THE VALUE OF PRIOR INFORMATION.
Abstract
A natural question that arises in Bayesian statistics, is how inference might be affected by a change in the prior distribution, i.e., by the availability of more or less information on the parameters of the assumed model. A criterion for evaluating this is suggested and used to compare the value of prior information, when sampling is from the normal, negative binomial, poisson and exponential distributions. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1968
- Accession Number
- AD0679928
Entities
People
- Irwin Guttman
- Norman Richard Draper
Organizations
- University of Wisconsin–Madison