I have conducted psm using kernel density matching. Since this produces multiple matches i need to adjust for non-independence as here: https://www.google.com/url?sa=t&sour...=1563368902329. I'm not at my computer so i can't refresh my memory, but as i recall, altvariance is not available for kernel density matching. I can't remember if ai(int) is, but even if it is the integer specification cannot be a constant since the number of matches varies. Any ideas on how to adjust the SEs? Bootstrapping seems to be contraindicated.