multilevel moderated mediation(2-1-1 with a level 1 moderator in the second stage)

244 views

Skip to first unread message

ping bao

unread,

Apr 15, 2022, 3:00:54 AM4/15/22

to lavaan

Hello,

I'm trying to do a 2-1-1 moderated mediation analysis. independent variable(team leadership, level 2), mediator(individual psychologal safety, level 1), outcome variable(individual voice, level1), and the moderator(leader-member exchange, level 1) moderates the relationship between psychological safety and voice. However, I found no relevant resource in R. The problem is that I don't know how to test the moderated mediation hypothesis.

Could you please tell me how to construt the MSEM to test the model?

Best,

ping

Terrence Jorgensen

unread,

Apr 19, 2022, 5:24:03 AM4/19/22

to lavaan

I'm trying to do a 2-1-1 moderated mediation analysis

This is problematic in lavaan because there is no way to calculate product terms between the latent within-level or between-level components. You can do the decomposition yourself (manifest approach): calculate cluster means (between component), then calculate cluster-mean-centered data (within component), and calculate the product terms to model moderating effects. In your model syntax, only include the respective components in their appropriate level-specific models.

Terrence D. Jorgensen

Assistant Professor, Methods and Statistics

Research Institute for Child Development and Education, the University of Amsterdam