Thank you for the great MPLUS guide and discussion forum! I am hoping for your help with a current complex data set:
The data are affect and motivation ratings of 82 participants in 8 subjects characterized by different student quotas, thus subject ratings nested in person. The hypothesis is that gender (between) and quota in subject (within, but constant across persons) interactively influence motivation indirectly via affect.
What I did so far is this: ... DEFINE: int2 = gender*quota;
ANALYSIS: TYPE = TWOLEVEL; ESTIMATOR = ML;
MODEL: %WITHIN% posAff BY PA1-PA8; negAff BY NA1-NA8; Mot BY Mot1 Mot2; Mot ON posAff negAff int2 quota; posAff ON int2 quota; negAff ON int2 quota;
MODEL INDIRECT: Mot IND posAff int2; Mot IND negAff int2;
1. Question: Does this model make sense or do I need to use the growth model type despite not having several time points? Did I miss anything? 2. Question: Is it correct to include the cross-level interaction term at level 1 or is there a better way? 3. Question: Can I get indirect effects estimates for each gender separately?
If you have any hints for me, I'd greatly appreciate it.