Академический Документы
Профессиональный Документы
Культура Документы
Questions
Marks
10 Questions. All
compulsory.
6 Questions. Do any 5 only.
20
3 Questions. All
compulsory
Total
ECN 702
1
50
30
100
Trimester 3, 2014
[20 Marks]
There are 10 Questions in this Section and all questions are compulsory. Each is worth 2 marks.
1. What is econometrics and what is its importance in economics?
2. What is the difference between Time Series and Cross-sectional data?
3. What does coefficient of correlation show? Does the coefficient of correlation have any
capacity in showing which variable causes which?
4. How do you think the development of economic models is different from scientific findings in
a science lab?
5. How does the population regression function (PRF) differ from the sample regression function
(SRF)? Why do we develop a sample regression function?
6. What is goodness of fit? What variable is computed in econometrics that measures the
goodness of fit?
7. In econometrics, under the concept of estimation, what two types of estimation is carried out?
8. In econometrics, what is the difference between r2 and R2?
9. When would one conduct an auxiliary regression?
10. What is the difference between homoscedasticity and heteroscedasticity?
Section B: Descriptive Questions
[50 Marks]
There are 6 Questions in this Section. You are required to do any 5 only.
Question 1
Clearly outline and explain the 10 assumptions of the Gaussian Classical Linear Regression
Models.
(10 marks)
Question 2
The disturbance term ui is a surrogate for all those variables that are omitted from the model that
collectively affect Y. However, we do not introduce them into the model explicitly. The reasons are
many. Outline and explain at least 5 reasons we do not introduce them in the model explicitly.
(10 marks)
Question 3
ECN 702
2
Trimester 3, 2014
State in preceding order and explain at least 5 out of the 8 steps of the classical or traditional
methodology of econometrics.
(10 marks)
Question 4
There are at least 5 reasons for multicollinearity among the explanatory variables. State and
explain the 5 sources of multicollinearity.
(10 marks)
Question 5
There are several reasons why the error terms may be correlated in a series. State and explain at
least 5 reasons why the error terms would be correlated.
(10 marks)
Question 6
State whether the following equations are linear or not and why.
(10 marks)
a.
b.
c.
d.
[30 Marks]
Question 1
[13 marks]
There are two parts, PART A and PART B, to this question and both are compulsory.
PART A
Salary was regressed on experience using computer software and the following information was
obtained. Use the following information and answer the questions that follow:
Ordinary Least Squares Estimation
****************************************************************************
Dependent Variable is Y
****************************************************************************
Regressor
Coeficient
Standard Error
T-Ratio [Prob]
f
0.54021
0.00972
43.6623
0.000
C
12.2134
0.9323
91.2245
0.000
****************************************************************************
R-squared
0.97986
R-Bar-Squared
0.9793
ECN 702
3
Trimester 3, 2014
Required:
a) Report the Equation in the standard way.
(2 marks)
(2 marks)
c) What useful information will be conveyed by the estimated value of the parameters?
(2 mark)
d) What is the reason for presence of u?
(1 mark)
PART B
Suppose you are an economist at the Ministry of Labour and have been tasked with developing an
econometric model to see the factors that determine whether a person can get a suitable job in the
country or not. Use this information and answer the following questions:
a. What 4 variables will you consider in your model?
(2 marks)
b. What would be your prior expectation of the signs of each of the betas in your model?
(2 marks)
c. Identify at least two factors that would have a negative relationship with you dependent
variable in this model.
(2 marks)
[6 Marks]
Assume that you did a two variable linear regression and found that 1 = 50.4 , 2 = 0.6056,
se (2) = 0.0257. Your sample size has been 10. Assuming = 5%, that is, 95% confidence
coefficient answer the following questions:
a. Construct a 95% confidence interval and interpret it.
(3 marks)
b. Conduct a hypothesis test to see the validity of the sample findings if the belief is that
the true 2 = 0.4
(3 marks)
ECN 702
4
Trimester 3, 2014
[11 Marks]
Suppose you run a multiple regression where your dependent variable is child mortality (CMi) and
the three independent variables are per capita GNP (PGNPi ), Female literacy rate (FLRi ) and
weight at birth ((WABi ).
You get your Eveiws output as follows:
se
R2 = 0.8069
= (12.3259)
(0.0023)
(0.3121)
(0.0021)
R2 = 0.7241
Required:
a. Interpret the results of the multiple regression output.
(6 marks)
b. When you run another multiple regression using the same data but this time removing
the variable WAB, you get the following output:
a. CMi = 263.6416, 0.0087 PGNPi , 2.3263 FLRi ,
b. se
= (12.3259)
c. R2 = 0.7054
(0.0023)
(0.3121)
R2 = 0.6803
THE END
ECN 702
5
Trimester 3, 2014
Formulas
Confidence interval
Pr [2 t /2 se (2) 2 2 + t/2 se (2)] = 1
= 2 t/2 se (2)
ECN 702
6
Trimester 3, 2014