baggr: a new meta-analysis package using Stan
Baggr is a new Bayesian evidence aggregation and meta-analysis package for R using Stan, available on CRAN. The tool provides a user-friendly way to build and criticise models for researchers who want to use Bayesian methods but may not be technically proficient enough to use Stan. Baggr suggests the models and priors based on available data and provides a workflow for prior and posterior checks, cross-validation, model comparison and visualisation for typical meta-analysis problems.
In this talk we will present the current state of the package development and discuss future directions. We will focus on particular challenges of building accessible tools and supporting new audiences.
Witold Wiecek, PhD, is a statistician working on applied research in development economics, epidemiology, health economics and outcomes research. His academic work is focused on Bayesian meta-analysis methods and, most recently, modelling the impact of the novel coronavirus.
Witold is also a consultant at Certara (since 2014), a global drug development consultancy, where he uses Bayesian methods to model real-world effectiveness of treatments and in health technology assessment.