st: nnmatch question

["Zhu, Carolyn" <carolyn.zhu@mssm.edu>]

Hello,

 

I am trying to use to estimate an average treatment effect with two groups using nnmatch. However, one group is more than 5 times larger than the other. How do the differences in sample size affect the estimates? Does it matter at all? How should one report on the sample size in nnmatch?

 

Thanks in advance for your replies.

 

Carolyn