Change search
ReferencesLink to record
Permanent link

Direct link
Why small low-powered studies are worse than large high-powered studies and how to protect against "trivial" findings in research: comment on Friston (2012)
Stockholm University, Faculty of Social Sciences, Stress Research Institute.
2013 (English)In: NeuroImage, ISSN 1053-8119, E-ISSN 1095-9572, Vol. 81, 496-8 p.Article in journal (Refereed) Published
Abstract [en]

It is sometimes argued that small studies provide better evidence for reported effects because they are less likely to report findings with small and trivial effect sizes (Friston, 2012). But larger studies are actually better at protecting against inferences from trivial effect sizes, if researchers just make use of effect sizes and confidence intervals. Poor statistical power also comes at a cost of inflated proportion of false positive findings, less power to "confirm" true effects and bias in reported (inflated) effect sizes. Small studies (n=16) lack the precision to reliably distinguish small and medium to large effect sizes (r<.50) from random noise (α=.05) that larger studies (n=100) does with high level of confidence (r=.50, p=.00000012). The present paper presents the arguments needed for researchers to refute the claim that small low-powered studies have a higher degree of scientific evidence than large high-powered studies.

Place, publisher, year, edition, pages
2013. Vol. 81, 496-8 p.
National Category
Medical and Health Sciences
URN: urn:nbn:se:su:diva-99681DOI: 10.1016/j.neuroimage.2013.03.030PubMedID: 23583358Local ID: P3125OAI: diva2:687940
Available from: 2014-01-15 Created: 2014-01-15 Last updated: 2014-01-16Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full textPubMed

Search in DiVA

By author/editor
Ingre, Michael
By organisation
Stress Research Institute
In the same journal
Medical and Health Sciences

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 25 hits
ReferencesLink to record
Permanent link

Direct link