In multiple regression analysis, residuals (Y - Ŷ) should be ____________.
A) Qualitative variables
B) Significantly different from zero
C) Correlated
D) Normally distributed with a mean of zero

Question

Quizplus · Accepted Answer

Residuals in multiple regression analysis should follow a normal distribution with a mean of zero, indicating that the model is correctly predicting the outcome variable. Residuals that are significantly different from zero or correlated may indicate problems with the model's assumptions, such as nonlinearity or multicollinearity, that can affect the accuracy of the predictions. Qualitative variables are not directly related to residuals, but may affect the choice of regression model and interpretation of the results. Therefore, the correct answer is D, normally distributed with a mean of zero.

In Multiple Regression Analysis, Residuals (Y - Ŷ) Should Be