scientific article; zbMATH DE number 3954034
[For the entire collection see Zbl 0584.00020.] A test of a single hypothesis \(H:\theta \in \Omega_ H\) against the alternative \(K:\theta \in \Omega_ K=\Omega \setminus \Omega_ H\) with a test function \(\Phi_ H(x)\) is said to be unbiased if \[ \inf_{\theta \in \Omega_ K}E_{\theta}[(\Phi_ HX)]\geq \sup_{\theta \in \Omega_ H}E_{\theta}[(F_ KX)]. \] If each test in a multiple test satisfies this condition we say that the tests are individually unbiased. This is a restriction of different parameter points for the same test. Another reasonable unbiasedness concept for multiple test, which imposes instead restrictions on the different tests in the same point, is the following. A multiple test is said to be rejection unbiased if for all \(\theta\in \Omega\) each hypothesis being false in the point \(\theta\) has a rejection probability which is at least as great as the rejection probability for each hypothesis being true in that point. In this paper we investigate the consequences of these unbiasedness conditions together with the basic multiple test concepts in simple examples, in particular examples related to twosidedness of tests and confidence intervals.
- Rationality and Unbiasedness in Hypothesis Testing
- scientific article; zbMATH DE number 758458 (Why is no real title available?)
- scientific article; zbMATH DE number 922430 (Why is no real title available?)
- Undesirable optimality results in multiple testing?
- Power comparisons of the unbiased Berk-Jones test and the unbiased reversed Berk-Jones test
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3723477)