Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Philip Ender <ender97@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: Stumped...xtmixed and ANOVA F-stats not agreeing for balanced design |
| Date | Thu, 5 May 2011 09:00:05 -0700 |
Jared Saletin <jsaletin@berkeley.edu> wrote that his -anova- model and -xtmixed- model for balanced data did not produce the same results. Jared, I don't think that your -anova- and -xtmixed- models are parameterized the same. The -anova- model has three error terms: s#a, s#b and Residual. While the -xtmixed- model has four random effects: s:_cons, s:R.a, s:R.b and Residual. If you use the -var- option with -xtmixed- you should be able to lline up the mean squares from the -anova- with the random effects from the -xtmixed-. -- Phil Ender UCLA Statistical Consulting Group * * 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/