...
首页> 外文期刊>The American statistician >Correction to'Easy Multiplicity Control in Equivalence Testing Using Two One-Sided Tests'
【24h】

Correction to'Easy Multiplicity Control in Equivalence Testing Using Two One-Sided Tests'

机译:对“使用两个单面测试的等效测试中的简便多重性控制”的更正

获取原文
获取原文并翻译 | 示例
   

获取外文期刊封面封底 >>

       

摘要

Under the assumptions of Lauzon and Caffo (2009), if one is to adhere to the rule that only MNCs where θ ≤ △μ < 2θ will factor into the FWER, then the appropriate correction involves dividing the nominal error rate by [K~2/4]. The K - 1 divisor proposed in Lauzon and Caffo (2009) only holds for K = 3 or if the null hypothesis holds for all comparisons; it is otherwise too liberal, becoming increasingly so with K. The concept of MNCs is only necessary for simple conceptual proofs and approximate control of the FWER. The results of Roehmel (2011) show that arguments from closed testing procedures yields the same corrections, yet demonstrate exact control of the FWER. This exact correction has been proven for K = 3 and 4 could be extended to K = ∞ by analogous use of the arguments.
机译:在Lauzon和Caffo(2009)的假设下,如果要遵循这样一个规则,即只有θ≤△μ<2θ的MNC会计入FWER,则适当的校正包括将名义错误率除以[K〜2 / 4]。 Lauzon和Caffo(2009)提出的K-1除数仅适用于K = 3或所有比较均采用零假设的情况。 MNC的概念仅对于简单的概念证明和FWER的近似控制才是必需的。 Roehmel(2011)的结果表明,封闭测试程序的论点得出了相同的更正,但显示了对FWER的精确控制。已针对K = 3证明了这种精确的校正,并且可以通过类似使用自变量将其扩展为K =∞。

著录项

  • 来源
    《The American statistician》 |2013年第2期|115-116|共2页
  • 作者单位

    Department of Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD, 21205;

    Department of Electrical Engineering, Vanderbilt University, Nashville, TN 37325;

    Leibniz Institute for Prevention Research and Epidemiology Achterstrasse 30 D-28359 Bremen, Germany;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号