When I run PROCESS model 5 I have no significant interaction effect for condition x parasocial relationship on attitude towards the ad (p = 0.453):
- Process model 5.
- Y = attitude towards the ad
- X = Condition
- Mediator(s): Host credibility, Ad credibility
- Moderator variable W: Parasocial relationship.
Out of curiosity, I had also conducted MANOVA and here I did get a significant interaction effect for condition x parasocial relationship on attitude towards the advertisement (p = 0.027):
- Y = Attitude toward brand , Attitude toward the ad
- Fixed factors = condition and parasocial relationship.
Does anyone know how this is possible? Actually, you are doing the same thing, aren't you? Because although mediation is included in process 5, it is separate from moderation.
Greetings, Roos
Hello Roos,
On the surface, it sounds to me as if you do not have an "apples to apples" comparison.
1. PROCESS model is univariate (1 DV); MANOVA is multivariate (2 DVs);
2. PROCESS model includes mediator; no mention of that in the MANOVA model.
So, different models can and do yield different results with respect to a specific effect, and whether that effect does or does not contribute significantly to the explanatory power of the model.
Good luck with your work.
Does your PROCESS model include multiple DVs as part of a Structural Equation Model?
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude