EFFICIENT RANK SET SAMPLING ESTIMATORS OF MEAN USING MEDIAN UNDER DOUBLE SAMPLING WITH TWO AUXILIARY VARIABLES

It is collectively understood that ranked set sampling (RSS) can be approved in certain applied circumstances when the measurement of a variable under study is pricey or takes an overwhelming amount of time to measure. It has been revealed that RSS is more efficient than the usual simple random sampling (SRS) outline in such situations. In this respect, we have put forward two estimators viz. regression and an exponential-ratio type estimators under double sampling which makes use of sample mean as well as sample median of two auxiliary variables in such a way that SRS has been employed in the first phase and RSS has been employed in the second phase to draw the preliminary and main samples, respectively. The estimator was found to be more efficient than the traditional RSS estimator of mean, the RSS regression estimator and other considered estimators provided certain conditions are met. The regression form of the estimator was observed to be the most efficient. An empirical study via simulated data has been done to substantiate the results as well as real life data on male infertility in the state of Meghalaya has been utilized.