If the residuals violate the assumption of normality, we expect that:
A)the sample was not taken randomly.
B)the residuals will not sum to zero.
C)least squares formulas will fail.
D)confidence intervals may be unreliable.

Question

Quizplus · Accepted Answer

If the residuals violate the assumption of normality, it means that the errors in the linear regression model are not normally distributed, leading to unreliable confidence intervals. This can be due to outliers, skewed distributions, or other factors. Therefore, the best choice is D - confidence intervals may be unreliable. It does not necessarily imply that the sample was not taken randomly, or that the residuals will not sum to zero or that least squares formulas will fail.

If the Residuals Violate the Assumption of Normality, We Expect