Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Najib Mozahem <najib.mozahem@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: Covariance structures for xtmelogit with a single random effect |
Date | Mon, 9 Sep 2013 17:55:39 +0300 |
Hi, I am using xtmelogit on my dataset. My model has only a random intercept because a likelihood test revealed that random coefficients do not contribute significantly to the model. I want to try out several covariance structures but whenever I try any structure I get the following: xtmelogit span1 i.type i.freq pages totalactive c.age##c.age c.time##c.time || id:, cov(unstructured) mle Note: single-variable random-effects specification in id equation; covariance structure set to identity On page 300 of "Multilevel and Longitudinal Modeling using stata" by Rabe-hesketh and Skrondal they fit a xtmixed model with just a random intercept using the residuals() option in order to specify an unstructured matrix, but this option does not exist for xtmelogit. Is there a way to try several covariance structures using xtmelogit? Thanks Najib * * 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/