ECON 104 Lecture Notes - Lecture 20: Scatter Plot
Document Summary
Look at x on y variable, why do we bother having multiple regressors if we are only interested in. Error u in population regression equation comes up because x and y might be related, but x isn"t only thing determining y. U = anything else that determines y, variables that are omitted from analysis. Sometimes omission of these variables can lead to bias and inconsistency in the ols estimator. If there is variable z that is omitted in analysis. If z is determinant of y (aka part of u) and it is correlated with regressor x -> there is an omitted variable bias. If you do your regression on r, you come up with beta 1 estimate, it won"t be correct. Could have infinitely many data points and se could be very small, very precise numbers, but the numbers won"t be what you"re looking for.
