# Time series and EViews

I have 3 questions on EViews:

m, x, and y are three series. I have found that $m \sim I(1)$, and $x \sim I(2)$,and $y \sim I(2)$

Firstly, can I generate first degree difference by writing " d(m) " in EViews? How do I generate second degree difference in EViews?

Second, I want to do a causality test. Two series are second degree differenced whereas one series is first degree differenced. Does testing these three series together lead to a problem? Or do all series have to be the same degree difference?

Third and the most important question is that can I use series at different degree diffrence by doing Johansen's cointegration test? Does this create a problem?

• Note that how to use particular software isn't on-topic here, so answers might well not address your first question. – Scortchi Apr 24 '15 at 8:45
• not important first question. i want to know second and third questions. Do you have any idea? @Scortchi – user315 Apr 24 '15 at 11:42
• To generate second order differences in Eviews, write d(d(m)). – Dr Nisha Arora Jun 2 '16 at 8:01