A new family of robust quantile-regression-based mean estimators using Sarndal approach

In presence of extreme-values or outliers, the ratio-type mean estimators are prominently developed by using the robust coefficients of regression and covariance matrices. In this article, we consider quantile regression coefficients while estimating the finite population mean. Incorporating Sarndal concept, we offer quantile-regression-based mean estimators under a simple random sample design. We calculate the mean squared error (MSE) for the suggested estimators and discover that they perform better than the existing estimators. Moreover, numerical representations and a simulation study support our discoveries in theory.