EViews.com

By using this estimation dy=d(y,0,4)
How can I check wether the result at level is non-stationary and at 1st difference is stationary?

I am new in using EViews.

Let's clear up some terminology first.

What do you mean by "estimation". There is no estimation there. You're just creating (presumably) a series that is equal to the 4 period difference in the other series.

If that's what you are doing, you can just do a unit root test, or whatever other test you would use for stationarity, on that new series.

ok.

You're still not making sense, unfortunately.

Hello : Should one always difference before log transforming variables for stationarity? My >5000 obs dataset of stock index variables are non-stationary but show stationarity when logged to obtain return variables. Do I still need to difference?

Thanks for suggestions.