New preprint: model checking for autoregressive vector models | R bloggers

Time series have become ubiquitous in psychological research, and Vector Autoregressive (VAR) models have become one of the most popular classes of models for studying within-person dynamics in such data. However, systematic checking of how well a VAR model fits the data is almost never performed. This is a problem because model misfit can lead to both incorrect interpretations of model parameters and missing effects in the data that would be theoretically interesting. We provide a tutorial explaining the theory behind model checking, introducing the most common forms of VAR model misspecification in the context of psychological time series, and introducing diagnostics for them, using plots and simulations. We then apply these tools to assess model goodness-of-fit for a multi-level VAR model estimated from a typical empirical dataset of three-week emotion measures from 179 individuals. We conclude by discussing three complementary areas of research that could improve psychological time series modeling in the future. The preprint is available here And here is a Github repository containing the R code to reproduce all analyzes shown in the article.

Related