Here's an example (simple mediation model with one mediator m and two groups):

TITLE: multigroup path analysis

DATA: data.dat;

VARIABLE: NAMES ARE x m y;

GROUPING IS g (1=male 2=female);

MODEL: y on x m;

m on x;

OUTPUT: SAMPSTAT STDYX;

In this example, all 3 paths would be estimated separately/freely between groups. If you wanted the paths to be constrained equal across groups, you would have to change the MODEL statement as follows:

MODEL: y on x (c)

m (b);

m on x (a);

The labels (a), (b), and (c) tell Mplus to hold the path coefficients equal across groups.

If you wanted to obtain the direct, indirect, and total effects from x to y, you would add the statement

MODEL INDIRECT: y ind x;

to the MODEL command.

Thank you so much! I am running the syntax~~@Christian Geiser

Prof. Geiser, could you send me an example of how to report multi-group mediation analysis in APA format? Thank you~~@Christian Geiser