论文信息 - A nonparametric confidence interval for the trimmed mean

A nonparametric confidence interval for the trimmed mean

We propose a nonparametric confidence interval for the trimmed mean. The confidence interval is based on the empirical likelihood interval for the sample mean. Its coverage probability is bounded below by the desired nominal level. Simulation results on its coverage probability and the power of the associated test show that it works well in small to moderate sample size situations.

Min Tsao | Julie Zhou | Julie Zhou | Min Tsao

[1] A. Owen. Empirical likelihood ratio confidence intervals for a single functional , 1988 .

[2] Thomas J. DiCiccio,et al. Empirical Likelihood is Bartlett-Correctable , 1991 .

[3] A. Owen. Empirical Likelihood Ratio Confidence Regions , 1990 .

[4] Yuichi Kitamura,et al. Empirical likelihood methods with weakly dependent processes , 1997 .

[5] G. Qin,et al. EMPIRICAL LIKELIHOOD RATIO CONFIDENCE INTERVAL FOR THE TRIMMED MEAN , 2002 .

[6] Peter Hall,et al. On the bootstrap and the trimmed mean , 1992 .

[7] Anna Clara Monti. Empirical likelihood confidence regions in time series models , 1997 .

[8] Peter Hall,et al. Methodology and algorithms of empirical likelihood , 1990 .