Ofer Eldad posted on Wednesday, September 14, 2016 - 9:40 am
Dear Profs. Muthen,
I'm running a multi-level mediation model with unbalanced data. Level-1 is different observations of a specific family and level-2 is the different families. There are a different number of observations per family and they are unevenly spaced. There is also some missing data.
Looking at the User's Guide it seems that MLR, MLF, and ML are available estimators for two-level modeling with random slopes and that MLR is the default/recommended estimator.
Is this also the case when we have mediation and unbalanced data?