search

insongkim / PanelMatch

118 stars 36 forks source link

Applying PanelMatch to binary and ordered categorial variables #101

Closed LuMesserschmidt closed 2 years ago

LuMesserschmidt commented 2 years ago

Dear @adamrauh and @insongkim,

I just want to be sure to not contradict any of PanelMatch´s assumptions: Is it possible to have a binary outcome variable instead of a continuous one? And although I doubt it, can your model be easily extended to handle ordered categorial DVs?

Thank you!

Luca

insongkim commented 2 years ago

Yes, you can apply it to binary outcomes: the effects should be interpreted on the probability scale accordingly.