Navigation auf


Epidemiology, Biostatistics and Prevention Institute

SEDA: Sensitivity diagnostics for Bayesian hierarchical models

Bayesian hierarchical models (BHM) are nowadays a well established statistical methodology used for decision making. BHMs have a unique ability to incorporate external prior information in the analysis. They can be conveniently estimated by Bayesian general-purpose software systems such as Stan, JAGS and R-INLA. The most intriguing aspect of BHMs is their sensitivity to assumed priors. Unfortunately, to-date a formal tool for sensitivity diagnostics in BHMs is missing. The project aims
to close this gap by developing and implementing a novel two-component sensitivity diagnostic tool for BHMs estimated by Stan, JAGS or INLA. A free accessible SEDA package in R will warn scientists about sensitivity issues in BHMs, supporting better self-control and model criticism. Two medical applications dealing with Breakpoints
for bacterial resistance and Material wear in 3D will answer questions relevant to microbiological and dental materials research.

Weiterführende Informationen


Sona Hunanyan
Malgorzata Roos (Project Leader)


Swiss National Science Foundation