HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Journal articles

Manipulating the Alpha Level Cannot Cure Significance Testing

David Trafimow 1 Valentin Amrhein 2, 3 Corson Areshenkoff 4 Carlos Barrera-Causil 5 Eric Beh 6 Yusuf Bilgiç 7 Roser Bono 8 Michael Bradley 9 William Briggs 10 Héctor Cepeda-Freyre 11 Sergio Chaigneau 12 Daniel Ciocca 13 Juan Correa 14 Denis Cousineau 15 Michiel de Boer 16 Subhra Dhar 17 Igor Dolgov 1 Juan Gómez-Benito 8 Marian Grendar 18 James Grice 19 Martin Guerrero-Gimenez 13 Andrés Gutiérrez 20 Tania Huedo-Medina 21 Klaus Jaffe 22 Armina Janyan 23 Ali Karimnezhad 24 Fränzi Korner-Nievergelt 3 Koji Kosugi 25 Martin Lachmair 26 Rubén Ledesma 27 Roberto Limongi 28 Marco Liuzza 29 Rosaria Lombardo 30 Michael Marks 1 Gunther Meinlschmidt 31 Ladislas Nalborczyk 32 Hung Nguyen 33 Raydonal Ospina 34 Jose Perezgonzalez 35 Roland Pfister 36 Juan Rahona 26 David Rodríguez-Medina 37 Xavier Romão 38 Susana Ruiz-Fernández 26 Isabel Suarez 39 Marion Tegethoff 31 Mauricio Tejo 40 Rens van de Schoot 41, 42 Ivan Vankov 23 Santiago Velasco-Forero 43 Tonghui Wang 1 Yuki Yamada 44 Felipe Zoppino 44 Fernando Marmolejo-Ramos 45
33 STARS - Spatio-Temporal Activity Recognition Systems
CRISAM - Inria Sophia Antipolis - Méditerranée
Abstract : 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.
Complete list of metadata

Contributor : Santiago Velasco-Forero Connect in order to contact the contributor
Submitted on : Monday, December 17, 2018 - 10:02:35 AM
Last modification on : Monday, May 9, 2022 - 10:26:05 AM

Links full text



David Trafimow, Valentin Amrhein, Corson Areshenkoff, Carlos Barrera-Causil, Eric Beh, et al.. Manipulating the Alpha Level Cannot Cure Significance Testing. Frontiers in Psychology, Frontiers Media, 2018, 9, ⟨10.3389/fpsyg.2018.00699⟩. ⟨hal-01957088⟩



Record views