Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Multi-level models with time-dependent covariates at multple levels


From   Joel Miller <dr.joel.miller@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Multi-level models with time-dependent covariates at multple levels
Date   Wed, 28 Oct 2009 23:39:24 -0400

Hi Statalist colleagues,

I'm struggling to think through a multi-level model specification. It
would be great if somebody could help clarify my thinking.

I want to run a three level model. Level 1 would be a series of
repeated measures for a subject, Level 2 would be the subject, and
Level 3 would be the neighborhood in which the subject lives.
Obviously, there are multiple subjects per neighborhood.

Here's my conceptual problem. I have time dependent variables,
measured for each individual, that obviously fit at level 1 of the
model. However, I also have time dependent variables measured at Level
3 - the neighborhood. So the neighborhood has different characterstics
at T1, T2, T3 etc.

Can conventional multilevel modeling (e.g. with GLLAMM) deal with
this? How? I'm guessing it's NOT ok to just treat the neighborhood
level variation as if it was a Level 1 variable.

Any help gratefully received!

Thanks in advance

Prof Joel Miller
School of Criminal Justice
Rutgers University
*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index