USING STATISTICAL PROCEDURE FOR TREATMENT MULTICOLLINEARITY PROBLEM.

Abstract

One of important assumptions in the linear regression is not found linear correlation between independent variables which lead to appearance multicollinearity problem. If this condition is not satisfied then the ordinary least square method has not efficiency property. Another statement have not the minimum variance. We use chi-square test to discover multicollinearity problem by using Ordinary least square, Ridge and principle component procedure in estimation by using statistical program (NCSS_2005) to get the result. The purpose of the research:- The aim of the research is to test and treatment multicollinearity problem by using Chi-square test, ordinary least square, principle component and ridge in estimation. The hypothesis of research:- 1. We have not multicollinearity problem at level significance (0.05, 0.01). 2. The effect of salary and investment upon out put are statistical significance. The frame of research:- The research was applied in battery Industry Company from 1992 to 2002.