Exercise

Residual analysis - II

In the previous exercise, you fit two different ARMA models to the logged and differenced varve series: an MA(1) and an ARMA(1,1) model. The residual analysis graphics are displayed in order of the run:

MA(1)
ARMA(1, 1)

Which of the following statements is FALSE (partially truthful statements are false - data analysis is not politics)?

Instructions

50 XP

Possible Answers

The residuals for the MA(1) model are not white noise.
The residuals for the ARMA(1, 1) model appear to be Gaussian white noise.
It is not a good idea to look at the residual analysis because it might tell you if your model is incorrect and you might have to stay late at work to figure out the correct model.