Coverage Inducing Priors in Nonstandard Inference Problems

• 成果类型: Article
• 署名作者: Muller, Ulrich K.; Norets, Andriy
• 署名单位: Princeton University; Brown University
• 刊物名称: JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
• ISSN/ISSBN: 0162-1459
• DOI: 10.1080/01621459.2015.1086654
• 发表日期: 2016
• 页码: 1233-1241
• 关键词: instrumental variables regression; MEDIAN-UNBIASED ESTIMATION; Purchasing power parity; confidence-intervals; AUTOREGRESSIVE MODELS; EFFICIENT TESTS; time-series; unit-root; parameter; break

摘要

We consider the construction of set estimators that possess both Bayesian credibility and frequentist coverage properties. We show that under mild regularity conditions there exists a prior distribution that induces (1 - alpha) frequentist coverage of a (1 - alpha) credible set. In contrast to the previous literature, this result does not rely on asymptotic normality or invariance, so it can be applied in nonstandard inference problems. Supplementary materials for this article are available online.