- CFA Exams
- CFA Level I Exam
- Topic 1. Quantitative Methods
- Learning Module 2. Multiple Regression
- Subject 10. Heteroskedasticity

###
**CFA Practice Question**

In her analysis of a multivariate regression constructed by one of her colleagues, Beth doubts if the regression residuals are positively correlated with several of the independent variables. She then regresses the squared residuals against the independent variables. The results of her hypothesis test indicate that the regression residuals, while uncorrelated within themselves, are strongly correlated with the independent variables.

The estimation bias described in this example is ______. The appropriate remedy to mitigate this form of estimation bias is ______.

A. Conditional heteroskedasticity; robust standard errors.

B. Serial correlation; Breusch-Pagan test.

C. Conditional heteroskedasticity; Durbin/Watson test.

**Explanation:**The estimation bias described in this example is conditional heteroskedasticity. There are two methods to correct for conditional heteroskedasticity: the computation of robust standard errors, and generalized least squares.

###
**User Contributed Comments**
2

User |
Comment |
---|---|

malawyer |
2 of the answers propose a test which cannot solve for an issue but only detect it |

chriswwu |
I think that is intentional to see whether people are paying attention |