Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

st: running a multilevel multinomial using gllamm


From   "Lima, Julie" <julie_lima@brown.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: running a multilevel multinomial using gllamm
Date   Wed, 20 Jul 2011 13:03:11 -0500

Hello,

I'm new to the list, so I apologize if this has already come up!   Any
advice would be appreciated.

We have nested data of 37,068 decedents within 2451 hospices within 9
regions and would like to run a random effects multilevel multinomial
regression model on an outcome with 4 non-ordered categories.  We have
covariates on the person and hospice levels, but none at the region
level.
The following 3-level code is currently running (so far without error)
on my system, but I was hoping to get some advice as to whether or not
I specified the model correctly.

#delimit ;
gllamm dem_deb_ftt prop_catr1
agecat_redo3 male race_3cat2
race_3cat3 xxcps6_nbf dxchf dxemcopd dxarthr any_cancer_nbf
 phbedft_nbf nh_shortstay  l_mp2885_3cat_1 l_mp2890_3cat_3
l_mp2850m_cohort fyodge2003
fisc_int1 fisc_int2 fisc_int4, i(l_hosp regionv2)
link(mlogit) family(binom)
 basecat(1) adapt trace;


Also, in my reading, I noticed that one could also run a 2-level
(person, hospice) model and then cluster hospices within regions.
Given that we do not have any covariates at the region level - would
this be more appropriate??

Thanks much,
Julie

--
*************************************
Julie Lima, MPH, PhD
Senior Research Analyst
Brown University
Center for Gerontology and Health Care Research
Providence, RI

*
*   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   |   Site index