Mediation with mixed model

[peeckhou]

Hello,

We have 146 dyads of mothers and fathers. We would like to test if perceiving the world as being more dangerous acts as a mediator between the gender of the parent and the fact of being overprotective, while keeping in mind that they are a couple. Mothers are in fact more overprotective than fathers, and we would like to test if mothers perceiving the world as being more dangerous are also more overprotective.

Our variables are :
- Identification (couples 1-146)
- gender (0 = mothers, 1 = fathers)
- world perception (means created with the transformation option)
- overprotection (means created with the transformation option)

We have downloaded the gamjl package to use the mixed model analysis but we don't understand how to make a mediation with this and how to interpret the data.

Thank you for your answer.

[mcfanda@gmail.com]

Mediation with mixed model is not yet implemented, mainly because it it not clear how to run it. You may notice that is also not doable in R (I mean with a R package. It will be, but not very soon.
In your case, however, you can save yourself from the trouble of using the mixed model for mediation thanks to the fact that your independent variable is a dichotomous variable, and use a set of simple regression. Here is well explained
Judd, C. M., Kenny, D. A., & McClelland, G. H. (2001). Estimating and testing mediation and moderation in within-subject designs. Psychological Methods, 6(2), 115–134. https://doi.org/10.1037/1082-989X.6.2.115

[peeckhou]

Thank you very much for your help !