@jprob for panel GMM equation (EV7)

[nadja123]

Hello there,

the command

Code: Select all

scalar myjprob = eq01.@jprob

does "successfully compute" the p-value, but the value is actually NA. (Doing it other way gives a numeric value). Is this a bug?

[EViews Gareth]

That's a really cryptic statement. What do you mean?

[nadja123]

How would I tell you... Suppose quation eq01.gmm(options) has been estimated in a balanced panel workfile. Then

Code: Select all

scalar myf*** = eq01.@jstat

computes the J-statistic. I want the p-value to J-statistic. In the help doc to EV7 there is mentioned

@jprob probability value of the J-statistic

as one of the equation data members. How to get that computed? Thank you.

[EViews Gareth]

Still not quite sure I follow your question, but perhaps the following will help:

Code: Select all

create u 100 rndseed 1 series y=nrnd series x=nrnd series z1=nrnd series z2=nrnd equation eq1.gmm y c x @ z1 z2 scalar auto = eq1.@jprob scalar manual = 1 - @cchisq(eq1.@jstat, 1)

You can see that the scalar calculated by eq1.@jprob is equal to the scalar calculated manually using the j-stat.

[EViews Glenn]

From what he says it might be in the panel equation estimator not the standard GMM. Can you send your workfile to support@eviews.com or post it here...Let us know what equation it is...

[edit]

We've replicated on this end. The @jprob keyword was added to EVIews 7 equations in the improvements to single equation GMM, but we neglected to update the earlier panel code. This will be fixed in the next update. For now, just use the @ncoefs, @jstat, @instrank members of the equation and the @chisq function to compute the p-val...