DOI: 10.5176/2251-1938_ORS18.27
Authors: Wilaiwan Panmuang, Supunnee Ungpansattawong
Abstract:
The propose of this paper was to compare the modified estimators of the population mean in paired double ranked set sampling (PDRSS) (Panmuang and Ungpansattawong: 2017) with the SRS estimator of Singh and Tailor (2003) in terms of the relative efficiency. The study was obtained via simulation by defined the parameters similar to the research of Al-Omari (2010) with bivariate normal distribution. The correlation coefficients of the variables x and y ( xy ) equal to ±0.99, ±0.90, ±0.80, ±0.70, and ±0.50, the sample sizes (n) are 2, 3, 4, 5, and 6. The simulation is repeated for 1,000 simulation trials in each situation. The result of the simulation study for a comparison of the performance measure shows that the PDRSS estimators are better than the SRS estimator of Singh and Tailor (2003). In case of the sample sizes (n) equal to 2, 3, and 5, the proposed estimators are the most efficient estimators when the correlation coefficients equals to -0.99. For the sample sizes (n) equal to 4 and 6, the proposed estimators are the best estimator when the correlation coefficient equals to 0.99.
Keywords:paired double ranked set sampling, ratio estimators, simulation.
Updating...