Different ATT results with outcomes coded as 0,1 and 1,2.

[Gayatri Krishna]

Hi all,
While running PSM with teffects psmatch in stata, I am getting different ATT when the exposure variable is coded as 0,1(No, Yes) and 1,2 (Yes, No). Why is it so? The same is happening with outcome variable too. How to resolve this?
Thanks in advance.

[Øyvind Snilsberg]

per the help file, the exposure variable should be coded as 0,1.

[Gayatri Krishna]

Thank you Sir.