Interaction effect PreviousNext
Mplus Discussion > Structural Equation Modeling >
 Ann Lee posted on Sunday, March 18, 2012 - 6:13 pm
Hi Dr. Muthen,

For Type=Random, supposing I have a moderator of x2 influencing the relationship between x1 and y.

x3 | x1 xwith x2,
y on x1 x3;

what is the default setting for the relationship between y and x2 in the estimated SEM? Is it true that if a model includes an A*B interaction, then both A and B must be in the model even if they are of no interest?

Thank you.
 Linda K. Muthen posted on Monday, March 19, 2012 - 6:25 pm
I would include both main effects along with the interaction.
 Anonymous posted on Tuesday, March 20, 2012 - 12:17 pm
I am doing the same thing (without TYPE = RANDOM), but how can you center the variables before computing the interaction (through DEFINE)? I know I would have to use SAVEDATA somewhere but which of all options (DIFFTEST, etc)?

I can't seem to be able to use XWITH
 Linda K. Muthen posted on Tuesday, March 20, 2012 - 2:27 pm
There is a CENTERING option for observed variables. It is not necessary to center latent variables.

I'm not sure what you mean by not being able to use XWITH. Perhaps you are using a version of the program before XWITH was added.
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