Which pair of variables are likely to cause multicollinearity (MC)?

Home, - What are the potential solutions of MC?

Question: - Suppose the regression function of GDP per capita = f (Saving, Credit, Education, Foreign Direct Investment). Which pair of variables are likely to cause multicollinearity (MC)? What are the potential solutions of MC?


When the independent variables are highly correlated with each other, then there exists a problem called Multicollinearity. In the regression analysis, it is assumed that the independent variables included in the model are not related with each other and when this assumption voids, then it creates a major problem in the estimated regression model.

Therefore, the regression coefficient interpreted in such a way that it captures the average change in the dependent variable for every one unit change in the independent variable, on holding other independent variables constant. Thus, when the independent variables are inter related, then, it would be difficult to provide a justification for the interpretation of the regression coefficients. This problem is known as Multicollinearity

One of the most common solution for Multicollinearity is

Ø  Remove the one or more highly correlated independent variables from the model

Ø  Merge those independent variables that are highly correlated with each other (like adding them together, etc)

Ø  Look for alternative statistical analysis such as Principal component analysis or partial least square analysis which will be effective for highly correlated independent variables.

The given regression function is

GDP per capita = f (Saving, Credit, Education,Foreign Direct Investment)

Here, it is expected that savings and foreign direct investment will certainly like to cause Multicollinearity as these two variables seems to be one and the same

Leave a comment


Related :-