1. What do you understand by heteroscedasticity What are the consequences of using OLS in the presence of heteroscedasticity? Explain the Breusch-Pagan-Godfrey test to detect it.

2. Given the following data points, estimate a linear probability model using OLS.

x 0 1 1 1
Y 0 0 0 1 1 1

Use the estimated model to classify individuals into 2 categories. Calculate the number of correct classifications using the following classification rule: <img src='./qimages/9914-2.jpg'>

3. Explain the underlying ideas behind the probit model. Explain how probit model can be estimated?

4. What do you understand by identification of a simultaneous equation system? Find out the identifiability of the following model.

Ct =ao a1 Yt ut
It =bo bl Yt b2 Yt-1 u2
Yt =Ct It Gt Y and I are endogenous variables)

5. Explain the concept of BLUE. Prove that OLS estimates are BLUE.

6. What is multicollinearity . How is the problem of multicollinearity tackled in a data set?

7. Consider the relationship between expenditure and income in 2 time periods as
E1 a1 a2Y1 u1
E2 b1 b2Y2 u2
State how would you apply the Chow test for structural stability between the 2 periods.

8. How is a dynamic model different from the Koyck model? (Specify both the models and explain the difference).

9. What is indirect least squares method? Explain the steps followed in estimation of this method.

10. Explain the following concepts:
Adjusted R^2 One-tail and two-tail tests

11. When do you encounter the problem of autocorrelation in data? Explain how would you detect the presence of autocorrelation.

12. For the regression model Yt a bxt ut, where b is known, show that the error variance of the forecast will be 0^2 1 where 0^2 is the population variance.