B0393
Title: Bayesian modeling and causal inference for multivariate longitudinal data with R package dynamite
Authors: Santtu Tikka - University of Jyvaskyla (Finland) [presenting]
Abstract: Panel data are ubiquitous in scientific domains such as sociology and econometrics. Various modelling approaches have been presented for the analysis of such data including dynamic panel models, cross-lagged panel models, and their extensions. Existing panel data modelling approaches typically impose some restrictive assumptions on the data-generating process, such as Gaussian errors, effects that are constant in time, or univariate responses. The dynamic multivariate panel model (DMPM) is presented that supports both time-varying and time-invariant effects, multiple simultaneous responses across a wide variety of distributions, arbitrary dependency structures of lagged responses of any order, and latent factors. A Bayesian approach is taken to the estimation of the model parameters and leverages state-of-the-art Markov chain Monte Carlo methods. It is shown how the posterior predictive distributions can be used to evaluate long-term counterfactual predictions which take into account the dynamic structure of the assumed causal graph of the system. The use of DMPMs is demonstrated by applying the model to both real and synthetic data. Finally, an overview of a new R package dynamite for Bayesian inference is given for panel data.