4. Heteroskedasticity and Serial Correlation Suppose you estimate the following time series regression: qoil,td=+oilpoil,t+poppopt+ where qoil,td is the quantity demanded of oil, poil,t is the price of oil, and popp is population (in millions). The results from this regression are shown in the table below. 10. Because this is a time series model, you are also worried that the model may suffer from serial correlation. Based on the following graph of the residuals (error term) over time, do you think the model suffers from serial correlation? 11. Regardless of your answer to (10.), you decide to use the Durbin-Watson Test to test for positive serial correlation. The value of the d-Stat is 0.194. If the critical values are DI=1.29 and DU=1.38, are we able to reject the null hypothesis of no positive serial correlation? 12. Based on your answers to (7.) through (11.), do you think heteroskedasticity, serial correlation, or both are an issue in this model? 13. Based on your answer to (12.), what should you do to correct for heteroshedasticity, serial correlation, or both? Hint: There is one thing you should do as a best practice and other things you could do. 14. Would you report conventional, Newey-West, or robust standard errors if you estimated a distributed lag (DL) model qoil,td=+oil1poil,t+oil2poil,t1+pop1pop+pop2popt1+ ? 15. Would you report conventional, Newvey-West, or robust standard errors if you estimated an autoregressive distributed lag (ADL) model qoil,td=+1qoil,t1d+oil1poil,t+oil2poil,t1+ pop1popt+pop2p2opt1+ ? 4. Heteroskedasticity and Serial Correlation Suppose you estimate the following time series regression: qoil,td=+oilpoil,t+poppopt+ where qoil,td is the quantity demanded of oil, poil,t is the price of oil, and popp is population (in millions). The results from this regression are shown in the table below. 10. Because this is a time series model, you are also worried that the model may suffer from serial correlation. Based on the following graph of the residuals (error term) over time, do you think the model suffers from serial correlation? 11. Regardless of your answer to (10.), you decide to use the Durbin-Watson Test to test for positive serial correlation. The value of the d-Stat is 0.194. If the critical values are DI=1.29 and DU=1.38, are we able to reject the null hypothesis of no positive serial correlation? 12. Based on your answers to (7.) through (11.), do you think heteroskedasticity, serial correlation, or both are an issue in this model? 13. Based on your answer to (12.), what should you do to correct for heteroshedasticity, serial correlation, or both? Hint: There is one thing you should do as a best practice and other things you could do. 14. Would you report conventional, Newey-West, or robust standard errors if you estimated a distributed lag (DL) model qoil,td=+oil1poil,t+oil2poil,t1+pop1pop+pop2popt1+ ? 15. Would you report conventional, Newvey-West, or robust standard errors if you estimated an autoregressive distributed lag (ADL) model qoil,td=+1qoil,t1d+oil1poil,t+oil2poil,t1+ pop1popt+pop2p2opt1+