Estimation of the population median of symmetic and asymmetric distributions using double robust extreme ranked set sampling

Double robust extreme ranked set sampling (DRERSS) and its properties for estimating the population median are suggested. The performance of DRERSS with respect to simple random sampling (SRS), ranked set sampling (RSS) and extreme ranked set sampling (ERSS) schemes is considered. Real data set that consist of heights of 346 students are used to evaluate the DRERSS method. It is found that when the underlying distribution is symmetric, the DRERSS estimators are unbiassed of the population median. When the sample size is odd, DRERSS is more efficient than SRS, ERSS and RSS. When the sample size is even, DRERSS is more efficient than RSS and ERSS if the underlying distribution is symmetric and for asymmetric distributions, DRERESS is more efficient than RSS and ERSS for k>0.