Dear Statalisters,
Does svy imply that I have robust SE?
A relative risk in a simpel random sampling crossectional study using
Poisson regression is
calculated like this where vce(robust) takes care of distribution mismatch
poisson y x1 x2 x3 , irr vce(robust)
I have a crossectional survey that is stratified and weighted. The problem
is that i can´t use vce(robust) together with svy prefix. I think I read
somewhere that the difference between vce(robust) and svy is that the
latter do not include finite population correction. Is this true? Is it OK
to skip the vce(robust) and use svy (given that I can skip the fpc)? Where
do I find a literature reference for that?
Hopefully someone knows this
Thanks
/Tomas
*
* 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/