Hello,
I have a general doubt about PSM. The analyses of the graph (using teffects psmatch) are much better comparing before and after matching, the trend of the curves (treatment and control) are similar. But only until the propensity score is equal 0.8. This is happening for my 5 different years data.
Is it possible to use or somehow delimit the use of the propensity score from 0 to 0.8 so that I have a better matching?
What would you suggest? What can be the problem of this part of density (from 0.8 to 1)?
Thanks in advance,
Jacque
Related Posts with Delimit Propensity Score to obtain a better matching
Help with re-formatting a dataset for choice modelling?Hi folks. I have data in the following format. Each observation represents a zip code, n_sites denot…
Test for differences between Cumulative Incidence Function curvesHello, I have a two-arm trial where an event can either occur or does not occur because of a compet…
Accounting for imperfect diagnostic test results on data from an entire populationHello, I have data from a seroprevalence study which collected samples from an entire population (a…
Correct estimation of panel dataHello, I am making model to analyze migration from third countries (together 20) to Slovakia, Hunga…
Multiple twoway lines in same graph region.Dear All, I have a 50 US state panel by day. A policy of interest is implemented in the different s…
Subscribe to:
Post Comments (Atom)
0 Response to Delimit Propensity Score to obtain a better matching
Post a Comment