Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Alexandru Cojocaru <scojocaru@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: estimated dependent variable with xtmixed |
Date | Fri, 21 Jun 2013 09:21:06 -0400 |
Thanks, Jeph! Will try to look for friends with HLM connections. cheers, Sandu On Thu, Jun 20, 2013 at 8:33 PM, Jeph Herrin <stata@spandrel.net> wrote: > Unfortunately, xtmixed does not support these kind of weights. This model is > called "V-known" and the only way I have found to estimate it is to use the > software package HLM. I have heard that this model is equivalent to putting > SE(Y) in the model as a covariate, but I haven't been able to convince > myself of that. > > hth, > Jeph > > > > On 6/20/2013 5:17 PM, Alexandru Cojocaru wrote: >> >> I would like to estimate a model of the following form: >> >> Y_ij=X_ij*b+u_i+e_ij, >> where Y_ij is the municipal poverty rate (from a household survey) >> X_ij are municipal characteristics (from a population Census) >> u_i are the region level random effects (municipalities are nested >> within regions). >> >> With xtmixed, this would look like: xtmixed mun_pov X_ij || region: >> Since municipal poverty rates are sample estimates, they have >> corresponding sampling variances. How would one properly incorporate >> this uncertainty into the xtmixed setup? >> >> Any advice would be much appreciated. >> >> cheers, >> Sandu >> * >> * 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/ >> >> > * > * 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/ * * 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/