Joint hypothesis test with unit root test

[upani]

Dear All,

I want to do a joint hypothesis test as in Dickey and Fuller (1981) for choosing over randomwalk vs intercept vs trend. The procedure is explained in page no 183 (Chapter 4 model with trend) of Walter Enders 2nd edition.

I need your suggestion

With sincere regards,
Upananda

[TomDoan]

That looks like it's now the following example from the 3rd edition:

http://www.estima.com/textbooks/enders_3/endersp269.rpf

[upani]

Dear Sir,

In the following thread, you have mentioned SET TREND = T-1, But in the example from Enders it is SET TREND = T, which one is correct method to include Trend ?

http://www.estima.com/forum/viewtopic.p ... ller#p4535
With regards,
Upananda

[TomDoan]

Dear Sir,

In the following thread, you have mentioned SET TREND = T-1, But in the example from Enders it is SET TREND = T, which one is correct method to include Trend ?

http://www.estima.com/forum/viewtopic.p ... ller#p4535
With regards,
Upananda

It doesn't matter. The difference between the two washes into the constant, which isn't of interest in that form of the test.