On the construction of shortest confidence intervals and Bayesian highest posterior density intervals
- 1 December 1991
- journal article
- Published by Wiley in Journal of Veterinary Pharmacology and Therapeutics
- Vol. 14 (4) , 395-399
- https://doi.org/10.1111/j.1365-2885.1991.tb00853.x
Abstract
In this note it is argued that the principal characteristic of the confidence intervals proposed by Bartoszynski & Powers (1990) is not primarily the fact that they are of minimum length but that they are Bayesian highest posterior density intervals. A simple iterative process for determining the ends of the interval is presented.Keywords
This publication has 20 references indexed in Scilit:
- How Appropriate Are Popular Sample Size Formulas?The American Statistician, 1989
- CLINICAL TRIAL NUMBERS AND CONFIDENCE INTERVALS OF PRE-SPECIFIED SIZEThe Lancet, 1988
- Applications of Bayesian Software: Two ExamplesJournal of the Royal Statistical Society: Series D (The Statistician), 1987
- Bayesianly Justifiable and Relevant Frequency Calculations for the Applied StatisticianThe Annals of Statistics, 1984
- The Assessment of the Subjective Opinion and its Use in Relation to Stopping Rules for Clinical TrialsJournal of the Royal Statistical Society: Series D (The Statistician), 1983
- Applications of a Method for the Efficient Computation of Posterior DistributionsJournal of the Royal Statistical Society Series C: Applied Statistics, 1982
- The Future of Statistics: A Bayesian 21st CenturyAdvances in Applied Probability, 1975
- Optimal Confidence Intervals of Fixed WidthThe American Statistician, 1971
- Shortest Confidence IntervalsThe American Statistician, 1969