Dear all,
I have a question regarding evaluating multiple outcomes using psmatch2. Shall I run the psmatch2 separately for each of the outcome? or shall I use a general propensity score model??
I did a study using propensity score matching technique and I have 6 different outcomes for the same propensity score model (matched on the same variables). I ran psmatch2 separately for each outcome and generated kernel density plots of propensity score models before and after matching. I had nearly identical distributions on the outcomes—my question is how can the groups differ on the outcomes if we have almost identical distributions (after matching) for the 6 outcomes?
I have a question regarding evaluating multiple outcomes using psmatch2. Shall I run the psmatch2 separately for each of the outcome? or shall I use a general propensity score model??
I did a study using propensity score matching technique and I have 6 different outcomes for the same propensity score model (matched on the same variables). I ran psmatch2 separately for each outcome and generated kernel density plots of propensity score models before and after matching. I had nearly identical distributions on the outcomes—my question is how can the groups differ on the outcomes if we have almost identical distributions (after matching) for the 6 outcomes?
