Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Maarten buis <maartenbuis@yahoo.co.uk> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: parallel process model in Stata |
Date | Tue, 1 Feb 2011 08:08:34 +0000 (GMT) |
--- On Tue, 1/2/11, Shubhabrata Mukherjee wrote: > Is there a any way to do a parallel process model in > Stata? > > I have variables y_a and y_b measured over 5 time > points. I can perform a mixed effects model on each of > the two variables separately. > > xtmixed y_a time || id: , cov(uns) > > xtmixed y_b time || id: , cov(uns) > > I want to see the effect of one on the other in one model. > That is, I am interested in finding out the effect of the > intercept for A on the slope of B and vice versa. Is there > any routine in Stata to do this? In such models you first need to worry about identification. Typically, you would need variables that influence y_a but not y_b and variables that influence y_b but not y_a in order to identify such bi-directional effects. Another strategy that can sometimes help is to let y_a[t] influence y_b[t+1] and y_b[t] influence y_a[t+1]. Regardless on how you try to identify such a model, you'll probably end up using -gllamm- for such a model. You can get it by typing in Stata -ssc install gllamm-, and lots of information on it is available from <http://www.gllamm.org>. Hope this helps, Maarten -------------------------- Maarten L. Buis Institut fuer Soziologie Universitaet Tuebingen Wilhelmstrasse 36 72074 Tuebingen Germany http://www.maartenbuis.nl -------------------------- * * 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/