Re: st: xtmixed:comparing strengths of coefficients

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Tue, 9/3/10, saurav pathak wrote:
> I am using xtmixed in a multilevel analysis. Level 1 is
> individual level and Level 2 in social reference group
> level. In my model I have level-1 variables as well as
> level-2 variables. I have hypothesized that level-2
> variables will have a stronger effect than level-1
> variables on the dependant variable.  <snip>
> How can I say (statistically) looking at level-2 varaible's
> coefficient that they have a stronger effect? The results 
> that I get do show greater magnitude of level-2 coefficients,
> but they are not standardized coeff. 

If the two variables have the same unit, then you can directly
compare the coefficients. For example it is quite common to
add the average of a individual level variable as a level-2 
variable, in which case the two variables have the same unit.
If not, you could standardize the variables you run your model,
this will make the units in some sense similar, i.e. in that 
case the unit of both variables will be standard errors.

Hope this helps,
Maarten

PS. It is usually discouraged to contact members of the list 
directly, see the Statalist FAQ point 5 for an explanation why:
<http://www.stata.com/support/faqs/res/statalist.html#private>

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   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/