Varyansların Heterojen Olması Durumunda K-İstatistiği KANOVA ile ANOVA F Testinin Gerçekle şen 1.Tip Hata Olasılıkları Bakımından Karşılaştırılması

Abstract

Bu çal ışmada K-istatisti ğ i ile ANOVA F testinin 50000 simülasyon denemesi sonunda gerçekle şen 1.Tip hata olas ı l ı kları bak ı m ı ndan karşı laşt ı rı lmas ı yap ı lm ışt ı r. Yap ı lan karşı laşt ı rmalar sonunda, K-istatistiğ inin, özellikle örneklerde 10 ve daha fazla gözlemin bulunmas ı durumlar ı nda ANOVA F testine göre bir çok deneme ko ş ulunda daha iyi sonuçlar verdi ğ i gözlenmiştir. Ancak bu testin.küçük gözlem kombinasyonlar ı ndan oldukça olumsuz yönde etkilendi ğ i ve bu olumsuz etkinin ise özellikle grup say ı s ı ve varyanslar ı n heterojenliğ inin artmas ı na paralel olarak daha da belirginleştiği görülmü ştür.

Keywords

• varyanslar ı n heterojenli ğ i
• 1.Tip hata olas ı l ığı
• KANOVA
• varyans analizi

The Comparison of K-Statistic KANOVA with ANOVA F Test in Terms of Actual Type I Error Rate When Variances are Heterogeneous

Abstract

In this study, K statistic was compared with ANOVA F test in terms of realized type I error rate at the end of 50000 simulation trials. At the end of these comparisons-under various experimental conditions it was observed that K statistic is better than ANOVA F test particularly when each groups have ten or more observations. However, K statistic was affected negatively by small observation combinations. This negative effect was more evident particularly when number of groups and heterogeneity of variance were increased

Keywords

• heterogeneity of variance
• Type I error rate
• KANOVA
• analysis of variance

References

1. Anonymous, 1994. FORTRAN Subroutines for Mathematical Applications. IMSL MATH/LIBRARY. Vol. 1-2. Visual Numerics, Inc., Houston, USA.
2. Brown, M. B. and A. B. Forsythe, 1974. The small sample behavior of some statistics which test the equality of several means. Technometrics, 16, 129-132.
3. Dijkstra, J. B. and P. S. Werter, 1981.Testing the equality of several means when the population variances are unequal. communications in Statistics, Simulation and computation, 10, 557-569.
4. Edgington, E. S. 1974. A new tabulation of statistical procedures used in APA journals. American Psychologist, 29, 25-26.
5. Ghost, M. and Y. S. Kim, 2001. The Behrens-Fisher problem revisited: A Bayes -frequentist synthesis. The Canadian Journal of Statistics, 29, No:1 .
6. Glass, G. V., P. D. Peckman and J. R. Sanders, 1972. Consequences of failure to meet assumptions underlying the fixed effects analyses of variance and covariance. Review of Educational Research, 43, 237-288.
7. Krutchkoff, R. G. 1988. One-way Fixed Effects Analysis of Variance when the Error Variances May be Unequal.J.Statist.Comput.Simul, 30,259-271.
8. Sokal, R. R. and F. J. Rohlf, 1995. Biometry. The Principles and Practice of Statistics in Biological Research. Third Ed. W.H. Freeman and Co., 887, New York.

9. Welch, B. L. 1951. On the comparison of several mean values: An alternative approach. Biometrika, 38, 330-336.
10. Zar, J. H. 1999.Biostatistical Analysis. Fourth Edition. Simon & Schuster/A Viacom Company, New Jersey. USA. 683 S.