Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: multilevel analysis: cross-level interactions in gllamm

From   Sophia Rabe-Hesketh <>
Subject   Re: st: multilevel analysis: cross-level interactions in gllamm
Date   Tue, 26 Nov 2002 13:42:16 +0000

Hi Gerben,

By cross-level interactions, do you mean for example an
interaction between var2 and var3?
In that case, you need to form a new variable equal 
to the product of the two and include that in the model:

gen var23 = var2*var3
gllamm var1 var2 var3 var23, i(group region) offset(ln_practice_size) allc

It's probably better if you use the 'adapt' option
for continuous responses.

Best wishes,


At 01:34 PM 11/26/02 +0100, you wrote:
>I used the gllamm command for a multilevel analysis in which I try to 
>explain why some general practitioners (GP) order more diagnostic tests 
>than others. There are explanatory variables at the level of single 
>general practitioners, GP-groups in which they collaborate, and regions 
>in which these groups work and order tests with a one diagnostic 
>facility per region. So far I kept it simple:
>gllamm var1 var2 var3, i(group region) offset(ln_practice_size) allc
>,where var1 is a variable at the GP level, var2 a variable at the group 
>level, and var3 a variable at the region level. How do I implement 
>cross-level interactions in this model, that is by using which syntax?
>Gerben ter Riet, MD PhD, Academic Medical Center
>Dept. General Practice, Room J3-354, 1105 AZ Amsterdam
>Tel. ++31 (0)20 566 4640; Fax: ++31 (0)20 566 9186 (Mo thru' Wed)
>Wiss. Mitarbeiter, Horten Zentrum, Univ Zurich; (Fri)
>Hon. Sen. Res. Fellow, Univ. Birmingham (UK)
>*   For searches and help try:
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index