ECONOMETRICS (ECON2061) - Complete Question Bank
100+ Questions with Detailed Answers
Sections: Multiple Choice | Short Answer | True/False | Spot the Error
SECTION A: Multiple Choice Questions (40 Questions)
Topic 1: OLS Regression Basics
Q1. In the regression model Y = β₀ + β₁X + u, the term u represents:
A) The predicted value of Y
B) The error term capturing unobserved factors
C) The slope coefficient
D) The correlation between X and Y
Q2. OLS estimation minimizes:
A) The sum of residuals (Σûᵢ)
B) The sum of squared residuals (Σûᵢ²)
C) The sum of absolute residuals (Σ|ûᵢ|)
D) The variance of the error term
̂
Q3. If β₁ = 2.5 in the model Wage = β₀ + β₁Education + u, this means:
A) Education causes a £2.50 increase in wages
B) One additional year of education is associated with £2.50 higher wages, on average
C) 2.5% of wage variation is explained by education
D) The correlation between education and wages is 2.5
Q4. The difference between the error term (u) and the residual (û) is:
A) There is no difference
B) The error is observable; the residual is not
C) The error is unobservable; the residual is observable
D) The residual is always larger than the error
Topic 2: Measures of Fit
Q5. R² = 0.65 means:
A) X causes 65% of Y
B) 65% of the variation in Y is explained by the regression
C) The correlation between X and Y is 0.65
D) The model is 65% accurate
Q6. Which statement about R² is FALSE?
A) R² always increases when you add more regressors
B) R² can never be negative
C) A high R² proves causation
D) R² = 1 - SSR/TSS
Q7. Adjusted R² differs from R² because it:
A) Is always higher than R²
B) Penalizes for adding useless variables
C) Can only be used with time series data
D) Measures causation rather than correlation
Q8. TSS = ESS + SSR. In this equation, ESS represents:
, A) Σ(Yᵢ - Ȳ)²
B) Σ(Ŷᵢ - Ȳ)²
C) Σ(Yᵢ - Ŷᵢ)²
D) Σûᵢ²
Topic 3: LSA Assumptions
Q9. LSA 1 states E(u|X) = 0. This assumption is violated when:
A) The sample size is small
B) There is omitted variable bias
C) The residuals are normally distributed
D) R² is low
Q10. What is the consequence of violating LSA 1 (E(u|X) = 0)?
A) Inefficient but unbiased estimates
B) Biased and inconsistent estimates
C) Incorrect standard errors only
D) OLS cannot be computed
Q11. Perfect multicollinearity means:
A) Two regressors are highly correlated
B) One regressor is an exact linear function of another
C) The error term is correlated with X
D) R² equals 1
Q12. The "dummy variable trap" occurs when:
A) You use too many dummy variables
B) You include all categories of a categorical variable plus an intercept
C) You forget to include dummy variables
D) The dummy variable coefficient is insignificant
Topic 4: Hypothesis Testing
Q13. The t-statistic for testing β₁ = 0 is:
̂ ̂
A) t = β₁ × SE(β₁)
̂ ̂
B) t = β₁ / SE(β₁)
̂ ̂
C) t = SE(β₁) / β₁
̂ ̂
D) t = β₁ - SE(β₁)
Q14. At the 5% significance level (two-sided), you reject H₀ if:
A) |t| > 1.645
B) |t| > 1.96
C) |t| > 2.576
D) |t| < 1.96
Q15. A 95% confidence interval for β₁ is:
̂ ̂
A) β₁ ± 1.645 × SE(β₁)
̂ ̂
B) β₁ ± 1.96 × SE(β₁)
̂ ̂
C) β₁ ± 2.576 × SE(β₁)
̂ ̂
D) β₁ ± SE(β₁)
Q16. The F-test is used to:
A) Test a single coefficient
B) Test joint hypotheses about multiple coefficients
C) Test for heteroskedasticity
D) Calculate R²
100+ Questions with Detailed Answers
Sections: Multiple Choice | Short Answer | True/False | Spot the Error
SECTION A: Multiple Choice Questions (40 Questions)
Topic 1: OLS Regression Basics
Q1. In the regression model Y = β₀ + β₁X + u, the term u represents:
A) The predicted value of Y
B) The error term capturing unobserved factors
C) The slope coefficient
D) The correlation between X and Y
Q2. OLS estimation minimizes:
A) The sum of residuals (Σûᵢ)
B) The sum of squared residuals (Σûᵢ²)
C) The sum of absolute residuals (Σ|ûᵢ|)
D) The variance of the error term
̂
Q3. If β₁ = 2.5 in the model Wage = β₀ + β₁Education + u, this means:
A) Education causes a £2.50 increase in wages
B) One additional year of education is associated with £2.50 higher wages, on average
C) 2.5% of wage variation is explained by education
D) The correlation between education and wages is 2.5
Q4. The difference between the error term (u) and the residual (û) is:
A) There is no difference
B) The error is observable; the residual is not
C) The error is unobservable; the residual is observable
D) The residual is always larger than the error
Topic 2: Measures of Fit
Q5. R² = 0.65 means:
A) X causes 65% of Y
B) 65% of the variation in Y is explained by the regression
C) The correlation between X and Y is 0.65
D) The model is 65% accurate
Q6. Which statement about R² is FALSE?
A) R² always increases when you add more regressors
B) R² can never be negative
C) A high R² proves causation
D) R² = 1 - SSR/TSS
Q7. Adjusted R² differs from R² because it:
A) Is always higher than R²
B) Penalizes for adding useless variables
C) Can only be used with time series data
D) Measures causation rather than correlation
Q8. TSS = ESS + SSR. In this equation, ESS represents:
, A) Σ(Yᵢ - Ȳ)²
B) Σ(Ŷᵢ - Ȳ)²
C) Σ(Yᵢ - Ŷᵢ)²
D) Σûᵢ²
Topic 3: LSA Assumptions
Q9. LSA 1 states E(u|X) = 0. This assumption is violated when:
A) The sample size is small
B) There is omitted variable bias
C) The residuals are normally distributed
D) R² is low
Q10. What is the consequence of violating LSA 1 (E(u|X) = 0)?
A) Inefficient but unbiased estimates
B) Biased and inconsistent estimates
C) Incorrect standard errors only
D) OLS cannot be computed
Q11. Perfect multicollinearity means:
A) Two regressors are highly correlated
B) One regressor is an exact linear function of another
C) The error term is correlated with X
D) R² equals 1
Q12. The "dummy variable trap" occurs when:
A) You use too many dummy variables
B) You include all categories of a categorical variable plus an intercept
C) You forget to include dummy variables
D) The dummy variable coefficient is insignificant
Topic 4: Hypothesis Testing
Q13. The t-statistic for testing β₁ = 0 is:
̂ ̂
A) t = β₁ × SE(β₁)
̂ ̂
B) t = β₁ / SE(β₁)
̂ ̂
C) t = SE(β₁) / β₁
̂ ̂
D) t = β₁ - SE(β₁)
Q14. At the 5% significance level (two-sided), you reject H₀ if:
A) |t| > 1.645
B) |t| > 1.96
C) |t| > 2.576
D) |t| < 1.96
Q15. A 95% confidence interval for β₁ is:
̂ ̂
A) β₁ ± 1.645 × SE(β₁)
̂ ̂
B) β₁ ± 1.96 × SE(β₁)
̂ ̂
C) β₁ ± 2.576 × SE(β₁)
̂ ̂
D) β₁ ± SE(β₁)
Q16. The F-test is used to:
A) Test a single coefficient
B) Test joint hypotheses about multiple coefficients
C) Test for heteroskedasticity
D) Calculate R²
