Dear Statalist,
I have 4 treatment groups and 1 control group in my dataset. This gives me 5 different groups I need to match across.
The traditional psmatch2 and teffects psmatch do not seem to work for multiple treatments. I read from another forum that teffects ra / teffects ipw / teffects ipwra work for multilevel treatments, but from there, how can I proceed with matching variables between different groups?
I also tried cem package, but the code takes forever to execute. I ran the code yesterday evening and discovered that the code was not executed until this morning. It could be that the datasize is over 4.4 million. Even after testing with matching on only one variable (with cem x1, treatment(groups)), the code is not being executed.
Any help would be really appreciated. Please help!
I have 4 treatment groups and 1 control group in my dataset. This gives me 5 different groups I need to match across.
The traditional psmatch2 and teffects psmatch do not seem to work for multiple treatments. I read from another forum that teffects ra / teffects ipw / teffects ipwra work for multilevel treatments, but from there, how can I proceed with matching variables between different groups?
I also tried cem package, but the code takes forever to execute. I ran the code yesterday evening and discovered that the code was not executed until this morning. It could be that the datasize is over 4.4 million. Even after testing with matching on only one variable (with cem x1, treatment(groups)), the code is not being executed.
Any help would be really appreciated. Please help!
