Cited by 256Academic Writing and PublishingAcademic Publishing and Open AccessPublishing and Scholarly Communication

Reproducibility: A tragedy of errors

Open access
Allison, David B., Brown, Andrew W., George, Brandon J., Kaiser, Kathryn A.·2016·Nature

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/530027a

Cites

9 of 9 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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?

  • Reproducibility issues usually stem from inappropriate study design or baseline comparisons, and miscalculation.

    “This usually stems from inappropriate study design or baseline comparisons, and miscalculation.”

    From Artificial Intelligence in Detecting Statistical Errors: Implications for Authors, Reviewers, and Editors (2025) · cites this paper as [19]

    The opening section describes a mathematical model that overestimated an effect by more than tenfold, and ‘What can be done?’ recommends statistical expertise for experimental design and analysis. But our copy does not identify inappropriate baseline comparisons as a cause or establish that these issues ‘usually’ cause reproducibility problems; the authors say their review was not systematic and they do not know the rate of errors. The referenced ‘Three common errors’ textbox is not present in the passages returned from our copy. (LLM, conf=0.82)

    Experimental — accuracy not yet measured
  • Sometimes, reproducibility is not feasible given a vague description of the methodology.

    “Sometimes, reproducibility is not feasible given a vague description of the methodology.19”

    From Artificial Intelligence in Detecting Statistical Errors: Implications for Authors, Reviewers, and Editors (2025) · cites this paper as [19]

    The section ‘Six Problems > No standard mechanism exists to request raw data’ describes difficulty obtaining data to check suspected errors, not an inability to reproduce work because its methodology is vaguely described. Searches for insufficient methodological detail and descriptions needed to replicate a study found no passage addressing that point in our copy. (LLM, conf=0.70)

    Experimental — accuracy not yet measured
  • Similar errors continue today, are sometimes severe enough to call entire studies into question, and may occur with nontrivial frequency.

    “Similar errors continue today, are sometimes severe enough to call entire studies into question (11), and may occur with nontrivial frequency (12–14).”
  • Correction of errors often requires an arduous process that highlights limitations of the self-correcting nature of science.

    “Unfortunately, correction of errors often requires an arduous process that highlights limitations of the self-correcting nature of science (11).”
  • The process of getting errors corrected, even through proper channels with journals, is often some combination of absurdly slow, inept, confusing, costly, time-intensive, and unsatisfying.

    “As we reported elsewhere (11), the process of getting errors corrected, even when going through proper channels with journals, is often some combination of absurdly slow, inept, confusing, costly, time-intensive, and unsatisfying.”

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

1 retracted

Flagged first, then by citations

1 of the 10 references couldn’t be matched to a known paper.

Research works printed but not linked: 1

  1. [7]George BJ Obesity in the press

Cited by

Sorted by citations

Authors
Allison, D. B., Brown, A. W., George, B. J., & Kaiser, K. A.
Affiliations
University of Alabama at Birmingham (US)
Corresponding author
Allison, David B.
Venue
Nature
Published
2016-02-03
Volume
530
Issue
7588
Pages
27-29
ISSN
  • 0028-0836 (ISSN-L)
  • 1476-4687
OpenAlex
W2263593861
Cited by
255
References
10
Citations by year
2026: 7 · 2025: 10 · 2024: 48 · 2023: 19 · 2022: 11 · 2021: 29 · 2020: 30 · 2019: 27 · 2018: 25 · 2017: 26 · 2016: 22 · 2015: 1
FWCI
175.39
Type
journal article
Language
en
Open access
Locations
  • Nature (publishedVersion)
  • PubMed (publishedVersion)
  • Europe PMC (PubMed Central) (acceptedVersion)
  • PubMed Central (acceptedVersion)
  • RePEc: Research Papers in Economics (acceptedVersion)
Funders
National Institute of Diabetes and Digestive and Kidney Diseases (P30DK056336)
Topics
  • Social Sciences › Arts and Humanities › History and Philosophy of Science › Academic Writing and Publishing
  • Social Sciences › Decision Sciences › Information Systems and Management › Academic Publishing and Open Access
  • Social Sciences › Arts and Humanities › History and Philosophy of Science › Publishing and Scholarly Communication
Keywords
reproducibility, research errors, peer review, scholarly publishing, research integrity, error correction
MeSH terms
Humans, Periodicals as Topic, Research Design, Retraction of Publication as Topic, Reproducibility of Results, Peer Review, Research, Access to Information, Editorial Policies

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.

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