The DOI matches its Crossref record; 110 of the 110 references it cites were found in a registry or an index; 11 of the 39 sources its claims cite were checked for support.
Power failure: why small sample size undermines the reliability of neuroscience
Open accessAbstract
A study with low statistical power has a reduced chance of detecting a true effect, but it is less well appreciated that low power also reduces the likelihood that a statistically significant result reflects a true effect. Here, we show that the average statistical power of studies in the neurosciences is very low. The consequences of this include overestimates of effect size and low reproducibility of results. There are also ethical dimensions to this problem, as unreliable research is inefficient and wasteful. Improving reproducibility in neuroscience is a key priority and requires attention to well-established but often ignored methodological principles.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
This review argues that low statistical power is widespread in neuroscience and that it makes published findings less likely to be true, more likely to overstate effects, and harder to replicate. Across meta-analyses from several subfields, median power was often below 31%, with some estimates as low as 8%.
Key findings
- In a sample of 48 articles, 49 meta-analyses and 730 primary studies, the median statistical power in neuroscience was 21%; excluding seven unusually high-powered neurological meta-analyses, it fell to 18%.
- The observed number of nominally significant studies was 349, compared with 254 expected under the meta-analytic summary effects, indicating excess significance (p < 0.0001).
- For brain volume studies, median power was 8% across 461 individual studies from 41 meta-analyses in eight articles; for animal water-maze and radial-maze studies, median power was 18% and 31%, respectively, with average sample sizes of 22 and 24 animals.
- With pre-study odds of 0.25 and α = 0.05, 20% power yielded a positive predictive value of 0.50, whereas 80% power yielded 0.80; simulations suggested initial effect estimates from studies powered between about 8% and 31% are inflated by 25% to 50%.
Methods
- The authors searched Web of Science on 2 February 2012 for 2011 articles using the terms “neuroscience” and “meta-analysis,” screened records independently, and included 48 articles containing 49 meta-analyses and 730 primary studies.
- They extracted summary effect sizes and study-level sample sizes or event counts, then calculated each primary study’s achieved power to detect its corresponding meta-analytic summary effect at α = 5% using G*Power, reporting mean and median power and testing for excess significance.
- Because the main sample underrepresented neuroimaging and animal-model research, they added representative meta-analyses from those areas outside the 2011 sampling frame.
Limitations
- The paper states that true effect sizes are unknown, so meta-analytic summary effects were used as proxies; because those summaries may themselves be inflated by publication and other biases, the power estimates are likely optimistic.
- The authors note that the main analysis underrepresented neuroimaging and animal-model subfields, and that the animal-model findings rest on only two meta-analyses, so extrapolation should be cautious.
- Evident from the text: the search was limited to 2011 Web of Science records and to meta-analyses with extractable data, and the excess-significance test depends on the assumption that the summary effects approximate the true effects.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
45 Own interpretation · 15 No basis given · 12 Own result · 9 No full text of the sources · 5 Partly supported · 4 Sources not found · 3 Supported · 3 Supported so far · 3 Own method · 1 Not checked yet · accuracy not yet measured
Every source the paper cites for this claim was checked and supports it.
In the paper
“As a consequence, researchers have strong incentives to engage in research practices that make their findings publishable quickly, even if those practices reduce the likelihood that the findings reflect a true (that is, non-null) effect 4 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Abstract
What the sources say
- Supportedcites [4] Nosek, B. A., Spies, J. R. & Motyl, M. Scientific utopia: II. Restructuring incentives and practices to promote truth over publishability. Perspect. Psychol. Sci. 7, 615–631 (2012). source
The Abstract says publication incentives encourage research decisions that produce positive results and inflate the rate of false effects. The section “Strategies That Will Accelerate the Accumulation of Knowledge” contrasts the immediate payoff of publishing with the longer-term goal of accuracy, supporting the claim’s emphasis on publishing quickly. (LLM, conf=0.92)
boosteddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“When studies in a given field are designed with a power of 20%, it means that if there are 100 genuine non-null effects to be discovered in that field, these studies are expected to discover only 20 of them 11 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Abstract
What the sources say
- Supportedcites [11] Sterne, J. A. & Davey Smith, G. Sifting the evidence — what's wrong with significance tests? BMJ 322, 226–231 (2001). source
Table 3 defines 20% power as rejecting the null hypothesis 20% of the time when it is false. Thus, among 100 genuine non-null effects, the expected number detected is 20. The supplied copy does not specify a page number. (LLM, conf=0.93)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Second, the lower the power of a study, the lower the probability that an observed effect that passes the required threshold of claiming its discovery (that is, reaching nominal statistical significance, such as p < 0.05) actually reflects a true effect 1,12 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Abstract
What the sources say
- Supportedcites [1] Ioannidis, J. P. Why most published research findings are false. PLoS Med. 2, e124 (2005). This study demonstrates that many (and possibly most) of the conclusions drawn from biomedical research are probably false. The reasons for this include using flexible study designs and flexible statistical analyses and running small studies with low statistical power. source
Corollary 1 states that, other factors being equal, the probability that a research finding is true decreases as study power decreases. Table 4 identifies this probability as the positive predictive value of a finding under a nominal significance threshold of α = 0.05. (LLM, conf=0.94)
- Not checked — no full text of the sourcecites [12] Ioannidis, J. P. A., Tarone, R. & McLaughlin, J. K. The false-positive to false-negative ratio in epidemiologic studies. Epidemiology 22, 450–456 (2011). source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Previous analyses of studies using animal models have shown that small studies consistently give more favourable (that is, 'positive') results than larger studies 78”
§ Power failure: why small sample size undermines the reliability of neuroscience > Animal model studies.
What the sources say
- Partly supportedcites [78] Sena, E. S., van der Worp, H. B., Bath, P. M., Howells, D. W. & Macleod, M. R. Publication bias in reports of animal stroke studies leads to major overstatement of efficacy. PLoS Biol. 8, e1000344 (2010). sourceExperimental — accuracy not yet measured
The Results (Prevalence of Publication Bias; page not provided) report that all 16 datasets had an excess of imprecise studies with large effect sizes, supporting the favourable-results pattern in animal stroke studies. But the Discussion (page not provided) cautions that precision in these studies relates more to biological variability and measurement error than to study size. The paper therefore does not establish the claimed comparison of small studies with larger studies. (LLM, conf=0.86)
boostedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“study quality is inversely related to effect size [79][80][81][82]”
§ Power failure: why small sample size undermines the reliability of neuroscience > Animal model studies.
What the sources say
- Not checked — no full text of the sourcecites [79] Ioannidis, J. P. Extrapolating from animals to humans. Sci. Transl. Med. 4, 151ps15 (2012). source
- Not checked — no full text of the sourcecites [80] Jonasson, Z. Meta-analysis of sex differences in rodent models of learning and memory: a review of behavioral and biological data. Neurosci. Biobehav. Rev. 28, 811–825 (2005). source
- Partly supportedcites [81] Macleod, M. R. et al. Evidence for the efficacy of NXY-059 in experimental focal cerebral ischaemia is confounded by study quality. Stroke 39, 2824–2829 (2008). sourceExperimental — accuracy not yet measured
The Results report that estimated efficacy was highest in lower-quality studies of NXY-059 in animal models of stroke. This supports the inverse relationship in the studies analyzed, but does not establish it across animal model studies generally. (LLM, conf=0.84)
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“A requirement for sample size and power calculation is included in the Animal Research: Reporting In Vivo Experiments (ARRIVE) guidelines 84”
§ Power failure: why small sample size undermines the reliability of neuroscience > Ethical implications.
What the sources say
- Partly supportedcites [84] Kilkenny, C., Browne, W. J., Cuthill, I. C., Emerson, M. & Altman, D. G. Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research. PLoS Biol. 8, e1000412 (2010). sourceExperimental — accuracy not yet measured
Table 2, item 10 (Sample size), asks authors to report animal numbers, explain how they were decided, and provide details of any sample-size calculation used. It does not require a calculation or mention a power calculation; the accompanying text also says the guidelines are not intended to be mandatory. (LLM, conf=0.89)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our results indicate that the median statistical power in neuroscience is 21%.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Results.
hedgedn = 730- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption: The 49 meta-analyses included are representative of neuroscience research.
In the paper
“Our results indicated that the median statistical power of these studies was 8% across 461 individual studies contributing to 41 separate meta-analyses, which were drawn from eight articles that were published between 2006 and 2009.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Neuroimaging studies.
What the sources say
- Not checked — no full text of the sourcecites [73] Ioannidis, J. P. Excess significance bias in the literature on brain volume abnormalities. Arch. Gen. Psychiatry 68, 773–780 (2011). source
hedgedn = 461- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“However, seven meta-analyses comprised studies with high (>90%) average power 24,26,31,57,63,68,71 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Results.
What the sources say
- Not checked — no full text of the sourcecites [24] Babbage, D. R. et al. Meta-analysis of facial affect recognition difficulties after traumatic brain injury. Neuropsychology 25, 277–285 (2011). source
- Not checked — no full text of the sourcecites [26] Bjorkhem-Bergman, L., Asplund, A. B. & Lindh, J. D. Metformin for weight reduction in non-diabetic patients on antipsychotic drugs: a systematic review and meta-analysis. J. Psychopharmacol. 25, 299–305 (2011). source
- Not checked — no full text of the sourcecites [31] Chen, C. et al. Allergy and risk of glioma: a meta-analysis. Eur. J. Neurol. 18, 387–395 (2011). source
- Not checked — no full text of the sourcecites [57] Tian, Y., Kang, L. G., Wang, H. Y. & Liu, Z. Y. Meta-analysis of transcranial magnetic stimulation to treat post-stroke dysfunction. Neural Regen. Res. 6, 1736–1741 (2011).
- Not checked — no full text of the sourcecites [63] Witteman, J., van Ijzendoorn, M. H., van de Velde, D., van Heuven, V. J. & Schiller, N. O. The nature of hemispheric specialization for linguistic and emotional prosodic perception: a meta-analysis of the lesion literature. Neuropsychologia 49, 3722–3738 (2011). source
- Not matchedcites [68] Yuan, H. et al. Meta-analysis of tau genetic polymorphism and sporadic progressive supranuclear palsy susceptibility. Neural Regen. Res. 6, 353–359 (2011).
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not checked — no full text of the sourcecites [71] Zhu, Y., He, Z. Y. & Liu, H. N. Meta-analysis of the relationship between homocysteine, vitamin B(12), folate, and multiple sclerosis. J. Clin. Neurosci. 18, 933–938 (2011). source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“If we exclude these 'outlying' meta-analyses, the median statistical power falls to 18%.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Results.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- It may state two separate claims as one. 21
- Probably not a substantial claim. 19
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Such practices include using flexible study designs and flexible statistical analyses and running small studies with low statistical power 1,5 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Abstract
What the sources say
- Supportedcites [1] Ioannidis, J. P. Why most published research findings are false. PLoS Med. 2, e124 (2005). This study demonstrates that many (and possibly most) of the conclusions drawn from biomedical research are probably false. The reasons for this include using flexible study designs and flexible statistical analyses and running small studies with low statistical power. source
Corollary 4 says that flexibility in study designs and analytical methods makes findings less likely to be true. Corollary 1 says that smaller studies have lower statistical power and that the probability a finding is true falls as power decreases. Together, these support the practices attributed to the paper. (LLM, conf=0.94)
- Not checked — no full text of the sourcecites [5] Simmons, J. P., Nelson, L. D. & Simonsohn, U. False-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychol. Sci. 22, 1359–1366 (2011). This article empirically illustrates that flexible study designs and data analysis dramatically increase the possibility of obtaining a nominally significant result. However, conclusions drawn from these results are almost certainly false. source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“It has been claimed and demonstrated that many (and possibly most) of the conclusions drawn from biomedical research are probably false 1 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Abstract
What the sources say
- Supportedcites [1] Ioannidis, J. P. Why most published research findings are false. PLoS Med. 2, e124 (2005). This study demonstrates that many (and possibly most) of the conclusions drawn from biomedical research are probably false. The reasons for this include using flexible study designs and flexible statistical analyses and running small studies with low statistical power. source
The Abstract says there is increasing concern that most published research findings are false and that simulations show a research claim is more likely to be false than true for most study designs and settings. The paper’s discussion of biomedical research makes the citing sentence a fair paraphrase of its argument. (LLM, conf=0.92)
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Second, publication bias, selective data analysis and selective reporting of outcomes are more likely to affect low-powered studies.”
- Also stated in: § Low power in the presence of other biases · hedged
“Publication bias and selective reporting of outcomes and analyses are also more likely to affect smaller, underpowered studies 17 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Low power in the presence of other biases
What the sources say
- Partly supportedcites [17] Dwan, K. et al. Systematic review of the empirical evidence of study publication bias and outcome reporting bias. PLoS ONE 3, e3081 (2008). sourceExperimental — accuracy not yet measured
The Discussion (page unavailable) reports that positive or statistically significant studies are more likely to be published and significant outcomes more likely to be fully reported. The paper does not establish that publication bias or selective reporting is more likely in smaller, low-powered studies; the Results mention power calculations but do not make that comparison. Its Introduction describes selective reporting of analyses but does not assess whether it varies by study power. (LLM, conf=0.83)
hedgedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Also stated in: § Low power in the presence of other biases · hedged
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Although previously almost all of the proposed candidate gene associations from small studies were false 99 (with some exceptions 100 ), collaborative consortia have substantially improved power, and the replicated results can be considered highly reliable.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Incentivizing replication.
What the sources say
- Not checked — no full text of the sourcecites [99] Ioannidis, J. P., Tarone, R. & McLaughlin, J. K. The false-positive to false-negative ratio in epidemiologic studies. Epidemiology 22, 450–456 (2011). source
- Partly supportedcites [100] Siontis, K. C., Patsopoulos, N. A. & Ioannidis, J. P. Replication of past candidate loci for common diseases and phenotypes in 100 genome-wide association studies. Eur. J. Hum. Genet. 18, 832–837 (2010). sourceExperimental — accuracy not yet measured
Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“However, these policies are uncommonly adhered to 95 , and thus the ability for independent experts to repeat published analysis remains low 96 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Improving availability of materials and data.
What the sources say
- Supportedcites [95] Alsheikh-Ali, A. A., Qureshi, W., Al-Mallah, M. H. & Ioannidis, J. P. Public availability of published research data in high-impact journals. PLoS ONE 6, e24357 (2011). source
The Discussion states that more than half of publications subject to journal data-availability instructions did not adhere to them. This supports the policy-adherence clause attributed to this source; the separate claim about repeating analyses carries citation 96. (LLM, conf=0.93)
- Not checked — no full text of the sourcecites [96] Ioannidis, J. P. et al. Repeatability of published microarray gene expression analyses. Nature Genet. 41, 149–155 (2009). source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our results indicate that the median statistical power for the water maze studies and the radial maze studies to detect these medium to large effects was 18% and 31%, respectively (TABLE 2).”
§ Power failure: why small sample size undermines the reliability of neuroscience > Animal model studies.
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The average sample size in these studies was 22 animals for the water maze and 24 for the radial maze experiments.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Animal model studies.
plainn = 22- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In order to achieve 80% power to detect, in a single study, the most probable true effects as indicated by the meta-analysis, a sample size of 134 animals would be required for the water maze experiment (assuming an effect size of d = 0.49) and 68 animals for the radial maze experiment (assuming an effect size of d = 0.69); to achieve 95% power, these sample sizes would need to increase to 220 and”
§ Power failure: why small sample size undermines the reliability of neuroscience > Ethical implications.
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“the total numbers of animals actually used in the studies contributing to the meta-analyses were even larger: 420 for the water maze experiments and 514 for the radial maze experiments.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Ethical implications.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The meta-analyses (n = 7) that comprised studies with high average power in excess of 90% had their broadly neurological subject matter in common.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
plaincorrelationaln = 7- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“First, low power, by definition, means that the chance of discovering effects that are genuinely true is low.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Abstract
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
A claim is checked only against the cited paper’s full text, never its abstract, and none is available for its sources.
In the paper
“Given that these publishing biases are pervasive across scientific practice, it is possible that false positives heavily contaminate the neuroscience literature as well, and this problem may affect at least as much, if not even more so, the most prominent journals 9,10 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Abstract
What the sources say
- Not checked — no full text of the sourcecites [9] Fang, F. C. & Casadevall, A. Retracted science and the retraction index. Infect. Immun. 79, 3855–3859 (2011). source
- Not checked — no full text of the sourcecites [10] Munafo, M. R., Stothart, G. & Flint, J. Bias in genetic association studies and impact factor. Mol. Psychiatry 14, 119–120 (2009). source
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Pre-registration -and, ultimately, full reporting of analysis plans -clarifies the distinction between confirmatory and exploratory analysis, encourages well-powered studies (at least in the case of confirmatory analyses) and reduces the file-drawer effect.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Registration of confirmatory analysis plan.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“The principles of the 'three Rs' in animal research (reduce, refine and replace) 83 require appropriate experimental design and statistics -both too many and too few animals present an issue as they reduce the value of research outputs.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Ethical implications.
What the sources say
- Not matchedcites [83] Russell, W. M. S. & Burch, R. L. The Principles of Humane Experimental Technique (Methuen, 1958).
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- Not matched
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“All of the articles that were identified via this electronic search were screened independently for suitability by two authors (K.S.B. and M.R.M.).”
§ Power failure: why small sample size undermines the reliability of neuroscience > Methods.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Low power therefore has an ethical dimensionunreliable research is inefficient and wasteful.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Ethical implications.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.
- Rests on an assumption: Low power is a problem in practice because of the normative publishing standards for producing novel, significant, clean results and the ubiquity of null hypothesis significance testing as the means of evaluating the truth of research findings.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“The results described in this section are based on only two meta-analyses, and we should be appropriately cautious in extrapolating from this limited evidence.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Animal model studies.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Nevertheless, it is notable that the results are so consistent with those observed in other fields, such as the neuroimaging and neuroscience studies that we have described above.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Animal model studies.
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“False positives occur more frequently and go unnoticed when degrees of freedom in data analysis and reporting are undisclosed 5 .”
§ Power failure: why small sample size undermines the reliability of neuroscience > Conclusions and future directions
What the sources say
- Not checked — no full text of the sourcecites [5] Simmons, J. P., Nelson, L. D. & Simonsohn, U. False-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychol. Sci. 22, 1359–1366 (2011). This article empirically illustrates that flexible study designs and data analysis dramatically increase the possibility of obtaining a nominally significant result. However, conclusions drawn from these results are almost certainly false. source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Making research materials available will improve the quality of studies aimed at replicating and extending research findings.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Improving availability of materials and data.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Data extraction was performed independently by K.S.B. and M.R.M. or C.M. and verified collaboratively.”
- Also stated in: § Figure 2 | Flow diagram of articles selected for inclusion. · plain
“Data extraction was independently performed by K.S.B. and either M.R.M. or C.M. and verified collaboratively.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Figure 2 | Flow diagram of articles selected for inclusion.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Also stated in: § Figure 2 | Flow diagram of articles selected for inclusion. · plain
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Meta-analysis provides the best estimate of the true effect size, albeit with limitations, including the limitation that the individual studies that contribute to a meta-analysis are themselves subject to the problems described above.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Empirical evidence from neuroscience
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As we have already discussed, several other biases are also likely to reduce the probability that a research finding reflects a true effect.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Implications for the likelihood that a research finding reflects a true effect.
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Inflated effect estimates make it difficult to determine an adequate sample size for replication studies, increasing the probability of type II errors.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Null findings locked in file drawers bias the literature, whereas exploratory analyses are only useful and valid if you acknowledge the caveats and limitations.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Disclose methods and findings transparently
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Assuming, as one might in a power calculation, that the initially observed effect we are trying to replicate reflects the true effect, the potential distribution of these replication effect estimates would be similar to the distribution of the original study (dashed green curve).”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
hedgedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- The authors state this as an assumption: The initially observed effect that a replication study attempts to replicate reflects the true effect.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Some fields are confronting the problem of the poor reliability of research findings that arises from low-powered studies.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Concluding remarks.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We can increase the power of the replication study (coloured area under the orange curve) by increasing the sample size so as to reduce the sem.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“For example, in genetic epidemiology sample sizes increased dramatically with the widespread understanding that the effects being sought are likely to be extremely small.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Concluding remarks.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“If a study finds evidence for an effect at p = 0.05, then the difference between the mean of the null distribution (indicated by the solid blue curve) and the mean of the observed distribution (dashed blue curve) is 1.96 × sem.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Making raw data available will improve data aggregation methods and confidence in reported results.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Improving availability of materials and data.
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Combining data increases the total sample size (and therefore power) while minimizing the labour and resource impact on any one contributor.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“By contrast, large, high-powered studies can readily detect both small and large effects and so are less biased, as both over-and underestimations of the true effect size will pass the threshold for 'discovery'.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“The winner's curse refers to the phenomenon that studies that find evidence of an effect often provide inflated estimates of the size of that effect.”
§ Power failure: why small sample size undermines the reliability of neuroscience > Work collaboratively to increase power and replicate findings
What the sources say
- Not matchedcites REF. 103
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
- Not matched
The authors state this as an assumption.
The summary effect size reported in each study is close to the true effect size.
“Importantly, these calculations assume that the summary effect size reported in each study is close to the true effect size, but it is likely that they are inflated owing to publication and other biases described above.”
The authors state this as an assumption.
The initially observed effect we are trying to replicate reflects the true effect.
“Assuming, as one might in a power calculation, that the initially observed effect we are trying to replicate reflects the true effect, the potential distribution of these replication effect estimates would be similar to the distribution of the original study (dashed green curve).”
The paper relies on this without support.
The 49 meta-analyses included are representative of neuroscience research.
“Our results indicate that the median statistical power in neuroscience is 21%.”
The paper relies on this without support.
Most structural and volumetric MRI studies are very small and have minimal power to detect differences between compared groups.
“Most structural and volumetric MRI studies are very small and have minimal power to detect differences between compared groups (for example, healthy people versus those with mental health diseases).”
The paper relies on this without support.
The effects now sought in neuroscience experiments are often smaller and more subtle than before.
“A consequence of the remarkable growth in neuroscience over the past 50 years has been that the effects we now seek in our experiments are often smaller and more subtle than before as opposed to when mostly easily discernible 'low-hanging fruit' were targeted.”
The paper relies on this without support.
Computational analysis of very large datasets is now relatively straightforward, so that an enormous number of tests can be run in a short time on the same dataset.
“At the same time, computational analysis of very large datasets is now relatively straightforward, so that an enormous number of tests can be run in a short time on the same dataset.”
The paper relies on this without support.
These dramatic advances in the flexibility of research design and analysis have occurred without accompanying changes to other aspects of research design, particularly power.
“These dramatic advances in the flexibility of research design and analysis have occurred without accompanying changes to other aspects of research design, particularly power.”
The paper relies on this without support.
Low power is a problem in practice because of the normative publishing standards for producing novel, significant, clean results and the ubiquity of null hypothesis significance testing as the means of evaluating the truth of research findings.
“Low power is a problem in practice because of the normative publishing standards for producing novel, significant, clean results and the ubiquity of null hypothesis significance testing as the means of evaluating the truth of research findings.”
The authors state this as an assumption.
The initially observed effect that a replication study attempts to replicate reflects the true effect.
“Assuming, as one might in a power calculation, that the initially observed effect we are trying to replicate reflects the true effect”
The paper relies on this without support.
In an exploratory research field such as much of neuroscience, the pre-study odds are often low.
“Notably, in an exploratory research field such as much of neuroscience, the pre-study odds are often low.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“Due to differences in age and sample size between protocols and phases, effect sizes could not be directly compared (Button et al., 2013), but the directions of associations with age were largely consistent for ADNI2 and ADNI3 phases separately, and each ADNI3 protocol (Figures 1, 2).”
From Diffusion MRI Indices and Their Relation to Cognitive Impairment in Brain Aging: The Updated Multi-protocol Approach in ADNI3 (2019) · cites this paper as Button et al., 2013
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“but not very different from studies in neuroscience in general (average power = 21%) (14).”
From Statistical and Methodological Considerations for the Interpretation of Intranasal Oxytocin Studies (2016) · cites this paper as [14]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“First, the sample size is small, which precludes the detection of small or medium-sized effects.43”
From Effects of Insufficient Sleep on Pituitary–Adrenocortical Response to CRH Stimulation in Healthy Men (2017) · cites this paper as [43]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The lower response rate at week 26 may also result in inflated effects sizes and lower reproducibility of results [63].”
From Benchmark findings from a veteran electronic patient-reported outcomes evaluation from a chronic pain management telehealth program (2024) · cites this paper as [63]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This may benefit studies in particular that focus on neuroimaging, which are costly and tend to have small sample sizes (Button et al. 2013), those that employ electrophysiological techniques that might be extremely time-consuming, or those that include genetics.”
From New directions in hypnosis research: strategies for advancing the cognitive and clinical neuroscience of hypnosis (2017) · cites this paper as Button et al. 2013
Experimental — accuracy not yet measured
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The current study is limited in that the sample sizes and sample demographics available for each protocol vary, complicating direct comparison of the protocols (Button et al., 2013).”
From Diffusion MRI Indices and Their Relation to Cognitive Impairment in Brain Aging: The Updated Multi-protocol Approach in ADNI3 (2019) · cites this paper as Button et al., 2013
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“This might indicate that results from previous studies do not always reflect a true effect (Button et al. 2013)”
From Terror Management Theory in the Consumer Domain: A Systematic Review and Meta‐Analysis on Mortality Salience Driven Consumer Responses (2025) · cites this paper as Button et al. 2013
The cited source says what the claim attributes to it.
In the citing paper
“Another limitation was the small sample size, due to the duration of the project (just over a year) and the availability of families to take their children for evaluation within this period. This could have reduced the statistical power of the study, potentially lowering the probability of detecting a true effect if one exists and reducing the likelihood that a statistically significant result”
From Enhancing Ecological Validity: Virtual Reality Assessment of Executive Functioning in Children and Adolescents with ADHD (2024) · cites this paper as [46]
The cited source says what the claim attributes to it.
In the citing paper
“Selective reporting, questionable research practices, and low-powered research all contribute to the unreliability of findings in a range of disciplines”
From Investigating the replicability of preclinical cancer biology (2021) · cites this paper as Button et al., 2013
The cited source says what the claim attributes to it.
In the citing paper
“Given the tendency for small sample sizes to produce inflated effect size estimates (Button et al., 2013), future studies should include larger samples in order to produce more precise effect size estimates.”
From Working Memory Training Does Not Improve Performance on Measures of Intelligence or Other Measures of “Far Transfer”: Evidence From a Meta-Analytic Review (2016) · cites this paper as Button et al., 2013
The cited source says what the claim attributes to it.
In the citing paper
“Fourth, due to the typical practices in neuroscience research of obtaining and relying on small and underpowered samples (Button et al., 2013).”
From Is the Putative Mirror Neuron System Associated with Empathy? A Systematic Review and Meta-Analysis (2021) · cites this paper as Button et al., 2013
The cited source says what the claim attributes to it.
In the citing paper
“Sample sizes of included studies were, on average, relatively small, which is consistent with the recent findings of Button et al. (2013) that suggest such sample sizes are typical in neuroscience research.”
From Is the Putative Mirror Neuron System Associated with Empathy? A Systematic Review and Meta-Analysis (2021) · cites this paper as Button et al. (2013)
The cited source says what the claim attributes to it.
In the citing paper
“Potentially problematic practices include selective reporting, selective analysis, and insufficient specification of the conditions necessary or sufficient to obtain the results ( 1223 ).”
From Estimating the reproducibility of psychological science (2015) · cites this paper as [14]
The cited source says what the claim attributes to it.
In the citing paper
“An important qualification about this result is the possibility that the original studies have inflated effect sizes due to publication, selection, reporting, or other biases ( 9, 1223 ).”
From Estimating the reproducibility of psychological science (2015) · cites this paper as [14]
The cited source says what the claim attributes to it.
In the citing paper
“Concerns about reproducibility are widespread across disciplines ( 921 ).”
From Estimating the reproducibility of psychological science (2015) · cites this paper as [14]
The cited source says what the claim attributes to it.
In the citing paper
“There is good evidence that low power of research designs combined with publication bias favoring positive results over negative results produces a literature with upwardly biased effect sizes ( 14, 16, 30, 31 ).”
From Estimating the reproducibility of psychological science (2015) · cites this paper as [14]
The cited source says what the claim attributes to it.
In the citing paper
“and has contributed, in part, to the replication crisis in other fields (Button et al., 2013; Szucs and Ioannidis, 2017; Ioannidis et al., 2017).”
From With Little Power Comes Great Responsibility (2020) · cites this paper as Button et al., 2013
The cited source says what the claim attributes to it.
In the citing paper
“when significant effects are found, they are likely to be noisier and have lower positive predictive value (Button et al., 2013).”
From With Little Power Comes Great Responsibility (2020) · cites this paper as Button et al., 2013
The cited source says what the claim attributes to it.
In the citing paper
“Low power has three effects: first, within the experiment, real effects are more likely to be missed; second, when an effect is detected, this will often be an overestimation of the true effect size [24]; and finally, when low power is combined with publication bias, there is an increase in the false positive rate in the published literature [36].”
From Reporting animal research: Explanation and elaboration for the ARRIVE guidelines 2.0 (2020) · cites this paper as [36]
The cited source says what the claim attributes to it.
In the citing paper
“The reproducibility of scientific findings has been called into question.”
From Evaluating replicability of laboratory experiments in economics (2016) · cites this paper as [6]
In the citing paper
“Furthermore, sample sizes are often underpowered (4-6), at times leading to conclusions that do not hold up with larger-sized replications or metaanalyses (7).”
From Generic language in scientific communication (2019) · cites this paper as [5]
In the citing paper
“Testing more rather than fewer participants will make inferences more robust and more widely applicable as this can reduce random variation, increasing the probability that a statistically significant result reflects a true effect also likely to occur in retesting with similar groups [5].”
From Hasty generalizations and generics in medical research: A systematic review (2024) · cites this paper as [5]
In the citing paper
“Less clear, perhaps, is that low power can also increase the false discovery rate and the likelihood that reported effect sizes are inflated, due to their reduced ability to mute stochastic noise [13,63,64].”
From The natural selection of bad science (2016) · cites this paper as [63]
In the citing paper
“For most published research findings to be false in psychology and neuroscience, for example, one must assume that the hypotheses tested in these disciplines are correct much less than 50% of the time (14, 18).”
From Is science really facing a reproducibility crisis, and do we need it to? (2018) · cites this paper as [18]
In the citing paper
“Indeed, power analyses show that resting-state functional connectivity for individual links in people is not robust with small samples (Button et al., 2013; Poldrack, 2019), especially in populations with psychiatric disorders, as we also have found.”
From The orbitofrontal cortex: reward, emotion and depression (2020) · cites this paper as Button et al., 2013
In the citing paper
“For example, in neuroscience, even with lenient assumptions, the average power of an experiment is about 20%13,14.”
From How to design preclinical studies in nanomedicine and cell therapy to maximize the prospects of clinical translation (2018) · cites this paper as [13]
In the citing paper
“Overall the sample size was relatively small, which may limit reproducibility of results (Button et al., 2013), but was comparable to other relevant fMRI studies (Christakou et al., 2013, Richey et al., 2012, Rubia et al., 2009b).”
From Differentiating neural reward responsiveness in autism versus ADHD (2014) · cites this paper as Button et al., 2013
In the citing paper
“even if a lack of statistical power is a major cause of concern in neuroscience and may be imputed for several studies (Button et al., 2013)”
From What visual illusions teach us about schizophrenia (2014) · cites this paper as Button et al., 2013
In the citing paper
“Many studies included relatively small sample sizes potentially limiting the ability to observe true effects (Button et al., 2013).”
From Neural networks involved in adolescent reward processing: an activation likelihood estimation meta-analysis of functional neuroimaging studies (2015) · cites this paper as Button et al., 2013
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Power failure: why small sample size undermines the reliability of neuroscience > Animal model studies.
“Studies of this size can only detect very large effects (d = 1.20 for n = 22, and d = 1.26 for n = 24) with 80% power”
§ Power failure: why small sample size undermines the reliability of neuroscience > Ethical implications.
“the average sample size of 22 animals for the water maze experiments was only sufficient to detect an effect size of d = 1.26 with Table 1 (cont.)| Characteristics of included meta-analyses Study k N Summary effect size Power Refs Median (range) Cohen's d OR Random or fixed effects Median (range) Yang 3 51 (18-205) 0.67 NA 0.65 (0.27-1.00) 67 Yuan 14 116.5 (19-1178) 4.98 Fixed 0.92”
Cited works that were retracted
Works this paper cites that were later retracted, from the retraction notices publishers deposit. Major when the notice came out before this paper was published.
A note, not an issue: for example, a cited work that was retracted only after this paper was published, which its authors could not have known.
In the paper
“Fang, F. C. & Casadevall, A. Retracted science and the retraction index. Infect. Immun. 79, 3855–3859 (2011).”
Cited work 10.1128/iai.05661-11 · retraction date not known · this paper published 2013-04-10
Citations
Cites
1 retracted
Retracted Science and the Retraction Index
Fang, F. C., & Casadevall, A. (2011) · Infection and Immunity · doi:10.1128/iai.05661-11 · Cited by 405
RetractedFaul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007) · Behavior Research Methods · doi:10.3758/bf03193146 · Cited by 65,971
Why Most Published Research Findings Are False
Ioannidis, J. P. A. (2005) · PLoS Medicine · doi:10.1371/journal.pmed.0020124 · Cited by 10,505
Self-citationSterne, J. A. C., Sutton, A. J., Ioannidis, J. P. A., Terrin, N., Jones, D. R., Lau, J., Carpenter, J., Rucker, G., Harbord, R. M., Schmid, C. H., et al. (2011) · BMJ · doi:10.1136/bmj.d4002 · Cited by 7,709
Self-citationImproving Bioscience Research Reporting: The ARRIVE Guidelines for Reporting Animal Research
Kilkenny, C., Browne, W. J., Cuthill, I. C., Emerson, M., & Altman, D. G. (2010) · PLoS Biology · doi:10.1371/journal.pbio.1000412 · Cited by 7,300
Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011) · Psychological Science · doi:10.1177/0956797611417632 · Cited by 6,567
Improving bioscience research reporting: The ARRIVE guidelines for reporting animal research
Kilkenny, C., Browne, W. J., Cuthill, I. C., Emerson, M., & Altman, D. G. (2010) · Journal of Pharmacology and Pharmacotherapeutics · doi:10.4103/0976-500x.72351 · Cited by 6,277
The Principles of Humane Experimental Technique
(1960) · The Medical Journal of Australia · doi:10.5694/j.1326-5377.1960.tb73127.x · Cited by 5,045
Why Most Published Research Findings Are False
Ioannidis, J. P. A. (2005) · CHANCE · doi:10.1080/09332480.2005.10722754 · Cited by 4,702
Self-citationBelief in the law of small numbers.
Tversky, A., & Kahneman, D. (1971) · Psychological Bulletin · doi:10.1037/h0031322 · Cited by 3,246
Research works printed but not linked: 2
- David, S. P. et al. Potential reporting bias in small fMRI studies of the brain. PLoS Biol. (in the press).
- Ioannidis, J. P., Tarone, R. & McLaughlin, J. K. The false-positive to false-negative ratio in epidemiologic studies. Epidemiology 22, 450–456 (2011). · DOI · Paper page
Cited by
Sorted by citations
Items 2,601–2,700 of 7,720
Valenzuela‐Fernández, L., Escobar‐Farfán, M., Guerra-Velásquez, M., & García-Salirrosas, E. E. (2023) · International Journal of Environmental Research and Public Health · doi:10.3390/ijerph20043330 · Cited by 25
“80 of the minimums required, which allows us to affirm that there is no problem with distant sample size differences [98] (See Table 1).”
BackgroundKeating, C. T., Ichijo, E., & Cook, J. L. (2023) · Scientific Reports · doi:10.1038/s41598-023-39070-0 · Cited by 25
Nierula, B., Stephani, T., Bailey, E., Kaptan, M., Pohle, L.-M. G., Horn, U., Mouraux, A., Maeß, B., Villringer, A., Curio, G., et al. (2024) · PLoS Biology · doi:10.1371/journal.pbio.3002828 · Cited by 23
“In any case, considering that underpowered studies are a troubling issue in neuroscience [97], both experimental and clinical studies that could arise from this work would need to be well powered.”
Etienne, F., Lurier, T., Yugueros-Marcos, J., & Mateus, A. L. P. (2025) · International Journal of Antimicrobial Agents · doi:10.1016/j.ijantimicag.2025.107505 · Cited by 23
Dworkin, R. H., Turk, D. C., Peirce‐Sandner, S., He, H., McDermott, M., Hochberg, M. C., Jordan, J. M., Katz, N. P., Lin, A. H., Neogi, T., et al. (2014) · Arthritis & Rheumatology · doi:10.1002/art.38869 · Cited by 23
“However, it is also possible that these associations reflect more careful design or implementation of the trial protocol or the enrollment of patients who are more likely to respond to treatment (27,29,30).”
BackgroundHuman muscle fatigue, eccentric damage and coherence in the EMG
Héroux, M. E., & Gandevia, S. C. (2013) · Acta Physiologica · doi:10.1111/apha.12133 · Cited by 22
Reproducibility Issues: Avoiding Pitfalls in Animal Inflammation Models
Laman, J. D., Kooistra, S. M., & Clausen, B. E. (2017) · Methods in molecular biology · doi:10.1007/978-1-4939-6786-5_1 · Cited by 23
Decomposing the causes of the socioeconomic status-health gradient with biometrical modeling.
Garrison, S. M., & Rodgers, J. L. (2018) · Journal of Personality and Social Psychology · doi:10.1037/pspp0000226 · Cited by 21
“Specifically, although we replicated the shared-environmental correlation between SES and mental health, that effect could still be subject to the “winner’s curse” (Button et al., 2013).”
BackgroundKing, M. A., Wissow, L. S., & Baum, R. (2017) · Health Care Management Review · doi:10.1097/hmr.0000000000000169 · Cited by 22
“We entered into statistical analyses with a priori hypotheses and set data analysis plans to help guard against the Bvibration of effects[ to which studies with modest sample sizes are susceptible (Button et al., 2013).”
MethodBanerjee, S., Garrison, L. P., Flum, D. R., & Arterburn, D. E. (2017) · Obesity · doi:10.1002/oby.21927 · Cited by 22
Marino, M. J. (2017) · Biochemical Pharmacology · doi:10.1016/j.bcp.2017.12.011 · Cited by 25
“Button and colleagues [20] performed a meta-analysis to estimate the median statistical power in the field of neuroscience and came up with an ‘‘optimistic” estimate that power ranged from 0.”
BackgroundSteinfath, M., Vogl, S., Violet, N., Schwarz, F., Mielke, H., Selhorst, T., Greiner, M., & Schönfelder, G. (2018) · PLoS ONE · doi:10.1371/journal.pone.0202762 · Cited by 26
Takeuchi, H., Taki, Y., Nouchi, R., Yokoyama, R., Kotozaki, Y., Nakagawa, S., Sekiguchi, A., Iizuka, K., Hanawa, S., Araki, T., et al. (2018) · Brain Structure and Function · doi:10.1007/s00429-018-1747-5 · Cited by 23
“2009) or more likely to yield false positives (Button et al. 2013).”
Schnell, S. D. (2018) · Bulletin of Mathematical Biology · doi:10.1007/s11538-018-0500-9 · Cited by 25
“…science that include bias in hypothesis testing (Fanelli 2010, Fanelli et al 2017), poor experimental design, statistical problems (John et al 2012, Button et al 2013) and lack of standardization to report and share experimental protocols, results and data (Wicherts et al. 2006, Nature Editorial…”
BackgroundPreregistration and reproducibility
Strømland, E. (2019) · Journal of Economic Psychology · doi:10.1016/j.joep.2019.01.006 · Cited by 26
Giehl, K., Tahmasian, M., Eickhoff, S. B., & van Eimeren, T. (2019) · Parkinsonism & Related Disorders · doi:10.1016/j.parkreldis.2019.02.015 · Cited by 20
Basu, N., Kaplan, C. M., Ichesco, E., Larkin, T. E., Schrepf, A., Murray, A. D., Clauw, D. J., Waiter, G. D., & Harris, R. E. (2019) · Rheumatology · doi:10.1093/rheumatology/kez132 · Cited by 24
How, M.-L. (2019) · Big Data and Cognitive Computing · doi:10.3390/bdcc3030046 · Cited by 24
“The Bayesian network (BN) [45–47] approach is suitable for analyzing non-parametric data from a small number of participants, because it does not require the underlying variables of a model to assume or have a normal parametric distribution [42,48,49].”
Backgroundvan Elk, M., & Snoek, L. (2019) · European Journal of Neuroscience · doi:10.1111/ejn.14563 · Cited by 21
“The field of neuroscience is plagued with many low‐powered studies and accordingly the literature abounds with many false‐positive findings, resulting in an overall inconsistent and scattered pattern of results (Button et al., 2013).”
BackgroundEmergency braking at intersections: A motion-base motorcycle simulator study
Kovácsová, N., Grottoli, M., Celiberti, F., et al. (2019) · Applied Ergonomics · doi:10.1016/j.apergo.2019.102970 · Cited by 22
Intrinsic Functional Connectivity of Dentate Nuclei in Autism Spectrum Disorder
Anteraper, S. A., Guell, X., Taylor, H. P., D’Mello, A. M., Whitfield, S., & Joshi, G. (2019) · Brain Connectivity · doi:10.1089/brain.2019.0692 · Cited by 22
“Behavioral correlations in studies with modest sample sizes are inherently limited (Button et al., 2013).”
BackgroundReliable measurement in sport psychology: The case of performance outcome measures
Schweizer, G., Furley, P., Rost, N., & Barth, K. (2020) · Psychology of sport and exercise · doi:10.1016/j.psychsport.2020.101663 · Cited by 27
“For example, future research is supposed to benefit from appropriate sample sizes 42 (Button et al., 2013; Fraley & Vazire, 2014; Schönbrodt & Perugini, 2013; Schweizer & Furley, 43 2016), methodological advances and new software1 (e.g., Wagenmakers et al., 2018; Love et al., 44 2019),…”
BackgroundSituational experience around the world: A replication and extension in 62 countries
Lee, D. I., Gardiner, G., Baranski, E. N., Members of The International Situations Project, & Funder, D. C. (2020) · Journal of Personality · doi:10.1111/jopy.12558 · Cited by 21
“Extending to fields outside of psychology, (Button et al., 2013; Camerer et al., 2016; Cova et al., in press) replication projects have provided insights into what we know, or what we thought we knew, about human behavior (Simmons, Nelson, & Simonsohn, 2011; Open Science Collaboration, 2015).”
BackgroundInsiders and Outsiders: Lessons for Neuroethics from the History of Bioethics
Chiong, W. (2020) · AJOB Neuroscience · doi:10.1080/21507740.2020.1778118 · Cited by 22
“One is the “reproducibility crisis” (Button et al. 2013), which may require a thorough examination of incentives and pressures in academic neuroscience.”
BackgroundAdvancing the Cognitive Science of Religion through Replication and Open Science
Hoogeveen, S., & van Elk, M. (2020) · Journal for the Cognitive Science of Religion · doi:10.1558/jcsr.39039 · Cited by 21
Social cognition in the blind brain: A coordinate‐based meta‐analysis
Arioli, M., Ricciardi, E., & Cattaneo, Z. (2020) · Human Brain Mapping · doi:10.1002/hbm.25289 · Cited by 25
“These constraints are known to increase the likelihood of false negatives (Button et al., 2013), thus pushing researchers toward procedures which, conversely, might promote false positives (Eklund, Nichols, & Knutsson, 2016; Muller et al., 2018).”
BackgroundLewis, G., Srinivasan, R., Roiser, J. P., Blakemore, S., Flouri, E., & Lewis, G. (2021) · Psychological Medicine · doi:10.1017/s0033291720005000 · Cited by 24
Quarshie, E. N., Shuweihdi, F., Waterman, M. G., & House, A. O. (2021) · BMJ Open · doi:10.1136/bmjopen-2020-041609 · Cited by 24
Richards, A. L., Inslicht, S. S., Yack, L. M., Metzler, T. J., Huie, J. R., Straus, L. D., Dukes, C., Hubachek, S. Q., Felmingham, K. L., Mathalon, D. H., et al. (2021) · SLEEP · doi:10.1093/sleep/zsab271 · Cited by 23
Brain Processing of Emotional Scenes in Aging: Effect of Arousal and Affective Context
Mathieu, N. G., Gentaz, É., Harquel, S., Vercueil, L., Chauvin, A., Bonnet, S., & Campagne, A. (2014) · PLoS ONE · doi:10.1371/journal.pone.0099523 · Cited by 22
“8 [62].”
Development of Posterior Medial Frontal Cortex Function in Pediatric Obsessive-Compulsive Disorder
Fitzgerald, K. D., Liu, Y., Johnson, T. D., Moser, J. S., Marsh, R., Hanna, G. L., & Taylor, S. F. (2018) · Journal of the American Academy of Child & Adolescent Psychiatry · doi:10.1016/j.jaac.2018.02.016 · Cited by 22
“These inconsistencies may relate to small sample sizes (n 1⁄4 <25 per group), which increase variability and reduce experimental power.(20) In addition, both pediatric and adult OCD cohorts often include patients taking selective serotonin reuptake inhibitors, which have an impact on neural engagement during performan…”
P-Curve and Effect Size: Correcting for Publication Bias Using Only Significant ResultsPreprint
Simonsohn, U., Nelson, L. D., & Simmons, J. P. (2014) · SSRN Electronic Journal · doi:10.2139/ssrn.2377290 · Cited by 22
7 Deadly Sins in Educational Research
Picho, K., & Artino, A. R. (2016) · Journal of Graduate Medical Education · doi:10.4300/jgme-d-16-00332.1 · Cited by 22
“In addition to missing a true difference between groups, low power also reduces the likelihood that a statistically significant result represents a true effect rather than a spurious finding.(13) Both of these issues weaken the reliability of findings in a given field.”
BackgroundLevitas, D. J., Hayashi, S., Vinci‐Booher, S., Heinsfeld, A. S., Bhatia, D., Lee, N., Galassi, A., Niso, G., & Pestilli, F. (2024) · Scientific Data · doi:10.1038/s41597-024-02959-0 · Cited by 23
“Data sharing is of great importance to the neuroimaging community for addressing the longstanding issue of under-powered studies and improving reproducibility 3,32–34 .”
BackgroundSantel, M., Neuner, F., Berg, M., Steuwe, C., Jobes, D. A., Driessen, M., & Beblo, T. (2023) · Frontiers in Psychiatry · doi:10.3389/fpsyt.2023.1038302 · Cited by 21
An implementation trial of ACT-based bibliotherapy for irritable bowel syndrome
Gillanders, D., Ferreira, N. B., Angioni, E., Carvalho, S. A., & Eugenicos, M. P. (2017) · Journal of Contextual Behavioral Science · doi:10.1016/j.jcbs.2017.04.006 · Cited by 21
“Small samples can also lead to spurious non-null findings and to overestimation of the magnitude of effects (Button et al., 2013).”
BackgroundMalik-Moraleda, S., Jouravlev, O., Taliaferro, M., Mineroff, Z., Cucu, T., Mahowald, K., Blank, I., & Fedorenko, E. G. (2024) · Cerebral Cortex · doi:10.1093/cercor/bhae049 · Cited by 32
Rahapsari, S., & Levita, L. (2024) · Trauma Violence & Abuse · doi:10.1177/15248380241286812 · Cited by 31
“Consequently, the smaller number of studies on working memory and cognitive flexibility may limit the ability to detect significant associations with ACEs in these domains (Button et al., 2013), potentially skewing the apparent strength of the relationship toward inhibitory control.”
Degen, P. M., & Medo, M. (2025) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011630 · Cited by 25
Heitland, I., Groenink, L., van Gool, J. M. G., Domschke, K., Reif, A., & Baas, J. M. P. (2015) · Genes Brain & Behavior · doi:10.1111/gbb.12276 · Cited by 23
Speed of Visual Search in Old Age: 1950 to 2016
Rabbitt, P. (2016) · The Journals of Gerontology Series B · doi:10.1093/geronb/gbw097 · Cited by 21
“Some cases are due to inadequate statistical power of comparisons (Button et al., 2013) and to “p-hacking”—post hoc selection of some significant among many nonsignificant possible comparisons within data sets (Simmons, Nelson, & Simonsohn, 2011).”
ResultRe-examining the influence of attention and consciousness on visual afterimage duration.
Travis, S. L., Dux, P. E., & Mattingley, J. B. (2017) · Journal of Experimental Psychology Human Perception & Performance · doi:10.1037/xhp0000458 · Cited by 21
“Thus, a statistically significant finding from a low-powered study is more likely to overestimate the effect size when compared with a statistically significant finding from a high-powered study (Button et al., 2013).”
Resultde Lara, C. M. R., Navas, J. F., Soriano‐Mas, C., Sescousse, G., & Perales, J. C. (2018) · International Gambling Studies · doi:10.1080/14459795.2018.1448427 · Cited by 25
“And fourth, although samples are larger than usual in GD neurobiological research, it is also true that they could be considered not large enough to overcome some recent critiques to neuroimaging research (Button et al., 2013).”
BackgroundDepardon, E., Kanoun, S., Humbert, O., et al. (2018) · PLoS ONE · doi:10.1371/journal.pone.0199529 · Cited by 21
Ankle angle variability during running in athletes with chronic ankle instability and copers
Wanner, P., Schmautz, T., Kluge, F., Eskofier, B. M., Pfeifer, K., & Steib, S. (2018) · Gait & Posture · doi:10.1016/j.gaitpost.2018.11.038 · Cited by 21
Cui, X., Yang, Q., Li, B., Tang, J., Zhang, X., Li, S., Li, F., Hu, J., Lou, Y., Qiu, Y., et al. (2019) · Frontiers in Pharmacology · doi:10.3389/fphar.2019.00127 · Cited by 21
“, 2006; Subramanian, 2016), there is an urgent need tomaximally enlarge the sample size and in turn enhance the statistical power of a given metabolomics study (Button et al., 2013).”
BackgroundStatistical Consequences of using Multi-armed Bandits to Conduct Adaptive Educational Experiments
Rafferty, A. N., Ying, H., & Williams, J. J. (2019) · Journal of Educational Data Mining · doi:10.5281/zenodo.3554749 · Cited by 22
“Overestimation of effect sizes is known to occur in low-powered studies (see, e.g., Button et al. 2013), but here, the overestimation is not due to filtering out nonsignificant simulations.”
BackgroundPower(ful) guidelines for experimental economists
Vasilaky, K., & Brock, J. (2020) · Journal of the Economic Science Association · doi:10.1007/s40881-020-00090-5 · Cited by 22
Cortical morphometry and neural dysfunction in Huntington’s disease: a review
Tan, B., Shishegar, R., Poudel, G. R., Fornito, A., & Georgiou‐Karistianis, N. (2020) · European Journal of Neurology · doi:10.1111/ene.14648 · Cited by 24
“A challenge for longitudinal neuroimaging investigations is recruiting a sample that is sufficiently powered to detect statistically significant differences [150,151].”
BackgroundKoenis, M. M. G., Durnez, J., Rodrigue, A., Mathias, S. R., Alexander‐Bloch, A. F., Barrett, J. A., Doucet, G. E., Frangou, S., Knowles, E. E. M., Mollon, J., et al. (2020) · Human Brain Mapping · doi:10.1002/hbm.25324 · Cited by 21
“First, many studies use small sample sizes with limited statistical power, which reduces the chance of detecting a true effect and decreases the likelihood that a statistically significant result reflects a true effect (Button et al., 2013).”
BackgroundCortical thickness in Parkinson's disease: a coordinate-based meta-analysis
Sheng, L., Zhao, P., Ma, H., Raduà, J., Yi, Z., Shi, Y., Zhong, J., Dai, Z., & Pan, P. (2021) · Aging · doi:10.18632/aging.202368 · Cited by 21
“A small sample size results in low statistical power, making it difficult to detect subtle differences and undermining the reliability of results [103].”
BackgroundFeelings of shame and guilt are associated with distinct neural activation in youth
Bastin, C., Rakesh, D., Harrison, B. J., Davey, C. G., Allen, N. B., Muller, S., & Whittle, S. (2021) · Biological Psychology · doi:10.1016/j.biopsycho.2021.108025 · Cited by 23
Diving into the deep end: a personal reflection on the MyConnectome study
Poldrack, R. A. (2020) · Current Opinion in Behavioral Sciences · doi:10.1016/j.cobeha.2020.10.008 · Cited by 20
Randomized Manipulation of Early Cognitive Experience Impacts Adult Brain Structure
Farah, M. J., Sternberg, S., Nichols, T. A., Duda, J., Lohrenz, T. M., Luo, Y., Sonnier‐Netto, L., Ramey, S. L., Montague, P. R., & Ramey, C. T. (2021) · Journal of Cognitive Neuroscience · doi:10.1162/jocn_a_01709 · Cited by 20
“The trend toward larger samples has been motivated in part by the realization that they reduce the risk of false positives, in addition to the more obvious reduction in risk of false negatives (Button et al., 2013).”
BackgroundA systematic review of neuroimaging and acute cannabis exposure in age-of-risk for psychosis
Cupo, L., Plitman, E., Guma, E., & Chakravarty, M. M. (2021) · Translational Psychiatry · doi:10.1038/s41398-021-01295-w · Cited by 21
The C-BIG Repository: an Institution-Level Open Science Platform
Das, S., Abou-Haidar, R., Rabalais, H., Sun, S. D. L. W., Rosli, Z., Chatpar, K., Boivin, M., Tabatabaei, M., Rogers, C., Legault, M., et al. (2021) · Neuroinformatics · doi:10.1007/s12021-021-09516-9 · Cited by 20
“Furthermore, as machine and deep learning techniques are beginning to mature, the need for larger sample sizes of easily accessible data is becoming stronger (Button et al. 2013; Turner et al. 2018).”
BackgroundLai, L., Gu, W., Hu, M., Wang, W., Wang, S., Huai, Y., Mei, J., & Wang, C. (2021) · Aging · doi:10.18632/aging.203100 · Cited by 23
Hu, H., Xu, Z., Zhu, Q., Liu, X., Jiang, S., & Zheng, J. (2021) · Frontiers in Immunology · doi:10.3389/fimmu.2021.669398 · Cited by 22
Wegener, D., Fabrigar, L., Pek, J., & Hoisington-Shaw, K. J. (2021) · Personality and Social Psychology Bulletin · doi:10.1177/01461672211030811 · Cited by 16
“These false findings have also been discussed as contributing to disappointing replication rates in psychology and other disciplines (e.g., Button et al., 2013; Christley, 2010; Ioannidis, 2005; Pashler & Harris, 2012).”
BackgroundGrosselin, F., Breton, A., Yahia‐Cherif, L., Wang, X., Spinelli, G., Hugueville, L., Fossati, P., Attal, Y., Navarro‐Sune, X., Chávez, M., et al. (2021) · Scientific Reports · doi:10.1038/s41598-021-96893-5 · Cited by 21
The Drunk Utilitarian Revisited: Does Alcohol Really Increase Utilitarianism in Moral Judgment?
Paruzel‐Czachura, M., Pypno‐Blajda, K., Everett, J. A. C., Białek, M., & Gawronski, B. (2021) · Personality and Social Psychology Bulletin · doi:10.1177/01461672211052120 · Cited by 24
“Based on these findings, the influence of alcohol on inhibitory control seems much more complex than presumed in research on the effects of alcohol on moral judgment, including this study (see Button et al., 2013).”
BackgroundResultStern, J., & Shiramizu, V. K. M. (2021) · Hormones and Behavior · doi:10.1016/j.yhbeh.2021.105103 · Cited by 21
“Therefore, all between-subjects studies investigating the CPH so far seem to be heavily underpowered, which can lead to false negative, but also to false positive findings (Button et al., 2013).”
Carter, A. R., Harrison, S., Gill, D., Smith, G. D., Taylor, A. E., Howe, L. D., & Davies, N. M. (2022) · International Journal of Epidemiology · doi:10.1093/ije/dyac002 · Cited by 22
“Low statistical power reduces the chance of detecting a true effect should one exist.(46) Although some power calculators have been developed to calculate power in gene*environment interaction analyses,(47) to our knowledge, none has been developed for use with PGSs.”
BackgroundManeejuk, P., Yamaka, W., & Srichaikul, W. (2022) · Mathematics · doi:10.3390/math10050723 · Cited by 21
Webb, C., Bennett, D. L., & Bywaters, P. (2022) · Social Policy and Society · doi:10.1017/s147474642200001x · Cited by 21
“While this side-steps problems caused by the lack of independence of observations within judgements of the same local authority and inspection framework over time, it curtails the power of models to identify multiple or small-to-moderate effects (Button et al. , 2013).”
de Ridder, J. (2022) · Studies in History and Philosophy of Science Part A · doi:10.1016/j.shpsa.2022.02.003 · Cited by 24
Cooperation with Strangers: Spillover of Community Norms
Molina, M. D., Nee, V. G., & Holm, H. J. (2022) · Organization Science · doi:10.1287/orsc.2021.1521 · Cited by 19
“We believe that this is due to the small sample size used to estimate this effect and that a larger sample size should replicate our fi nding (Button et al. 2013, Nuzzo 2014).”
ML4STEM Professional Development Program: Enriching K-12 STEM Teaching with Machine Learning
Tang, J., Zhou, X., Wan, X., Daley, M., & Bai, Z. (2022) · International Journal of Artificial Intelligence in Education · doi:10.1007/s40593-022-00292-4 · Cited by 29
“First , our study held a small sample size (N = 18), which constrains statistical power and generalizability of the findings to make inferences for a broader context (Button et al., 2013).”
BackgroundTyagi, P., Dhyani, M., Khattri, S., Tejan, V., Tikka, S. K., & Garg, S. (2022) · Asian Journal of Psychiatry · doi:10.1016/j.ajp.2022.103176 · Cited by 21
Mertens, Y. L., Manthey, A., Sierk, A., Walter, H., & Daniels, J. K. (2022) · BJPsych Open · doi:10.1192/bjo.2022.65 · Cited by 23
Lessons from a catalogue of 6674 brain recordings
Makin, A. D. J., Tyson‐Carr, J., Rampone, G., Derpsch, Y., Wright, D., & Bertamini, M. (2022) · eLife · doi:10.7554/elife.66388 · Cited by 23
Blueprints for measuring natural behavior
Puścian, A., & Knapska, E. (2022) · iScience · doi:10.1016/j.isci.2022.104635 · Cited by 22
“…and, at the same time, most problematic aspect of conventional behavioral testing is its notorious irreproducibility (Figure 1) (Mandillo et al., 2008; Button et al., 2013; Gilmore et al., 2017; Kafkafi et al., 2018; Bodden et al., 2019; Richter, 2020; Saré et al., 2021; Andrews et al., 2018).”
BackgroundTybur, J. M., Fan, L., Jones, B. C., Holzleitner, I. J., Lee, A. J., & DeBruine, L. M. (2022) · Evolution and Human Behavior · doi:10.1016/j.evolhumbehav.2022.01.003 · Cited by 22
Huang, M., Zhang, X., Chen, X., Mai, Y., Wu, X., Zhao, J., & Feng, Q. (2022) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2022.3188560 · Cited by 25
Harmonization-Information Trade-Offs for Sharing Individual Participant Data in Biomedicine
Espin, A. T. (2022) · Harvard Data Science Review · doi:10.1162/99608f92.a9717b34 · Cited by 23
“In addition, most biomedical research studies are statistically underpowered due to small sample sizes (Button et al., 2013; Dumas-Mallet et al., 2017).”
BackgroundSolana, P., & Santiago, J. (2022) · Neuroscience & Biobehavioral Reviews · doi:10.1016/j.neubiorev.2022.104834 · Cited by 24
“P-curve analysis does not allow us to discern between these possibilities, but what is clear is that low power reduces research outcomes’ value, making science uninterpretable, non-replicable, and inefficient (Button et al., 2013).”
BackgroundInfluentialTen guidelines for designing motor learning studies
Ranganathan, R., Lee, M.-H., & Krishnan, C. (2022) · Brazilian Journal of Motor Behavior · doi:10.20338/bjmb.v16i2.283 · Cited by 25
Hessels, J., Kroon, S., Boerman, S., et al. (2022) · Journal of Clinical Medicine · doi:10.3390/jcm11185280 · Cited by 21
“The small sample size reduces the chance of detecting a true effect and reduces the likelihood that a statistically significant result reflects a true effect [28].”
BackgroundSudre, G. P., Norman, L. J., Bouyssi-Kobar, M., Price, J., Shastri, G. G., & Shaw, P. (2022) · Biological Psychiatry · doi:10.1016/j.biopsych.2022.09.021 · Cited by 21
Fraley, R. C., Chong, J. Y., Baacke, K. A., Greco, A. J., Guan, H., & Vazire, S. (2022) · Advances in Methods and Practices in Psychological Science · doi:10.1177/25152459221120217 · Cited by 24
“Historically, statistical power has been overlooked in the design and evaluation of empirical research (Button et al., 2013; Cohen, 1988, 1990, 1992; Maxwell, 2004; Szucs & Ioannidis, 2017).”
BackgroundMulti-Resolution Grids in Earthquake Forecasting: The Quadtree Approach
Asim, K. M., Schorlemmer, D., Hainzl, S., Iturrieta, P. C., Savran, W. H., Bayona, J. A., & Werner, M. J. (2022) · Bulletin of the Seismological Society of America · doi:10.1785/0120220028 · Cited by 23
Published correlational effect sizes in social and developmental psychology
Weinerová, J., Szűcs, D., & Ioannidis, J. P. A. (2022) · Royal Society Open Science · doi:10.1098/rsos.220311 · Cited by 27
Self-citationZhang, M., Wang, L., Li, X., Song, L., Luo, D., Li, Q., Wang, Y., Wan, Z., & Mei, S. (2023) · The Science of The Total Environment · doi:10.1016/j.scitotenv.2023.162935 · Cited by 22
Exercise as adjunctive therapy for systemic lupus erythematosus
Frade, S., O'Neill, S. G., Greene, D., Nutter, E., & Cameron, M. (2023) · Cochrane Database of Systematic Reviews · doi:10.1002/14651858.cd014816.pub2 · Cited by 23
Amd, M., & Roche, B. (2018) · The Psychological Record · doi:10.1007/s40732-018-0273-4 · Cited by 22
“997, the positive predictive value (PPV; Button et al., 2013) was calculated to be 91%, which is the probability of the regression equation being a true effect.”
BackgroundEculizumab for treating patients with paroxysmal nocturnal hemoglobinuria
Martí-Carvajal, A. J., Anand, V., Cardona, A. F., & Solà, I. (2014) · Cochrane Database of Systematic Reviews · doi:10.1002/14651858.cd010340.pub2 · Cited by 24
Three Paths to Drug Discovery in Psychiatry
Berk, M., & Nierenberg, A. A. (2015) · American Journal of Psychiatry · doi:10.1176/appi.ajp.2014.14070858 · Cited by 21
Miller, K., Margolin, G., Shapiro, L. S., & Timmons, A. C. (2016) · Journal of Research on Adolescence · doi:10.1111/jora.12250 · Cited by 21
“Fourth, our sample of 99 youth may have been somewhat underpowered to test all of the hypothesized two- and three-way interactions; underpowered studies increase the risk of producing both type I and type II errors (Button et al., 2013).”
Lindner, M. D., Vancea, A., Chen, M., & Chacko, G. W. (2015) · American Journal of Evaluation · doi:10.1177/1098214015582049 · Cited by 22
“For example, most studies published in the literature have grossly inadequate sample size and statistical power (Bakker, van Dijk, & Wicherts, 2012; Button et al., 2013), fail to adequately identify reagents and their sources (Vasilevsky et al.”
BackgroundWilliamson, M. J., Kavanagh, A. S., Noad, M. J., et al. (2016) · Marine Mammal Science · doi:10.1111/mms.12324 · Cited by 21
“As such, it is likely that statistical power was low, and, therefore, the likelihood that the nonsignificant results found in NTA groups reflect a true effect is reduced (Wisz et al. 2008, Button et al. 2013).”
ResultHow to Use Neuromeasures To Make Better Advertising Decisions
Kennedy, R., & Northover, H. (2016) · Journal of Advertising Research · doi:10.2501/jar-2016-019 · Cited by 21
McClinton, S. M., Collazo, C., Vincent, E., & Vardaxis, V. G. (2016) · Journal of Orthopaedic and Sports Physical Therapy · doi:10.2519/jospt.2016.6482 · Cited by 21
Smith, J. L., Dash, N. J., Johnstone, S. J., Houben, K., & Field, M. (2017) · Drug and Alcohol Dependence · doi:10.1016/j.drugalcdep.2016.12.018 · Cited by 24
Matheson, G. J., Stenkrona, P., Cselényi, Z., Plavén‐Sigray, P., Halldin, C., Farde, L., & Červenka, S. (2017) · NeuroImage · doi:10.1016/j.neuroimage.2017.04.031 · Cited by 21
Genon, S., Wensing, T., Reid, A., Hoffstaedter, F., Caspers, S., Grefkes, C., Nickl‐Jockschat, T., & Eickhoff, S. B. (2017) · NeuroImage · doi:10.1016/j.neuroimage.2017.05.053 · Cited by 18
“…concerns about adjustment for demographic covariates, there is an ongoing debate about the influence of sample size and power in structural brain behavior correlation and neuroimaging in general (e.g. (Friston, 2012; Button et al., 2013; Friston, 2013; Ingre, 2013; Carter et al., 2016)).”
Increased Global Interaction Across Functional Brain Modules During Cognitive Emotion Regulation
Brandl, F., Bratec, S. M., Xie, X., Wohlschläger, A. M., Riedl, V., Meng, C., & Sorg, C. (2017) · Cerebral Cortex · doi:10.1093/cercor/bhx178 · Cited by 21
Linnman, C., Catana, C., Petkov, M. P., Chonde, D. B., Becerra, L. R., Hooker, J. M., & Borsook, D. (2017) · NeuroImage Clinical · doi:10.1016/j.nicl.2017.11.011 · Cited by 23
Fernández, A. H., Mora, E., & Hernández, M. I. V. (2018) · Physiology & Behavior · doi:10.1016/j.physbeh.2018.04.023 · Cited by 21
Archer, D. P., McCann, S., Walker, A. M., Premji, Z., Rogan, K. J., Hutton, M. J. H., & Gray, L. J. (2018) · British Journal of Anaesthesia · doi:10.1016/j.bja.2018.07.024 · Cited by 21
It’s all about gains: Risk preferences in problem gambling.
Ring, P., Probst, C. C., Neyse, L., Wolff, S., Kaernbach, C., van Eimeren, T., Camerer, C. F., & Schmidt, U. (2018) · Journal of Experimental Psychology General · doi:10.1037/xge0000418 · Cited by 20
Metadata
- Authors
- Button, K. S., Ioannidis, J. P. A., Mokrysz, C., Nosek, B. A., Flint, J., Robinson, E. S. J., & Munafò, M. R.
- Affiliations
- University of Bristol (GB)
- Stanford Medicine (US)
- University of Virginia (US)
- Venue
- Nature Reviews Neuroscience
- Published
- 2013-04-10
- Volume
- 14
- Issue
- 5
- Pages
- 365-376
- ISSN
- 1471-003X (ISSN-L)
- 1471-0048
- OpenAlex
- W2105824687
- Semantic Scholar
- c2bc880156c10fc8d1752cd3d528e4ef96cbaac4
- Cited by
- 7,641
- References
- 106
- Citations by year
- 2026: 584 · 2025: 611 · 2024: 606 · 2023: 553 · 2022: 658 · 2021: 692 · 2020: 752 · 2019: 683 · 2018: 679 · 2017: 597 · 2016: 486 · 2015: 352 · 2014: 268 · 2013: 109
- FWCI
- 82.23
- Type
- journal article
- Language
- en
- Update notices
- erratum (2013-04-15)
- Open access
- bronze
- doi.org
- Locations
- Nature reviews. Neuroscience (publishedVersion)
- PubMed (publishedVersion)
- Explore Bristol Research (submittedVersion)
- UCL Discovery (University College London) (submittedVersion)
- Funders
- Wellcome Trust (090532)
- United Kingdom Clinical Research Collaboration
- Cancer Research UK
- National Institute for Health and Care Research
- British Heart Foundation
- Medical Research Council
- Economic and Social Research Council
- Topics
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Psychology Research and Bibliometrics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
- Keywords
- statistical power, small sample size, effect size overestimation, reproducibility, research ethics, research waste, research practices, brain research
- MeSH terms
- Humans, Neurosciences, Probability, Reproducibility of Results, Sample Size
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
