ADF Test .Please help

[anza.ch]

1) In ADF test results, do we just follow the rule that if t-stat>t-tab, reject Ho of non stationarity. my question is that in ADF, do we only compare the T-stat with the critical values or we should look at the respective p-value as well ?

2) second question, a dependent variable which is stationary at level, but i want it to be stationary at first difference. is there any method of doing so??

Please help someone

[aasif127]

Doesn't matter. I would rather like to have close look at p-values because they will tell you at what level your data is stationary which you can not get from t values. Your second question is not clear to me.

[anza.ch]

Thank you. Please guide if The ADF test suggests that X(t) is I(1), and the Phillip Perron test that X(t) is I(0). In short both these unit root tests give opposite results. The sample is from 1970 to 2011. What should I decide?