In general an -if- condition in Stata means
that all observations not meeting
that condition are ignored, absolutely.
There is no question of anything superseding anything
else, or vice versa.
Nick
[email protected]
Hixson, Eric
> I am using the -robust cluster()- option with -logit- to
> address lack of independence of observations. My question
> has to do with how Stata performs the operation when an -if-
> statement is also included.
>
> I have a dataset composed of subjects (id) that may have
> multiple observations. For each observation, the subject
> could have one of two levels of a certain characteristic
> (type = 0/1).
>
> If I specify "logit outcome var2 if type==1, robust
> cluster(id)", would observations for a given subject where
> type==0 be included in the determination of the robust
> standard errors [per the -cluster(id) statement] or would the
> sample first be limited by the -if type==1- before any
> modeling is done?
>
> In other words, does -if- supercede -cluster()- or visa
> versa? If this should be obvious, go easy on the caning and
> I'll keep reading!
*
* 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/
