2-1-2 ML-SEM mediation model and powe... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 Thomas Dodson posted on Saturday, April 21, 2018 - 12:18 pm
I have a more general question. Is it possible in Mplus to assess mediation with a multilevel structural equation model (ML-SEM) analysis when my mediating variable is a repeated measure (i.e., participants complete questions 1 time a day for 7 days), and both my independent variable and outcome variable are only assessed one time? Thus, the repeated measure is clustered within participants. Also, all of the variables are continuous.

Based on my readings, I am thinking it would be a 2-1-2 ML-SEM design. I have seen some discussion from Preacher et al. (2010, 2011) about this model, but I was not sure if Mplus could accommodate this procedure. Also, are there suggested power analysis tools for this model?
 Bengt O. Muthen posted on Sunday, April 22, 2018 - 10:20 am
You can analyze it in wide format with data in columns:

X M1 ... M7 Y

and specify a growth model for M1-M7. Then it is really 1-1-1.
 Thomas Dodson posted on Sunday, April 22, 2018 - 11:21 am
Thank you Dr. Muthen.

I am considering making the repeated measure indicators of a latent variable. Would this be possible or will they need to remain observed variables?


 Bengt O. Muthen posted on Monday, April 23, 2018 - 4:53 pm
This is possible.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message