Re: st: Non-nested models and LRTEST

[Richard Williams <Richard.A.Williams.5@ND.edu>]

At 05:52 AM 8/25/2006, Marcello Pagano wrote:
> Rho YH wrote:
>
> >If I make a logistic model 1) A = B C D E F G and another model
> >2) A = B C D E F H I , which are not necessarily nested each other
> >(but one has more variables than the other), are the results of
> >LRTEST still relevant ?
> >For example, if LRTEST 1) 2) is significant (p<0.05), then is model
> >2) better explaining than model 1)?
> >Do my ideas make sense?
If the models aren't nested you shouldn't do a chi-square contrast 
between them.  Consider using a BIC test instead, which you can get 
by adding the stats option, e.g.

lrtest m1 m2, stats

There is a brief discussion of BIC and AIC measures on pp. 4-6 of

http://www.nd.edu/~rwilliam/xsoc694/x05.pdf


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX:    (574)288-4373
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW (personal):    http://www.nd.edu/~rwilliam
WWW (department):    http://www.nd.edu/~soc 
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/