Is it possible to use Mplus to implement a X-M-Y mediation model, where the mediator (M) and the outcome (Y) are both categorical latent variables (defined by Latent Class Analysis)?
If so, could you provide references where such syntax is available?
You could do this in Mplus by specifying a latent class model with two latent class variables (one for M and one for Y) and a covariate (X). The syntax would be similar as for a 2-time point latent transition analysis with a covariate. See examples 8.13 and/or 8.14 in the Mplus User's Guide:
https://www.statmodel.com/HTML_UG/chapter8V8.htm
https://statmodel.com/usersguide/chap8/ex8.13part1.html
https://statmodel.com/usersguide/chap8/ex8.14.html
What would probably be complicated is the interpretation of the mediated (indirect) effect in that situation.
You can have a look at the following for some ideas depending on how many categories you have
https://journals.sagepub.com/doi/full/10.1177/0022243719881618
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India