Research Ethics: Principles and Practices

Abstract

Research ethics constitutes the normative framework governing the conduct of systematic enquiry, balancing the pursuit of knowledge against fundamental rights and human dignity. This article examines the historical development of research ethics from post-war codes through contemporary challenges including the replication crisis, dual-use research dilemmas, and conflicts of interest. We explore core principles—respect for persons, beneficence, and justice—and their application to informed consent, vulnerable populations, and research misconduct. Particular attention is given to forensic investigation contexts where evidentiary standards intersect with ethical obligations, demonstrating how rigorous ethical practice strengthens rather than constrains analytical validity.

Historical Development

The Nuremberg Legacy

Modern research ethics emerged directly from systematic atrocity. The Nuremberg Code (1947), formulated during the prosecution of Nazi physicians who conducted experiments on concentration camp prisoners, established ten foundational principles (Shuster, 1997). The first principle—"The voluntary consent of the human subject is absolutely essential"—represented a categorical rejection of utilitarian justifications for research harm. The Code required that subjects possess legal capacity to consent, be free from coercion, and comprehend the research nature and risks.

Yet the Code's impact proved limited initially. Conceived as legal precedent rather than professional guidance, it was often dismissed as relevant only to war crimes rather than ordinary research practice (Ghooi, 2011). This interpretation permitted continued ethical failures across democratic nations conducting ostensibly legitimate research.

The Tuskegee Syphilis Study

Between 1932 and 1972, the United States Public Health Service observed the natural progression of untreated syphilis in 399 African American men in Alabama, withholding effective treatment even after penicillin became standard care in the 1940s (Brandt, 1978). Participants were told they were receiving "special free treatment" for "bad blood"—deliberate deception that violated emerging consent standards. The 40-year study exemplified structural racism in research, treating economically disadvantaged Black men as a permanent research underclass.

When Jean Heller exposed the study in 1972, public outrage catalysed fundamental reform (Heller, 1972). The Tuskegee revelations demonstrated that the Nuremberg Code's protections were urgently needed within democratic institutions, not merely as safeguards against totalitarian excess.

Declaration of Helsinki

The World Medical Association adopted the Declaration of Helsinki in 1964, creating profession-specific guidance for physician-researchers. Unlike the Nuremberg Code's absolute statements, Helsinki acknowledged the complexity of clinical research, distinguishing therapeutic research (intended to benefit participants) from non-therapeutic research (purely knowledge-generating). The Declaration has undergone seven major revisions, reflecting evolving ethical understanding whilst maintaining core commitments to participant welfare and informed consent (World Medical Association, 2013).

Helsinki introduced the concept of independent ethical review, requiring that research protocols be submitted to specially appointed committees—the foundation of modern institutional review boards (IRBs) and research ethics committees (RECs). This external oversight acknowledged that researchers' commitment to scientific progress might compromise their judgement regarding participant protection.

The Belmont Report

Commissioned in response to Tuskegee, the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research produced the Belmont Report (1979), which articulated three fundamental ethical principles:

1. Respect for Persons: Individuals should be treated as autonomous agents, and persons with diminished autonomy are entitled to protection. This principle grounds informed consent requirements and special protections for vulnerable populations.

2. Beneficence: Research should maximise benefits and minimise harms. This extends beyond non-maleficence (avoiding harm) to an obligation to secure participant wellbeing, requiring systematic risk-benefit analysis.

3. Justice: Research burdens and benefits should be distributed fairly. This principle challenges exploitation of disadvantaged groups for research that primarily benefits the privileged—the Tuskegee pattern.

The Belmont principles remain foundational to research ethics globally, translated into operational requirements through regulations like the United States Common Rule and the UK Research Governance Framework.

Conceptual Foundation

Informed consent represents the practical expression of respect for persons. Valid consent requires three elements (Faden & Beauchamp, 1986):

Information: Participants must receive comprehensive disclosure of research purpose, procedures, risks, benefits, alternatives, and their right to withdraw. The information standard is not merely technical completeness but genuine comprehension.

Comprehension: Information must be presented in language and format accessible to the intended population. Researchers bear responsibility for ensuring understanding, not merely providing disclosure.

Voluntariness: Consent must be given freely, without coercion, undue influence, or manipulation. This proves particularly challenging where power asymmetries exist—physician-patient relationships, teacher-student dynamics, or institutional settings.

Capacity Considerations

Decision-making capacity exists on a continuum, not as a binary state. The functional approach to capacity assessment examines whether an individual can:

Understand relevant information
Appreciate how information applies to their situation
Reason about treatment options
Express a choice

Capacity is decision-specific, not global. An individual may possess capacity for low-risk decisions whilst lacking capacity for complex, high-risk choices (Appelbaum, 2007). Fluctuating capacity—common in psychiatric conditions, neurodegenerative disease, or critical illness—requires ongoing reassessment rather than single-point determination.

Documentation and Process

Written consent forms serve evidentiary purposes but do not constitute consent itself. The ethical core is the conversation through which understanding develops and autonomous choice becomes possible. Documentation should capture this process, not replace it.

The UK Research Governance Framework requires that consent forms:

Use clear, non-technical language
Explain purpose, methods, duration, and demands
Identify reasonably foreseeable risks and burdens
Describe potential benefits to participant and society
Explain confidentiality protections and limits
Clarify the voluntary nature and withdrawal rights
Provide contact information for questions and concerns

Consent is not a singular event but a continuing process, particularly in longitudinal research. As research evolves, new information emerges, or participants' circumstances change, researchers must revisit consent. The doctrine of continuous consent recognises that autonomous choice requires ongoing opportunity for reassessment (Doyal, 1997).

Privacy and Confidentiality

Conceptual Distinction

Privacy concerns individuals' control over access to themselves—the right to control information about oneself. Confidentiality concerns researchers' obligations regarding information entrusted to them. Privacy is a right participants possess; confidentiality is a duty researchers assume (Wiles et al., 2008).

Research routinely requires that participants waive some privacy, sharing personal information or allowing observation. This makes confidentiality protections essential. The UK Data Protection Act 2018 and UK GDPR establish both rights and obligations governing research data.

Anonymisation and Pseudonymisation

Anonymisation removes identifying information such that individuals cannot be re-identified through reasonable means. True anonymisation is increasingly difficult given data linkage capabilities, particularly with genomic data, location data, or detailed health records (Ohm, 2010).

Pseudonymisation replaces direct identifiers with codes, maintaining a key linking codes to identities. This allows data linkage whilst protecting privacy, but requires robust key security. The UK GDPR recognises pseudonymisation as a safeguard but not as removing data from regulatory scope.

Limits to Confidentiality

Confidentiality is not absolute. Researchers face disclosure obligations when:

Safeguarding concerns arise: Credible evidence of child abuse or vulnerable adult maltreatment triggers reporting duties under the Children Act 1989 and Care Act 2014.
Serious crime is disclosed: Though no general duty to report crime exists in England and Wales, specific offences (terrorism, money laundering) impose reporting obligations.
Court orders compel disclosure: Researchers cannot guarantee confidentiality against judicial compulsion. Participants should be informed of these limits prospectively.

The ethical approach requires that participants be informed of confidentiality limits before disclosure occurs, allowing informed decisions about what to reveal.

Vulnerable Populations

Defining Vulnerability

Vulnerability in research ethics refers to increased likelihood of harm or decreased ability to protect one's interests. The Belmont Report identified prisoners, children, pregnant women, and mentally disabled persons as vulnerable, but vulnerability is context-dependent rather than categorical (Luna, 2009).

Situational factors creating vulnerability include:

Cognitive impairment: Compromising consent capacity
Institutional dependence: Creating coercion risk (prisoners, psychiatric patients)
Economic disadvantage: Enabling undue inducement
Social marginalisation: Limiting voice and redress
Serious illness: Fostering therapeutic misconception

Additional Protections

Research involving vulnerable populations requires enhanced protections:

Children: Assent from the child (where developmentally appropriate) alongside parental consent. As children mature, their voice should carry increasing weight. The UK Research Governance Framework requires that research not otherwise possible in adults be required before enrolling children.

Prisoners: Particular scrutiny for coercion and undue influence, given the inherently coercive nature of incarceration. Research must address prison conditions or prisoner health, not merely exploit a captive population.

Cognitive Impairment: Where individuals lack consent capacity, substitute decision-makers may authorise participation if research serves the individual's interests or addresses their condition and involves minimal risk. The Mental Capacity Act 2005 provides specific protections for research involving incapacitated adults.

Avoiding Exploitation

Justice requires that vulnerable populations not bear research burdens disproportionate to benefits they receive. Historically, disadvantaged groups provided research subjects for interventions primarily benefiting the privileged. Contemporary standards require:

Fair subject selection: Burdens and benefits should be distributed equitably
Community engagement: Affected communities should participate in research design
Benefit sharing: Research should address health priorities of study populations
Capacity building: Research should strengthen local research infrastructure

Research Misconduct

Fabrication, Falsification, and Plagiarism

The canonical triumvirate of research misconduct comprises:

Fabrication: Inventing data or results. This fundamentally betrays the research enterprise, introducing fiction as evidence.

Falsification: Manipulating materials, equipment, or processes, or changing/omitting data such that research is not accurately represented. This includes selective reporting of results, undisclosed data cleaning, and p-hacking.

Plagiarism: Appropriating another's ideas, processes, results, or words without appropriate credit. This violates intellectual property norms and misrepresents research provenance.

The UK Research Integrity Office (2019) defines research misconduct as "fabrication, falsification, plagiarism, or other practices that seriously deviate from those commonly accepted within the research community for proposing, conducting, or reporting research." This recognises that misconduct extends beyond the FFP triumvirate.

Questionable Research Practices

Between outright fraud and exemplary practice lies a grey zone of questionable research practices (QRPs) that compromise research integrity without constituting formal misconduct (Martinson et al., 2005). These include:

HARKing: Hypothesising After Results are Known, presenting exploratory findings as confirmatory
P-hacking: Running multiple analyses until achieving statistical significance, then reporting only significant results
Publication bias: Failing to publish null results, distorting meta-analytic evidence
Inadequate supervision: Senior researchers claiming authorship on work they did not adequately oversee
Insufficient record-keeping: Failing to maintain adequate documentation for verification and replication

QRPs prove difficult to address because they often reflect institutional incentives (publication pressure, funding competition) rather than individual malfeasance. Structural reform—rewarding transparency over novelty, valuing replication, registering protocols—proves more effective than individual sanctions.

Investigation Procedures

The UK Concordat to Support Research Integrity (2019) establishes institutional responsibilities for investigating misconduct allegations. Procedures must balance fairness to accused individuals with commitment to research integrity.

Preliminary assessment determines whether allegations fall within misconduct definitions and whether formal investigation is warranted. Malicious or frivolous complaints should be filtered, whilst legitimate concerns proceed.

Formal investigation involves an independent panel examining evidence, interviewing relevant parties, and determining whether misconduct occurred and appropriate responses. Investigations should conclude within 120 days absent compelling reasons for delay.

Outcomes range from finding no misconduct, to requiring correction or retraction of published work, to dismissal and referral to professional bodies. Transparency requires public disclosure of investigation outcomes, balanced against privacy considerations.

Conflicts of Interest

Definition and Types

Conflicts of interest exist when professional judgement concerning a primary interest (research integrity, participant welfare) may be unduly influenced by a secondary interest (financial gain, career advancement, personal relationships) (Thompson, 1993). COIs need not involve actual bias—the risk that judgement could be compromised suffices.

Financial COIs: Research funding from entities with financial stakes in outcomes; equity holdings in commercial entities related to research; consulting relationships with industry.

Non-financial COIs: Personal relationships; institutional loyalties; ideological commitments; career advancement tied to particular findings; intellectual investments in theories under investigation.

The critical distinction is between COIs (situations posing risk of bias) and bias itself (actual distortion of judgement). Eliminating all COIs would prove impossible—researchers always have career interests in novel findings. The question is which COIs pose sufficient risk to warrant disclosure or management.

Disclosure Requirements

The International Committee of Medical Journal Editors (ICMJE) requires disclosure of all financial and personal relationships that could inappropriately influence work, defined as relationships within 36 months of publication. This includes:

Research funding sources
Employment and consultancies
Stock/equity interests
Patents and royalties
Expert testimony fees
Advisory board memberships
Travel expenses from industry

Disclosure does not eliminate COI risks but allows readers to assess potential influence. The limitation is that disclosure may provide false assurance—readers may underestimate bias risk once disclosure occurs, the "moral licensing" effect (Loewenstein et al., 2011).

Management Strategies

Beyond disclosure, institutions employ COI management strategies:

Recusal: Individuals with significant COIs abstain from relevant decisions (grant review, data interpretation, regulatory determinations).

Independent oversight: Independent data monitoring committees oversee trials where researchers have financial interests.

Blinding: Keeping researchers unaware of outcomes or group assignments limits opportunity for bias.

Prohibition: Some COIs prove unmanageable, requiring prohibition. Regulatory bodies often prohibit financial relationships between researchers and regulatory entities they study.

Publication Ethics

COPE Guidelines

The Committee on Publication Ethics (COPE) provides guidance for ethical publishing, addressing authorship, peer review, conflicts of interest, and misconduct (COPE, 2022). Core principles include:

Authorship: All named authors should have made substantial contributions and approved the final version. Substantial contribution requires involvement in conception/design, data acquisition/analysis, or drafting/revision. "Gift authorship" (adding senior colleagues who did not contribute) and "ghost authorship" (omitting corporate or professional writers) violate authorship standards.

Peer review integrity: Reviewers should declare COIs, maintain confidentiality, provide substantive critique, and avoid appropriating authors' ideas. Authors should disclose all relevant work, including negative findings.

Corrections and retractions: Errors should be corrected promptly through corrigenda. Retractions are required when findings are unreliable due to misconduct or error, or when research was unethical. Retractions should clearly identify retracted work to prevent continued citation.

The Replication Crisis

Across psychology, medicine, and social sciences, systematic replication attempts reveal that substantial proportions of published findings do not replicate (Open Science Collaboration, 2015). The replication crisis reflects multiple failures:

Publication bias: Journals favour novel, statistically significant findings, creating incentives to pursue novelty over rigour and suppress null results.

Insufficient statistical power: Small studies with inadequate samples generate false positives proliferated through publication bias (Button et al., 2013).

Analytical flexibility: Multiple defensible analytical choices allow researchers to try various approaches until achieving desired results, inflating false positive rates (Simmons et al., 2011).

Lack of transparency: Opaque methods prevent scrutiny and replication. Insufficient detail about procedures, analysis decisions not fully reported, and proprietary data preclude verification.

Reform Movements

The open science movement addresses replication challenges through:

Pre-registration: Committing to hypotheses, methods, and analyses before data collection, preventing HARKing and p-hacking.

Open data: Making data publicly available enables verification and re-analysis.

Open materials: Sharing protocols, measures, and analysis code facilitates replication.

Registered reports: Journals commit to publishing studies based on methods before results are known, eliminating results-based publication bias.

Replication studies: Valuing and publishing direct replications establishes which findings prove robust.

Dual-Use Research

The Dilemma

Dual-use research of concern (DURC) generates knowledge that could be misused to threaten public health or safety. The 2012 controversy over H5N1 influenza research exemplifies the challenge: gain-of-function research created transmissible avian flu strains, advancing pandemic preparedness but potentially providing bioterrorism blueprints (Fouchier et al., 2012).

The tension is fundamental. Openness and transparency constitute scientific virtues, enabling verification and cumulative knowledge development. Yet some knowledge may pose risks exceeding benefits if widely disseminated. Censorship conflicts with scientific norms whilst potentially preventing catastrophic misuse.

The UK Framework

The UK Biological Security Strategy addresses DURC through:

Research council oversight: UKRI requires institutions to identify and assess DURC, implementing risk mitigation before funding approval.

Publication review: Research raising dual-use concerns undergoes additional review examining whether benefits of publication outweigh risks.

Redaction options: In exceptional cases, methodological details may be redacted from publications whilst maintaining sufficient information for scientific assessment.

International coordination: Biological security requires global cooperation, given that knowledge restrictions in one jurisdiction prove ineffective when research occurs globally.

Ethical Analysis

The precautionary principle suggests that where credible risk of catastrophic harm exists, restrictions may be justified despite uncertainty. Yet over-restriction of research may itself cause harm by impeding beneficial research. The approach requires case-by-case analysis balancing:

Magnitude of potential harm from misuse
Likelihood of misuse
Scientific and public health benefits
Availability of alternative research approaches
Effectiveness of access restrictions
Impact on scientific norms and progress

Institutional Review and Governance

Research Ethics Committees

RECs provide independent ethical review of research involving human participants, ensuring that research design respects ethical principles and complies with regulatory requirements. The UK Health Research Authority coordinates NHS RECs, reviewing health and social care research.

REC responsibilities include assessing:

Scientific validity: Unscientific research cannot be ethical, as it cannot generate benefits to justify participant burdens.

Risk-benefit ratio: Whether potential benefits justify risks, and whether risks have been minimised.

Informed consent processes: Adequacy of information provision, consent documentation, and capacity assessment.

Confidentiality protections: Data security measures and plans for data retention and destruction.

Fair participant selection: Whether vulnerable populations are appropriately protected and burdens equitably distributed.

The UK Research Integrity Office

UKRIO provides independent guidance on research integrity, supporting institutions in developing and implementing integrity standards. Unlike regulatory bodies with enforcement powers, UKRIO operates through guidance, training, and advice.

UKRIO's Code of Practice for Research (2019) establishes standards across the research lifecycle:

Research design: Studies should be based on sound principles, use appropriate methodology, and receive ethical approval where required.

Research conduct: Researchers should maintain detailed records, report findings accurately and objectively, and acknowledge all contributors appropriately.

Dissemination: Research should be disseminated openly, allowing scrutiny and verification, with appropriate protections for confidential information.

Research culture: Institutions should foster integrity through training, mentoring, clear policies, and fair investigation procedures.

Application to Forensic Investigation

Evidentiary Standards and Research Ethics

Forensic investigation occupies a unique position relative to research ethics frameworks. When investigation aims not at generalizable knowledge but at establishing particular facts for legal proceedings, traditional research ethics requirements may not directly apply. Yet ethical principles remain highly relevant.

Consent in forensic contexts: Document analysis for litigation typically does not require participant consent—documents are obtained through disclosure procedures, and analysis constitutes legal work rather than research. However, when investigation involves interviews, recording, or intrusion into private matters, consent or legitimate legal basis becomes ethically essential.

Confidentiality obligations: Legal privilege and data protection create confidentiality obligations analogous to research settings. Forensic investigators must protect sensitive information, disclosing only what litigation requires.

Vulnerable populations: Children, cognitively impaired persons, and trauma survivors require particular care in forensic settings. The same capacity considerations relevant to research consent apply to forensic interview and evidence gathering.

Evidence Integrity

Research ethics' emphasis on preventing fabrication, falsification, and selective reporting directly translates to forensic standards. The credibility of forensic evidence depends on:

Complete documentation: Contemporary, comprehensive notes capturing all relevant observations, not merely those supporting preferred narratives.

Methodological transparency: Clear articulation of analytical methods, allowing scrutiny and assessment of appropriateness.

Neutral analysis: Avoiding confirmation bias, actively seeking disconfirming evidence, and representing findings objectively rather than advocacy.

Provenance maintenance: Chain of custody for documents, clear record of when and how evidence was obtained and analysed.

These standards mirror research integrity requirements. Forensic investigation that selectively reports findings, obscures methodology, or reconstructs evidence post hoc violates principles equally applicable to research misconduct.

Conflicts of Interest in Expert Testimony

Expert witnesses face particular COI challenges. Financial interests—being retained by one party—create obvious pressures toward advocacy. The Civil Justice Council's Guidance for the Instruction of Experts (2014) requires that experts' duties to the court override any obligation to instructing parties. Yet psychological evidence demonstrates that even well-intentioned experts show bias toward parties retaining them (Murrie et al., 2013).

Transparency proves essential. Expert reports should disclose:

Retaining party and compensation arrangements
Prior relationships with parties or counsel
Other cases involving similar issues where expert testified
Professional or ideological commitments relevant to opinions

Beyond disclosure, experts should employ bias mitigation strategies: blind review where possible, actively seeking disconfirming evidence, distinguishing levels of certainty, and acknowledging limitations.

Publication of Case Material

When forensic investigation generates insights of broader research relevance, publication considerations arise. Legal proceedings often generate substantial confidential information about vulnerable individuals. Publication requires balancing knowledge advancement against privacy protection.

Anonymisation proves difficult—even with names removed, case details may enable identification. Consent for publication becomes problematic when obtained from parties with litigation interests. The ethical approach requires:

Necessity assessment: Publication should offer significant contribution beyond individual case resolution
Minimal disclosure: Include only information necessary for research purposes
Delayed publication: Await case conclusion to avoid prejudicing proceedings
Consultation: Seek ethics committee or professional body guidance for novel cases

Connection to Phronesis Platform

The Phronesis platform operationalises research ethics principles within forensic investigation contexts. Its engine architecture reflects ethical commitments:

Evidence Integrity: The Documentary Analysis and Omission Detection engines identify selective reporting patterns—the forensic equivalent of publication bias and questionable research practices. By systematically comparing source materials to final reports, these engines detect fabrication (claims without documentary support), falsification (misrepresentation of evidence), and plagiarism (unauthorised use of others' analysis).

Transparency and Replicability: The Systematic Adversarial Methodology (S.A.M.) emphasises methodological transparency through its four-phase structure (Anchor, Inherit, Compound, Arrive). Each analytical claim requires documentation of provenance—source document, specific passage, and inferential chain. This mirrors open science commitments to transparency, allowing verification and challenge.

Conflict of Interest Detection: The Professional Tracker and Accountability Audit engines monitor expert witnesses and institutional actors for patterns suggesting conflicts of interest. Repeated testimony for prosecution, financial relationships with referring agencies, or methodological inconsistencies across cases reveal potential bias sources.

Vulnerable Population Protection: The platform's emphasis on outcome consequences—the "Arrive" phase documenting catastrophic outcomes—reflects beneficence and justice principles. When institutional failures harm vulnerable individuals (children in care, persons with disabilities, accused persons lacking resources), the platform systematically documents these harms, holding institutions accountable for breached duties.

Informed Consent Verification: The Documentary Analysis engine can verify consent claims in research or media contexts, comparing claimed consent to documentary evidence of refusals, objections, or capacity limitations.

The platform demonstrates that rigorous ethical standards strengthen rather than constrain analytical validity. Fabricated evidence, selective reporting, undisclosed conflicts, and vulnerable population exploitation ultimately undermine evidentiary credibility. Ethical practice and robust methodology converge—both demand transparency, completeness, objectivity, and accountability.

Conclusion

Research ethics evolved through tragic failure—experiments on concentration camp prisoners, decades-long observation of treatable disease, radiation exposure of pregnant women without consent. These failures generated frameworks protecting human dignity whilst enabling knowledge advancement.

Contemporary challenges—replication crises, dual-use research, pervasive conflicts of interest—demonstrate that ethical vigilance remains essential. No framework eliminates ethical complexity or guarantees right action. What frameworks provide is structure for recognising and analysing ethical challenges, principles for reasoning through dilemmas, and procedures for accountability when failures occur.

For forensic investigators, research ethics principles prove directly applicable. Evidence integrity mirrors research integrity. Consent requirements in research translate to legitimate basis requirements for evidence gathering. Vulnerability protections apply equally whether participants are research subjects or litigation parties. Conflicts of interest compromise forensic opinions as surely as they distort research findings.

The fundamental insight is that ethics and rigour prove complementary rather than competing. Sound methodology requires ethical practice—transparency, completeness, objectivity, accountability. Ethical failures are methodological failures. By operationalising ethical principles through systematic methodology, forensic investigation achieves both moral legitimacy and analytical validity.

References

Appelbaum, P. S. (2007). Assessment of patients' competence to consent to treatment. New England Journal of Medicine, 357(18), 1834-1840.

Beauchamp, T. L., & Childress, J. F. (2013). Principles of biomedical ethics (7th ed.). Oxford University Press.

Brandt, A. M. (1978). Racism and research: The case of the Tuskegee Syphilis Study. Hastings Center Report, 8(6), 21-29.

Button, K. S., Ioannidis, J. P., Mokrysz, C., Nosek, B. A., Flint, J., Robinson, E. S., & Munafò, M. R. (2013). Power failure: Why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience, 14(5), 365-376.

Civil Justice Council. (2014). Guidance for the instruction of experts in civil claims. Ministry of Justice.

Committee on Publication Ethics. (2022). Core practices. https://publicationethics.org/core-practices

Doyal, L. (1997). Informed consent in medical research: Journals should not publish research to which patients have not given fully informed consent—with three exceptions. BMJ, 314(7087), 1107.

Faden, R. R., & Beauchamp, T. L. (1986). A history and theory of informed consent. Oxford University Press.

Fouchier, R. A., García-Sastre, A., Kawaoka, Y., Barclay, W. S., Bouvier, N. M., Brown, I. H., ... & Palese, P. (2012). Pause on avian flu transmission research. Science, 335(6067), 400-401.

Ghooi, R. B. (2011). The Nuremberg Code—A critique. Perspectives in Clinical Research, 2(2), 72-76.

Heller, J. (1972, July 26). Syphilis victims in U.S. study went untreated for 40 years. New York Times.

Loewenstein, G., Sah, S., & Cain, D. M. (2011). The unintended consequences of conflict of interest disclosure. JAMA, 307(7), 669-670.

Luna, F. (2009). Elucidating the concept of vulnerability: Layers not labels. International Journal of Feminist Approaches to Bioethics, 2(1), 121-139.

Martinson, B. C., Anderson, M. S., & de Vries, R. (2005). Scientists behaving badly. Nature, 435(7043), 737-738.

Murrie, D. C., Boccaccini, M. T., Guarnera, L. A., & Rufino, K. A. (2013). Are forensic experts biased by the side that retained them? Psychological Science, 24(10), 1889-1897.

National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979). The Belmont Report: Ethical principles and guidelines for the protection of human subjects of research. US Government Printing Office.

Ohm, P. (2010). Broken promises of privacy: Responding to the surprising failure of anonymization. UCLA Law Review, 57, 1701-1777.

Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), aac4716.

Shuster, E. (1997). Fifty years later: The significance of the Nuremberg Code. New England Journal of Medicine, 337(20), 1436-1440.

Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22(11), 1359-1366.

Thompson, D. F. (1993). Understanding financial conflicts of interest. New England Journal of Medicine, 329(8), 573-576.

UK Research Integrity Office. (2019). Code of practice for research: Promoting good practice and preventing misconduct. https://ukrio.org/publications/code-of-practice-for-research/

Universities UK. (2019). The Concordat to Support Research Integrity. https://www.universitiesuk.ac.uk/topics/research-and-innovation/concordat-support-research-integrity

Wiles, R., Crow, G., Heath, S., & Charles, V. (2008). The management of confidentiality and anonymity in social research. International Journal of Social Research Methodology, 11(5), 417-428.

World Medical Association. (2013). World Medical Association Declaration of Helsinki: Ethical principles for medical research involving human subjects. JAMA, 310(20), 2191-2194.