Comparison of efficiency of two-layer median ranked set sampling using concomitant variables

In ranked set sampling (RSS), ranking is done on the basis of a concomitant variable which is functionally related with the response variable. Chen and Shen (2003) developed two-layer RSS procedure with two concomitant variables. In addition, Muttlak (1997) proposed on modifications to RSS procedure which is called median ranked set sampling (MRSS). In this study, we extend median ranked set sampling design in terms of concomitant variables and proposed the two-layer median ranked set sampling (TMRSS). The performance of MRSS and TMRSS are evaluated for different distributions in terms of mean estimator and regression estimators by simulation studies. Also, the efficiency of TRSS method is compared with respect to corresponding estimators in simple random sampling (SRS) and RSS methods.