We don’t know the exact affect X1 has on the dependent variable. So, in this case we cannot exactly trust the coefficient value (m1). Because of this relationship, we cannot expect the values of X2 or X3 to be constant when there is a change in X1.
In our Loan example, we saw that X1 is the sum of X2 and X3. What are the problems that arise out of multicollinearity? Which means predicted expense will increase by 23240 if the person is a smoker, and reduces by 23,240 if the person is a non-smoker (provided all other variables are constant). In case of smoker, the coefficient is 23,240. If you look at the equation, you can see X1 is accompanied with m1 which is the coefficient of X1.įor Linear Regression, coefficient (m1) represents the mean change in the dependent variable (y) for each 1 unit change in an independent variable (X1) when you hold all of the other independent variables constant.įor example, in the previous article, we saw the equation for predicted medical expense to be i.e We shouldn’t be able to derive the values of this variable using other independent variables.Īs we have seen in the previous articles, The equation of dependent variable with respect to independent variables can be written as One of the conditions for a variable to be an Independent variable is that it has to be independent of other variables. Lets see what Multicollinearity is and why we should be worried about it.
Our goal in regression is to find out which of the independent variables can be used to predict dependent variable.Independent variable is the one that is used to predict the dependent variable.Dependent variable is the one that we want to predict.Check this post to find an explanation of Multiple Linear Regression and dependent/independent variables. One of the important aspect that we have to take care of while regression is Multicollinearity. This post will answer questions like What is multicollinearity ?, What are the problems that arise out of Multicollinearity? When do I have to fix Multicollinearity?