I am conducting a moderated mediation analysis using an experiment. There are six established mediators (measured using 3 items each) that I want to test. Participants will be randomly divided into 3 groups (given different stimuli) and further divided into 2 groups (low v/s high) in each group. I want to test the interactions on the mediators. Further the six mediators will be regressed onto the dependent variable. If you are familiar with Hayes Process models, then this would be the Hayes process model 8.

I am struggling to figure out the minimum sample size in each of the three groups (assuming participants in each group will be equally divided between the low and high conditions) that will provide sufficient power (80%).