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 at the end of May, and its replacement, statalist.org is already up and running.


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

st: gllamm level 1 variance in dyadic analysis


From   Jennifer Collins <collins.969@buckeyemail.osu.edu>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   st: gllamm level 1 variance in dyadic analysis
Date   Thu, 21 Mar 2013 20:42:42 +0000

Hello,

I am using Stata/SE 12.1 for Windows (64 bit)

I am using gllamm to run a 2-level model on dyads in longitudinal survey data where the dependent variable has a Poisson distribution.  The responses of the individuals in the dyad are Level 1 units nested within the level 2 unit (the dyad).  In the method I am using, the level 1 part of the model including 2 dummy coded variables that indicate the unique intercepts for person 1 and person 2.  The level 1 model also includes a unique "time" variable (coded as 0, 1, 2) for each person, so that each person in the dyad can have their own trajectory of change.  All the other variables are level 2 variables to predict the said intercepts and time trajectories.  I have run these kinds of models in HLM, but am relatively new to Stata.  My code is below.  I want to know the variance for the two intercepts and two time variables, and if the variance is significant.

eq het: person1int person2int person1time person2time

gllamm depvariable person1int person2int person1time person2time var1 var2 var3 var4 var5......, i(dyad id) noconstant  family(poisson) nip(27) pweight(weight) s(het) link(log) adapt robust cluster(clustervar) eform

gllamm immediately states "s() option ignored because families do not include dispersion parameters"

If I run this code without specifying family(poisson), I eventually get results, and am then able to run "display exp(2*(lns1))person1int)" and "nlcom (s_person1int: exp(2*(lns1)person1int))" to get the level 1 variance and its significance.  PBut by not specifying the poisson family, gllamm doesn't know that my DV has a poisson distribution, right?  How can I get the level 1 variance for the 2 intercepts and 2 time variables for a Poisson distribution? I would like to avoid a 3-level model.

Thank you for your consideration.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


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