Re: st: comparing two mim gllamm logistic models

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

--- On Fri, 4/6/10, jl591164@albany.edu wrote:
> I use -mim, cat(fit):gllamm- to fit three-level logistic
> regression models. Now i need to compare the model fit of two nested
> models, e.g. randome intercept models with and without a quadratic term
> of time.

Create the time and time^2 using -orthpoly-, than include those variables
in your model, the test for the "square" term gives you the test whether
or not the quadratic term improves the model fit. For more, see 
-help orthog-.

> Another example of two nested models is a randome intercept model and a
> randome intercept and random slope model. Do i compare the -2logL, AIC
> and BIC of the two models of concern? If so, how do i obtain these model
> fit statistics after fitting mim:gllamm models?

This is a very hard problem, even without -mim-, as this is a test of a 
hypothesis on the edge of the parameter space. See -help j_chibar-.

Hope this helps,
Maarten

--------------------------
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/