Sophie Ward posted on Saturday, February 02, 2019 - 3:36 am
I am currently looking to analyse a complex model as part of my PhD and would really appreciate your advice.
The model contains: 1 IV (4 subconstructs) / Work Frustrators 1 DV (4 subconstructs) / Coping Strategies 1 mediator (single) / Frustration 2 moderators (21 subconstructs) (4 subconstructs) / EI and Social Support
I have looked at multilevel modelling however as I am not assessing between-groups/at different levels I dont think this is appropriate.
I have played around with PROCESS although I am aware you cannot add more than 2 moderator variables /multiple levels of mod.
I also have MPlus and have been looking at models 58 and 75 and the aplicable code. Although again I am not sure how to add the multiple sub-constructs I have into the model, or if I even can. Looking at some similar papers they have run multiple versions of the model using each sub-construct as a moderator, however I am hesitant to do this due to issues surrounding statistical power.
I need to assess the subconstructs, particularly at the moderator level to gain a more in-depth understanding of the role different facets of EI and social support play in the relationship overall. The DV also needs to be split due to the differences in coping strategy being assessed.
Any advice would be greatly appreciated as I am new to such complex modelling.
Do you see the subconstructs as indicators of (being influenced by/reflecting) a factor (like a second-order construct)? Question is if the subconstruct relations with other variables can be summarized into such a single factor's relations.
You mention multilevel modeling - are the observations clustered within groups (hierarchical data)? If so, how many groups?