gms | German Medical Science

65th Annual Meeting of the German Association for Medical Informatics, Biometry and Epidemiology (GMDS), Meeting of the Central European Network (CEN: German Region, Austro-Swiss Region and Polish Region) of the International Biometric Society (IBS)

06.09. - 09.09.2020, Berlin (online conference)

Article

Send article

Calibration of Bayesian analyses including historical data: The perspective matters

Meeting Abstract

Search Medline for

  • Christian Röver - Universitätsmedizin Göttingen, Göttingen, Germany
  • Tim Friede - Universitätsmedizin Göttingen, Göttingen, Germany

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 65th Annual Meeting of the German Association for Medical Informatics, Biometry and Epidemiology (GMDS), Meeting of the Central European Network (CEN: German Region, Austro-Swiss Region and Polish Region) of the International Biometric Society (IBS). Berlin, 06.-09.09.2020. Düsseldorf: German Medical Science GMS Publishing House; 2021. DocAbstr. 496

doi: 10.3205/20gmds031, urn:nbn:de:0183-20gmds0312

Published: February 26, 2021

© 2021 Röver et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License. See license information at http://creativecommons.org/licenses/by/4.0/.

Outline

Text

Bayesian methods yield posterior probabilities that are “calibrated” in a specific sense, which may also be expressed as a long-run frequency. We revisit the calibration concept and investigate the roles of informative and vague priors. In particular, we point out the implications for sequential analyses on accumulating data (utilizing informative priors from earlier stages) and robustification approaches in the context of analyses including historical data. We argue that Bayesian methods provide proper answers to particular questions. But the context needs to be kept in mind when interpreting posterior probabilities, especially when the focus is on long-run coverage probabilities.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.