Using Latent Variable as a Control Va... PreviousNext
Mplus Discussion > Structural Equation Modeling >
 JHS posted on Wednesday, February 24, 2016 - 5:08 pm

I have a question about how to use a latent variable as a control variable in a mediation in MPlus. I am looking at how state sadness after completing a lab task (M2) mediates the relationship between a personality variable (Y) and a symptom variable (X). I would like to control for baseline sadness before the task (M1).

Would this code be correct?

Y BY y1 y2 y3;
M1 BY m11 m12;
M2 BY m21 m22;
X BY x1 x2 x3;

M2 on X M1;
Y on X M2;

Should I correlate or regress M1 with any of the other variables? Should I correlate the indicators of M1 and M2?

Thank you!
 Bengt O. Muthen posted on Wednesday, February 24, 2016 - 6:21 pm
Q1. I would include M1 in the Y regression.

Q2. You should explore that in a separate initial analysis of the measurement models.
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