Question

5. Which one of the following problems will not cause the endogeneity problem in a linear regression model? (a) Omitting relevant variables (b) Including irrelevant variables. (c) Errors-in-variables. (d) Simultaneous equations (e) Models with lagged dependent variables and autocorrelated errors. 6. Consider the following linear regression model y=Bo+Bi +B22e where r2 is an endogenous regressor. With two qualified instrumental variables z1 and 22 for r2, you want to estimate ( B, B2) using 2SLS. Then, the first-stage estimation is: (a) r2 on 1, r1, 1, and z2 (b) 2 on 1, 21, and z2. (c) r on 1, 2, 21, and z2 (d) y on 1, r1, 21, and z2 (e) y on 1, z1, and z2. Consider a standard AR(p) model 0001-1 .+Opyt-p +v yt Which one of the following statements is not correct? (a) The del can be estimated using OLS. (b) Becausethe autocorrelation, homosked suc OLS standard errors are not valid. (c) One can chose the value of p usingegular hypothesis tests. (d) One can choose he value of psing adjusted R2 (e) One can choose the walue of p using AIC 8. Suppose that a time seri can be modeled by the following stationary AR(1) process +01t-1+ iid N(0,Which one of the llowing statements is true? where (a) 00 0 mut hold. (b) l01 (c) Ey001yt-1 (d) V(y)<Var (yt-1) if -1 <0 1 (e Cov(yt, yt-)| > |Cov(ya, y-1)| if t<s. mus be less than 1

Answer 1

Solution 5: The main reasons for Endogeneity problem are omitted variables, simultaneous equation or error in variable. But if we include irrelevant variable, it is not acceptable at all. All above described cause endogeneity problem.

But last option Model with lagged dependent variable and auto correlated errors does not cause endogeneity problem.It causes another problem which can be solved by monte-carlo method.

So option (E) is correct.