The only real difference I'm aware of is that the current DPD-Ox
version does not have the Blundell-Bond (1998) estimator
where both differenced and level equations are used.
The new version of DPD is being finalized at the moment, and will
be released next week.
The sample output abest1.out, for example, lists the Sargan test
output for two-step estimation. This requires that specification
tests are switched on with SetTest().
The outcome for the two-step Sargan test will be slightly different
in the new version, which will be explained in the documentation
(basically the ratio of the one-step to the two-step variance).
oxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxox
Dr Jurgen A Doornik
Nuffield College
Oxford OX1 1NF, UK
tel. +44-1865-278610 fax +44-1865-278557
http://www.nuff.ox.ac.uk/Users/Doornik/
oxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxox
----- Original Message -----
From: Christopher F Baum <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, 03 June, 1999 13:33
Subject: Re: DPD for Ox
> I have been told by a colleague who uses GAUSS DPD (who reviewed the
latest
> Ox DPD documentation) that the Ox version is considerably out of date
> vis-a-vis the latest GAUSS version. This may be an artifact of that
> phenomenon. Will it be possible to get the latest version of DPD for Ox,
so
> that we don't have to go back to GAUSS?
>
> Thanks
> Kit Baum
> Boston College Economics
>
> --On Thu, Jun 3, 1999 11:26 +0000 Maurice Bun <[log in to unmask]> wrote:
>
> > Currently I am working with DPD for Ox and it appears to me that only
> > a one step version of the Sargan test is programmed even if this test
> > statistic is invoked after two step estimation.
> >
> > Is my conjecture correct and if so, is there any code available for the
> > two step version of this test statistic?
> >
> > Maurice Bun
> > University of Amsterdam
>
>
>
>
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|