Moderation or Moderated mediation?

29 July 2021 3 2K Report

Dear researchers,

I would like to ask about the differences between moderation and moderated mediation analysis? And how to make hypotheses in a moderated mediation relationship?

For example, in a model like this (see attached file), if I am curious about whether corporate image moderates the relationship between customer satisfaction(the moderator) and repurchase intention, do I only do a moderation analysis or a second stage moderated mediation analysis?

If I am going to do moderated mediation, am I testing whether corporate image moderates the mediation effect of customer satisfaction on repurchase intention? I cannot relate to any practical implications of this moderated mediation analysis? Moreover, if making hypotheses, should I say: corporate image moderates the indirect effect of customer experience on repurchase intention through customer satisfaction?

My point is if I am only interested in whether corporate image moderates the relationship between customer satisfaction and repurchase intention, as customer satisfaction is a mediator, do I have to do moderated mediation analysis? what's its advantage compared with merely moderation analysis?

Could anyone recommend some papers that conducted moderated mediation analysis? I want to see how they make hypotheses, and how to interpret the findings. Although I have read the paper by Andrew F. Hayes (2015) "An Index and Test of Linear Moderated Mediation", I still feel not very clear.

Thank you all

Best Regards

Kara

Imran Anwar

@Dear Kara Ma Simple moderation only accounts for enhancing/buffering (interaction effect) the effect of IV on DV (example: the effect of customer satisfaction on repurchase intention is stronger when corporate image is high and vice versa). Whereas, in moderated mediation (when moderation is only at first level i.e., IV >>> Mediator or second level i.e., Mediator >>> DV) we account for conditional indirect effect i.e., varying indirect effect at different levels of the moderator (example: the indirect effect of customer experience on repurchase intention (via customer satisfaction) would be stronger when corporate image is high and vice versa).

SmartPLS: What should I do if discriminant validity HTMT>0.90?

I got 02 moderator values defining Johnson-Neyman significance regions. Is it an error?

Do I need to do linear regression before running mediation analysis using PROCESS MACRO by Hayes ?

Mediation path analysis - Can I run two seperate models?

SEM or Multiple Regression for an analysis without Latent Variables?

What is the original citation for the commonly used activity factors for converting basal metabolic rate to total caloric need?

How do you interpret multiple meta-regression in STATA?

How can we estimate sample size for a moderated mediation analysis with more than two mediators?

What does it mean if in mediation analysis the indirect effect is significant but is a larger b value than that of the direct effect?

Can we have a model where A affects B which affects C but B is not a mediator?

What is the step by step process involved in interpreting the results of mediation analysis done using PROCESS macro?