Of the following conditions, which one can fail yet produce a regression model that is reasonably good, in many cases, if the sample size is large?
A)Normality of errors
B)Independence of errors
C)Linearity
D)None of these

Question

Quizplus · Accepted Answer

With a large sample size, the Central Limit Theorem suggests that the distribution of errors will approximate normality, making the normality of errors less critical for the validity of regression results.

Of the Following Conditions, Which One Can Fail Yet Produce