Degrees of Freedom

[ifu06636]

I am aware that I can do this in eviews, but i wanted to check how do i actually decide the degrees of freedom when doing a chi square test.

i have a very simply mean equation of Returns = Constant + Residuals of conditional mean

The variance equation is more complex and I have 6158 observations, So is my degrees of freedom n-1 i.e. 6157?

I guess not but Im not certain, would someone be able to help, Im sure its a simple answer!

[EViews Glenn]

A chi-square test for what?

And for what purpose do you wish to compute the d.f.?

[ifu06636]

Hi, apologies I should have been more specific.
It is to do a white test, for homoskedasticity before doing a GARCH model,
I am not sure how to find the degrees of freedom for the LM Stat to use with a table

[startz]

If you do View/Residual diagnostics/Heteroskedasticity you get the chi-squared statistic and EViews tells you the degrees of freedom

[ifu06636]

Is there a specific way in which i should be setting the Lag lenth for testing for Arch effects, Specifically the LM Stat?