Page 1 of 1

Mediation other than MedMod package

Posted: Fri Mar 19, 2021 1:04 am
by jiwonkim
Hello,
I have been trying to check if there is a mediation effect on my dataset.
At first, I tried to do it using linear regression. (check if there is a direct effect between dependent variable and independent variable. And then if there is a direct effect between mediator and independent, and then total effects using three variables..)
At that time, I could see there was mediation effect. But when I used the MedMod package, it turned out there was no mediation effect.
My questions are..
1) Is there a way to legitimately check the mediation effect besides MedMod package?
2) If so, how?
3) Can anyone pinpoint what I have been doing wrong, if I have?
Thank you very much!

Warm regards,
JK

Re: Mediation other than MedMod package

Posted: Fri Mar 19, 2021 12:25 pm
by mcfanda@gmail.com
Hi
you can try jAMM in the jamovi library, but the results will be the same as MedMod for a three-variable model. The question is: How do you establish mediation with the three regressions approach and what are the differences you see with the medmod module?