Title: Robust measurement errors method
Authors: Claudio Agostinelli - University of Trento (Italy)
Fatemah Alqallaf - Kuwait University (Kuwait) [presenting]
Abstract: The Simulation-Extrapolation methods (SIMEX) is widely used in presence of measurement errors. Unknown parameters and regression coefficients are estimate using algorithms based on generalized least squares, and the method, as a whole, is very sensitive to outliers. To overcome this problem, we propose a robust technique based on robust regression methods, such as MM-estimators. We focus our attention on the estimation of the correlation coefficient and on the coefficient of determination of the regression. Illustrations based on Monte Carlo simulations and real examples are provided.