1. Consider the following Residuals vs Fits plot and residuals\' probability plo
ID: 3044295 • Letter: 1
Question
1. Consider the following Residuals vs Fits plot and residuals' probability plot (along with the Anderson-Darling test results) of a linear regression study, y~x. Versus Fits (response is Y) Probability Plot of RES 0.75 0.25 0.00 Mean -142109E-15 0.3286 30 0.584 0117 StDev AD 7.5 9.5 0.8 0.6 0.4 0.2 0.0 0.2 40608 Fitted Value a. State whether the assumption E(4) = 0 is violated or not. Why? b. According to the Anderson-Darling result, is the normality assumption justified at any reasonable significance level? Answer by considering Ho: Random error terms are normal, and by using p-value. c. By considering the Residuals vs Fits plot, do the residuals have a constant variance, i.e., are the distributions of the errors identical?Explanation / Answer
Answer:
a).
The assumption is violated because most of the points in the Residuals vs Fit plot are above the 0 value.
b).
The calculated AD statistic 0.584, P=0.117 which is > 0.0.5 level of significance. Ho is not rejected. We conclude that normality assumption is not violated.
c).
The residuals look like fan out pattern, the constant variance assumption is violated. The distribution of errors are not identical.
Related Questions
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.