Hi, I have been knocking my head against the wall of statistics (so to speak) for uncounted hours and just can't figure this out. I cannot remember feeling this dumb. Okay, so the problem is the Park Test, testing heteroskedasticity.
I am using Eviews
The OLS is RD c Sales
From which I get residuals resid01
Now, for the Park test, I am running another ols with the form:
log(resid01^2) c log(Sales)
However, this doesn't yield the right answer :shakehead
The White test works out just fine, so the data is fine and the residuals must also be fine. But the Park test is not fine and I am not fine because of it. Myself and the Park test are not independent, there apparently is colinearity and the Park Test's stochastic behavior is making my mood very very stochastic. Please Help!!!! :yup: