Re: st: Clustered, longitudinal data with ordinal outcome

[Stas Kolenikov <skolenik@gmail.com>]

As such, GLLAMM sees data as flat. There's a country-specific effect,
but that's it. It is essentially a random effect only model. Putting
something like AR(1) structure in the residuals is extremely
cumbersome.

On Thu, Jun 9, 2011 at 3:09 PM, Wayne Sandholtz <wayne.sandholtz@uci.edu> wrote:
> I have panel data for about 140 countries, with yearly observations for each
> country.  The panels are unbalanced (more years for some countries than
> others), but for each country the years are continuous (no missing years in
> the middle).  The dependent variable is ordinal (1 - 5).
>
> I am using GLLAMM to take into account the within-group dependence of
> observations.  My question is, does GLLAMM take into account the
> longitudinal nature of the within-group data?  Or do I need to include a
> time variable in the model in order for the estimation to take into account
> the time-ordered nature of the observations?
>
> Or is there another model that I should consider?
>
> Many thanks for any help,
>
> Wayne Sandholtz
> *
> *   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/
>



-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.

*
*   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/