This post starts the new chapter in our econometrics course. From now on, we move to regression of several variables, instead of just one variable. Once again we derive OLS estimates or normal equations in a multiple variable model ( precisely, 3 variable model).

Then, we pick up hypothesis testing in multiple linear regression model. How to test several coefficients jointly? How to test linear combination of coefficients?

Last video, explains, yet another very important concept : omitted variable bias. What happens, if you remove one very important variable from the regression model? How does that affect OLS estimates, are they still unbiased? Continue reading →