Open Access
Preprint of Too good to be false: Nonsignificant results revisited
Reads0
Chats0
TLDR
This article examined evidence for false negatives in nonsignificant results in three different ways and concluded that false negatives deserve more attention in the current debate on statistical practices in psychology, and they also proposed the adapted Fisher test to detect the presence of at least one false negative in a set of statistically nonsignificantly results.Abstract:
Due to its probabilistic nature, Null Hypothesis Significance Testing (NHST) is subject to decision errors. The concern for false positives has overshadowed the concern for false negatives in the recent debates in psychology. This might be unwarranted, since reported statistically nonsignificant findings may just be ‘too good to be false’. We examined evidence for false negatives in nonsignificant results in three different ways. We adapted the Fisher test to detect the presence of at least one false negative in a set of statistically nonsignificant results. Simulations show that the adapted Fisher method generally is a powerful method to detect false negatives. We examined evidence for false negatives in the psychology literature in three applications of the adapted Fisher method. These applications indicate that (i) the observed effect size distribution of nonsignificant effects exceeds the expected distribution assuming a null-effect, and approximately two out of three (66.7%) psychology articles reporting nonsignificant results contain evidence for at least one false negative, (ii) nonsignificant results on gender effects contain evidence of true nonzero effects, and (iii) the statistically nonsignificant replications from the Reproducibility Project Psychology (RPP) do not warrant strong conclusions about the absence or presence of true zero effects underlying these nonsignificant results. We conclude that false negatives deserve more attention in the current debate on statistical practices in psychology. Potentially neglecting effects due to a lack of statistical power can lead to a waste of research resources and stifle the scientific discovery process.read more
Citations
More filters
Journal ArticleDOI
Mechanisms underlying training-induced cognitive change
Claudia C. von Bastian,Sylvie Belleville,Robert Udale,Alice Reinhartz,M. Essounni,Tilo Strobach +5 more
TL;DR: The authors provide an overview of cognitive training to induce transfer to untrained tasks, with a particular focus on the theoretical mechanisms that have been proposed to underlie training and transfer effects, and conclude that there is relatively little evidence that training enhances cognitive capacity, that is, the overall cognitive resources available to an individual.
Posted ContentDOI
Contributions towards understanding and building sustainable science
TL;DR: This dissertation argues that "after-the-fact" research papers do not help alleviate issues of access, selective publication, and reproducibility, but actually cause some of these threats because the chronology of the research cycle is lost in a research paper.
Statistical Analyses for Studying Replication: Meta-Analytic Perspectives.
Larry V. Hedges,Jacob M. Schauer +1 more
TL;DR: The authors provide alternative conceptual frameworks that lead to different statistical analyses to test hypotheses about replication, including whether the burden of proof is placed on replication or nonreplication, whether replication is exact or allows for a small amount of "negligible heterogeneity," and whether the studies observed are assumed to be fixed (constituting the entire body of relevant evidence) or are a sample from a universe of possibly relevant studies.
Journal ArticleDOI
Published correlational effect sizes in social and developmental psychology
TL;DR: This article evaluated 12 412 manually collected correlation effect sizes (Sample 1) and 31 157 computer-extracted correlation effect size (Sample 2) published in journals focused on social or developmental psychology and found that large sample sizes were associated with smaller reported correlations.
Journal ArticleDOI
Estudios de casos y controles: Propuesta de robustez de análisis para ciencias de la conducta
Jesua Iván Guzmán-González,Franco Giordano Sánchez-García,Luis M. Sánchez-Loyo,Saúl Ramírez-De Los Santos +3 more
TL;DR: In el campo de las ciencias de la conducta resulta indispensable realizar interpretaciones adecuadas de los resultados obtenidos que permitan el rechazo de la hipótesis nula as mentioned in this paper .
References
More filters
Book
Statistical Power Analysis for the Behavioral Sciences
TL;DR: The concepts of power analysis are discussed in this paper, where Chi-square Tests for Goodness of Fit and Contingency Tables, t-Test for Means, and Sign Test are used.
Journal ArticleDOI
The file drawer problem and tolerance for null results
TL;DR: Quantitative procedures for computing the tolerance for filed and future null results are reported and illustrated, and the implications are discussed.
Journal ArticleDOI
Estimating the reproducibility of psychological science
Alexander A. Aarts,Joanna E. Anderson,Christopher J. Anderson,Peter Raymond Attridge,Peter Raymond Attridge,Angela S. Attwood,Jordan Axt,Molly Babel,Štěpán Bahník,Erica Baranski,Michael Barnett-Cowan,Elizabeth Bartmess,Jennifer S. Beer,Raoul Bell,Heather Bentley,Leah Beyan,Grace Binion,Grace Binion,Denny Borsboom,Annick Bosch,Frank A. Bosco,Sara Bowman,Mark J. Brandt,Erin L Braswell,Hilmar Brohmer,Benjamin T. Brown,Kristina G. Brown,Jovita Brüning,Jovita Brüning,Ann Calhoun-Sauls,Shannon P. Callahan,Elizabeth Chagnon,Jesse Chandler,Jesse Chandler,Christopher R. Chartier,Felix Cheung,Felix Cheung,Cody D. Christopherson,Linda Cillessen,Russ Clay,Hayley M. D. Cleary,Mark D. Cloud,Michael Conn,Johanna Cohoon,Simon Columbus,Andreas Cordes,Giulio Costantini,Leslie Cramblet Alvarez,Ed Cremata,Jan Crusius,Jamie DeCoster,Michelle A. DeGaetano,Nicolás Delia Penna,Bobby Den Bezemer,Marie K. Deserno,Olivia Devitt,Laura Dewitte,David G. Dobolyi,Geneva T. Dodson,M. Brent Donnellan,Ryan Donohue,Rebecca A. Dore,Angela Rachael Dorrough,Angela Rachael Dorrough,Anna Dreber,Michelle Dugas,Elizabeth W. Dunn,Kayleigh E Easey,Sylvia Eboigbe,Casey Eggleston,Jo Embley,Sacha Epskamp,Timothy M. Errington,Vivien Estel,Frank J. Farach,Jenelle Feather,Anna Fedor,Belén Fernández-Castilla,Susann Fiedler,James G. Field,Stanka A. Fitneva,Taru Flagan,Amanda L. Forest,Eskil Forsell,Joshua D. Foster,Michael C. Frank,Rebecca S. Frazier,Heather M. Fuchs,Philip A. Gable,Jeff Galak,Elisa Maria Galliani,Anup Gampa,Sara García,Douglas Gazarian,Elizabeth Gilbert,Roger Giner-Sorolla,Andreas Glöckner,Andreas Glöckner,Lars Goellner,Jin X. Goh,Rebecca M. Goldberg,Patrick T. Goodbourn,Shauna Gordon-McKeon,Bryan Gorges,Jessie Gorges,Justin Goss,Jesse Graham,James A. Grange,Jeremy R. Gray,Chris H.J. Hartgerink,Joshua K. Hartshorne,Fred Hasselman,Timothy Hayes,Emma Heikensten,Felix Henninger,Felix Henninger,John Hodsoll,Taylor Holubar,Gea Hoogendoorn,Denise J. Humphries,Cathy On-Ying Hung,Nathali Immelman,Vanessa C. Irsik,Georg Jahn,Frank Jäkel,Marc Jekel,Magnus Johannesson,Larissa Gabrielle Johnson,David J. Johnson,Kate M. Johnson,William J. Johnston,Kai J. Jonas,Jennifer A. Joy-Gaba,Heather Barry Kappes,Kim Kelso,Mallory C. Kidwell,Seung K. Kim,Matthew W. Kirkhart,Bennett Kleinberg,Bennett Kleinberg,Goran Knežević,Franziska Maria Kolorz,Jolanda J. Kossakowski,Robert Krause,Job Krijnen,Tim Kuhlmann,Yoram K. Kunkels,Megan M. Kyc,Calvin K. Lai,Aamir Laique,Daniel Lakens,Kristin A. Lane,Bethany Lassetter,Ljiljana B. Lazarević,Etienne P. Le Bel,Key Jung Lee,Minha Lee,Kristi M. Lemm,Carmel A. Levitan,Melissa Lewis,Lin Lin,Stephanie C. Lin,Matthias Lippold,Darren Loureiro,Ilse Luteijn,Sean P. Mackinnon,Heather N. Mainard,Denise C. Marigold,Daniel P. Martin,Tylar Martinez,E. J. Masicampo,Joshua J. Matacotta,Maya B. Mathur,Michael May,Michael May,Nicole Mechin,Pranjal H. Mehta,Johannes M. Meixner,Johannes M. Meixner,Alissa Melinger,Jeremy K. Miller,Mallorie Miller,Katherine Moore,Katherine Moore,Marcus Möschl,Matt Motyl,Stephanie M. Müller,Marcus R. Munafò,Koen Ilja Neijenhuijs,Taylor Nervi,Gandalf Nicolas,Gustav Nilsonne,Gustav Nilsonne,Brian A. Nosek,Brian A. Nosek,Michèle B. Nuijten,Catherine Olsson,Catherine Olsson,Colleen Osborne,Lutz Ostkamp,Misha Pavel,Ian S. Penton-Voak,Olivia Perna,Cyril Pernet,Marco Perugini,R. Nathan Pipitone,Michael C. Pitts,Franziska Plessow,Franziska Plessow,Jason M. Prenoveau,Rima-Maria Rahal,Rima-Maria Rahal,Kate A. Ratliff,David Reinhard,Frank Renkewitz,Ashley A. Ricker,Anastasia E. Rigney,Andrew M Rivers,Mark A. Roebke,Abraham M. Rutchick,Robert S. Ryan,Onur Sahin,Anondah R. Saide,Gillian M. Sandstrom,David Santos,David Santos,Rebecca Saxe,René Schlegelmilch,René Schlegelmilch,Kathleen Schmidt,Sabine Scholz,Larissa Seibel,Dylan Selterman,Samuel Shaki,William B. Simpson,H. Colleen Sinclair,Jeanine L. M. Skorinko,Agnieszka Slowik,Joel S. Snyder,Courtney K. Soderberg,Carina Sonnleitner,Nick Spencer,Jeffrey R. Spies,Sara Steegen,Stefan Stieger,Nina Strohminger,Gavin Brent Sullivan,Thomas Talhelm,Megan Tapia,Anniek M. te Dorsthorst,Manuela Thomae,Manuela Thomae,Sarah L. Thomas,Pia Tio,Frits Traets,Steve N.H. Tsang,Francis Tuerlinckx,Paul J. Turchan,Milan Valášek,Anna E. Van't Veer,Robbie C. M. van Aert,Marcel A.L.M. van Assen,Riet van Bork,Mathijs Van De Ven,Don van den Bergh,Marije van der Hulst,Roel van Dooren,Johnny van Doorn,Daan R. van Renswoude,Hedderik van Rijn,Wolf Vanpaemel,Alejandro Vásquez Echeverría,Melissa Vazquez,Natalia Vélez,Marieke Vermue,Mark Verschoor,Michelangelo Vianello,Martin Voracek,Gina Vuu,Eric-Jan Wagenmakers,Joanneke Weerdmeester,Ashlee Welsh,Erin C. Westgate,Joeri Wissink,Michael J. Wood,Andy T. Woods,Andy T. Woods,Emily M. Wright,Sining Wu,Marcel Zeelenberg,Kellylynn Zuni +290 more
TL;DR: A large-scale assessment suggests that experimental reproducibility in psychology leaves a lot to be desired, and correlational tests suggest that replication success was better predicted by the strength of original evidence than by characteristics of the original and replication teams.
Journal ArticleDOI
The Kolmogorov-Smirnov Test for Goodness of Fit
TL;DR: In this paper, the maximum difference between an empirical and a hypothetical cumulative distribution is calculated, and confidence limits for a cumulative distribution are described, showing that the test is superior to the chi-square test.