The help file says that -ssm- is a wrapper for -gllamm-, and -command-
option of -ssm- will produce the -gllamm- syntax that has been used.
You would need to take that syntax and change it to your needs. I
don't know if -gllamm- is fully -svy-compatible (in the sense of
supporting say -subpop- option which you might need in place of the
-if- qualifier), but certainly gets close. In all likelihood, the
error message is caused by the need of -gllamm- to have a weight
variable for each level in your model, and you only have it for level
1 units there. But you would more likely want to use that -gllamm-
syntax and tailor it to your sample selection needs.
On 1/19/09, Erica Ryu <[email protected]> wrote:
> Hello,
>
> I am trying to do a logit model with sample selection and am using the
> following command (since I cannot use the svy prefix with ssm, I use "if"
> and pweight):
>
> ssm yvar x1 x2 x3 if prot==1 [pweight=pawt2], s(frd=participation
> rtrad_white ca_8 q7b) q(8) fam(bin) link (logit) sel
>
> which then produces an "invalid" error message. I think there is something
> wrong with the way I included the pweight, because when I take out
> [pweight=pawt2], the model runs fine.
>
> Please advise. Thank you!
>
> Erica
> *
> * 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/
