Multicollinearity Diagnostics in Statistical Modeling & Remedies to deal with it using SAS

Regression modeling is one of the most widely used statistical techniques in clinical trials. Many a times, when we fit a multiple regression model, results may seem paradoxical. For instance, the model may fit the data well, even though none of the predictors has a statistically significant impact on explaining the outcome variable. How is this possible? This happens when multicollinearity exists between two or more predictor variables. If the problem of multicollinearity is not addressed properly, it can have a significant impact on the quality and stability of the fitted regression model. The aim of the proposed paper is to explain the issue of multicollinearity, effects of multicollinearity, various techniques to detect multicollinearity and the remedial measures one should take to deal with it. The paper will focus on explaining it theoretically as well as using SAS ® procedures such as PROC REG and