Three-level multigroup and indirect e... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 Tina Kretschmer posted on Monday, October 31, 2016 - 11:53 pm
Hi Bengt and Linda,

Hopefully you can advise on this problem:
We're trying to fit a three-level model (participants nested in classrooms nested in schools) where one of the steps involves a multigroup analysis (say, boys versus girls) combined with testing an indirect effect.
Till now, we specified school and classroom as cluster variables and used TYPE=TWOLEVEL COMPLEX. None of the model variables were on the between-level so the model part is specified using %within% only.

The problem in the multiple group analysis is that people in the same cluster are assigned to different groups, something that is apparently not possible in Mplus: **For TYPE=TWOLEVEL COMPLEX with multiple groups, cluster IDs for the last cluster variable cannot appear in more than one group.**

Elsewhere in the forum you suggested using KNOWNCLASS and TYPE=TWOLEVEL MIXTURE but now the trouble is that the mediation part won't run: ** MODEL INDIRECT is not available for TYPE=TWOLEVEL MIXTURE.

Does you have advice on dealing with this? Many thanks in advance!!
 Bengt O. Muthen posted on Wednesday, November 02, 2016 - 5:28 pm
Have a look at Mplus Web Note 16 on our website.
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