Investigation of the efficiency of alternative estimators in multiple linear regression models in case of Multicollinearity and Outlier Observation

In this study, alternative estimators were examined for problems with multicollinearity and outlier observations in multiple linear regression models. In the presence of multicollinearity problems among the independent variables in the model, the variances may be larger than they actually are, although the parameter estimates obtained by the Least Squares (LCM) method maintain their unbiased properties. Ridge and Liu estimators in the literature are frequently used to solve this situation. The outlier observation problem arises in parameter estimates and hypothesis testing with LCM estimators. M estimators for the outlier observation problem are frequently used in the literature. In this study, alternative estimation methods were examined for situations where the mentioned problems exist simultaneously.