I have an occupational cohort (n=866, cases=222) and would like to generate
cases and controls from the data. I'd like to have as many controls as
possible, with the two requirements:
1) The control must have been employed when the employee became a case (risk
set sampling)
2) The control must have started employement around the same time as the
case (up to 6 months before but not after).
It is also OK to use, as a control, someone who becomes a case at a later
time.
I have come across the command -sttocc- but it only takes care of the first
requirement. I do not know how to incorporate the second constraint into
the command.
Are there other stata commands that I'm not aware of that will be of help
here?
Also, I want to know whether -sttocc- samples the controls with replacement?
It seems to be the case sine I have set the option number(#) to 30, just to
see what happens and it was possible to generate 30 controls for each case -
which seems odd given the cohort sample size.
Is it possible that a subject can be a control for more than one case? Is
there anything wrong with this?
Any help will be greatly appreciated!!
Thank you,
Cornelia
_________________________________________________________________
Want to block unwanted pop-ups? Download the free MSN Toolbar now!
http://toolbar.msn.co.uk/
