Sensitivity of Bayes Inference with Ignorable but Data-Dependent Stopping Rules.
Abstract
It is sometimes argued that Bayesian inference is unaffected by data dependent stopping rules. Although this is formally true for ignorable rules, there is likely to be heightened sensitivity of inference to prior assumptions when using data dependent rules rather than stopping rules that do not depend on the data. This point is illustrated in a simple example. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1982
- Accession Number
- ADA125306
Entities
People
- Donald B. Rubin
- Paul R. Rosenbaum
Organizations
- University of Wisconsin–Madison