Which of the following is not a possible indicator of multicollinearity?
A) significant correlations between pairs of independent variables
B) non-significant t-tests for individual β parameters when the F-test for overall model adequacy is significant
C) signs opposite from what is expected in the estimated β parameters
D) non-random patterns in the plot of the residuals versus the fitted values

Question

Quizplus · Accepted Answer

Non-random patterns in the plot of the residuals versus the fitted values are typically indicators of model misspecification or heteroscedasticity, not multicollinearity. Multicollinearity is more related to the relationships between independent variables.

Which of the Following Is Not a Possible Indicator of Multicollinearity