Abstract

In many researchs and economic studies the multicollinearity problem may exist when many of explanatory variables depend upon one response variable.
If this problem a valiable, this means we can not get a good estimator for the parameters in linear model that leads to the OLS method failuer, and the properties of the estimator are absent . This problem appears when there is a relationship between explanatory variable (as equations), then to treat this problem estimating the parameters which suffer from this problem we use Ridge regression estimator insted of ordinary least squares estimator.
The aim of this research is to use the Ridge regression when the data suffer from multicollinearty after testing the data by using Frrar-Gluber test to areal data.