摘要翻译：
经济学和金融学中的许多实证研究都涉及同时检验多个假设。本文通过将MinP检验统计量的最小集扩展为包含全局检验（如似然比检验）的$P$%值，提出了一种扩展的MinP（EMinP）检验。我们表明，与MinP检验相比，EMinP检验可以显著地提高拒绝所有单个假设交叉的全局能力。与封闭测试相比，EMinP测试具有计算上的优势，共享了MinP测试的逐步下降过程，并且比用于构造封闭测试的测试具有更好的全局能力。此外，我们认为EMinP测试可以被视为一种工具，当两个具有不同全局能力的竞争测试被利用时，可以防止数据窥探。最后，将所提出的测试应用于一个测试运动效果的实证应用。
---
英文标题：
《Extended MinP Tests of Multiple Hypotheses》
---
作者：
Zeng-Hua Lu
---
最新提交年份：
2019
---
分类信息：

一级分类：Economics        经济学
二级分类：Econometrics        计量经济学
分类描述：Econometric Theory, Micro-Econometrics, Macro-Econometrics, Empirical Content of Economic Relations discovered via New Methods, Methodological Aspects of the Application of Statistical Inference to Economic Data.
计量经济学理论，微观计量经济学，宏观计量经济学，通过新方法发现的经济关系的实证内容，统计推论应用于经济数据的方法论方面。
--

---
英文摘要：
  Much empirical research in economics and finance involves simultaneously testing multiple hypotheses. This paper proposes extended MinP (EMinP) tests by expanding the minimand set of the MinP test statistic to include the $p$% -value of a global test such as a likelihood ratio test. We show that, compared with MinP tests, EMinP tests may considerably improve the global power in rejecting the intersection of all individual hypotheses. Compared with closed tests EMinP tests have the computational advantage by sharing the benefit of the stepdown procedure of MinP tests and can have a better global power over the tests used to construct closed tests. Furthermore, we argue that EMinP tests may be viewed as a tool to prevent data snooping when two competing tests that have distinct global powers are exploited. Finally, the proposed tests are applied to an empirical application on testing the effects of exercise.
---
PDF链接：
https://arxiv.org/pdf/1911.04696