Abstract
In this article, we use the integral form of the binomial distribution to derive saddlepoint approximations for sample quantiles. As an application, we present the calculation of the tail probability of the empirical log-likelihood ratio statistic for quantiles. Simulation results are also given to show that our approximations are extremely accurate.
Mathematics Subject Classification:
Acknowledgment
The authors would like to thank a referee for his/her comments and criticisms which helped to improve the article a great deal.