st: statistical test for interactions in svyregress models

[Michelle Garrison <garrison@u.washington.edu>]

Hi -- I'm looking for a way to statistically test for interactions 
(effect modification) in a svyregress model.  If I was just using 
regress, I'd just use lrtest to compare nested models with and without 
the interaction terms -- but since lrtest can't be used with svy 
commands, I'm not sure how to go about this. Is there a way in stata to 
get this information from a Wald test?

If it makes a difference, the interaction of interest is between a 
categorical variable with 3 values (race) and an ordered variable with 
4. Race also happens to be the one that the survey weights are based on.

Comparing the results of regressions stratified by race, it seems pretty 
clear that interaction is occuring, but we'd like to have a statistical 
test we can cite as saying so.

Thanks in advance for any help people have!


--
___________________________________________________

 Michelle Garrison, M.P.H. -- Research Consultant
 Child Health Institute, University of Washington
 office: (206) 616-1203    fax: (206) 616-4623
 campus mailbox 354920
 e-mail: garrison@u.washington.edu 
___________________________________________________

*
*   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/