Questions: Testing for Heteroskedasticity

5 questions to test your understanding

Score: 0 / 5
Question 1 Multiple Choice

After running a regression of consumption on income, you plot residuals against fitted values and see a clear funnel shape — variance increasing with fitted values. Which formal test is most appropriate to confirm this, and what would rejection indicate?

AAn F-test for joint significance; rejection means some regressors are not significant
BA Breusch-Pagan test; rejection means error variance is systematically related to the regressors
CA Durbin-Watson test; rejection means residuals are correlated across observations
DA Hausman test; rejection means the regressors are endogenous
Question 2 Multiple Choice

A researcher uses White's test instead of Breusch-Pagan. What is the main reason White's test might be preferred, and what is its key limitation?

AWhite's test is preferred because it has higher power in all sample sizes; its limitation is it requires specifying the functional form for variance
BWhite's test is preferred because it does not require assuming which regressors drive variance; its limitation is lower power in small samples due to using more degrees of freedom
CWhite's test is preferred for detecting serial correlation; its limitation is it cannot detect heteroskedasticity in cross-sectional data
DWhite's test is preferred because it is computationally simpler; its limitation is it only works for continuous regressors
Question 3 True / False

The Breusch-Pagan test works by regressing squared OLS residuals on the original explanatory variables and testing whether the coefficients are jointly zero.

TTrue
FFalse
Question 4 True / False

Because OLS coefficient estimates are unbiased under heteroskedasticity, detecting heteroskedasticity requires no change to the estimation or inference procedure.

TTrue
FFalse
Question 5 Short Answer

What is the null hypothesis of the Breusch-Pagan test, and what does rejection tell you about the need to change your estimation strategy?

Think about your answer, then reveal below.