What is the stata command for Epanechnikov's kernel matching?

[Shoaib khan]

Dear Statalist,

I want to use propensity score matching to match the treated with the control. I am using Kernel matching. I want to use Epanechnikov's kernel function and 0.05 as a bandwidth.
Please anybody could tell the kernel matching method command and some following method for kernel matching?
I would be highly grateful to you.
Thanks.

[Marcos Almeida]

Welcome to the Stata Forum/Statalist.

I gather you're talking about the user-written psmatch2.

Being this so, you may try to add in the options: kernel k(epan) bwidth(0.05).

I have more experience with teffects, rather than psmatch2.

Hopefully it helps.

[Shoaib khan]

Sir,
Thank you very much for your kind reply. I used the following command : psmatch2 t x1 x2, out(y)
I guess this command is only for Propensity score matchning method and not for kernel matching method. Please could you write the complete command for kernel matching.
Thanks a lot .

Regards,
Shoaib

[Marcos Almeida]

You may find examples here.

[Shoaib khan]

Thank you Marcos for sharing this information. I have questions regarding K nearest neighbor matchning.
when I use stata this command db psmatch2 ......There are several options are opening which i don't know what i should put their values for example : # neighbor and # of matches. My question is which number i should give to neighbor option and number of matches. In option variable list, should i include the dummy variables as well or only independent variables? In caliper option, which value is should put in this option?

Thanks a lot.

Regards,
Shaoib