Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power

MENDEŞ, Mehmet; AKKARTAL, ERKUT

Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power

Mehmet MENDEŞ, (Çanakkale Onsekiz Mart Üniversitesi, Ziraat Fakültesi, Zootekni Bölümü, Çanakkale, Türkiye)

Erkut AKKARTAL (Yeditepe Üniversitesi, İstanbul, Türkiye)

Kafkas Üniversitesi Veteriner Fakültesi Dergisi

14 1

Yıl: 2010 Cilt: 16 Sayı: 5 Sayfa Aralığı: 711 - 716 Metin Dili: İngilizce İndeks Tarihi: 29-07-2022

Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power

Öz:

Bu çalışmada Varyans analizi tekniği (F) ve Welch testi ile bunların permutasyon versiyonları (PF ve PW) 1.Tip hata ve testin gücü bakımından karşılaştırılmıştır. Söz konusu karşılaştırmalar Monte Carlo simulasyon tekniği kullanılmıştır. Yapılan simülasyon çalışmaları sonucunda varyanslar homojen iken bu testlerin permutasyon versiyonlarının 1. Tip hata olasılığını koruma bakımından daha güvenilir sonuçlar verdikleri görülmüştür. Diğer taraftan varyansların heterojenleşmesinden bütün testlerin olumsuz yönde etkilendikleri görülmüştür. Varyansların heterojen ve dağılımların da çarpık (χ2 (3) ve Exp [0.75]), olması halinde örnek hacmi ve etki büyüklüğü ne olursa olsun PF testinin F testine göre biraz daha güçlü olduğu görülmüştür. Ancak dağılımlar simetrik iken (β (5.5)) PF ve F testlerinin güç değerleri benzerdir. W testi varyansların homojen olması halinde daha güçlü iken, PW testi varyansların homojen olmadığı ve örnek hacimlerinin dengesiz olduğu (mesela 5:10:15) durumda biraz daha güçlüdür.

Anahtar Kelime: Monte Carlo method livestock simulation models animal experiments mathematical models analysis of variance

Belge Türü: Makale Makale Türü: Araştırma Makalesi Erişim Türü: Erişime Açık

1. Schneider PJ, Penfield DA: Alexander and Govern’s approximation: Providing an alternative to ANOVA under variance heterogeneity. J Exp Educ, 65 (3): 271-286, 1997.
2. Mendeş M: The comparison of some parametric alternative tests to one-way analysis of variance in terms of Type I error rates and test power under non-normality and homogeneity of variance. PhD Thesis. Ankara University Graduate School of Natural and Applied Sciences Department of Animal Science (unpublished), 2002.
3. Mielke PW, Beny KJ: Permutation test for common locations among samples with unequal variances. J Educ Behav Stat, 19 (3): 217-236, 1994.
4. O'Gorman TW: An adaptive permutation test procedure for several common tests significance. Comput Stat Data An, 35, 335-350, 2001.
5. Peres-Neto PR, Olden JD: Assessing the robustness of randomization tests: Examples from behavioral studies. Anim Behav, 61, 79-86, 2001.
6. Anderson MJ, Robinson J: Permutation tests for linear models. Aust NZ Stat, 43 (1): 75-88, 2001.
7. Nichols TE, Holmes AP: Nonparametric permutation tests for functional neuro-imaging: A primer with examples. Hum Brain Mapp, 15, 1-25, 2001.
8. Önder H: Use and comparison of permutation tests in linear models. PhD Thesis. Çukurova University Graduate School of Natural and Applied Sciences Department of Animal Science (unpublished), 2005.
9. Koşkan O: Resampling approximation in group comparisons. PhD Thesis. Ankara University Graduate School of Natural and Applied Sciences Department of Animal Science (unpublished), 2007.
10. Feinstein AR: Permutation tests and statistical significance. MD Comput: Comput Med Prac, 10, 28-41, 1993.
11. Manly BFJ: Randomization, bootstrap and Monte Carlo methods in Biology, Second Ed., London: Chapman and Hall, 1997.
12. Opdyke JP: Fast permutation tests that maximize power under conventional Monte Carlo sampling for pairwise and multiple comparisons. J Mod Appl Stat Meth, 2 (1):27-49, 2003.
13. Anderson MJ, Legendre P: An empirical comparison of permutation methods for tests of partial regression coefficients in a linear model. J Stat Comput Simul, 62, 271-303, 1999.
14. Tussel F: A permutation test for randomness with power against smooth variation. Stat Comput, 11, 147-154.2001.
15. Maggini R, Guisan A, Cherix D: A stratified approach to modeling the distribution of a threatened ant species in the Swiss National Park. Biodivers Conserv, 11, 2117-2141, 2002.
16. Bracken MB: On stratification, minimization and protection against types 1 and 2 error. J Clin Epidemiol, 54, 104-107, 2001.
17. Tanizakm H: On small sample properties permutation tests: An independence test between two samples. Int J Pure Appl Math, 13 (2): 235-243, 2004.
18. Bohdan L: Resampling tests for regression coefficient. 2nd International Conference APLIMAT, 2003.
19. Mills D: Improved mathematical methods for drug design. http://www.cmd.port.ac.uk/immdd/doc/reports/permut.doc Accessed: 14.02.2005.
20. Anonymous: FORTRAN Subroutines for mathematical applications. IMSL MATH/LIBRARY. Vol.1-2, Houston: Visual Numerics, Inc., 1994.
21. Winer BJ, Brown DR, Michels KM: Statistical Principles in Experimental Design. McGraw-Hill Book Company, New York, USA, 1991.
22. Welch BL: On the comparison of several mean values: an alternative approach. Biometrika, 38, 933-943, 1951.
23. Mendeş M, Başpmnar E: Comparison of various test statistics calculated from the samples taken from nonnormal distributive populations. J Agri Sci, 9 (1): 23-28, 2003.
24. Routledge RD: P-values from permutation and F-tests. Comput Stat Data Anal, 24, 379-386, 1997.
25. Ludbrook J, Dudley H: Why permutation tests are superior to t and F tests in biomedical research. Am Stat, 52 (2): 127-132, 1998.
26. Corcoran CD, Mehta CR: Exact level and power of permutation, bootstrap and asymptotic tests of trend. J Modern Appl Stat Meth, 1, 42-51, 2002.
27. Balasubramani GK, Wisniewski SR, Zhank H, Eng HF: Development of an efficient SAS macro to perform permutation test for two independent samples. Comput Meth Prog Bio, 79, 179-187, 2005.
28. Huang Y, Xu H, Calian V, Hsu JC: To permute or not permute. Bioinformatics, 22 (18): 2244-2248, 2006.
29. Ergün, G, Aktaş, S: ANOVA modellerinde kareler toplaml yöntemlerinin karşllaştlrllmasl. Kafkas Univ Vet Fak Derg, 15 (3): 481-484, 2009.

APA	MENDEŞ M, AKKARTAL E (2010). Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. , 711 - 716.
Chicago	MENDEŞ Mehmet,AKKARTAL ERKUT Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. (2010): 711 - 716.
MLA	MENDEŞ Mehmet,AKKARTAL ERKUT Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. , 2010, ss.711 - 716.
AMA	MENDEŞ M,AKKARTAL E Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. . 2010; 711 - 716.
Vancouver	MENDEŞ M,AKKARTAL E Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. . 2010; 711 - 716.
IEEE	MENDEŞ M,AKKARTAL E "Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power." , ss.711 - 716, 2010.
ISNAD	MENDEŞ, Mehmet - AKKARTAL, ERKUT. "Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power". (2010), 711-716.

APA	MENDEŞ M, AKKARTAL E (2010). Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. Kafkas Üniversitesi Veteriner Fakültesi Dergisi, 16(5), 711 - 716.
Chicago	MENDEŞ Mehmet,AKKARTAL ERKUT Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. Kafkas Üniversitesi Veteriner Fakültesi Dergisi 16, no.5 (2010): 711 - 716.
MLA	MENDEŞ Mehmet,AKKARTAL ERKUT Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. Kafkas Üniversitesi Veteriner Fakültesi Dergisi, vol.16, no.5, 2010, ss.711 - 716.
AMA	MENDEŞ M,AKKARTAL E Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. Kafkas Üniversitesi Veteriner Fakültesi Dergisi. 2010; 16(5): 711 - 716.
Vancouver	MENDEŞ M,AKKARTAL E Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power. Kafkas Üniversitesi Veteriner Fakültesi Dergisi. 2010; 16(5): 711 - 716.
IEEE	MENDEŞ M,AKKARTAL E "Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power." Kafkas Üniversitesi Veteriner Fakültesi Dergisi, 16, ss.711 - 716, 2010.
ISNAD	MENDEŞ, Mehmet - AKKARTAL, ERKUT. "Comparison of ANOVA F and WELCH Tests with Their respective permutation versions in terms of Type I error rates and test power". Kafkas Üniversitesi Veteriner Fakültesi Dergisi 16/5 (2010), 711-716.