Hello,

I have been trying to test a moderated mediation model with two moderators using lavaan package in R. I read some of the answers about moderated mediation with SEM, but I still have a vague idea about what paths my model should specify.

Can anyone confirm if my model is sufficient to test a moderated mediation?

All variables are continuous.

Below is my model (Y: DV, X: IV, Med: mediator, Mod: moderators, Con: control variables):

# Direct Effect

Y ~ a1*X + a2*Mod1 + a3*Mod2 + a4*Con1 + a5*Con2 + a6*XMod1 + a7*XMod2 + a8*Mod1Mod2 + a9*XMod1Mod2

Med ~ b1*X + b2*Mod1 + b3*Mod2 + b4*Con1 + b5*Con2 + b6*XMod1 + b7*XMod2 + b8*Mod1Mod2 + b9*XMod1Mod2

Y ~ b*Med 

# Indirect Effect

ab := (a1 + a6 + a7 + a8 + a9)*b

# Total Effect

total := (a1 + a6 + a7 + a8 + a9) + ab

Thank you.

Similar questions and discussions