|
|
MGA with an interaction term PER group |
|
Message/Author |
|
|
I am trying to run a MGA for three different denominations embedded within one data file. I am trying to see if there will be group differences related to an interaction term, where each variable of the interaction term is a latent variable (i.e. religious education moderated by peer religiosity). I am trying to use Type=random because there is an interaction. I receive an error message with the correction/suggestion to use "ALGORITHM=INTEGRATION" in my analysis section, however, I received a new error message, "ALGORITHM=INTEGRATION is not available for multiple group analysis. Try using the KNOWNCLASS option for TYPE=MIXTURE." However, the type=mixture gives me a new error message since I do not have any categorical dependent variables. I guess my question is whether or not it's possible to run a MGA with an interaction term PER group. btw: It is QUITE possible my entire syntax for the model is wrong. I am trying to figure this all out on my own using your manual and the book, "Structural Equation Modeling with MPlus" by Barbara Byrne. Would it be possible to send you my syntax in a private message? Thank you for your time and assistance. |
|
|
With TYPE=RANDOM, the GROUPING option is not available. You would need to use TYPE=MIXTURE RANDOM, the CLASSES option, and the KNOWNCLASS option. When all classes are known, this is the same as multiple group analysis. See Example 7.21. If you need further help, send the relevant files and your license number to support@statmodel.com. |
|
|
Thank you. I will see if I can figure things out from here.
|
|
Back to top |
|
|