Release of a new R package for analyzing intensive longitudinal data
mlts simplifies the specification and estimation of multilevel time series models with latent and manifest variables using a fully Bayesian approach via Rstan.
Logo des mlts Pakets
Meldung vom:
With mlts, you can effortlessly:
- Fit popular precompiled multilevel time series models (AR, VAR) and dynamic structural equation models (DSEM) with different time lags.
- Specify measurement models at both the within- and between-person levels (up to 9 latent variables).
- Handle missing data and utilize the tinterval functionality (akin to Mplus) for approximating individually varying time intervals.
- Include outcome variables as well as time-stable covariates with ease.
- Estimate person-specific random effects for autoregressive effects, cross-lagged effects, innovation variances, and innovation covariances.
- Estimate standardized effects.
- Plot results of fixed and person-specific estimates
- Visualize and display the specified models through equations and path models.
mlts has been particularly designed for applied researchers to embrace the complexities of multilevel latent time series modeling with ease and flexibility.
Koslowski K, Münch F, Koch T, Holtmann J (2024). “mlts: Multilevel Latent Time Series Models with R and Stan.” https://github.com/munchfab/mltsExterner Link.
Get started with following demonstrations of two simple modeling variants using mlts. Detailed tutorials are going to follow soon.