Hello,
I am trying to conduct a Random Intercept Latent Transition Analysis (RI-LTA), which is available in Mplus, using R. I was wondering if there is a specific R package for this, and whether it is as straightforward to implement as in Mplus. I am aware of the LMest package, but my data are continuous and LMest does not currently support a true RI-LTA structure with both continuous outcomes and a random intercept factor, so it cannot fully replicate the Mplus model.
Any guidance or suggestions would be greatly appreciated.
No as of now R doesn't have like a turnkey package for riltas with continuous indicators + subject-level random intercepts.
What you could do is a bayesian multilevel HMM with categorical/continuous/count emissions and subject-level random effects on transition and emission parameters -> This captures between-person heterogeneity, but it is not the rilta factor that loads 1 on all repeated indicators. Still, it’s the closest “off-the-shelf” RStudio approach for continuous data here
Thanks Marius Ole Johansen
That makes a lot of sense. I appreciate you pointing me toward the Bayesian multilevel HMM approach; it’s good to know there’s at least a partial alternative in R. I’ll look into the implementation details for this approach, but if it turns out not to match the RI-LTA structure, I’ll probably have to stick with Mplus for now to get the exact model specification I need.
Thanks again for taking the time to respond and for sharing your insight.
- Badges
- Science topic