## Abstract The problem of defining Pβvalues for twoβsided tests is considered when the null density __f__~0~ of the associated test statistic is strictly unimodal. It is argued that the usual practice of defining Pβvalues for twoβsided tests as twice the oneβtail Pβvalue is only justifable when __
P-values For Discrete Test Statistics
β Scribed by Assoc. Professor W. E. Franck
- Publisher
- John Wiley and Sons
- Year
- 1986
- Tongue
- English
- Weight
- 227 KB
- Volume
- 28
- Category
- Article
- ISSN
- 0323-3847
No coin nor oath required. For personal study only.
β¦ Synopsis
This paper considered the relative merits of the P-value and the mid-P-value. It is shown that inference based onthe mid-P-value is in a certain sense on firmer ground. In particular the expected mid-P-value does not change under an irrelevant breakup in the test statistic.
π SIMILAR VOLUMES
Discrete-time models of continuous-time plants are commonly required owing to the popular use of computers to implement control and estimation algorithms. When stochastic design techniques such as the discrete-time Kalman filter are utilized, it is necessary to determine the equivalent discrete-time
Scientific Data Gathering -- Displaying And Summarizing Data -- Logic, Probability, And Uncertainty -- Discrete Random Variables -- Bayesian Inference For Discrete Random Variables -- Continuous Random Variables -- Bayesian Inference For Binomial Proportion -- Comparing Bayesian And Frequentist Infe