Ranked Set Sampling (RSS) is a popular sampling method in recent years. The procedure of RSS was suggested by McIntyre (1952) for estimation population mean of pasture yields in Australia. RSS is using when the variable of interest is difficult and expensive to measure and visiual ranking can be done easily. The mean estimator of RSS is more efficient than the mean estimator of Simple Random Sampling (SRS). In the last years many authors have suggested different modifications of the RSS which are Paired RSS, Extreme RSS, Median RSS, Double RSS, LRSS and Truncation Based RSS and used it in wide applications. The aim of this study is to estimate the regression estimators and to compare relative eﬃciencies of mean square errors of the regression models, population mean and regression coeﬃcients for different modified ranked set sampling methods. The Monte Carlo simulation study is performed via R Project with 10,000 repetitions. The performance of the estimators is constructed in terms of bias, mean squared error and relative efficiency for different levels of correlation coefficient, set and cycle sizes under Bivariate Normal Distribution. The results indicate that the regression estimators under modified RSS methods performs better than the regression estimators under SRS.
Anahtar Kelimeler: Ranked Set Sampling, Modified Ranked Set Sampling, Relative Efficiency