Normal ve Karma-Normal Dağılımlı Veri Grupları İçin İlişki Katsayılarının Eşitliğinde Permütasyon Testi
Abstract
İki bağımsız yığına ilişkin korelasyonların eşitliği hipotezinin testi için kullanılan parametrik testler, örnek çaplarının büyük ve tesadüfi değişkenlerin dağılımlarının normal olması varsayımını gerektirmektedir. Bu hipotezin testinde hem küçük örnek çapı durumunda hem de normallik varsayımının sağlanmaması durumunda kullanılabilen permütasyon testinin, birinci tip hata miktarının incelenmesi makalenin amacını oluşturmaktadır. Bu amaçla yapılan simülasyon çalışması örnek çapının küçük olması durumunda, hem normal hem de karma-normal dağılımlı veri üretilmesine dayalı olarak gerçekleştirilmiştir.
References
- [1] Anderson, T.W., An Introduction to Multivariate Statistical Analysis. 2nd Ed., Wiley, New York, (1984)
- [2] Fay, M.P. ve Follmann, D.A., Designing Monte Carlo implementations of permutation or bootstrap hypothesis tests. The American Statistician, 56, 1, (2002).
- [3] Sakaori, F., Permutation test for equality of correlation coefficients in two populations. Commun. Statist.- Simula., 31(4), 641-651, (2002).
Permutation Test For The Equality of Correlation Coefficients of Normal and Mixed-Normal Distributed Data Sets
Abstract
Parametric tests, using for the equality of correlation coefficients among two independent populations, requires the assumption of the normallity of variables and large sample sizes. For testing the hypothesis of the equality of correlation coefficients the permutation test can be used. These test is useful in case of small sample sizes and also non-normallity In this study we calculated the amount of the Type I errors. In order to do this, a simulation study is made in case of small sample sizes and the data sets are generated from normally and mixed-normal distributions.
Keywords
References
- [1] Anderson, T.W., An Introduction to Multivariate Statistical Analysis. 2nd Ed., Wiley, New York, (1984)
- [2] Fay, M.P. ve Follmann, D.A., Designing Monte Carlo implementations of permutation or bootstrap hypothesis tests. The American Statistician, 56, 1, (2002).
- [3] Sakaori, F., Permutation test for equality of correlation coefficients in two populations. Commun. Statist.- Simula., 31(4), 641-651, (2002).
Details
| Other ID | JA58AF83AE |
|---|---|
| Journal Section | Research Articles |
| Authors | |
| Publication Date | June 1, 2010 |
| Submission Date | June 1, 2010 |
| Published in Issue | Year 2010 Volume: 2 Issue: 36 |
Cite
Journal Owner: On behalf of Selçuk University Faculty of Science, Rector Prof. Dr. Metin AKSOY
Selcuk University Journal of Science Faculty accepts articles in Turkish and English with original results in basic sciences and other applied sciences. The journal may also include compilations containing current innovations.
It was first published in 1981 as "S.Ü. Fen-Edebiyat Fakültesi Dergisi" and was published under this name until 1984 (Number 1-4).
In 1984, its name was changed to "S.Ü. Fen-Edeb. Fak. Fen Dergisi" and it was published under this name as of the 5th issue.
When the Faculty of Letters and Sciences was separated into the Faculty of Science and the Faculty of Letters with the decision of the Council of Ministers numbered 2008/4344 published in the Official Gazette dated 3 December 2008 and numbered 27073, it has been published as "Selcuk University Journal of Science Faculty" since 2009.
It has been scanned in DergiPark since 2016.
Selcuk University Journal of Science Faculty is licensed under a Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License.