Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Nasos Roussias <nasos.roussias@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: Random Slopes and their Standard Errors with xtmixed |
Date | Wed, 28 Jul 2010 11:27:37 +0300 |
Dear List Members, I have a question about calculating 2nd-level random slopes and their significance levels for a multilevel model. I am calculating a two-level random slopes model using xtmixed with Stata 11.1. I allow one variable's effects to vary at the 2nd level (country-level). For this variable, i want to get the random slopes and their corresponding standard errors for each of the 2nd-level countries, so i can calculate whether each of the country slopes is significant or not. I can get the coefficient for each country slope by using the /predict ,reffects/ command and adding each country's random effects to the grand slope. I can also estimate whether each of the BLUPs themselves are significantly different from the grand slope using the BLUPs standard errors from the /predict ,reses/ command. However, i have not been able to figure how to combine the information from the BLUPs standard errors and the grand slope standard errors to get an estimate of each country's slope standard error. Can anyone help? Is there some canned command that does this? If not, can someone tell me how to calculate it manually? Thank you for your help Nasos Roussias * * 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/