Scribd
Upload
37 views2 pages

Problem Set 2

This problem set contains questions about ordinary least squares (OLS) estimation and hypothesis testing in simple linear regression models. It asks students to: 1) Identify which statements about OLS are correct; 2) Determine which statement does not describe the R-squared statistic; and 3) Interpret regression results and test significance for a model of how years of schooling affects wages, considering potential issues.

Uploaded by

dxd032
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
37 views2 pages

Problem Set 2

This problem set contains questions about ordinary least squares (OLS) estimation and hypothesis testing in simple linear regression models. It asks students to: 1) Identify which statements about OLS are correct; 2) Determine which statement does not describe the R-squared statistic; and 3) Interpret regression results and test significance for a model of how years of schooling affects wages, considering potential issues.

Uploaded by

dxd032
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Problem Set 2

This exercise is based on material found in Section 2 of the lecture notes booklet. The questions
are about OLS and hypothesis testing in the simple regression setting.

1. Which of the following statements about the Ordinary Least Squares (OLS) estimation of a
linear regression model is/are correct?

(a) OLS uses the method of minimising the sum of squared residuals;
(b) In a bivariate regression setting, OLS is a procedure that will find the best fitting line
through the scatter plot of the two variables;
(c) Under the classical assumptions, OLS is unbiased;
(d) Under the classical assumptions, OLS is the most efficient estimator amongst all unbiased
estimators;
(e) All of the above.

2. Which of the following statements does not describe a property of the R2 statistic?

(a) The R2 statistic is a measure of the “goodness of fit” of a regression model;

(b) The R2 statistic is a good statistic to use when comparing models with different dependent
variables;

(c) 0  R  1 ;
2

(d) The closer the R2 is to 1, the better the fit of the model;
(e) None of the above.

3. Consider the following bivariate regression that highlights how years of schooling affects an
individual’s hourly wage:

𝑊𝑖 = 𝛽0 + 𝛽1 𝑆𝑖 + 𝜀𝑖 for i = 1,...,100

where Wi is the individual’s hourly wage and S i is how many years of schooling they have had.
The OLS estimated regression takes the form, with standard errors in parentheses:

Wˆi = 7.50 + 1.25Si


(0.147) (0.077)

Interpret the parameter on the years of schooling variable and test its significance at the 5%
level. Can you think of reasons why such a model might be poorly specified?
Q4 is mildly difficult, as you need to derive the OLS estimator(s) for each model.

4. Consider the following regression models

i) 𝑌𝑖 = 𝛽0 + 𝜀𝑖
ii) 𝑌𝑖 = 𝛽1 𝑋𝑖 + 𝜀𝑖
iii) 𝑌𝑖 = 𝛽0 + 𝛽1 𝑋𝑖 + 𝜀𝑖

for i = 1, , n and where  i is a random disturbance term which satisfies the classical
assumptions. Derive an expression for the Ordinary Least Squares (OLS) estimators of the
unknown parameters in each model and show that these expressions minimise the sum of
squared errors, i.e. check the second-order conditions.

Q5 is also difficult and is based on the proofs in Section 2.

5. Consider the following regression model

𝑌𝑖 = 𝛽0 + 𝛽1 𝑋𝑖 + 𝜀𝑖
where 𝜀𝑖 is a random disturbance term which satisfies the classical assumptions.

a) Show that 𝛽̂0 and 𝛽̂1 are unbiased.


b) Derive the expression for the variance of 𝛽̂0 and 𝛽̂1.

You might also like