摘要: SUMMARY Conditional and marginal likelihoods constructed from parameter-dependent functions of the data have an additional parameter dependence that changes with choice supporting metric for corresponding densities. We consider constructing conditional by using densities expressed in terms intrinsic support measure, based on either Euclidean or constant information metric. Barndorff-Nielsen's (1983) approximation to distribution maximum likelihood estimate is used approximate these likelihoods. The resulting are closely related modified profile Barndorff-Nielsen Cox & Reid (1987). normal circle problem examined some detail.