|
|
Longitudinal mediation with LCA outcome |
|
Message/Author |
|
Gaëlle Cyr posted on Saturday, January 27, 2018 - 1:41 pm
|
|
|
Dear Drs. Muthén, I performed a LCA, which allowed me to classify my participants within four classes. Seeing as I have access to longitudinal data, I would now like to test a cross-lagged panel mediation model predicting which class my participants belong to. In other words, my outcome variable would be a categorical variable (1-2-3-4) representing the class each participant belongs to. My indepedant and mediation variables are continuous. Is it possible to perform this type of analysis in Mplus and, if so, do you have any syntax example and/or articles on this subject to suggest ? Thank you very much, Gaëlle |
|
|
Assuming you arrange your longitudinal data in single-level, wide format you have to decide which of your variables influence the latent class variable. UG ex 7.19 can be used as a starting point. |
|
Back to top |
|
|