Addressing common inferential mistakes when failing to reject the null-hypothesis.

Q2 Pharmacology, Toxicology and Pharmaceutics
F1000Research Pub Date : 2025-04-01 eCollection Date: 2024-01-01 DOI:10.12688/f1000research.158434.2
Amand Schmidt
{"title":"Addressing common inferential mistakes when failing to reject the null-hypothesis.","authors":"Amand Schmidt","doi":"10.12688/f1000research.158434.2","DOIUrl":null,"url":null,"abstract":"<p><p>Failure to reject a null-hypothesis may lead to erroneous conclusions regarding the absence of an association or inadequate statistical power. Because an estimate (and its variance) can never be exactly zero, traditional statistical tests cannot conclusively demonstrate the absence of an association. Instead, estimates of accuracy should be used to identify settings in which an association and its variability are sufficiently small to be clinically acceptable, directly providing information on safety and efficacy. Post-hoc power calculations should be avoided, as they offer no additional information beyond statistical tests and p-values. Furthermore, post-hoc power calculations can be misleading because of an inability to distinguish between results based on insufficient sample size and results that reflect clinically irrelevant differences. Most multiple testing procedures unrealistically assume that all positive results are false positives. However, in applied settings, results typically represent a mix of true and false positives. This implies that multiplicity corrections do not effectively differentiate between true and false positives. Instead, considering the distributions of p-values and the proportion of significant results can help to identify bodies of evidence unlikely to be driven by false-positive results. In conclusion, rather than attempting to categorize results as true or false, medical research should embrace established statistical methods that focus on estimation accuracy, replication, and consistency.</p>","PeriodicalId":12260,"journal":{"name":"F1000Research","volume":"13 ","pages":"1488"},"PeriodicalIF":0.0000,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11928781/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"F1000Research","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.12688/f1000research.158434.2","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/1/1 0:00:00","PubModel":"eCollection","JCR":"Q2","JCRName":"Pharmacology, Toxicology and Pharmaceutics","Score":null,"Total":0}
引用次数: 0

Abstract

Failure to reject a null-hypothesis may lead to erroneous conclusions regarding the absence of an association or inadequate statistical power. Because an estimate (and its variance) can never be exactly zero, traditional statistical tests cannot conclusively demonstrate the absence of an association. Instead, estimates of accuracy should be used to identify settings in which an association and its variability are sufficiently small to be clinically acceptable, directly providing information on safety and efficacy. Post-hoc power calculations should be avoided, as they offer no additional information beyond statistical tests and p-values. Furthermore, post-hoc power calculations can be misleading because of an inability to distinguish between results based on insufficient sample size and results that reflect clinically irrelevant differences. Most multiple testing procedures unrealistically assume that all positive results are false positives. However, in applied settings, results typically represent a mix of true and false positives. This implies that multiplicity corrections do not effectively differentiate between true and false positives. Instead, considering the distributions of p-values and the proportion of significant results can help to identify bodies of evidence unlikely to be driven by false-positive results. In conclusion, rather than attempting to categorize results as true or false, medical research should embrace established statistical methods that focus on estimation accuracy, replication, and consistency.

解决在未能拒绝零假设时常见的推理错误。
未能拒绝零假设可能导致关于缺乏关联或统计能力不足的错误结论。由于估计值(及其方差)永远不可能完全为零,传统的统计检验无法最终证明相关性的缺失。相反,准确度估计应该用于识别关联及其变异性足够小到临床可接受的情况,直接提供有关安全性和有效性的信息。应避免事后功率计算,因为它们除了统计检验和p值之外没有提供额外的信息。此外,事后功率计算可能会产生误导,因为无法区分基于不足样本量的结果和反映临床无关差异的结果。大多数多重检测程序不切实际地假设所有阳性结果都是假阳性。然而,在应用环境中,结果通常代表真阳性和假阳性的混合。这意味着多重校正不能有效区分真阳性和假阳性。相反,考虑p值的分布和显著结果的比例可以帮助识别不太可能由假阳性结果驱动的证据体。总之,医学研究不应试图将结果归类为对或错,而应采用既定的统计方法,重点关注估计的准确性、可重复性和一致性。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
F1000Research
F1000Research Pharmacology, Toxicology and Pharmaceutics-Pharmacology, Toxicology and Pharmaceutics (all)
CiteScore
5.00
自引率
0.00%
发文量
1646
审稿时长
1 weeks
期刊介绍: F1000Research publishes articles and other research outputs reporting basic scientific, scholarly, translational and clinical research across the physical and life sciences, engineering, medicine, social sciences and humanities. F1000Research is a scholarly publication platform set up for the scientific, scholarly and medical research community; each article has at least one author who is a qualified researcher, scholar or clinician actively working in their speciality and who has made a key contribution to the article. Articles must be original (not duplications). All research is suitable irrespective of the perceived level of interest or novelty; we welcome confirmatory and negative results, as well as null studies. F1000Research publishes different type of research, including clinical trials, systematic reviews, software tools, method articles, and many others. Reviews and Opinion articles providing a balanced and comprehensive overview of the latest discoveries in a particular field, or presenting a personal perspective on recent developments, are also welcome. See the full list of article types we accept for more information.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信