The evaluation of the p -value as an estimator for the null hypothesis in the exponential distribution
- 31 May 2022
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Simulation and Computation
- Vol. 53 (6) , 2633-2644
Abstract
This paper is concerned with investigating the adequacy of using the p-value as an estimator for the set specified by the null hypothesis in the Exponential distribution. It is shown that the p-value is an admissible estimator in the one-sided test of the location parameter. When the one-sided test of the scale parameter is considered, the p-value is found to be a generalized Bayes estimator with infinite Bayes risk. However, it is very difficult to find an estimator that dominates it. When the parameter space is restricted, the modified p-value is an admissible estimator in the one-sided test of the scale parameter and performs better than the usual p-value. Although the usual p-value is generally inadmissible in the two-sided test, it can be useful as an estimator in this type of test for the scale parameter.Keywords
This publication has 12 references indexed in Scilit:
- Modified p-values for one-sided testing in restricted parameter spacesStatistics & Probability Letters, 2007
- Improved estimation of accuracy in simple hypothesis versus simple alternative testingJournal of Multivariate Analysis, 2004
- A NOTE ON THE ADMISSIBILITY OF P-VALUE FOR THE ONE-SIDED HYPOTHESIS TEST IN THE NEGATIVE BINOMIAL MODELTaiwanese Journal of Mathematics, 1997
- Estimation of Accuracy in TestingThe Annals of Statistics, 1992
- DISCUSSING TRUTH OR FALSITY BY COMPUTING A Q-VALUE1989
- Bayesian testing of an exponential point null hypothesisCommunications in Statistics - Theory and Methods, 1988
- Testing Precise HypothesesStatistical Science, 1987
- Reconciling Bayesian and Frequentist Evidence in the One-Sided Testing ProblemJournal of the American Statistical Association, 1987
- Statistical Decision Theory and Bayesian Analysis1985
- A STATISTICAL PARADOXBiometrika, 1957