Research Article
BibTex RIS Cite

Behrens-Fisher Problemi ve Çözüm Yöntemleri

Year 2007, Volume: 5 Issue: 1, 25 - 34, 13.07.2007

Abstract

N(µ1, σ12), N(µ2, σ22) şeklinde normal dağılan iki kitlenin ortalamaları arasındaki farklılık araştırılırken, bu iki kitlenin varyansları bilinmiyorsa ve eşit değilse söz konusu iki kitlenin ortalamalarının testi, Behrens-Fisher problemi olarak bilinir. Behrens-Fisher probleminin çözümünde çok sayıda yöntemler geliştirilmiştir. Welch-t testi, permütasyon testi v.b. bu testlerden bazılarıdır. Bu çalışmada Behrens-Fisher probleminin çözümünde kullanılan Welch-t testi, permütasyon testi ve sayısal yöntemler tartışılmıştır. Monte Carlo sonuçlarına değinilerek Welch-t testi ve permütasyon testi karşılaştırılmıştır. Son olarak sayısal yöntemler için Cressi ve Whitford tarafından önerilen istatistikleri üreten bir bilgisayar yazılımı uygulanmıştır.

References

  • Good, P., 1994. Permutation Tests, Practical Guide to Resampling Methods for Testing Hypothesis, Springer Series in Statistics, Springer, Berlin.
  • Janssen, A., 1997. Studentized Permutation Tests for non-i.i.d. Hypotheses and the Generalized Behrens-Fisher Problem, Statistics & Probability Letters, 36, 9-21.
  • Jay, D. and Roxy, P., 1993. The Exploration and Analysis of Data, Duxbury Press, Belmont, California.
  • Lehmann, E.L., 1959. Testing Statistical Hypotheses, Willey Publication in Statistics, New York.
  • Lehmann, E.L., 1975. Nonparametrics: Statistical Methods Based on Ranks, McGraw-Hill International Book Company, San Fransisco.
  • Mehta, J. S. and Srinivasan, R., 1970. On the Behrens-Fisher Problem, Biometrica, 57, 649-655.
  • Pfanzagl, J., 1974. On the Behrens-Fisher Problem, Biometrica, 61, 39-47.
  • Reed III, J.F., 2003. Solutions to the Behrens-Fisher Problem, Computer Methods and Programs in Biomedicine, 70, 260-261.
  • [http://www.maths.gmw.ac.uk/~bb/CTS_chapter2_students.pdf, Erişim Tarihi:10.03.2005]

Behrens-Fisher Problem and Solution Methods

Year 2007, Volume: 5 Issue: 1, 25 - 34, 13.07.2007

Abstract

While difference of means of two population of which distiributions are N(µ1, σ12) and N(µ2, σ22) has been searched, if variances of the two population aren’t known and equal, the test of means of two populations has been known as Behrens-Fisher Problem. A lot of methods have been improved for the solution of Behrens-Fisher problem. Some of those methods are Welch-t test and permutation test. In this study, Welch-t test, permutation test and numerical methods that are used for solution of Behrens-Fisher problem were discussed Welch-t test and permutation test compared. Lastly, the computer program that produced statistics proposed by Cressi and Whitford has been carried out.

References

  • Good, P., 1994. Permutation Tests, Practical Guide to Resampling Methods for Testing Hypothesis, Springer Series in Statistics, Springer, Berlin.
  • Janssen, A., 1997. Studentized Permutation Tests for non-i.i.d. Hypotheses and the Generalized Behrens-Fisher Problem, Statistics & Probability Letters, 36, 9-21.
  • Jay, D. and Roxy, P., 1993. The Exploration and Analysis of Data, Duxbury Press, Belmont, California.
  • Lehmann, E.L., 1959. Testing Statistical Hypotheses, Willey Publication in Statistics, New York.
  • Lehmann, E.L., 1975. Nonparametrics: Statistical Methods Based on Ranks, McGraw-Hill International Book Company, San Fransisco.
  • Mehta, J. S. and Srinivasan, R., 1970. On the Behrens-Fisher Problem, Biometrica, 57, 649-655.
  • Pfanzagl, J., 1974. On the Behrens-Fisher Problem, Biometrica, 61, 39-47.
  • Reed III, J.F., 2003. Solutions to the Behrens-Fisher Problem, Computer Methods and Programs in Biomedicine, 70, 260-261.
  • [http://www.maths.gmw.ac.uk/~bb/CTS_chapter2_students.pdf, Erişim Tarihi:10.03.2005]
There are 9 citations in total.

Details

Primary Language Turkish
Subjects Statistics
Journal Section Research Articles
Authors

Süleyman Günay

Semra Türkan This is me

Publication Date July 13, 2007
Published in Issue Year 2007 Volume: 5 Issue: 1

Cite

APA Günay, S., & Türkan, S. (2007). Behrens-Fisher Problemi ve Çözüm Yöntemleri. İstatistik Araştırma Dergisi, 5(1), 25-34.