Thanks for the very useful package! In my analysis I am using a model averaging approach, where I calculated average coefficient estimates from equally supported models using the model.avg from MuMIn. This function returns an object of the class "averaging". Is there a straight-forward way to adapt or_glm to work on this type of object?
I understand if this is beyond the scope of the package. If you feel this to be statistically inappropriate, an explanation of why would be helpful. To the best of my understanding, it would be reasonable to interpret ORs based on averaged coefficients as we interpret the averaged log-odds.
Thanks for the very useful package! In my analysis I am using a model averaging approach, where I calculated average coefficient estimates from equally supported models using the
model.avg
from MuMIn. This function returns an object of the class "averaging". Is there a straight-forward way to adaptor_glm
to work on this type of object?I understand if this is beyond the scope of the package. If you feel this to be statistically inappropriate, an explanation of why would be helpful. To the best of my understanding, it would be reasonable to interpret ORs based on averaged coefficients as we interpret the averaged log-odds.