Re: st: basic question about multi-level data

[Rose Medeiros <rosem@cisunix.unh.edu>]

Reading information on xtmixed (note that xtmixed is new to Stata 9) or 
gllamm will show you how to estimate the models, but very little about 
whether you should estimate them (i.e. are they appropriate, the 
assumptions, etc.). Stata press does have a book titled "Multilevel and 
Longitudinal Modeling Using Stata" which might be of some help (I 
haven't actually seen the book). You might also refer to "Hierarchical 
Linear Modeling" by Raudenbush and Byrk, or "Multilevel Modeling" by J. 
Hox .

Richard Sherman wrote:


> Dear Statalist,
>
> A student of mine is working with a type of data with which I have no  
> experience. The key feature of the data is that the unit of  
> observation/analysis is multi-level, with some observations measured  
> at the level of specific neighborhoods, others at a region larger  
> than a neighborhood, and some at the whole city (in this case  
> Baghdad). That is, the same variables are measured over different units.
>
> Ordinarily when I get close to this type of data, I start moving in  
> the other direction without delay. But I know vaguely of the  
> existence of multi-level techniques, and I thought I'd ask you all  
> for where in Stata I should be looking to start finding out about how  
> and whether to do this type of analysis.  Thanks for your help,
>
> Richard
>
> -------------------------------------------------
> Richard Sherman
> Department of Political Science, Leiden University
> Leiden, Netherlands
> http://homepage.mac.com/richard.sherman
>
> *
> *   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/
*
*   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/