Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Mixed model: fixed intercept and random-coefficient


From   Trelle Sven <strelle@ctu.unibe.ch>
To   statalist@hsphsun2.harvard.edu
Subject   st: Mixed model: fixed intercept and random-coefficient
Date   Mon, 14 Dec 2009 22:38:05 +0100

Dear all,
I want to fit a random-coefficient/slope model with a binary outcome
(logit) for a meta-analysis. We tried to use xtmelogit. However,
xtmelogit seems to automatically assume a random intercept whereas we
want the intercept to be a fixed effect. Does anybody know whether it is
possible to use xtmelogit/poisson with a random-coefficient only and if
yes, how this is specified (we were not able to find anything on the
web/statalist archive)?

We also tried two different approaches in GLLAMM which gave quite
different results. The first approach is shown below:
 
Approach based on gllamm manual (example 9.5):
	gen cons=1
	eq int: cons
	eq slope: treatment
	gllamm success treatment, i(id) nrf(2) eqs(int slope) l(logit)
(binom) weight(wt) adapt nocor trace allc
 
However, this approach also includes a random intercept. Does anybody
know whether it is possible to specify a fixed-intercept and
random-coefficient model in gllamm (the noconstant option doesn't seem
right because it gave implausible results)?

Any help is appreciated
Thanks in advance
Sven



*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index