Maharshi Dayanand Saraswati University (MDSU) 2008 M.A Economics thetical and Econometrics - Question Paper

ques. Paper Mathematical Economics and Econometrics M.A.(Economics) 2008
Mathematical Economics and Econometrics – II
Economics (Optional)(Mathematical Economics and Econometrics – II)
Note: (1) All ques. are compulsory. (2) every ques. carries equal marks. (3) Simple calculator can be used for computations.

1. (a) In the situation of K variable general linear regression model, prove that under
certain assumptions the ordinary lowest savares (OLS) method estimators are BLU
(Best Linear Unbiased).

(b) We have 5 observations on the hours of work and production.
(X) Hours 10 12 13 15 20
(Y) Production Units 15 25 25 30 50
With the help of the provided data :

(i) Estimate the parameters of the linear regression formula ^Y = ^a + ^bx and
discuss what does the slope ‘b’ indicate.

(ii) Estimate the parameters of the reverse regression formula ^ X = ^c + ^ dy.

(iii) compute coefficient of correlation and coefficient of determination.

(iv) Estimate the -R

2 (R-bar-square) and the avg. elasticity of output with
respect to labour.

OR

(a) Derive formulae for the parameters of the K-variable general linear regression
model through maximum likelyhood method.

(b) explain the ‘distributed lay’ model developed by Koyeck.

2. (a) What is meant by the issue of ‘Heteroscedasticity’ ? elaborate its consequences
on the properties of the estimators ? explain any 2 methods for detecting this
problem.

(b) What is meant by ‘Perfect multicollinearity’ ? Prove that under such
circumstances the estimates of the parameters become meaningless and their
standard errors are infinite.

OR

(a) ‘When can it be stated that the simultaneous formula model is identified ? discuss
the necessary and sufficient conditions for this purpose.

(b) discuss under what circumstances, and how you would use the 2 Stage lowest
Squares (2SLS) method to estimate the simultaneous formula.

3. (a) After explaining the causes of the issue of ‘autocorrelation’ explain the
Durbin-Watson (D – W) method for detecting this issue.

(b) discuss the importance and uses of the Generalised lowest Squares (GLS) method.

OR

(a) discuss in brief the ARMA and ARIMA models used for forecasting purpose.
Why these models have become more popular ?

(b) If the regression suffers from the Co-integration issue how does it affect the
estimates of the parameters ? explain any 1 method for detecting this issue.

4. Write short notes (any two) :
(a) Distributed lag model provided by Almon.
(b) issue of the ‘unit root’ and the Dicuy-Fuller method for detecting this issue.
(c) Uses and limitations of the LPM (Linear Probability Model)
(d) Properties of a good estimator in the situation of both small and large samples.

Earning:   Approval pending.