Mediation and moderated mediation with control variables

General help and assistance with jamovi. Bug reports can be made at our issues page: https://github.com/jamovi/jamovi/issues . (If you're unsure feel free to discuss it here)
Post Reply
MaxAalsmeer
Posts: 1
Joined: Thu May 19, 2022 5:45 pm

Mediation and moderated mediation with control variables

Post by MaxAalsmeer »

HI!

I'm having some trouble with the Jamovi JAMM (advanced mediation module) and I am not sure if I am performing my 2 analyses the correct way. Unfortunately, I cannot find any examples or guides for the mediation or moderated mediation model that includes control variables. I hope someone can help!

I added a link with screenshots that visualize my file better:
https://docs.google.com/document/d/e/2P ... pUDYJsWbMu
C4nE6KhMJ2mW1D57wgGvPGGi/pub

These are my 2 problems/questions

1) Am I doing the following hayes model 4 mediation analyses inc. control variables the correct way?

I investigate workplace loneliness (IV) on work-family conflict (DV) with mediating variable psychological stress. All three variables are continuous. I control for age, tenure, hours worked per week, commute time (all continuous) and educational level and gender (nominal). I want to use the hayes model 4 for this.

When I set up the GLM mediation model I place work-family conflict in DV box, psychological stress in the mediator box and all control variables and workplace loneliness (predictor) in the covariates box (and my 2 nominal control variables in factors). (see attatchment)

In "models for mediators" I have workplace loneliness
In model terms I have psychological stress, workplace loneliness and my control variables.

2) I am trying to do the Hayes model 7 (moderated mediation) and once again I am unsure if I am doing this the correct way. In this case, I want to add the moderator Leader charisma on the relationship between the IV workplace loneliness and the mediator psychological stress

As you can see in the attachment I add work-family conflict as de DV, psych. stress as the mediator, and all continuous control variables in covariates, nominal control variables in factors and main variable workplace loneliness in covariates since it is continuous.

In models for mediators, I have leader charisma, workplace loneliness and workplace loneliness * leader charisma
In model term I have psychological stress, leader charisma, workplace loneliness and my control variables
In mediated effects I have leader charisma.

I know this is a lot of text, but it is easily visualized with the attachment I added.

I hope someone can help me find out how to conduct these 2 analyses!

With warm regards

Max
User avatar
mcfanda@gmail.com
Posts: 457
Joined: Thu Mar 23, 2017 9:24 pm

Re: Mediation and moderated mediation with control variables

Post by mcfanda@gmail.com »

Hi
almost right. You just need to add the control variable also to the mediator models (the green path should become black). This is because to control for them (keep them constant) they have to predict both the mediator and the dependent variable. For the interaction (moderation), it is fine like you did
Post Reply