I don't know.
Recently, however, Bauer (2003), Curran (2003), and Mehta and Neale (2005) have shown how multilevel models can be fitted as structural equation models by treating the lower level units as indicators for latent variables involved in a multilevel model. By treating the macrolevel units as
the units of analysis and the microlevel units as indicators, the micro-macro multilevel model can be turned into a structural equation model that can be fitted by any structural equation modeling software package.
Bauer, D. J. (2003). Estimating multilevel linear models as structural equation models. Journal of Educational and Behavioral Statistics, 28, 135�167.
Curran, P. J. (2003). Have multilevel models been structural equation models all along? Multivariate Behavioral Research, 38, 529�569.
Mehta, P. A., & Neale, M. C. (2005). People are variables too: Multilevel structural equations modeling. Psychological Methods, 10, 259�284.
Nicola
At 02.33 11/09/2007 -0400, Zhong Liu wrote:
>Dear Statalist:
>
>Could you please tell me whether xtmixed can be applied to a simultaneous
>equations system (including two equations) in a way like two-stage-least
>square method?
>
>That is to first regress the endogenous variable in the right-side against
>all exogenous variables in the two equations buy using xtmixed. The second
>step is to obtain the predicted values of the aforementioned endogenous
>variable. The third step is to regress the endogenous variable in the left
>side with the obtained predicted values at the second step, and the relevant
>exogenous variables in each equation by using xtmixed again.
>
>Thanks in advance,
>
>Zhong Liu
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
