Which of the following is true when testing for normality of errors?
A) Normality is verified by inspecting for a bell-shaped distribution.
B) It is easier to evaluate normality with small sample sizes.
C) A scatter diagram of the whole data is always used to verify normality.
D) Errors are normally distributed when the scatter diagram shows a straight-line distribution.

Question

Quizplus · Accepted Answer

Normality of errors in regression analysis is often verified by inspecting the distribution of residuals for a bell-shaped curve, which indicates a normal distribution. This can be done visually using a histogram or a Q-Q plot. Other options mention methods that are not standard or accurate for testing normality. For example, small sample sizes make it harder, not easier, to evaluate normality due to less data to accurately represent the distribution. A scatter diagram of the whole data is not specifically used to verify normality of errors; it's more about the relationship between variables. Lastly, a straight-line distribution in a scatter diagram does not indicate normal distribution of errors; it might indicate a linear relationship between variables.

Which of the Following Is True When Testing for Normality