Comparison of stratified diagonal ranked set sampling and stratified simple random sampling estimators

Abstract

Stratified Diagonal Ranked Set Sampling (SDRSS) was compared with Stratified Simple Random Sampling (SSRS) for estimating the unknown population mean. Simulated data generated twice were used to check the performance of SDRSS for estimating the mean and variance. The selected diagonal elements for the first and second independentlysimulated data were stratified into two strata using students’ performance as a stratifying factor each with stratum sample size of 10 and 14 for the first data set, also 11 and 13 for the second data set. Statistical properties of SDRSS and SSRS were examined. The mean square error for SDRSS in the first and second data sets were smaller compared to SSRS. Hence SDRSS is more efficient than SSRS.