I used psmatch2 to match cases and controls in a sample and now I am trying to get a standard mean differences for the matched sample. I don't know if I'm using the stddiff command correctly. Here's what I did:
Did I use the stddiff code correctly? I've tried comparing it to pstest and I'm receiving wildly divergent answers. (I can tell more, but I just wanted to make sure I got the code right first.)
Code:
psmatch2 cohort age gender, out(clinicperday) ties caliper(0.2) stddiff _pscore, by (cohort) abs
