Manipulating the Alpha Level Cannot Cure Significance Testing
2 Unibas - Université de Bâle = University of Basel = Basel Universität
3 Swiss Ornithological Institute
4 Queen's University [Kingston, Canada]
5 Metropolitan Technological Institute
6 UoN - University of Newcastle [Callaghan, Australia]
7 SUNY Geneseo - State University of New York at Geneseo
8 UB - Universitat de Barcelona
9 UNB - University of New Brunswick
10 Chercheur indépendant
11 BUAP - Benemérita Universidad Autónoma de Puebla
12 Universidad Adolfo Ibáñez = University Adolfo Ibañez [Santiago]
13 CCT - Science and Technology Center [Mendoza, Argentina]
14 UNAL - Universidad Nacional de Colombia [Bogotà]
15 LIX - Laboratoire d'informatique de l'École polytechnique [Palaiseau]
16 VU - Vrije Universiteit Amsterdam [Amsterdam]
17 IIT Kanpur - Indian Institute of Technology Kanpur
18 Univerzita Komenského = Comenius University [Bratislava]
19 OSU - Oklahoma State University [Stillwater]
20 Universidad Santo Tomas [Bogota]
21 UCONN - University of Connecticut [Storrs]
22 USB - Universidad Simon Bolivar
23 New Bulgarian University
24 University of Ottawa [Ottawa]
25 Senshu University
26 Leibniz-Institut für Wissensmedien [Tübingen]
27 UNMdP - Universidad Nacional de Mar del Plata [Mar del Plata]
28 PUCV - Pontificia Universidad Católica de Valparaíso
29 UMG - Università degli Studi "Magna Graecia" di Catanzaro = University of Catanzaro
30 Università degli studi della Campania "Luigi Vanvitelli" = University of the Study of Campania Luigi Vanvitelli
31 University Hospital Basel [Basel]
32 LPNC - Laboratoire de Psychologie et NeuroCognition
33 STARS - Spatio-Temporal Activity Recognition Systems
34 UFPE - Universidade Federal de Pernambuco [Recife]
35 Massey University
36 JMU - Julius-Maximilians-Universität Würzburg = University of Würzburg [Würsburg, Germany]
37 UAM - Universidad Autónoma Metropolitana [Mexico]
38 Universidade do Porto = University of Porto
39 Universidad del Norte, Barranquilla
40 Universidad de Playa Ancha
41 Universiteit Utrecht / Utrecht University [Utrecht]
42 NWU - North-West University [Vaal Triangle Campus, Vanderbijlpark, South Africa]
43 CMM - Centre de Morphologie Mathématique
44 Kyushu University
45 University of Adelaide
- Fonction : Auteur
- PersonId : 953486
- Fonction : Auteur
- PersonId : 793536
- ORCID : 0000-0001-7435-7262
- Fonction : Auteur
- PersonId : 182947
- IdHAL : ladislas-nalborczyk
- ORCID : 0000-0002-7419-9855
- IdRef : 241927544
- Fonction : Auteur
- PersonId : 793537
- ORCID : 0000-0002-1709-1506
- Fonction : Auteur
- PersonId : 779361
- ORCID : 0000-0001-5579-0140
- Fonction : Auteur
- PersonId : 4198
- IdHAL : velascoforero
- ORCID : 0000-0002-2438-1747
- IdRef : 169108589
Résumé
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.