Examples
Worked examples
- Is an instance
Removing outlier data points to make a non-significant result reach p < 0.05 without disclosing the exclusion.
Counter-examples
Looks similar, but isn't
- Not an instance
Pre-registered exclusion of outliers based on a documented protocol applied identically across conditions.
Editorial commentary
Falsification is one of the three components of research misconduct under the US federal FFP definition (42 CFR Part 93), alongside fabrication and plagiarism. Where fabrication means making up data or results outright, falsification means an experiment or observation genuinely happened but its record was altered so it no longer accurately represents what occurred — manipulating research materials, equipment, or processes, or changing or omitting data or results, such that the research record misrepresents the actual results. See the falsification vs fabrication comparison for a fuller side-by-side.
Common forms include selective omission of inconvenient data points without disclosure, undisclosed image adjustments that alter a figure’s interpretation (see image manipulation), changing recorded measurement values, and misrepresenting the protocol actually followed. As with the broader misconduct definition, intent and disclosure are what separate falsification from legitimate practice: a pre-registered exclusion rule applied identically across every condition and disclosed in the methods is normal analysis, not falsification of the same data.
Detection is largely forensic and after the fact. Image-forensic tools (ImageTwin, Proofig, and comparable services) screen figures for duplication and splicing; statistical forensic methods — GRIM (granularity-related inconsistency of means), SPRITE, and Benford’s-law-style digit analysis — flag reported summary statistics that are mathematically inconsistent with the stated sample size or an implausible underlying distribution. None of these tools proves intent on their own; a flagged inconsistency triggers an inquiry, not an automatic finding, since 42 CFR Part 93 requires the conduct to have been committed intentionally, knowingly, or recklessly, not merely to look statistically unusual.
Falsification findings under US federal jurisdiction can carry consequences up to debarment from federal funding; see research misconduct for how the jurisdictional standard is applied and how it differs outside the United States.
References
- US Office of Research Integrity, 42 CFR 93.212 (Falsification) and 93.103 (Requirements for findings of research misconduct). Note (checked against eCFR, August 2026): 42 CFR Part 93 has been restructured — the fabrication, falsification and plagiarism definitions now sit at 93.211, 93.212 and 93.227 respectively, not at 93.103(a)–(c) as in the earlier text that some institutional SOPs still cite; 93.103 now sets only the three requirements for a finding
- Bik, Casadevall & Fang (2016) ‘The prevalence of inappropriate image duplication in biomedical research’, mBio 7(3)
Also known as
data falsification · data manipulation
Machine-readable encodings
Use in your systems
<role vocab="credit"
vocab-identifier="https://casrai.org/dictionary/"
vocab-term="Falsification"
vocab-term-identifier="https://casrai.org/dictionary/term/falsification" />{
"@context": "https://schema.org",
"@type": "DefinedTerm",
"@id": "https://casrai.org/dictionary/term/falsification",
"name": "Falsification",
"identifier": "https://casrai.org/dictionary/term/falsification",
"description": "Manipulating research materials, equipment, processes, or data such that the research record does not accurately represent the actual results. A modification qualifies if it is undisclosed and changes the conclusions a reader would draw.",
"inDefinedTermSet": "https://casrai.org/dictionary/domain/research-integrity#set",
"url": "https://casrai.org/dictionary/term/falsification",
"alternateName": [
"data falsification",
"data manipulation"
],
"license": "https://creativecommons.org/licenses/by/4.0/",
"publisher": {
"@id": "https://casrai.org/#organization"
},
"author": {
"@id": "https://casrai.org/#editorial-team"
},
"datePublished": "2026-05-21T02:22:51",
"dateModified": "2026-09-05T17:31:48",
"inLanguage": "en-GB",
"isAccessibleForFree": true
}
