Date:         Tue, 29 Mar 2005 11:15:28 -0800
Reply-To:     cassell.david@EPAMAIL.EPA.GOV
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "David L. Cassell" <cassell.david@EPAMAIL.EPA.GOV>
Subject:      Re: Proc TSCSREG question
In-Reply-To:  <200503291227.j2TCRQlE003435@listserv.cc.uga.edu>
Content-type: text/plain; charset=US-ASCII

Thushyanthan Baskaran <tbaskaran@DIW.DE> replied:
> The relevant part of the code is quite simple:
>
> proc tscsreg data = hallo.tradeprod3;
> where keep = 1; /*needed to deal with the unbalanced nature of the
dataset*/
> id ccode_x year;
> model logFlow = logProdGdp logProdPGdp dist contig comlang_off colony
> EU_both EFTA_both / fixtwo; /*a gravity type model*/
> run;
>
> The panel was unbalanced, but I applied a macro that seemingly solved
the
> problem. I have trimmed down the dataset a little, however I left some
> varibales in the dataset that were not used in the regression.

With more information, we really can't tell what's going
wrong.  But you say that the "panel was unbalanced, but I applied
a macro that seemingly solved the problem."  How are you fixing
the problem?  This seems fairly problematic to me.  If the panels
are unbalanced, what sort of macro are you using to 'fix' your
problem, and how are you 'fixing' things?  Is it possible that your
macro is making things go berserk?  Have you checked that the output
of the macro is a set of balanced panels, with the right properties
for your analysis?  I would check the macro output first, and the
assumptions for the resulting panel structure second.

And BTW, are you sure you want to use the Fuller-Battese method here,
when you're stating you want a 'gravity type' model?

HTH,
David
--
David Cassell, CSC
Cassell.David@epa.gov
Senior computing specialist
mathematical statistician