CASRAI Dictionary

Tag: scientific misconduct

  • Stanford President Research Misconduct Timeline

    Marc Tessier-Lavigne resigned as Stanford University president on 19 July 2023, days after a Special Committee-commissioned scientific panel completed a seven-month review of 12 papers he had co-authored. The panel found repeated data manipulation in laboratories he ran but cleared him of personally falsifying data. His stanford president research misconduct case remains one of the clearest recent examples of a university running a named, transparent misconduct review to its conclusion.

    Research misconduct, as defined by the US Office of Research Integrity, is “fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results.” Stanford’s inquiry into its own president tested that definition against a public figure, under public scrutiny, in real time — which is what makes its process, rather than its verdict alone, instructive for research administrators.

    What Triggered the Stanford Research Misconduct Investigation?

    The trigger was student journalism, not an internal audit. On 29 November 2022, The Stanford Daily, reported by then-freshman Theo Baker, published allegations that several papers co-authored by Tessier-Lavigne contained manipulated images, drawing on critiques that had circulated anonymously on the post-publication review site PubPeer since 2015.

    The same week, The EMBO Journal confirmed it had separately opened its own inquiry into a 2008 paper Tessier-Lavigne co-authored — a detail largely absent from mainstream coverage of the case, and a reminder that journal-level scrutiny and institutional review are distinct, parallel tracks that do not wait on each other.

    How Did Stanford’s Special Committee Investigate the Allegations?

    Stanford’s Board of Trustees announced in December 2022 that a Special Committee would review the claims. It appointed Mark Filip, a former federal judge and partner at Kirkland & Ellis, to lead the fact-finding, supported by a five-member scientific panel: Hollis Cline (Scripps Research), Kafui Dzirasa (Duke University), Steven Hyman (Harvard University, provost emeritus), Randy Schekman (UC Berkeley, former editor-in-chief of PNAS), and Shirley Tilghman (Princeton University, former president).

    Between roughly January and June 2023, the panel examined 12 papers, interviewed lab members and co-authors, and reviewed underlying research data rather than relying on published figures alone — a distinction that matters because most disputed images had already passed peer review once.

    Date Event
    2015 onward Anonymous image-manipulation critiques accumulate on PubPeer
    29 Nov 2022 Stanford Daily publishes first report; EMBO Journal opens a parallel inquiry
    Dec 2022 Board of Trustees announces a Special Committee review
    Jan–Jun 2023 Mark Filip and a five-member scientific panel investigate 12 papers
    17 Jul 2023 Panel delivers 95-page final report to the Board
    19 Jul 2023 Tessier-Lavigne announces resignation, effective 31 Aug 2023
    31 Aug 2023 Resignation takes effect; Richard Saller becomes interim president
    4 Apr 2024 Jonathan Levin named Stanford’s 13th president
    1 Aug 2024 Levin assumes office

    What Did the Scientific Panel’s Report Conclude?

    The panel’s 95-page report, released on 17 July 2023, concluded that laboratories Tessier-Lavigne ran had engaged in “manipulation of research data” across several papers, and that he was principal author on five of the 12 papers reviewed. It found no evidence he personally fabricated or falsified data, but concluded he had not taken sufficiently decisive steps to correct the scientific record once problems were flagged.

    Following the report, Tessier-Lavigne said he intended to retract at least three papers and correct two more, all originally published between 1999 and 2009 — before his Stanford presidency began. Two papers were formally retracted from Science on 31 August 2023, according to Retraction Watch.

    • 12 co-authored papers examined by the scientific panel
    • 5 papers on which Tessier-Lavigne was principal author
    • At least 3 papers slated for retraction, 2 for correction
    • 1999–2009: publication window for the papers in question

    The Resignation Timeline: From Report to Successor

    Tessier-Lavigne announced his resignation on 19 July 2023, two days after receiving the panel’s findings, saying Stanford needed a president “whose leadership is not hampered by such discussions.” The resignation took effect on 31 August 2023, ending a seven-year presidency.

    Richard Saller, a classicist and former Stanford provost, served as interim president from September 2023. Stanford’s trustees named Jonathan Levin, dean of the Graduate School of Business, as the university’s 13th president on 4 April 2024; Levin assumed office on 1 August 2024 — just over a year after the resignation was announced.

    Answer-First Q&A on the Tessier-Lavigne Case

    When did Marc Tessier-Lavigne resign as Stanford president?

    Marc Tessier-Lavigne announced his resignation on 19 July 2023, two days after Stanford’s Special Committee released its scientific panel’s final report. His resignation took effect on 31 August 2023, and Richard Saller became interim president the same day.

    What did Stanford’s investigation find about the research?

    The scientific panel found that laboratories Tessier-Lavigne led had engaged in repeated data manipulation across several papers examined, out of 12 reviewed in total. It attributed the manipulation to others in the labs, not to Tessier-Lavigne personally, but faulted his oversight and slow correction of the record.

    Was Tessier-Lavigne found to have committed research misconduct himself?

    No. The panel cleared Tessier-Lavigne of personally fabricating or falsifying data. It concluded he was unaware of the manipulation at the time of publication but should have acted more decisively once concerns were raised, particularly regarding papers on which he was principal author.

    What counts as research misconduct under research-integrity standards?

    Research misconduct is generally defined as fabrication, falsification, or plagiarism in proposing, performing, or reporting research — the definition used by the US Office of Research Integrity. Authorship disputes, honest error, and differences of scientific judgement are explicitly excluded from this definition.

    Implications for Institutional Research-Integrity Response

    The Tessier-Lavigne case is a rare instance where a governing board investigated its own chief executive using outside legal counsel and a named, independent scientific panel, then published the resulting report in full. That combination — external fact-finders, disclosed panel membership, and a public report — is closer to the process integrity that bodies such as the Committee on Publication Ethics (COPE) recommend for misconduct investigations than the closed-door reviews typical of many institutions.

    For research administrators, the case demonstrates that reputational and governance consequences can follow from oversight failures even where personal fabrication is not established. It also shows that journal-level inquiries (EMBO Journal) and institutional inquiries can run in parallel without either resolving the other, meaning research administration teams should track both tracks rather than treating a clean institutional finding as the final word. A decade-old PubPeer thread, in this case, outlasted the papers’ authors’ careers before triggering formal review — underscoring why routine, proactive image-integrity screening matters more than reactive response once allegations become public.

  • Causes of Research Misconduct: Pressures, Metrics and Prevention

    What counts as research misconduct?

    Rising retraction counts and a steady stream of high-profile data-integrity cases have kept research misconduct on the agenda for research offices, funders and publishers well into 2026. The causes of research misconduct are rarely a single bad actor acting alone; they are usually a combination of individual choices and the incentive structures institutions build around publication, funding and promotion.

    The internationally recognised core definition, set out by the US Office of Research Integrity (ORI), covers three deliberate acts: fabrication (inventing data), falsification (manipulating data, materials or processes to misrepresent results), and plagiarism (using others’ ideas, words or results without credit) – together known as FFP. The UK Research Integrity Office (UKRIO) uses a broader definition that also captures breaches of ethical or legal obligations, such as unauthorised use of confidential data or failure to obtain proper approvals.

    FFP is distinct from questionable research practices (QRPs) – selective reporting, inappropriate authorship credit, or p-hacking – and from honest error. The distinction matters because prevention strategies differ: FFP requires deterrence and detection, while QRPs respond better to training, culture change and transparent reporting standards.

    The structural and incentive-driven causes of research misconduct

    A 2017 National Academies of Sciences, Engineering, and Medicine report, Fostering Integrity in Research, grouped the drivers of misconduct into six overlapping categories: career and funding pressures, institutional failures of oversight, commercial conflicts of interest, inadequate training, erosion of mentoring standards, and misconduct as part of a wider pattern of deviant behaviour. Subsequent survey research has consistently pointed to the same structural pressures rather than isolated moral failure.

    Holtfreter et al. (2020), surveying academics on the perceived causes of misconduct, found that professional strains and stressors – particularly the pressure to secure competitive grant funding – were cited most often, ahead of individual psychological factors. This lines up with a widely cited earlier synthesis: Fanelli’s 2009 meta-analysis of survey data found that around 2% of scientists admitted to fabricating or falsifying data at least once, while up to a third admitted other questionable research practices – and both figures rose substantially when respondents were asked to estimate colleagues’ behaviour rather than report on their own.

    “Publish or perish” culture and metrics gaming sit at the centre of the structural explanation. When journal impact factor, h-index, publication counts and grant income are used as proxies for quality in hiring, tenure and national assessment exercises, researchers face direct incentives to inflate output rather than rigour. Davis (2003) categorised the underlying factors into three levels, which remains a useful frame for institutional leaders diagnosing where their own controls are weakest.

    Causal level Example factors Where responsibility sits
    Individual Career ambition, financial pressure, poor ethics training, psychological stress Researcher, supervisor
    Organisational Weak oversight, inadequate mentoring, metrics-driven promotion criteria, under-resourced integrity offices Institution, department
    Systemic Publish-or-perish funding models, journal impact-factor incentives, low probability of detection, weak sanctions Funders, publishers, national assessment bodies

    Two systemic factors deserve particular attention from research administrators: low detection probability and weak penalties. A web-search-grounded synthesis of current literature commissioned for this article converged on the same point – academics themselves believe that a low likelihood of investigation, combined with inconsistent sanctions once misconduct is confirmed, is a significant driver of continued misconduct. This is an institutional-design problem, not only an ethics-training problem.

    Quick answers: types, drivers and consequences

    What are the three main types of research misconduct?

    The three internationally recognised categories are fabrication (inventing data or results), falsification (manipulating data, materials or processes to misrepresent findings), and plagiarism (using others’ ideas or words without credit). Together these form the “FFP” definition used by ORI and most national integrity bodies.

    What are the reasons for unethical research?

    Reported reasons include career and funding pressures, institutional failures of oversight, commercial conflicts of interest, inadequate training in research ethics, erosion of mentoring standards, and – in a minority of cases – misconduct forming part of a broader pattern of deviant behaviour, per the National Academies’ 2017 analysis.

    What are the 5 unethical practices in conducting research?

    Commonly cited categories are falsification of data, failure to credit others, plagiarism, undisclosed conflicts of interest, and biased design or interpretation driven by outside influence. Authorship misconduct – including honorary and ghost authorship – is frequently added as a sixth practice in institutional policies.

    What are the 5 main ethical issues in research?

    Beyond FFP itself, institutions most often flag informed consent failures, conflicts of interest, data management and privacy breaches, authorship disputes, and inadequate oversight of research involving human or animal subjects as recurring ethical issues requiring governance attention.

    Evidence-based prevention strategies for institutional leaders

    Because the causes are structural as well as individual, effective prevention combines training with changes to incentive design. Institutional leaders following frameworks from COPE, UKRIO and the UK Concordat to Support Research Integrity typically prioritise the following:

    • Decouple assessment from raw output metrics. Reduce reliance on publication counts and journal impact factor in hiring, tenure and internal funding decisions, in line with responsible-metrics initiatives such as DORA.
    • Fund and empower a dedicated integrity office. A resourced office that can investigate allegations promptly – and is seen to do so – directly addresses the “low detection probability” driver identified in the literature.
    • Make authorship transparent and auditable. Structured, taxonomy-based contributor statements reduce opportunities for honorary and ghost authorship. CASRAI originated the CRediT contributor role taxonomy in 2014; the standard is now stewarded by NISO as ANSI/NISO Z39.104-2022, and its adoption by journals makes individual contributions explicit rather than assumed.
    • Strengthen mentoring and mandatory ethics training for early-career researchers, who are disproportionately exposed to supervision gaps.
    • Protect whistleblowers with clear, enforced anti-retaliation policies – a precondition for any self-reporting culture to function.
    • Apply consistent, proportionate sanctions once misconduct is confirmed, closing the gap between policy and enforcement that researchers themselves identify as a weakness.

    Implications for institutions, funders and publishers

    The practical implication is that misconduct prevention cannot sit solely within research ethics training. In 2023, Crossref and Retraction Watch partnered to integrate more than 43,000 retraction records into open, machine-readable metadata – a structural fix that makes retraction status discoverable at the point of citation, rather than relying on researchers to notice a correction years later. That kind of infrastructure-level intervention complements, rather than replaces, institutional oversight.

    For research administrators, the actionable shift is from a compliance mindset (“train researchers, then police them”) to a design mindset: audit which internal metrics reward speed over rigour, resource integrity offices adequately, and make authorship and contribution as transparent as data availability statements already are. Consult the CASRAI Dictionary for precise definitions when drafting or updating institutional misconduct policy, and review authorship guidance where disputes over credit are a recurring source of allegations.

    None of this suggests misconduct is inevitable. It suggests that where institutions have reduced metrics pressure, resourced oversight and made contribution transparent, the same literature that identifies the causes also points to measurable, achievable prevention.