Explanation:
The risk consultant is facing the classic multicollinearity problem: statistical significance is indicated, but none of the explanatory variables are. A potential solution is to delete explanatory variables one by one and then rerun the regression to check significance at the independent variable level.
(Book 2, Module 20.1, LO 20.c)
Ultimate access to all questions.
Question 67
A risk consultant has been having difficulty with a particular regression performed for her supervisor at the Department of Economic Statistics. Specifically, the F-test calculated for the regression analysis indicates the regression appears to be statistically valid, but none of the explanatory variables are able to reject the null that they are zero. Which of the following indicates the apparent problem facing the consultant, as well as the potential solution she can use to address the matter?
A
Heteroskedasticity. Add more explanatory variables.
B
Heteroskedasticity. Delete explanatory variables one-by-one and rerun the regression.
C
Multicollinearity. Delete explanatory variables one-by-one and rerun the regression.
D
Multicollinearity. Add more explanatory variables.
No comments yet.