Journal of Statistics Applications & Probability


The inefficiency of the ordinary least square estimator for the parameter estimation of a linear regression model with multicollinearity problem has led to the development of various ridge regression estimators. These estimators are recently classified as one-parameter and two-parameter ridge-type estimators. This paper proposes a new two-parameter estimator following a newly developed one-parameter ridge estimator to handle multicollinearity in the linear regression model. Theoretical and simulation results show that, under some conditions, the proposed estimator performs better than some popular existing estimators in that it has a smaller mean square error. Furthermore, we used real-life data to illustrate the papers findings establishes the same results from theory and simulation.

Digital Object Identifier (DOI)