Robust statistical significance for Gravitational Wave inference

Abstract

The field of gravitational-wave data analysis heavily relies on Bayes factors to make statements about astrophysical effects in events detected by the LIGO-Virgo-KAGRA collaboration. This is true on an event level, for example when discussing the potential detection of precession or eccentricity, as well as on a population level. For the latter, perhaps the main limiting factor is the reliance on the output from event-level nested sampling analyses, which is known to introduce growing uncertaint…