The Effect of High Leverage Points on VIF Measures on Noncollinear Data
Multicollinearity is a case of multiple regression in which the predictor variables are highly correlated among themselves. The problem will get more complicated when multicollinearity exists together with high leverage points. The usage of classical VIF for multicollinearity diagnostics is not reliable as it is not resistant to the presence of high leverage points. In this study, we proposed RVIF which is based on the MM estimator in the detection of multicollinearity due to the high leverage point. The computation of RVIF is based on robust coefficient determination which is called RR2 (MM). We denote this estimator as RVIF (MM). The numerical results and Monte Carlo simulation study indicate that the CVIF performs poorly in the presence of high leverage point and the proposed RVIF is very resistant to the high leverage point and unable to detect the multicollinearity in the data.