Book chapter
323Slicing and Dicing a Path Through the Fiducial Forest
Handbook of Bayesian, Fiducial, and Frequentist Inference, pp.323-343
CRC Press
2024
DOI: 10.1201/9780429341731-15
Abstract
Of the three inference approaches, Bayesian, Frequentist, and Fiducial, the last is arguably the most controversial and least understood. R.A. Fisher (1930) introduced fiducial inference to reach his goal of deriving a prior-free post-data distribution for, or of, the unknown parameters. Paraphrasing L.J. Savage (1961), Fisher boldly attempted to "make the Bayesian omelet [compute a post-data distribution] without breaking the Bayesian eggs [without specifying a prior]." According to writers on its historical development (e.g. Dempster, 1964, Pedersen, 1978, Seidenfeld, 1992, Zabell, 1992, Efron, 1998, Hannig, 2009), the fiducial idea has had a checkered past, with its share of detractors and proponents. At times the fiducial approach betrayed glimpses of remarkable usefulness (see, for example, the recent work by Hannig and colleagues, and remarks in Efron, 1998), but at other times, owing to misunderstandings and missteps, it invited controversy.
Details
- Title: Subtitle
- 323Slicing and Dicing a Path Through the Fiducial Forest
- Creators
- Joseph B. Lang - University of Iowa
- Contributors
- James Berger (Editor)Xiao-Li Meng (Editor)Nancy Reid (Editor)Min-ge Xie (Editor)
- Resource Type
- Book chapter
- Publication Details
- Handbook of Bayesian, Fiducial, and Frequentist Inference, pp.323-343
- Publisher
- CRC Press; New York
- DOI
- 10.1201/9780429341731-15
- Alternative title
- Handbook of Bayesian, Fiducial, and Frequentist Inference
- Language
- English
- Date published
- 2024
- Academic Unit
- Statistics and Actuarial Science; Biostatistics
- Record Identifier
- 9984548480802771
Metrics
1 Record Views