University Of Basrah Topic : Econometrics

Collage Of Economic & Admin. Time : 3 hour’s

Statistic Department Date : 5/9/2020
Questions Of Final Course Of Master Degree 2019 – 2020 (Second Treatment )
Q 1/
Note : Answer Only 2 . ( 50 marks )
A/ (1) Consider the Keynesian model given by:
𝐶𝑡 = 𝑎0 + 𝑎1 𝑌𝑡 - 𝑎2 𝑇𝑡 +U1 Where : C : consumption
𝐼𝑡 = 𝑏0 + 𝑏1 𝑌𝑡−1 + U2 Y : Income , I : Investment
𝑇𝑡 = 𝑐0 + 𝑐1 𝑌𝑡 + U3 T: rate of Tax
𝑌𝑡 = 𝐶𝑡 + 𝐼𝑡 + 𝐺𝑡 G : government expenditures
(a) Find Identification for first and second equation .
(b) Find reduced form for this model .
(2) Given this information : Ŷ = -9.532+0.617 X1 + 0.348 X2 , N =17
S.e (8.844)
Var(b1) = 0.0212 , F= 96.65 , 𝑅2 = 0.93
∑ 𝑒𝑡2 =1878.04 , ∑∆𝑒𝑡2 =3910.08 , du = 1.54 , dl = 1.02
Testing this results , and discus the results about using for Prediction .
B /
(1) Explain , why introduce Random Variable in the econometrics model .
(2) For this information : Ŷ = - 8.256 + 0.871 X 𝑅2 = 0.91 , N = 17
s.e (9.588)
t ( 2.2)
1
 (a) Testing the significance of intercept coefficient .
(b) calculate the variance of independent coefficient .
(c)Testing the significance of this model .
(3) Discuss briefly the Rational Expectation used your information report .
Q2/
A / Answer (T) OR (F). ( 10 marks )
1 - This hypothesis E (𝑋𝑖 𝑋𝐽 ) = 0 mean , the values of independent variable have
positive ,negative and zero values .
2 - Used the lagged variable in the model as the explanatory variables its caused
multicollinearity problem because its changes together .
3- If you know that 𝑅2 for this equation Y =𝐵𝑂 +𝐵1 𝑋 2 is smaller than 𝑅 2 for this
equation Ln y =𝑎𝑜 +𝑎1 Ln x you can say the first equation is the best .
4- If we found the perfect multicollinearity in the data then the parameters
estimated with OLS method may be unbiased but not efficient .
5 - Accept this hypothesis : 𝐻𝑂 : 𝐵1 =𝐵2 = 𝐵3 = 0 that s mean all the
independent variables are very important to inference in Y .
6 - If the aim of estimated model is forecasting about some economic phenomena
that 𝑅2 is criteria .
7- If there is no relationship between the explanatory variables , we did not need
to use multiple regression , and you can use simple liner model .
8- In case found multicollinearity problem , the estimated parameters stay
unbiased and her variance unbiased too .
9 – For this regression Y = 𝐵0 + 𝐵1 𝑋1 +𝐵2 𝑋2 + 𝑈𝑖 the var (𝐵2 ) = Ƃ2 / ∑𝑥22 (1- 𝑟12
2
).
10 -When we used Dummy variables ( dependent or independent ) in the model
the effect explained on intercept only .
2
B / Chose the correct answer . ( 10 marks )
1- In sample of 50 data , with 4 explanatory variables and durbin- watson = 1
when 𝑑𝐿 = 1.2 , 𝑑𝑢 = 1.4 the model have :
(a) positive autocorrelation . (b) negative autocorrelation .
(c) durbin - watson test is failing .
2 – If it was 𝑅2 = 1 ,this mean the regression line passes through ( ӯ ) .
3 - If that 𝑅2 is high and F - test is good but t - test for some parameters is not
significant that’s mean problem of :
(a) multicollinearity (b) hetroscedastisty (c)autocorrelation .
4- The effect of perfect multicollinearity problem is :
(a) parameters unbiased but not efficiency . (b) Biased and not efficiency.
(c) you can determine the effect of explanotary variables on dependent
variable Y .

5 - If this hypothesis E (𝑋𝑖 𝑢𝑗 ) ≠ 0 ; the problem is :

(a) Multicollinearity. (b) Hetroscdastisty . (c) Autocorrelation .
6 – For this information, ∑𝑑 2 =169.5 , n = 10 , Spearman Rank Correlation
(𝑟𝑠 ) =?
(a) = - 0.27 (b) = - 0.027 (c) = 0.027

7 – If this hypothesis E (𝑋𝑖 𝑋𝑗 ) ≠ 0 ; the problem is :

(a) Multicollinearity (b) Hetroscdastisty (c) Autocorrelation .
8 - The expected value of error term equal zero when :
(a) mean did not correlated with the values of independent variable .
3
 (b) mean that no correlation between 𝑢𝑖 𝑢𝑗 .
(c ) mean some values negative ,some positive and other zero values
Q 3 / Short Answers : ( 30 marks )
Note : Answer Only One ( A or B )
A /
(1) What is the meant of Rational Expectation
(2) What is the desired properties in any economics model .
(3) What is the meant of dummy variables , explain .
(4) For this model : Ŷ = 19 5 + 0.9 x , n = 12 , 𝑅2 = 0.99
S.e (4.4)
t (2.1)
(a) Estimating variance for independent variable . (b) Test the significance
for 𝐵0 . (c) Testing significance estimating model where F= 4 .1 .
B/
(1) What is the procedure of Goldfeld – Quandt test to indicator of
hetroscedasticity.
(2) Explain the types of lagged variables .
(3) If we have this model for specified as :
𝑌1 = 𝑌2 - 2 𝑋1 + 𝑋2 + 𝑈1 …………(1)
𝑌2 = 𝑌3 + 𝑋3 + 𝑈2 …………..(2)
𝑌3 = 𝑌1 - 𝑌2 - 2𝑋3 + 𝑈3 …………..(3)

4
 a- Find identification conditions . When :Y1 , Y2 , Y3 dependent variables,
X : independent variables.
b–Find reduced form for this model .
(4) Given this information :
Ŷ = 0.98 + 0.35 𝑋1 + 0.81 𝑋2 n = 25 , F = 13 , 𝑅2 = 0.90 , D.W =0.131
s.e (0.09) (0.24) du = 1.32 , dl = 1.2
1 – Testing and discus this results .
2- using scatter plot .

GOOD LUCK
Dr. waded A. Wadi Dr. Ali Nasir
Teacher Head of statistic Dept.