Three regression ratio estimators are used to estimate the population mean Y under simple random sampling. Using two auxiliary variables X and Z, it should be conditioned that there is high correlation between Y and X, Y and Z, whereas there mayor may not be any correlation between X dan Z. These estimators are recommended by Singh, Upadhyaya and Premchandra [5] which is a review of the article “An Improved Version of Regression Ratio Estimator with Two Auxiliary Variables in Sample Surveys.” All estimators are biased. The efficient estimator is one with the least Mean Square Error (MSE), determined by comparing each type of estimator.
Copyrights © 2014