The Korean Journal of Applied Statistics (응용통계연구)

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

Asymptotic Relative Efficiency for New Score Functions in Rank Regression Models

순위회귀모형의 새로운 스코어 함수의 효율성 연구

  • 최영훈 (한신대학교 정보통계학과)
  • Published : 2004.07.01
Download PDF
⟨ Previous Next ⟩

Abstract

We explore the selection of r and s that provides improvement over the Wilcoxon scores under the asymmetric distributions we encounter in practice. We select 0 〈 r 〈 1, s 〉 1 for right-skewed distribution and r 〉 1,0 〈 s 〈 1 for left-skewed distributions from the perspective plots. We also study the association between the desirable r and s and the test statistic for skewness.

본 연구를 통하여 순위를 이용한 선형회귀모형의 가정된 분포형태가 우리가 실질적으로 많이 접하게 되는 비대칭분포이면서 만일 가정된 분포가 오른쪽으로 늘어진 경우일 때에는, 본 논문에서 제안된 스코어의 가능한 한 0 보다 크고 1 보다 작은 r 및 1 보다 큰 s 를 선택 (0 〈 r 〈 1, s 〉 1) 하는 것이 윌콕슨 스코어보다 높은 효율성을 나타낸다. 이와 반대로 만일 가정된 비대칭분포가 왼쪽으로 늘어진 경우일 때에는, 제안된 스코어의 가능한 한 1 보다 큰 r 및 0 보다 크고 1 보다 작은 s 를 선택 (r 〉 1, 0 〈 s 〈 1) 하는 것이 윌콕슨 스코어보다 높은 효율성을 나타낸다. 아울러 바람직한 r 과 s 를 결정하기 위한 가정된 분포의 대칭성 검정기법도 제시한다.

Keywords

References

  1. The American Statistician v.50 A class of Mann-Whitney-Wilcoxon type statistics Ahmad, I. A. https://doi.org/10.2307/2684929
  2. Communications in Statistics v.27 A study of the power of the rank transform test in a 2(3) factorial experiment Choi, Y. H.
  3. Statistics & Probability Letters v.57 A new class of score generating functions for regression models Choi, Y. H.;Ozturk, O. https://doi.org/10.1016/S0167-7152(02)00061-5
  4. Robust Nonparametric Statistical Methods Hettmansperger, T. P.;McKean J. W.
  5. Nonparametric Statistical Methods Hollander, M;Wolfe, D. A.
  6. The Annals of Mathematical Statistics v.43 Estimating regression coefficients by minimizing the dispersion of residuals Jaeckel, L. A. https://doi.org/10.1214/aoms/1177692377
  7. The Annals of Mathematical Statistics v.42 Asymptotic linearity of a rank statistics in regression parameter Jureckova, J. https://doi.org/10.1214/aoms/1177693245
  8. The Annals of Mathematical Statistics v.42 Nonparametric estimate of regression coefficients Jureckova, J. https://doi.org/10.1214/aoms/1177693245
  9. Technometrics v.31 Rank scores suitable for analyses of linear models under asymmetric error distributions McKean, J. W.;Sievers, G. L. https://doi.org/10.2307/1268818
  10. An Introduction to Statistical Methods and Data Analysis Ott, L.
  11. Journal of Nonparametric Statistics v.10 Two-sample inference based on one-sample ranked set sample sign statistics Ozturk, O https://doi.org/10.1080/10485259908832760
  12. Australian and New Zealand Journal of Statistics v.43 no.1 A generalization of Ahmad's class of Mann-Whitney-Wilcoxon statistics Ozturk, O https://doi.org/10.1111/1467-842X.00155
  13. Statistics & Probability Letters v.29 Almost fully efficient and robust simultaneous estimation of location and scale parameters: a minimum distance approach Ozturk, O;Hettmansperger, T. P. https://doi.org/10.1016/0167-7152(95)00178-6
  14. Biometrika v.84 Generalized weighted Cramer-Von Mises distance estimators Ozturk, O;Hettmansperger, T. P. https://doi.org/10.1093/biomet/84.2.283
  15. Journal of Nonparametric Statistics v.5 Influence functions for rank based procedures in the linear models Witt, L. D.;Naranjo, J. D.;McKean, J. W. https://doi.org/10.1080/10485259508832653
  16. Nonparametric Statistics v.12 Generalizing the Mann-Whitney-Wilcoxon statistic Xie;Priebe https://doi.org/10.1080/10485250008832827