Confidence intervals for the binomial parameter: some new considerations

摘要: Several methods have been proposed to construct confidence intervals for the binomial parameter. Some recent papers introduced "mean coverage" criterion evaluate performance of and suggested that exact methods, because their conservatism, are less useful than asymptotic ones. In these studies, however, were always represented by Clopper-Pearson interval (C-P). Now we focus on Sterne's interval, which is also known be better C-P in two-sided case. Introducing a computer intensive level-adjustment procedure allows constructing terms mean coverage, demonstrate performs best intervals, even coverage context. Level adjustment improves as well, which, with an appropriate level adjustment, becomes equivalent mid-P interval. Finally show behaviour method far poorer generally expected.