I'm interested in running a moderated mediated model with Hayes' PROCESS to test the hypothesis that hunger (x) influences healthy choices (y), and how this relationship is mediated by self-control (m). Particularly, I'm interested in testing how the link between self-control and healthy choices could be moderated by the presence of different kinds of contextual cues (no cues, social proof, visual cue). I've seen dichotomized moderators before but not sure how to proceed when my moderator has 3 levels. Any advice would be appreciated!
Dear Tracy,
Try serial multipel mediation
http://core.ecu.edu/psyc/wuenschk/MV/multReg/Serial_Multiple_Mediation6.pdf
I've had a look in Hayes book and while he describes a dichotomous categorical variable as the moderator in a mediated relationship he does not appear to rule out the possibility of using a multi level moderator. Perhaps it's worth trying in PROCESS and checking your output (using bootstrap sampling) and see what you get? It should be model 59.
- Badges
- Science topic
- Similar topics
- Psychology
- Social Psychology