, J. V. 1978. Robustness? British Journal of Mathematical and Statistical Psychology, 31, 144-152.
  • Bratley, P., B. L. Fox and L. E. Schrage, 1987. A guide to simulation (2nd ed.). New York: Springer-Verlag.
  • Chambers, J. M. 1977. Computational methods for data analysis. New York: John Wiley & Sons.
  • Cohen, J. 1988. Statistical Power Analysis for the Behavioral Sciences, 2nd ed. Hillsdale, NJ: Erlbaum.
  • Darlington, R. B. 1996. Estimating the true accuracy of regression predictions. Mid-Western Educational Researcher, 9(4), 29- 31.
  • Diaz-Emparanze, I. 2002. Is small Monte Carlo analysis a good analysis? Chacking the size power and consistency of a simulation based test. Statistical Papers, 43 (4), 567-577.
  • Fishman, G. S., L. R. Moore, 1982. A statistical evaluation of multiplicative congruential random number generators with modulus 231-1. Journal of the American Statistical Association, 77, 129-136.
  • Halperin, S. 1976. Design of Monte Carlo studies. Paper presented at the meeting of the American Educational Research Association, San Francisco, CA. (ERIC Document Reproduction Service No. ED 121 850)
  • Harwell, M. R. 1990. Summarizing Monte Carlo results in methodological research. Paper presented at the meeting of the American Educational Research Association, Boston, MA. (ERIC Document Reproduction Service No. ED 319 775)
  • Karian, Z. A., E. J. Dudewicz, 1991. Modern statistical systems, and GPSS simulation: The first course. New York Computer Science Press.
  • Kennedy, W. J., Jr., J. E. Gentle, 1980. Statistical computing. New York Marcel Dekker.
  • Kleijnen, J. P. C. 1974. Statistical techniques in simulation (Part I). New York Marcel Dekker.
  • Kromrey, J. D., C. C. Hines, 1995. Use of empirical estimates of shrinkage in multiple regression: A caution. Educational and Psychological Measurement, 55: 901-925.
  • L'Ecuyer, P. 1988. Efficient and portable combined random number generators. Communications of the ACM, 31: 742- 774.
  • Mendes, M. 2002. Normal dağılım ve varyansların homojenliği ön şartlarının gerçekleşmediği durumlarda varyans analizi tekniği yerine kullanılabilecek bazı parametrik alternatif testlerin I.tip hata ve testin gücü bakımından irdelenmesi. Ankara Üniv. Fen Bil. Enstitüsü. Doktora Tezi. (Basılmamış)
  • Mendes, M., E. Başpınar, 2003. Normal Olmayan Dağılımlı Populasyonlardan Alınan Örneklerde Hesaplanan Çeşitli Test İstatistiklerinin 1.Tip Hata Olasılıkları Bakımından Karşılaştırılması. Ankara Üniv. Ziraat Fak. Tarım Bilimleri Dergisi, 9(1): 23-28.
  • Mendes, M., A. Pala, 2004. Evaluation of four tests when normality and homogeneity of variance assumptions are violated. Pakistan Journal of Information and Technology, 4 (1): 38-42.
  • Micceri, T. 1989. The unicorn, the normal curve, and other improbable creatures. Psychological Bulletin, 105 (1): 156- 166.
  • Mooney, C. Z. 1997. Monte Carlo simulation (Sage University Paper series on Quantitative Applications in the Social Sci., series no. 07-116). Thousand Oaks, CA: Sage.
  • Morgan, B. J. T. 1984. Elements of simulation. New York Chapman and Hall.
  • Nash, J. C. 1990. Compact numerical methods for computers: Linear algebra and function minimisation (2nd ed.). New York Adam Hilger.
  • Olson, C. L. 1972. A monte carlo investigation of the robustness of multivariate analysis of variance. Unpublished doctoral dissertation, University of Toronto, Toronto, Canada.
  • Park, S. K., K. W. Miller, 1988. Random number generators: Good ones are hard to find. Communications of the ACM, 31: 1192-1201.
  • Ripley, B. D. 1987. Stochastic simulation. New York: John Wiley & Sons.
  • Robey, R. R., R. S. Barcikowski, 1992. Type I error and the number of iterations in Monte Carlo studies of robustness. British Journal of Mathematical and Statistical Psychology, 45: 283-288.
  • Rubinstein, R. Y. 1981. Simulation and the Monte Carlo method. New York John Wiley & Sons.
  • Sawilowsky, S. S., R. C. Blair, 1992. A more realistic look at the robustness and type II error properties of the t test to departures from population normality. Psychological Bulletin, 111 (2): 352-360.
  • Wilcox, R. R. 2002. Comparing variances of two independent groups, British Journal of Mathematical and Statistical Psychology, 55: 169-175.