Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: How do I interpret random coefficient parameters (SD) using xtmixed


From   Owen Gallupe <ogallupe@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: How do I interpret random coefficient parameters (SD) using xtmixed
Date   Sun, 17 Jul 2011 14:43:53 -0700

Hi,

I'm no expert on the subject, but this is my understanding. Unless I'm
mistaken, those are standard deviations across the different countries
in your analysis. This tells you whether or not there is much
difference in your variables across countries. So, sd(livsit1) = 3.346
tells you that the standard deviation across countries is 3.346 on
livsit1...it's up to you to interpret whether or not this is large.

sd(_cons) is the standard deviation of the intercept across countries.
If this is substantially large, it justifies running a random effects
model.

By using an unstructured covariance matrix [if appropriate - xtmixed
w_dur livsit1 || country: livsit1, mle cov(un)], you would also get
the correlation between the random intercepts and random slopes (in
your case, between sd(livsit1) and sd(_cons)) which can be
informative. For example, if corr(livsit1, _cons) is positive, that
would mean that countries with a higher intercept (higher mean scores
on the dependent variable) also tend to have a steeper slope on
livsit1. That is, the effect of livsit1 on w_dur tends to be higher
when the country has a higher level of livsit1.

A quick search turned up this page which might be useful:
http://dss.princeton.edu/training/Multilevel101.pdf

Hope this helps.

Owen



On Sat, Jul 16, 2011 at 1:06 PM, Zachary Zimmer
<zachary.zimmer@ipia.utah.edu> wrote:
> I am running random coefficient models using xtmixed assuming there are different intercepts and slopes for variables across different countries.
>
> For instance, using the STATA command:
> . xtmixed w_dur livsit1 || country: livsit1, mle
> the intercept and slope of livsit1 when estimating the dependent variable w_dur for is allowed to vary across county.
>
> The Random-effects Parameters are:
>
> sd(livsit1) = 3.346
> sd(_cons) = 1.889
> sd(Residual) = 28.601
>
> Can someone tell me how specifically to interpret these parameters?  What does the 3.346 mean?  Etc.
>
> Thanks for any help.
>
>
> Zachary Zimmer
> Professor, Department of Sociology
> Senior Scholar, Institute of Public and International Affairs
> University of Utah
> *
> *   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/
>

*
*   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   |   Site index