![Statalist](/includes/images/head_statalist.jpg) |
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: Re: RE: panel normalisation before xtgee
Thanks for your help Garry,
do you mean that I should change the nb(#) rather than using the normalisation procedure?
Allison
----- Original Message -----
From: Garry Anderson <[email protected]>
Date: Tuesday, March 31, 2009 5:39 pm
Subject: RE: panel normalisation before xtgee
To: [email protected]
> Hi Allison,
>
> On another matter, be aware that f(nb) is forcing your negative
> binomialparameter from xtgee to be k=1.
> Note that k is the recipricol of alpha from -nbreg-
> Variance = u(1+alpha.u) where alpha is reported from the -nbreg-
> command.
>
> f(nb #) with a suitable value of # may give a better fit. # is alpha
> Garry
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of
> [email protected]
> Sent: Tuesday, 31 March 2009 5:59 PM
> To: [email protected]
> Subject: st: panel normalisation before xtgee
>
> Dear Statalist,
>
> I am using xtgee (xtgee, f(nb) corr(ar1) vce(r)) to analyze an
> unbalanced panel data set (with 73 countries, 1980-2006). Someone
> suggested that I normalize my variables by subtracting the mean from
> each of my observations.
> e.g.
> YEAR VARIABLE ONE
> year 1980 observation 1 for variable 1 - mean of
> observation of
> variable 1 year 1981 observation 2 for variable 1 - mean
> of observation
> of variable 1 year 1982 observation 3 for variable 1 -
> mean of
> observation of variable 1
>
> Is this necessary?
> Any opinions on this subject would be great.
> Allison
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/