On Bootstrapping; Bartlett Adjusted Empirical Likelihood Ratio Statistic in Regression Analysis

  • Woochul Kim (Department of Computer Science and Statistics, Seoul National University, Seoul, 151-742, Korea.) ;
  • Duk-Hyun Ko (Department of Computer Science and Statistics, Seoul National University, Seoul, 151-742, Korea.) ;
  • Keewon Lee (Department of Statistics, Hallim University, Chunchon, 200-702, Korea)
  • Published : 1996.06.01

Abstract

The bootstrap calibration method for empirical likelihood is considered to make a confidence region for the regression coefficients. Asymptotic properties are studied regarding the coverage probability. Small sample simulation results reveal that the bootstrap calibration works quite well.

Keywords

References

  1. Annals of the Institute of Mathematical Statistics On the accuracy of empirical likelihood confidence region for linear model Chen, S. X.
  2. Annal of Statistics Empirical likelihood is Bartlett-correctable DiCiccio, T.J.;Hall, P.;Romano, J.P.
  3. The Bootstrap and Edgeworth Expansion Hall, P.
  4. Biometrika Empirical likelihood ration confidence intervals for a single functional Owen, A.
  5. Annal of Statistics Empirical likelihood ration confidence regions Owen, A.
  6. Annal of Statistics Empirical likelihood for linear models Owen, A.