Manipulating the Alpha Level Cannot Cure Significance Testing.
Manipulating the Alpha Level Cannot Cure Significance Testing.
复制标题
DOI：
10.3389/fpsyg.2018.00699
复制
发表时间：
2018
影响因子：
3.8
通讯作者：
Marmolejo-Ramos F
中科院分区：
文献类型：
--
作者：
Trafimow D;Amrhein V;Areshenkoff CN;Barrera-Causil CJ;Beh EJ;Bilgiç YK;Bono R;Bradley MT;Briggs WM;Cepeda-Freyre HA;Chaigneau SE;Ciocca DR;Correa JC;Cousineau D;de Boer MR;Dhar SS;Dolgov I;Gómez-Benito J;Grendar M;Grice JW;Guerrero-Gimenez ME;Gutiérrez A;Huedo-Medina TB;Jaffe K;Janyan A;Karimnezhad A;Korner-Nievergelt F;Kosugi K;Lachmair M;Ledesma RD;Limongi R;Liuzza MT;Lombardo R;Marks MJ;Meinlschmidt G;Nalborczyk L;Nguyen HT;Ospina R;Perezgonzalez JD;Pfister R;Rahona JJ;Rodríguez-Medina DA;Romão X;Ruiz-Fernández S;Suarez I;Tegethoff M;Tejo M;van de Schoot R;Vankov II;Velasco-Forero S;Wang T;Yamada Y;Zoppino FCM;Marmolejo-Ramos F
我们认为，对科学假设做出接受/拒绝的决定，包括最近要求将规范α水平从p = 0.05改变为p = 0.005，对发现新发现和科学进步是有害的。考虑到总括和可变α水平都是有问题的，完全免除显著性检验是明智的。有一些替代方法比显著性检验更直接地解决了研究设计和样本量问题;但没有一种统计工具应该被视为新的神奇方法，给出明确的机械答案。推论不应该基于单一的研究，而是基于多个独立研究的累积证据。在评估证据的强度时，我们应该考虑辅助假设、实验设计的强度以及对应用的影响。将所有这些归结为基于p值阈值0.05、0.01、0.005或其他任何值的二元决策是不可接受的。
We argue that making accept/reject decisions on scientific hypotheses, including a recent call for changing the canonical alpha level from p = 0.05 to p = 0.005, is deleterious for the finding of new discoveries and the progress of science. Given that blanket and variable alpha levels both are problematic, it is sensible to dispense with significance testing altogether. There are alternatives that address study design and sample size much more directly than significance testing does; but none of the statistical tools should be taken as the new magic method giving clear-cut mechanical answers. Inference should not be based on single studies at all, but on cumulative evidence from multiple independent studies. When evaluating the strength of the evidence, we should consider, for example, auxiliary assumptions, the strength of the experimental design, and implications for applications. To boil all this down to a binary decision based on a p-value threshold of 0.05, 0.01, 0.005, or anything else, is not acceptable.
登录
查看更多内容
地球是平的（p > 0.05）：重要性阈值和不可复制研究的危机。
影响因子：
2.7
作者：
Amrhein V;Korner-Nievergelt F;Roth T
通讯作者：
Roth T
影响因子：
13.6
作者：
Greenland S;Senn SJ;Rothman KJ;Carlin JB;Poole C;Goodman SN;Altman DG
通讯作者：
Altman DG
影响因子：
1.3
作者：
Ferrill, Mary J;Brown, Dana A;Kyle, Jeffrey A
通讯作者：
Kyle, Jeffrey A
影响因子：
2.3
作者：
Bradley, Michael T.;Brand, Andrew
通讯作者：
Brand, Andrew
影响因子：
2
作者：
GOODMAN, SN
通讯作者：
GOODMAN, SN