Exam Eco1 2022 01 17

EXAM: ECONOMETRICS I.

EUS – ECONOMICS MAJOR 17-1-2022

SURNAME……………………………….………………..……...........................................……NAME..............................…
……..……...............................…........................................DNI/PASSPORT…………………...............................
Continuous assessment: YES................................. Single assessment:
YES..........................
Important: In all tests performed in the exam, the student must define: (i) H0 and Ha, (ii) the test statistic, (iii) the distribution
followed by the test statistic and (iv) the decision rule. All these elements must be defined even though the exercise does not
involve any calculation. The student must write the answers on the space provided below each question.

1. We have a sample on 1954 firms with the following information: Exports (Percentage of foreign sales to total
sales); higheduc (Percentage of workers with higher education); firmsize (logarithm of the number of workers
in the firm); newprods (percentage of new products to total sales). We use higheduc, firmsize and newprods to
explain Exports (Model 1). The estimation results are the following:
1.a) We are interested in testing if the effect of a unitary increase in variable newprods on Exports is twice the
effect of a unitary increase in variable higheduc. Write the specification of the model that incorporates this
restriction. (0.75 points)
1.b) Test the restriction from question 1.a). State the null and alternative hypotheses and define the test statistic
and the distribution followed by this statistic. Indicate the conclusion from the test. (1 point)
1.c) The Variance Inflation Factors (VIF) from model 1 are:
higheduc firmsize newprods

1.028 1.043 1.021
Does model 1 have multicollinearity issues? Justify your answer. Calculate the R-squared and the Adjusted
R-squared from the following regression (1 point):
𝑓𝑖𝑟𝑚𝑠𝑖𝑧𝑒𝑖 = β1 + β2𝑛𝑒𝑤𝑝𝑟𝑜𝑑𝑠𝑖 + β3ℎ𝑖𝑔ℎ𝑒𝑑𝑢𝑐 + 𝑢𝑖

2. Now we estimate the following model:
where sq_firmsize is firmsize squared.
2.a) Consider the estimation results from model 2. If we performed a Reset test in model 1, what conclusion
would we draw from this Reset test? Justify your answer using the test that you consider adequate: state the
null and alternative hypotheses, define and calculate the test statistic and explain your conclusion. (1 point)
2.b) Considering the results from question 2.a), what are the properties of the OLS estimator in model 1? (0.5
points)
2.c) Consider the estimation results from model 2. What is the marginal effect of firmsize on Exports? What is
the value of firmsize with a zero effect on the variable Exports? (1 point)
3. We have a sample on 5,333 women with the following information: children (number of children),
schooling (number of years of schooling), age (age in years), weeklyhours (Number of hours worked per
week), believer (dummy variable that takes value 1 if the woman believes in God and 0 otherwise), and
sch_believer (interaction between schooling and believer). We estimate the following models:
3.a) Provide the economic interpretation of the coefficients of age and believer in Model 2. (0.5 points)
3.b) Consider the results from model 3. What is the partial effect of schooling for a woman who believes in
God? What is the partial effect of age for a woman who does not believe in God? (1 point)
3.c) Consider the results from model 3. What is the number of children predicted by the model for a
non-believer woman who is 48 years old, has 7 years of schooling and works 20 hours per week? (0.5 points)
3.d) Using the estimation results shown above, test if there is a structural break with respect to believer that
affects only the intercept. (0.5 points)
3.e) Using the model 1, explain how we can use the Chow test to analyze if there is a structural break with
respect to whether the woman lives in an urban or a rural area. In particular, explain the steps we should
follow to carry on this test, including the hypotheses and the formula of the test statistic. In addition, write the
specification of the model with dummy variables that is equivalent to this Chow test. Use this model to
explain what hypotheses we should test to get the same conclusion as in the Chow test. (1.25 points)
Now we estimate the following model:
3.f) Using the results from model 4, what is the average number of children for believers? And for
non-believers? Test if there are significant differences in the average number of children between believing
and non-believing women. (1 point)
-----------------------------------------------------------------------------------------------------------------------------------------------------
Critical values
F test (alpha=5%):
F(2, 1954)=3.00 F(1, 1954)=3.85 F(2, 1953)=3.00
F(1, 1953)=3.85 F(1, 1949)=3.85 F(2, 1949)=3.00
Two-sided t test (total alpha=5%) and sample size larger than 1000 observations:
t =1.96

Exam Eco1 2022 01 17

Uploaded by

Copyright:

Available Formats

Exam Eco1 2022 01 17

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Exam Eco1 2022 01 17

Uploaded by

Copyright:

Available Formats

EXAM: ECONOMETRICS I.

EUS – ECONOMICS MAJOR 17-1-2022

1.c) The Variance Inflation Factors (VIF) from model 1 are:

higheduc firmsize newprods

𝑓𝑖𝑟𝑚𝑠𝑖𝑧𝑒𝑖 = β1 + β2𝑛𝑒𝑤𝑝𝑟𝑜𝑑𝑠𝑖 + β3ℎ𝑖𝑔ℎ𝑒𝑑𝑢𝑐 + 𝑢𝑖

where sq_firmsize is firmsize squared.

F(2, 1954)=3.00 F(1, 1954)=3.85 F(2, 1953)=3.00

F(1, 1953)=3.85 F(1, 1949)=3.85 F(2, 1949)=3.00

You might also like