Inference for a canonical parameter in the presence of nuisance parameters usually requires high dimensional integrals to obtain the marginal or conditional tail probabilities. A simple and very accurate method is proposed to obtain any arbitrary level of significance for the parameter of interest. This method only requires a fine tabulation of the canonical parameter and the corresponding observed likelihood function, which can be either the full, marginal or conditional observed likelihood function, as input, and produces the left tail probabilities at the observed data value as output. Applications of this method to some widely used engineering statistical models will be discussed.
This article is from 'Computing Science and Statistics: Proceedings of the Symposium on the Interface Critical Applications of Scientific Computing: Biology, Engineering, Medicine, Speech Held in Seattle, Washington on 21-24 April 1991,' AD-A252 938, p141-147.