Dear All,
I was trying to test the sensitivity of treatment effect due to selection on unobserved variables (hidden bias). I have a continuous response (outcome) variable. I checked the 'rbounds' command. However, it implements sensitivity tests for matched (1x1) pairs only. In my case the treated and control group are not matched one to one (see below). In other words the matrix is not (1x1). Is there other command that I can use in such circumstances.
Thanks,
Menale
ATT estimation with Nearest Neighbor Matching method
(random draw version)
Bootstrapped standard errors
---------------------------------------------------------
n. treat. n. contr. ATT Std. Err. t
---------------------------------------------------------
370 180 874.543 262.418 3.333
---------------------------------------------------------
Note: the numbers of treated and controls refer to actual
nearest neighbour matches
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
