I am trying to run a 'conditional mediation' in Jamovi, and I normally run it via PROCESS 3.4 in SPSS.
In PROCESS, there is an option to get the pairwise contrasts, allowing you to compare the differences for each level of my moderator (I have a categorical moderator with 3 levels).
Does anybody have any suggestions how to do that? Apart from having to create extra dummies for each comparison, which takes time?
Thanks in advance.
Jan
Moderated mediation with categorical moderator
Re: Moderated mediation with categorical moderator
is this something that jAMM can do?
jonathon
jonathon
Re: Moderated mediation with categorical moderator
Yes, I use the add-on. But I can't find the option to calculate pairwise comparisons.
- mcfanda@gmail.com
- Posts: 462
- Joined: Thu Mar 23, 2017 9:24 pm
Re: Moderated mediation with categorical moderator
Hi
in jAMM you get the "Moderation effects (interactions)" table that show you the K-1 contrasts needed to test the overall interaction. Thus, If you inspect that table, you'll see K-1 interactions that inform about which difference is significant and which is not.
Assume you have X , M and Y and a moderarator Q with three groups. In the table you'll see
X:Q12->M (this tells you if the effect of X on M is difference between Group 1 and 2 of Q)
X:Q13->M (this tells you if the effect of X on M is difference between Group 1 and 3 of Q)
M:Q12->Y (this tells you if the effect of M on Y is difference between Group 1 and 2 of Q)
M:Q13->Y (this tells you if the effect of M on Y is difference between Group 1 and 3 of Q)
Knowing this, you can interpret which groups differs in the mediated effects.
There is no multiple comparisons, we do not feel it's a good idea to run post-hoc tests on mediated effects, mainly because there's almost no literature supporting this analysis. However, we may implement it in the future.
in jAMM you get the "Moderation effects (interactions)" table that show you the K-1 contrasts needed to test the overall interaction. Thus, If you inspect that table, you'll see K-1 interactions that inform about which difference is significant and which is not.
Assume you have X , M and Y and a moderarator Q with three groups. In the table you'll see
X:Q12->M (this tells you if the effect of X on M is difference between Group 1 and 2 of Q)
X:Q13->M (this tells you if the effect of X on M is difference between Group 1 and 3 of Q)
M:Q12->Y (this tells you if the effect of M on Y is difference between Group 1 and 2 of Q)
M:Q13->Y (this tells you if the effect of M on Y is difference between Group 1 and 3 of Q)
Knowing this, you can interpret which groups differs in the mediated effects.
There is no multiple comparisons, we do not feel it's a good idea to run post-hoc tests on mediated effects, mainly because there's almost no literature supporting this analysis. However, we may implement it in the future.