What is it about?

The literature is full of methods of checking a Bayesian model, including the prior distribution. Less is said about how to proceed with inference after a Bayesian model is found to be inadequate. This paper addresses that issue.

Featured Image

Why is it important?

Inference has to proceed in some way even after a Bayesian model is found to be inadequate. Should the researcher infer that no conclusions can be drawn? If not, what conclusion may be drawn and with what posterior probability does the conclusion hold? This paper provides answers to those questions.

Perspectives

This paper provides a broad framework, inviting many details to be filled in for specific applications. An application to the estimation of local false discovery rates is given as an example.

David R. Bickel
University of North Carolina at Greensboro

Read the Original

This page is a summary of: Inference after checking multiple Bayesian models for data conflict and applications to mitigating the influence of rejected priors, International Journal of Approximate Reasoning, November 2015, Elsevier, DOI: 10.1016/j.ijar.2015.07.012.
You can read the full text:

Read

Resources

Contributors

The following have contributed to this page