Quantile regression-ratio-type estimator of population mean in stratified successive sampling using calibrated weights

This work proposes a new method for estimation of the population mean in successive sampling over two occasions, even when the data may not be normally distributed. This method is robust to outliers, in the sense that it is less affected by extreme values compared to the ordinary least squares (OLS) methods. This approach utilizes quantile regression ratio-type estimators which leverage data from correlated auxiliary variables. The properties of the proposed estimators have been examined in terms of bias and mean squared error (MSE) and subsequently, optimum stratum weights have been derived using calibration techniques. Empirical studies with real and simulated data demonstrate that the proposed quantile regression estimators outperform the conventional estimators under simple random sampling without replacement (SRSWOR). Finally, the recommendations to the survey statisticians have been made for their real-life applications.