LOYOLA COLLEGE (AUTONOMOUS), CHENNAI –600 034
M.Sc., DEGREE EXAMINATION – STATISTICS
THIRD SEMESTER – NOVEMBER 2003
ST-3950/S919 – ECONOMETRICS
12.11.2003 Max:100 marks
1.00 – 4.00
SECTION-A
Answer ALL questions. (10X2=20 marks)
- What is meant by a generalized least square estimator?
- Explain auto regressive process.
- What are lagged variables?
- What are the properties of OLS estimators of a linear model Y =u?
- What is multi collinearity?
- Explain specification error.
- What is the auto correlation?
- What are the reasons for auto correlation disturbances?
- What are the sources of non spherical disturbances?
- Explain the homoscedasticity property.
SECTION-B
Answer any FIVE questions. (5X8=40 marks)
- Consider the linear Y =u with E (u) = 0 and E(uu!) = Prove that an unbiased estimate of is given by where r is the residual vector.
- Derive the MLE of the parameters of the linear regression model Y =
- Derive the variance – covariance matrix of the autocorrelated disturbance terms?
- Explain in detail the concept of multi cotnearity.
- Explain the effect of excluding the relevant variables is the linear model Y =
- Explain clearly the concept of hetroscedasticity property.
- Explain the concept of structural change.
- Write short notes on (i) dummy variables (ii) seasonal adjustment.
SECTION-C
Answer any TWO questions. (2X20=40 marks)
- For the general linear model Y =u, derive the least square estimator and
find Var .
- a) State and prove gauses Markov theorem.
- b) Derive the test procedures to test the linear hypothesis H:Rb = S for the general linear
model.
- a) What are the properties of OLS estimators under Non spherical disturbances?
- b) Explain the Drubin – Watson test to test for auto correlation. (8+12)
- a) Explain cochrane – orcutt iterative estimation procedure used in the presence of
autocorrelated disturbance.
- b) Describe the ALMON Lag model to estimate the parameters of a distributed Lag
model. (10+10)