Task 3 is an exercise on the CLRM:
The objective is to test whether or not a ‘structural break’ occurs as new data is acquired
Use the Stata multiple regression output you obtained in Task 2 above as the estimated regression output based on original data. Next, randomly select data on m more observations (different from those you had selected earlier) where m is an integer of your choice satisfying. You hypothesise that the additional data fits into the original model. You will now perform a Chow test at the 5% level significance whether or not your hypothesis is correct.
- Open the Excel file named Answers.xls and click open the sheet Task3.
- In cell A2 (colour coded blue), write down the value n (number of observations) allocated to you.
- Enter the value of RSSR and RSSU in cell D2 and E2 (colour coded blue), respectively
- In cell A4 (colour coded blue), write down the value of m you have selected.
- Write down the null hypothesis and the alternative hypothesis for this Chow Test (in the Word document); (3 marks)
- Enter the value of the F-statistic as per your calculations in cell E7 (colour coded green); (5 marks)
- Enter the critical F value for this test in cell E9 (colour coded green); (4 marks)
- Enter the number 1 in cell E12 (colour coded green) if you reject the null hypothesis based on your analysis – otherwise enter the number 0. (4 marks)
- Enter the number 1 in cell E15 (colour coded green) if your test is significant – otherwise enter the number 0. (4 marks)
- Enter the number 1 in cell E18 (colour coded green) if your conclusion is that there is a structural break – otherwise enter the number 0. (4 marks)
- Consider now that your sample data have a time dimension. Add a variable called year to your data starting in the past and ending in 2016 (your m additional observations are the most recent ones). Use this time variable and explain how you will conduct a dummy variable alternative to the Chow Test.
- Create the dummy variable
- Specify the multiple regression model
- Write the mean demand functions for the different time periods
- Report the differential intercept and the differential slope coefficients (in cells E21 and E22)
(8 marks)