Cointegration & Engle-Granger Test

[econometrics-lady]

Hello Eviews world,

I'm having some issues with a model I'm working on, and I'm at my wits end. I'm working on a time series model that tests whether financial development, leads to, or affects economic growth, and at present have 7 variables, including the dependent variable. I intend on using the Engle-Granger test to test for cointegration, before proceeding to Granger causality, however I'm not sure how to proceed with testing for co integration. I have verified are stationary at 1st diff I(1). How do I proceed with this information?

1. Do I run the original regression using OLS, inserting d(dependent variable, 1) at the beginning of the eqn, or
2. Do I run the original regression using OLS, but inserting d(variable, 1), for all variables (both dependent and independent)
3. Do I then run the ADF test on residuals acquired from either eqn 1, or 2? Will this be interpreted as other ADF test, or are there any special rules to it?
4. What exactly does the Engle-Granger test give me, as far as results? Is it the level to which two variables are cointegrated? Which means? The level to which one variable affects the other?

Finally, is Granger causality an ECM model? Or how can I derive an ECM model with this information. Anything would be much appreciated. Thank you.

[fredasitorus]

hi..
have u finished your paper? do you have the answer?
i'm also having a trouble with this test..
i hope you can post the answer.
thank you.