ECON321 Lecture Notes - Lecture 7: Birth Weight, Econometrics, Multicollinearity
Document Summary
Introduction to econometrics 321; spring 2013, university of waterloo. Week 5: solution i to omitted variable bias: multiple linear regression model. We can potentially get rid of ovb by adding the omitted variable in the regression. Remark: this solution assumes that we have (or can obtain) data on the variable causing the omitted variable bias. Often this unrealistic and hence we have to use other solutions (see e. g. chapters 10, 12 and 13). Yi = (cid:12)0 + (cid:12)1x1;i + (cid:12)2x2;i + (cid:1) (cid:1) (cid:1) + (cid:12)kxk;i + ui; i = 1; :::; n. Yi = (cid:12)0x0;i + (cid:12)1x1;i + (cid:12)2x2;i + (cid:1) (cid:1) (cid:1) + (cid:12)kxk;i + ui; i = 1; :::; n; where x0;i = 1 for all i = 1; :::; n: the regressor x0;i is thus a constant. Now e (yi jx1;i = x1; :::; xk;i = xn ) = (cid:12)0 + (cid:12)1x1 + (cid:1) (cid:1) (cid:1) + (cid:12)kxk: and therefore.