Regression Analysis-Moderation or Mediated Analysis

Please excuse my lack of understanding of Regression Analysis, but would it be more appropriate to run a moderated or mediated regression analysis for the following question: "Researchers theorized that metacognition for perception may underlie the relationship between source monitoring (imagined vs. real) and hallucination history and thus that once you control for metacognition, the latter relationship would reduce or disappear". :shakehead
your consideration is right. In this case,you should run a mediated(not moderated) model using sem software such as Amos or Mplus and so on. you can email to me( if you encounter some specific problems while running your model.


Less is more. Stay pure. Stay poor.
Agreed, given the problem wording, reduction or disappearance of effect would go with a mediation approach.