Question

ASSIGNMENT 1

Krupp and Pollard (1996) analyzed the effects of antidumping fillings by U.S. chemical industries on import of various chemicals from China. The import effect results to barium chloride is shown by table below.

            Dependent variable : log(chnimp)

Coefficient	OLS	Paris-Winsten
log(chempi)	3.12 (0.48)	2.94 (0.63)
log(gas)	0.196 (0.907)	1.05 (0.98)
log(rtwex)	0.983 (0.400)	1.13 (0.51)
befile6	0.060 (0.261)	-0.016 (0.322)
afdec6	-0.565 (0.286)	-0.577 (0.342)
intercept	-17.81 (21.05)	-37.08 (22.78)
ρ	-	0.293
Obs	131	131
R2	0.305	0.202

                        ( ) = standard error     

where;

chnimp    = volume of imports of barium chloride from China

chempi     = index of chemical production (as demand for barium chloride)

gas            = volume of gasoline production (as demand for barium chloride)

rtwex        = exchange rate index

befile6     = dummy variable: 1 = six months before filing

afdec6      = dummy variable; 1 = six months after the positive decision.       

1. Give the difference between the OLS and Prais-Winsten estimation.

(4 marks)

2. Which estimation is better? Give TWO(2) reasons.

(4 marks)

3. What the meaning of ρ and show how this coefficient is derived.

(4 marks)

4. Give an interpretation for R2 from the estimation.

(4 marks)

Answer 1

1.  The Prais-Winsten estimator takes into account AR(1) serial correlation of the errors in a linear regression model. This procedure recursively estimates the coefficients and the error autocorrelation of the specified model until sufficient convergence of the AR(1) coefficient is reached. All estimates are obtained by OLS.Prais–Winsten estimation is a procedure meant to take care of the serial correlation of type AR(1) in a linear model.

while OLS estimation is the procedure in which sum of square of error of linear regression model is minimised to find coefficients of the independant variable. OLS estimators are best linear unbiased estimator.

2. Paris -Winsten estimationis more efficient than ols estimation because it is based on Generalised Least Square Method which gives Unbiased and efficient estimator.

paris Winstern estimation can be executed in the absence of P-valune and it is better finite sample approach with the presence of serial correlation.

3.  the p-value is the probability of obtaining results as extreme as the observed results of a statistical hypothesis test, assuming that the null hypothesis is correct .It can also be said as the minimun lebel of rejecting the null hypothesis.

If your test statistic is positive, first find the probability that Z is greater than your test statistic (look up your test statistic on the Z-table, find its corresponding probability, and subtract it from one). Then double this result to get the p-value.

4. In ols estimation, R^2 = .305*100 = 30.5 % means that 30.5 % of variation in volume of imports of barium chloride from China is explained by all the independant variable.

In paris - Winstern, R^2 = .202*100 = 20.2% means that 20.2 % od variation in volume of imports of barium chloride from China is explained by all the independant variable.