Improved Family of Ratio Estimators in Simple and Stratified Random Sampling

This article proposes two improved family of estimators for estimating the finite population mean in simple random sampling (SRS) and stratified random sampling (StRS). The proposed estimators always perform better than a family of ratio estimators suggested by Khoshnevisan et al. (2007) in SRS and Koyuncu and Kadilar (2009a) in StRS. They also perform better than the ratio estimator given by Gupta and Shabbir (2008) in SRS and Koyuncu and Kadilar (2010) and Shabbir and Gupta (2011) in StRS. The expressions for bias and mean squared error (MSE) of considered estimators are obtained. The results are illustrated by real data sets.