In the output below, the ACME p value shows that it is significant. But the p-value for Prop. Mediated Average is >10%. I thought Prop. Mediate is the proportion of the total effect mediated by the mediating variable(cong_mesg). So I am confused why ACME is significant but not Prop. Mediated (even for control and treatment groups).Does this show evidence in support of mediation?

mediation ouput


1 Answer 1


This problem is a recurring question (with the same example) that has been treated elsewhere : here and here.

  • $\begingroup$ I think this is a different question that's really more about the difference between the indirect (mediating) effect and the proportion mediated, which conceptually are very similar. It's reasonable to be surprised that the p-values for them would differ when the tests seem to amount to the same thing conceptually. $\endgroup$
    – Noah
    Feb 27, 2021 at 8:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.