Dear All,
I am working on my MSc thesis and my hypotheses and research design imply a two-level, first-stage moderated mediation model.
I have two mediators in my model, which are two distinct paths and are not interchangeable. Accounting for two levels is needed because the employees in my sample are nested in work units and I am investigating the role of LMX differentiation.
First-stage moderated mediation occurs when the mediating process that links an independent variable to an outcome variable varies because the moderating variable either strengthens or attenuates the relationship between the independent variable and the mediator (Edwards & Lambert, 2007).
The fact that the moderator is at the unit-level of analysis complicates matters and I am not sure how to proceed.
Hence, I was wondering what would be the best (i.e., sophisticated) way to test multilevel moderated mediation?
One option I am looking into is multilevel structural equation modeling (MSEM) (Liu, Zhang, & Wang, 2012; Preacher, Zyphur, & Zhang, 2010).
Also, what software would you recommend? Amos, Mplus or another?
Any advice would be greatly appreciated.
Kind regards,
Radostina