The Korean Journal of Applied Statistics (응용통계연구)

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

A Monte Carlo Comparison of the Small Sample Behavior of Disparity Measures

소표본에서 차이측도 통계량의 비교연구

  • 홍종선 (성균관대학교 경제학부 통계학전공) ;
  • 정동빈 (강릉대학교 자연과학대학 정보통계학과) ;
  • 박용석 (성균관대학교 통계학과, 대학원)
  • Published : 2003.09.01
Download PDF
⟨ Previous Next ⟩

Abstract

There has been a long debate on the applicability of the chi-square approximation to statistics based on small sample size. Extending comparison results among Pearson chi-square Χ$^2$, generalized likelihood .ratio G$^2$, and the power divergence Ι(2/3) statistics suggested by Rudas(1986), recently developed disparity statistics (BWHD(1/9), BWCS(1/3), NED(4/3)) we compared and analyzed in this paper. By Monte Carlo studies about the independence model of two dimension contingency tables, the conditional model and one variable independence model of three dimensional tables, simulated 90 and 95 percentage points and approximate 95% confidence intervals for the true percentage points are obtained. It is found that the Χ$^2$, Ι(2/3), BWHD(1/9) test statistics have very similar behavior and there seem to be applcable for small sample sizes than others.

소표본 분할표 자료에서 적합도 검정통계량들의 카이제곱 근사 적용 가능에 대하여 많은 연구가 진행되었다. 소표본에서 세 가지 검정 통계량(피어슨 카이제곱 Χ$^2$, 일반화 가능도비 G$^2$, 그리고 역발산 Ι(2/3) 검정통계량)에 관하여 비교한 Rudas(1986)의 연구를 확장하여, 최근에 제안된 차이측도(BWHD(1/9), BWCS(1/3), NED(4/3) 검정통계량)를 포함시켜 비교 분석하였다. 독립모형의 이차원 분할표, 조건부 독립모형과 한 변수 독립 모형을 따르는 삼차원 분할표에 대한 모의실험을 통하여 생성된 90과 95 백분위수와 이에 대응하는 95% 신뢰구간을 살펴보고 실제 백분위수와 비교하였다. 그 결과 Χ$^2$, Ι(2/3), 그리고 BWHD(1/9) 검정통계량이 유사한 결과를 나타내었고 이 통계량들이 기존에 제안된 검정통계량들보다 적은 표본크기에서도 카이제곱 근사방법에 적용 가능함을 발견하였다.

Keywords

References

  1. Statistics and probability Letters v.19 On Disparity based goodness-of-fit tests for multinomial models Basu,A.;Sarkar,S. https://doi.org/10.1016/0167-7152(94)90181-3
  2. Biometrics v.10 Some methods for strengthening the common $\chi^ 2$ test Cohran,W.G. https://doi.org/10.2307/3001616
  3. Mathematical Methods of Statistics Cramer;Harald
  4. Journal of the Royal Statistical Society B v.46 Multinomial goodness-of-fit tests Cressie,N.A.C.;Read,T.R.C.
  5. Order Statistics David,H.A.
  6. Journal of the Royal Statistical Society B v.41 The use of chi-squared statistics for categorical data problems Fienberg,S.E.
  7. Statistical methods for Research Workers(8th ed.) Fisher,R.A.
  8. Journal of Statistical Computation and Simulation v.65 Negative exponential disparity family based goodness-of-fit tests for multinomial models Jeong,D.;Sarkar,S. https://doi.org/10.1080/00949650008811989
  9. Journal of the American Statistical Association v.78 On the accuracy of simulated percentage points Juritz,J.M.;Juritz,W.F.;Stephens,M.A. https://doi.org/10.2307/2288654
  10. The Advanced Theory of Statistics(5th ed.) v.1 Kendall,M.G.
  11. Journal of American Statistical Association v.73 Small-sample comparison of exact levels for chi-squared goodness-of-fit statistics Lantz,K. https://doi.org/10.2307/2286650
  12. The Annals of Statistics v.22 Efficiency versus Robustness: The Case for Minimum Hellinger Distance and Related Methods Lindsay,B.G. https://doi.org/10.1214/aos/1176325512
  13. Journal of American Statistical Association v.65 A comparison of minimum logit chi-square estimation and maximum likelihood estimation in 2×2×2 and 3×2×2 contingency tables: tests for interaction Odoroff,C.L.
  14. Goodness-of-fit(Colloquia Mathematica Societas J$\'{a}$nos Bolyai) v.45 Testing goodness-of-fit of log-lnear models based on small samples: a Monte Carlo study Rudas,T.
  15. Journal of the Statistical Computation and Simulation v.24 A monte Carlo comparison of the small sample behaviour of the Pearson, the Likelihood Ratio and the Cressie-Read Statistic Rudas,T. https://doi.org/10.1080/00949658608810894
  16. Communications in Statistics, Theory and Method v.25 Small sample comparisons for the blended weight chi-square goodness-of-fit test statistics Shin,D.W.;Basu,A.;Sarkar,S. https://doi.org/10.1080/03610929608831689
  17. Journal of American Statistical Association v.65 The minimum expectation in $\chi^ 2$ goodness of fit tests and the accuracy of approximation for the null distribution Yarnold,J.K. https://doi.org/10.2307/2284594