Introduction to Econometrics
Final Exam Review
Questions & Solutions
2025
©2025
, Question 1: Omitted Variable Bias
Scenario: A researcher estimates the effect of a job training program on
wages using a simple linear regression. However, the model omits “work
experience,” which is correlated with both training participation and
wages. Which consequence is most likely?
- A. The OLS estimates remain unbiased but become less efficient.
- B. The OLS estimates are biased and inconsistent.
- C. The R-squared will be artificially high.
- D. The standard errors will be understated.
ANS: B
Rationale: When a relevant variable that is correlated with both the
dependent variable and one of the independent variables is omitted, the
resulting OLS estimates become biased and inconsistent (known as
omitted variable bias).
---
Question 2: Detecting Heteroskedasticity
Scenario: After fitting an OLS regression, a researcher performs the
Breusch‑Pagan test and obtains a p‑value of 0.01. What does this result
indicate?
- A. Homoskedastic errors are present.
- B. There is evidence of heteroskedasticity.
- C. The model is correctly specified.
- D. There is autocorrelation in the residuals.
ANS: B
Rationale: A p‑value of 0.01 in the Breusch‑Pagan test leads to
rejection of the null hypothesis of constant variance (homoskedasticity),
indicating that heteroskedasticity is present in the error term.
---
©2025
, Question 3: Addressing Endogeneity
Scenario: In a study of education’s effect on earnings, a researcher
suspects that education is endogenous because unobserved ability
influences both education and earnings. Which estimation technique is
most appropriate to obtain consistent parameter estimates?
- A. Ordinary Least Squares (OLS)
- B. Two‑Stage Least Squares (2SLS)
- C. Maximum Likelihood Estimation (MLE)
- D. Ridge Regression
ANS: B
Rationale: Two‑Stage Least Squares (2SLS) is the instrumental variable
(IV) technique used to correct for endogeneity by using instruments that
are correlated with the endogenous regressor but uncorrelated with the
error term.
---
Question 4: Autocorrelation Testing
Scenario: A time series regression produces a Durbin‑Watson statistic of
1.2. What does this imply about the error terms?
- A. There is likely positive autocorrelation.
- B. There is likely negative autocorrelation.
- C. The error terms are homoskedastic.
- D. The model suffers from omitted variable bias.
ANS: A
Rationale: In the Durbin‑Watson test, a value significantly below 2
(such as 1.2) indicates the presence of positive autocorrelation among
the residuals.
---
Question 5: Cointegration in Time Series
Scenario: Researchers studying the long‑term relationship between GDP
©2025
Final Exam Review
Questions & Solutions
2025
©2025
, Question 1: Omitted Variable Bias
Scenario: A researcher estimates the effect of a job training program on
wages using a simple linear regression. However, the model omits “work
experience,” which is correlated with both training participation and
wages. Which consequence is most likely?
- A. The OLS estimates remain unbiased but become less efficient.
- B. The OLS estimates are biased and inconsistent.
- C. The R-squared will be artificially high.
- D. The standard errors will be understated.
ANS: B
Rationale: When a relevant variable that is correlated with both the
dependent variable and one of the independent variables is omitted, the
resulting OLS estimates become biased and inconsistent (known as
omitted variable bias).
---
Question 2: Detecting Heteroskedasticity
Scenario: After fitting an OLS regression, a researcher performs the
Breusch‑Pagan test and obtains a p‑value of 0.01. What does this result
indicate?
- A. Homoskedastic errors are present.
- B. There is evidence of heteroskedasticity.
- C. The model is correctly specified.
- D. There is autocorrelation in the residuals.
ANS: B
Rationale: A p‑value of 0.01 in the Breusch‑Pagan test leads to
rejection of the null hypothesis of constant variance (homoskedasticity),
indicating that heteroskedasticity is present in the error term.
---
©2025
, Question 3: Addressing Endogeneity
Scenario: In a study of education’s effect on earnings, a researcher
suspects that education is endogenous because unobserved ability
influences both education and earnings. Which estimation technique is
most appropriate to obtain consistent parameter estimates?
- A. Ordinary Least Squares (OLS)
- B. Two‑Stage Least Squares (2SLS)
- C. Maximum Likelihood Estimation (MLE)
- D. Ridge Regression
ANS: B
Rationale: Two‑Stage Least Squares (2SLS) is the instrumental variable
(IV) technique used to correct for endogeneity by using instruments that
are correlated with the endogenous regressor but uncorrelated with the
error term.
---
Question 4: Autocorrelation Testing
Scenario: A time series regression produces a Durbin‑Watson statistic of
1.2. What does this imply about the error terms?
- A. There is likely positive autocorrelation.
- B. There is likely negative autocorrelation.
- C. The error terms are homoskedastic.
- D. The model suffers from omitted variable bias.
ANS: A
Rationale: In the Durbin‑Watson test, a value significantly below 2
(such as 1.2) indicates the presence of positive autocorrelation among
the residuals.
---
Question 5: Cointegration in Time Series
Scenario: Researchers studying the long‑term relationship between GDP
©2025
