Why are heteroscedasticity, multicollinearity and serial correlation a problem in regression model reliability?
A. Their presence indicates that the standard error of the model is understated and the model is unreliable.
B. They indicate that the model may be subject to Type I error.
C. They indicate that the residuals are not normally distributed.
D. 1 and 2 above.
E. None of the above.