Biometrika · 1999 · 146 citations · 18 references
Natural SciencesMathematicsBayesian Data AnalysisProbability TheoryStatisticsBayesian InferenceBayesian StatisticsInferential StatisticsPrior SpecificationStatistical InferenceBayesian Hypothesis TestingBayesian Probability
this paper with approximations of the form p(/) = \\Phi 1 (r; Q) = \\Phi(r) + OE(r)(r \\Gamma Q ) (1:5) p(/) = \\Phi 2 (r; Q) = \\Phifr \\Gamma r log(r=Q)g : (1:6) where OE is the standard normal density function. The primary objective is a simple and widely applicable formula for Q that ensures the O(n ) accuracy of the p- value p(/). The frequentist version is recorded in (3.6) and is a generalisation of (1.1) and (2.6). The Bayesian version is also recorded in (3.6) and is a generalisation of (1.2) and (2.9)
18