Ch 9, page 10

Ch 9, page 10

by Jean Digitale -

"Because the number of treated subjects often far exceeds the number of outcomes, it may be possible to control for many more potential confounders in a propensity score model than in a model for outcome."

Here, does the number of outcomes mean the number of subjects with the outcome?

Thanks,
Jean

Re: Ch 9, page 10

by Michael Kohn -

Yes.  The assumption here is that the outcome is dichotomous (binary), each subject either has the outcome or doesn't.  So, the number of outcomes equals the number of subjects with the outcome.