Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Stas Kolenikov <skolenik@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: xtmelogit vs hlm |
Date | Tue, 16 Mar 2010 14:02:06 -0500 |
On Tue, Mar 16, 2010 at 6:32 AM, Neal Beck <nealinmadrid@gmail.com> wrote: > > As OP, turns out that on RTFMing I can easily get HLM and xtmelogit to within 10%, quite good enough for govt work given the differing numerical methods. Why Stata thinks that I might (as a default) want to estimate a mixed model with the random coefficient centered on 0 (that is, no fixed component entered automatically) is well beyond me. But not reading the FM is not a good excuse for assuming that all Stata syntax looks like R syntax (mutatis mutandis). Can you propose any good default value for a random effect other than zero??? You can suggest the developers to issue an error message, which is probably quite proper. A random effect without a fixed effect is somewhat weird, but you as a user might have a different idea. If R defaults to populating the fixed effects that way, then you'd really have to work around that to force them to be centered at zero. I'd be curious to learn more about where the mismatch was. -- Stas Kolenikov, also found at http://stas.kolenikov.name Small print: I use this email account for mailing lists only. * * 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/