Hi Matt,
gllamm6 is now very out-of-date. The current version is called gllamm
and can be obtained using
ssc install gllamm
see http://www.gllamm.org
The syntax seems fine. In the expand() option, the second argument
is the dummy that indicates which alternative has been selected
(see also gllamm manual, Chapter 9, availab le at above web site).
Sophia
On Fri, 30 Jul 2004 12:42:37 +0100
"Matt Dickson, CMPO" <[email protected]> wrote:
Hi,
I am trying to ascertain whether I can estimate a mixed multinomial logit model using gllamm6
using panel data and want to restrict the parameters to be equal across equations while allowing
for 3 random parameters. I think the syntax below is correct but am not sure and cannot see where
it is possible to change the distribution of the random parameters to be other than normal? Could
anyone offer any help?
Outcome here is the outcome variable, regressors are our x variables, pid is the individual
identifier:
gllamm6 <outcome> <regressors>, i(pid) nrf(3)-[we have three random parameters] eqs(the
variables that have random coefficients) family(binomial) link(mlogit) expanded(response options
o)
Any help would be very much appreciated.
Matt
----------------------
Matt Dickson
Leverhulme Centre for Market and Public Organisation
Tel: 0117 9289061
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
