Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Multilevel selection models using gllamm

From   Sophia Rabe-Hesketh <[email protected]>
To   [email protected]
Subject   Re: st: Multilevel selection models using gllamm
Date   Wed, 02 Jun 2004 17:38:23 -0700

Azman Shahadan,

gllamm has changed a bit since then. The equation name
for the random effects standard deviation is now id1_1,
so you should use

constraint define 1 [id1_1]i1=1

The best way to find out equation names is to use
the -trace- option in gllamm to get the
"General model information". You get something
like this:

***level 2 (id) equation(s):
   (1 random effect(s))

   lambdas for random effect 1
   id1_1l: i2
   standard deviation for random effect 1
   id1_1 : i1

The bit before the colon is the equation name
and the bit after is the column name.

You can also look at the equation and column names
of the parameter vector (shown in output with
the -trace- option), e.g.

         diet:      lns1:    id1_1l:     id1_1:
        _cons      _cons         i2         i1
r1   .5512129  -.6977461        1.1         .5


[email protected] wrote:

I am trying to use gllamm for multilevel Heckman selection model, but I am stuck which I have problem to specify the constraints;
I use command;
Constraint define 1 [id1]il=1 (as stated in the note by Sophia) but stata not recognise the variable id1.

Could anyone help me, I have note of stata user group meeting at Maastricht, but I could not understand how it work.

Azman Shahadan
Centre for applied statistics
Lancaster university
* For searches and help try:

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index