Bayesian statistical dose-response models

Abstract

In the coming years, NASA will collect nationally representative data to support efforts to develop international standards for permissible noise from commercial supersonic flight over land. The data will elucidate the relationship between sonic boom noise exposure and community response. This presentation introduces several Bayesian statistical models and a method to calculate a summary dose-response curve. Additionally, the statistical models illuminate the trade-off between sample size and precision of estimated quantities such as percent highly annoyed. This trade-off will inform sample size decisions for future community surveys. All models are fit to data from a recent NASA community response test, the Quiet Supersonic Flights 2018 (QSF18) test.