Ethics, Equity and Research Governance

Introduction

Qualitative health research is concerned with people’s meanings, experiences, relationships and interpretations of health and illness. Because it often involves personal narratives, sensitive settings and unequal relationships between researchers and participants, ethical practice is not limited to obtaining approval from a review committee. It is a continuing responsibility that shapes the research question, recruitment strategy, data collection, interpretation, dissemination and implementation of findings.

For health-care managers and researchers, ethics is simultaneously a moral, methodological and governance concern. A study may be scientifically valuable yet ethically unacceptable if participation is coerced, confidentiality is poorly protected, communities are misrepresented or benefits are distributed inequitably. Conversely, rigorous ethical engagement can improve data quality by strengthening trust, openness and participant agency.

This chapter examines informed consent, confidentiality, vulnerability, trauma, safeguarding, data ownership, conflicts of interest, ethics review, Indigenous and decolonizing principles, health-data governance and equitable research practice. It treats ethics as relational and context-dependent rather than as a checklist.

Learning objectives

By the end of this chapter, students should be able to:

1. Explain the ethical principles that guide qualitative health research.
2. Design consent, confidentiality and safeguarding procedures appropriate to sensitive research.
3. Analyse vulnerability, trauma and power in researcher–participant relationships.
4. Distinguish data ownership, stewardship, access and custodianship.
5. Apply ethics-review and health-data-governance requirements.
6. Integrate Indigenous, decolonizing and equity-oriented approaches into research design.
7. Identify and manage conflicts of interest.
8. Evaluate the ethical implications of digital technologies and artificial intelligence in qualitative research.

1. Ethics as an ongoing research process

The conventional principles of research ethics are often expressed as respect for persons, beneficence, non-maleficence and justice. Respect requires recognition of participants’ autonomy, dignity, cultural identity and right to decline participation. Beneficence requires seeking meaningful social or clinical value. Non-maleficence requires reducing foreseeable physical, psychological, social, legal and economic harms. Justice requires fair distribution of research burdens and benefits.

These principles must be interpreted through the realities of qualitative research. A participant may consent to an interview but not anticipate that a distinctive quotation could identify them. A community may agree to a project but reject publication of findings that portray it as deficient. A researcher may unintentionally influence disclosure through professional status, language, age, gender or institutional affiliation.

Ethical reflexivity is therefore essential. Reflexivity involves examining how the researcher’s identity, assumptions, interests and institutional position influence the research process. A reflexive journal can record decisions about recruitment, field relationships, interpretation, emotional responses and changes to the protocol. Reflexivity does not eliminate power differences; it makes them visible and open to review.

Key ethical questions across the research cycle

Ethics is also methodological. If participants do not trust the research team, they may withhold information; if recruitment excludes people with limited literacy or language access, findings may be systematically incomplete. Ethical quality and scientific quality are therefore mutually reinforcing.

2. Informed consent in qualitative research

Informed consent is a process through which a person voluntarily agrees to participate after receiving understandable information about the study. It is not merely a signed form. Valid consent normally requires adequate information, comprehension, voluntariness and decision-making capacity.

Participants should be told, in accessible language:

• the purpose and funder of the study;
• what participation involves, including recording, observation or collection of documents;
• the expected duration and possible future uses of data;
• foreseeable benefits, risks and inconveniences;
• confidentiality limits;
• the voluntary nature of participation and the right to withdraw;
• compensation and any costs;
• contact details for the research team and ethics body; and
• how findings will be shared.

Qualitative consent is often iterative. The research question, interview direction or field setting may develop during the study. Researchers should re-confirm consent when procedures change, when sensitive topics arise, when new uses of data are proposed or when a participant’s circumstances alter. In ethnographic work, consent may involve repeated discussion with individuals and organizations rather than one event.

Written consent is not always the most ethical form. A signature may create risk for undocumented migrants, survivors of violence or participants who fear institutional records. Where approved by the ethics committee, oral consent, witnessed consent or a brief consent script may be preferable. The alternative must still document that information was provided and understood.

Capacity and supported decision-making

Capacity is decision-specific and may fluctuate. Researchers should not equate disability, old age, mental illness, low literacy or limited English with incapacity. Appropriate support may include plain language, interpreters, additional time, trusted supporters and accessible formats. If a person lacks capacity to consent, local law and ethics requirements may permit consent from an authorized representative, while still requiring the person’s assent wherever possible.

Compensation and undue influence

Payment should compensate time, transport and inconvenience rather than purchase agreement. Excessive payment, access to essential services or recruitment by a dependent clinician may compromise voluntariness. Researchers should distinguish reimbursement from benefits that could constitute undue influence and should justify amounts to the review committee.

3. Confidentiality, anonymity and disclosure

Anonymity means that even the research team cannot link data to a person; this is uncommon in interview-based studies. Confidentiality means that the team knows participants’ identities but protects information from unauthorized disclosure. Pseudonymization replaces direct identifiers with codes, but coded data remain potentially identifiable if the key is retained or if combinations of details are distinctive.

Confidentiality plans should address:

1. Collection: conduct interviews in private locations; avoid unnecessary identifiers; explain recording procedures.
2. Processing: separate identity keys from transcripts; remove names, locations and rare occupations where appropriate.
3. Storage: use encrypted institutional systems, strong authentication and controlled permissions; avoid personal devices and unapproved cloud services.
4. Analysis: limit access according to role and document data-access decisions.
5. Publication: use composite cases or altered characteristics when necessary, while preserving analytic integrity.
6. Retention and destruction: follow the approved schedule and legal requirements.

A quotation may identify a participant through distinctive wording, a rare illness, a small community or a combination of demographic details. “Member checking” can support respectful interpretation, but it does not transfer responsibility for confidentiality to participants. Researchers must explain that complete anonymity cannot always be guaranteed, especially in small communities or group interviews.

Confidentiality has limits. Researchers may have duties under law or institutional policy concerning imminent serious harm, abuse, exploitation, threats to others or court orders. These limits must be explained before participation. Researchers should not promise absolute secrecy when they cannot provide it.

4. Vulnerability, trauma and safeguarding

Vulnerability is not a fixed characteristic of a person. It arises when circumstances limit autonomy, increase exposure to harm or reduce the ability to protect one’s interests. Relevant conditions may include poverty, displacement, detention, illness, dependency on services, cognitive impairment, language barriers, stigma, discrimination and unequal power relationships.

An equity-oriented approach avoids excluding every potentially vulnerable group. Exclusion may deny populations a voice and produce evidence that is less relevant. Instead, researchers should ask what additional protections are required and whether the study has adequate expertise, referral pathways and resources.

Trauma-informed research

Trauma-informed practice recognizes that participants may have experienced violence, neglect, war, racism, medical trauma or institutional abuse. It emphasizes safety, trust, choice, collaboration and empowerment. Practical measures include:

• explaining the topics and allowing participants to skip questions;
• avoiding pressure to provide detailed traumatic narratives;
• permitting breaks, pauses or termination without penalty;
• offering participant control over recording and quotations;
• training interviewers in distress recognition and response;
• preparing referral information before data collection; and
• arranging researcher debriefing and supervision.

Researchers are not automatically therapists. They should avoid promising treatment and should establish in advance what support can realistically be offered.

Safeguarding

Safeguarding concerns the prevention and response to abuse, neglect, exploitation and serious danger. A safeguarding plan should identify:

• foreseeable risks and warning signs;
• who receives disclosures;
• when confidentiality may be breached;
• reporting duties and escalation routes;
• emergency contacts and referral services;
• documentation standards; and
• responsibilities after the interview ends.

The plan must be adapted to jurisdiction. Researchers should never investigate allegations beyond their competence or make promises about outcomes they cannot control. In group research, participants must also be warned that the research team cannot guarantee that other group members will maintain confidentiality.

5. Equity, inclusion and justice

Equitable research is not achieved simply by recruiting equal numbers from demographic categories. It requires attention to who defines the problem, who controls resources, who bears research risks, whose knowledge is treated as credible and who benefits from the findings.

Inclusive design may require multilingual materials, accessible venues, disability accommodations, flexible scheduling, compensation for transport, childcare support and recruitment through trusted organizations. Researchers should monitor whether gatekeepers unintentionally exclude people with dissenting views or those who are less connected to services.

Researchers should distinguish equality from equity. Equality provides the same input to everyone; equity adjusts support according to barriers and needs. For example, providing the same online survey link to all participants is formally equal but inequitable if some lack internet access or require screen-reader compatibility.

Sampling should be justified analytically and ethically. Purposive sampling can include groups whose experiences are central to the question, while maximum-variation sampling can illuminate differences across age, gender, geography or service access. Researchers should avoid tokenism: a single participant should not be treated as representing an entire group.

Dissemination is part of justice. Findings should be returned in formats that participants can use, such as plain-language summaries, community meetings, translated reports or policy briefings. Researchers should explain how disagreement with interpretations will be considered and should avoid deficit narratives that locate structural problems within individuals or communities.

6. Indigenous and decolonizing principles

Indigenous research ethics cannot be reduced to adding an Indigenous participant or obtaining individual consent. In many Indigenous contexts, knowledge is relational, collective, place-based and connected to responsibilities across generations. Governance may involve community authorities, Elders, councils or cultural knowledge holders in addition to individual participants.

Decolonizing research challenges extractive patterns in which external researchers take knowledge, interpret it through dominant academic frameworks and return little control or benefit. Decolonizing approaches may involve:

• co-defining the research question;
• recognizing Indigenous authority over knowledge and cultural protocols;
• supporting community control over data access and interpretation;
• using Indigenous methodologies where appropriate;
• employing and training local researchers;
• negotiating benefit-sharing and intellectual-property arrangements;
• respecting restrictions on sacred, collective or gender-specific knowledge; and
• enabling community review before public dissemination.

A useful principle is relational accountability: researchers are accountable not only to an institution or funder but also to participants, communities, ancestors, future generations and the relationships that make the research possible. Indigenous data-governance frameworks, including the CARE principles—Collective Benefit, Authority to Control, Responsibility and Ethics—complement technical data standards. They emphasize that data should serve collective well-being and community authority, not merely institutional efficiency.

Researchers must avoid treating Indigenous peoples as homogeneous. Protocols differ across nations and communities, and the appropriate governance arrangements must be locally determined.

7. Data ownership, stewardship and governance

Data ownership is a legal and institutional concept, while stewardship describes the responsibility to care for data and use them appropriately. A researcher who collects interviews does not necessarily own the knowledge expressed in them. Rights may be held or shared by participants, communities, employers, health services, universities, funders or governments, depending on law and agreement.

A data-management plan should specify:

• what data will be collected and why;
• who is the data controller or responsible institution;
• who may access raw, coded and derived data;
• whether data will be transferred internationally;
• how long data will be retained;
• whether secondary use is permitted;
• how withdrawal requests will be handled; and
• how data will be destroyed or archived.

The principle of data minimization is important: collect only what is necessary for the stated purpose. De-identification reduces risk but does not guarantee anonymity. Risk increases when datasets are linked, when samples are small or when public information can be combined with research data.

Health-data governance combines ethical, legal, technical and organizational controls. It includes access authorization, encryption, audit logs, incident response, data-quality procedures, staff training and accountability. Data protection laws vary across jurisdictions, but common principles include purpose limitation, transparency, security, accuracy, retention control and rights of access or correction.

Secondary analysis requires particular care. A participant who consented to a study of maternity care may not have consented to unrestricted use of their interview for commercial profiling or unrelated genetic research. Broad consent may be ethically acceptable only when future uses, governance safeguards and withdrawal limits are clearly explained and approved.

Digital recording and artificial intelligence

Automatic transcription, cloud storage and generative artificial intelligence can create new risks. Audio files may contain voices, accents and contextual identifiers. External transcription platforms may process data in other countries or retain them for model training. Generative AI may hallucinate, omit culturally meaningful speech or reproduce confidential content.

Before using such tools, researchers should assess contractual terms, data location, encryption, human oversight, accuracy, bias and participant consent. Sensitive identifiable data should not be entered into public AI systems. If AI assists transcription, coding or translation, the research team remains responsible for verification, confidentiality and disclosure of the tool’s role.

8. Ethics review and research governance

Research ethics committees, institutional review boards or equivalent bodies assess whether a project’s risks are reasonable in relation to its value and whether protections are adequate. Review should occur before recruitment or data collection, unless an authorized exemption applies. Governance may also require site approval, clinical authorization, data-access agreements, community permission, insurance and safety assessments.

A strong application demonstrates alignment among the research question, methods, risks and protections. It should explain recruitment, consent, incentives, data management, researcher competence, safeguarding, dissemination and arrangements for adverse events or protocol deviations.

Ethics review is not a substitute for researcher judgment. Approved protocols do not anticipate every field situation. Researchers should report substantial changes, unanticipated problems and serious incidents promptly. Amendments may be required when sample groups, interview topics, locations, data uses or technologies change.

A useful governance distinction is:

• Ethics review: evaluates moral acceptability and participant protection.
• Scientific review: evaluates validity, relevance and methodological quality.
• Data governance: controls collection, access, security and reuse.
• Site governance: confirms operational permission and local accountability.
• Community governance: addresses collective authority, cultural safety and benefit.

These systems should communicate rather than operate as isolated administrative steps.

9. Conflicts of interest and research integrity

A conflict of interest exists when secondary interests could improperly influence professional judgment. Conflicts may be financial, professional, institutional, personal or relational. Examples include a funder that sells a health technology under study, a manager recruiting their own employees, or a researcher evaluating a program they designed.

A conflict is not necessarily misconduct. The ethical requirement is disclosure, assessment and management. Strategies include independent recruitment, separating care from research roles, independent data analysis, pre-specified dissemination commitments, transparent funding statements, recusal from decisions and oversight by an external committee.

Researchers must not suppress unfavorable findings, selectively quote participants or alter interpretations to satisfy sponsors. Data fabrication, falsification, plagiarism and deceptive authorship undermine both scientific and public trust. Participants should not be presented as endorsing conclusions that they did not make.

10. Applied case: a digital mental-health study

Background

A university–hospital partnership proposes interviews with young adults using a mobile mental-health application. The funder is the company that developed the application. The study aims to understand why some users discontinue treatment.

Problem

The research team plans to recruit through clinicians, record interviews online and upload recordings to a commercial transcription service. Some users report suicidal thoughts in the app. The protocol offers a small payment but does not explain whether quotations may appear in reports. The company requests access to raw transcripts.

Analysis

Several ethical concerns arise. Recruitment through clinicians may create perceived pressure and blur care and research. Online interviews may expose participants to privacy risks. Suicidality creates safeguarding obligations. The transcription service may create cross-border data-transfer and confidentiality risks. The funder’s request for raw transcripts is a conflict-of-interest and data-governance issue. Finally, payment and unclear quotation practices weaken informed consent.

Approach and solution

The team should:

1. recruit through an independent pathway where possible and state clearly that care will not be affected;
2. use an accessible consent process explaining recording, quotations, withdrawal limits and data transfers;
3. create a risk protocol for disclosures of imminent danger, including trained interviewers and referral pathways;
4. offer participants control over recording and the use of identifiable quotations;
5. use institutionally approved, secure transcription or locally processed transcription;
6. provide the company with aggregated findings rather than raw identifiable transcripts;
7. appoint an independent data-access committee and disclose the funding relationship;
8. involve young people in reviewing the interview guide and dissemination plan; and
9. submit all changes for ethics and governance approval before implementation.

Outcome

The revised project receives approval after the team separates clinical and research roles, introduces a safeguarding procedure, limits sponsor access and establishes participant and youth-advisory input. The findings identify usability problems and concerns about algorithmic recommendations that the company had not anticipated.

Learning points

The case demonstrates that consent, safeguarding, data governance, conflicts of interest and equity are interconnected. Ethical review is strongest when it examines the full socio-technical system rather than only the interview questions.

11. Practical ethical quality framework

Before beginning a qualitative health study, researchers can ask:

Purpose: Is the study necessary, socially valuable and methodologically credible?

People: Who may be harmed, excluded or pressured? Who has decision-making authority?

Process: Are consent, recruitment, compensation and withdrawal procedures understandable and voluntary?

Protection: Are confidentiality, safeguarding, trauma support and digital security adequate?

Power: How will researcher authority, institutional interests and community relationships be managed?

Partnership: Are participants or communities involved in design, interpretation and dissemination?

Publication: Could findings identify, stigmatize or commercially disadvantage anyone?

Persistence: What obligations continue after data collection and publication?

This framework should be documented, discussed with the research team and revisited during fieldwork. It is especially valuable when circumstances change unexpectedly.

Challenges and limitations

Ethical decision-making often involves competing goods. Protecting anonymity may reduce contextual richness; broad inclusion may increase risk; returning transcripts may empower participants but also expose them to distress; community review may improve accountability but create pressure to suppress critical findings. There is no universal solution independent of context.

Ethics committees may be slow, institutionally distant from communities or oriented toward biomedical models. Researchers must therefore pursue meaningful engagement beyond compliance. At the same time, flexible qualitative research cannot justify undocumented procedural changes. Adaptation requires transparency, proportionality and accountability.

Digital research creates further uncertainty because platforms, vendors and algorithms change faster than governance policies. Technical security cannot solve relational harms such as stereotyping, extractive interpretation or loss of community control. Ethical competence therefore requires legal literacy, cultural humility, methodological skill and organizational support.

Best-practice principles

• Treat consent as continuous and relational.
• Design for the least powerful participant, not the average participant.
• Collect the minimum necessary identifiable data.
• Explain confidentiality limits honestly.
• Prepare safeguarding and distress-response procedures before fieldwork.
• Separate research, care, managerial and commercial roles where possible.
• Involve communities in defining questions, interpreting findings and deciding dissemination.
• Respect Indigenous authority, cultural protocols and collective data rights.
• Record and manage conflicts of interest.
• Use secure, approved technologies and scrutinize AI suppliers.
• Report limitations and adverse events rather than concealing them.
• Make findings accessible to those who contributed the knowledge.

Chapter Summary

Ethics in qualitative health research is a continuing practice of respect, protection, accountability and justice. Informed consent must be understandable, voluntary and revisited when research circumstances change. Confidentiality requires more than removing names; distinctive narratives, small populations and linked datasets can remain identifiable. Vulnerability is contextual, and trauma-informed research provides choice, safety and appropriate referral without confusing research with therapy. Safeguarding plans must address serious risk and legal duties.

Equitable research examines who defines problems, who is included, who controls knowledge and who benefits. Indigenous and decolonizing approaches emphasize relational accountability, collective authority, cultural protocols and community benefit. Data governance requires clear responsibility, secure systems, purpose limitation, controlled access and careful decisions about secondary use. Ethics review should operate alongside scientific, institutional, site and community governance. Conflicts of interest must be disclosed and managed, especially when sponsors control technologies or services under study. Digital tools and AI can improve efficiency but introduce risks involving privacy, cross-border processing, bias, accuracy and accountability.

Key Terms

Informed consent: Voluntary agreement to participate based on understandable information.

Confidentiality: Protection of information entrusted to the research team from unauthorized disclosure.

Anonymity: A condition in which identity cannot be linked to research data.

Pseudonymization: Replacing identifiers with codes while retaining a possible re-identification key.

Vulnerability: A contextual condition that may limit autonomy or increase exposure to harm.

Trauma-informed research: Research designed around safety, choice, trust, collaboration and empowerment.

Safeguarding: Measures to prevent and respond to abuse, neglect, exploitation or serious danger.

Equity: Fairness achieved by addressing differing barriers, needs and power relations.

Reflexivity: Critical examination of how the researcher influences the research process.

Data stewardship: Responsible care, management and use of data.

Data governance: Rules and practices governing data collection, access, security and reuse.

Data minimization: Collecting only data necessary for the specified purpose.

Collective data rights: Rights and authority held by a community over knowledge concerning it.

Decolonizing research: Approaches that challenge extractive knowledge systems and redistribute authority.

Relational accountability: Responsibility to participants, communities and ongoing relationships.

Ethics committee: An independent body reviewing the ethical acceptability of research.

Conflict of interest: A secondary interest that could improperly influence professional judgment.

Undue influence: Excessive pressure or inducement that compromises voluntary choice.

Member checking: Inviting participants to comment on interpretations or representations.

Research integrity: Commitment to honest, accurate, transparent and responsible research.

Review Questions

Short Answer Questions

1. Distinguish confidentiality from anonymity.
2. Why is informed consent considered an ongoing process in qualitative research?
3. Identify three elements of trauma-informed interviewing.
4. What is meant by data stewardship?
5. Why should conflicts of interest be disclosed even when no misconduct has occurred?

Long Answer Questions

1. Explain how respect for persons, beneficence, non-maleficence and justice apply to qualitative health research.
2. Discuss the ethical challenges of interviewing people who may be vulnerable or have experienced trauma.
3. Compare equality and equity in the design of a health-services study.
4. Explain how Indigenous and decolonizing principles challenge extractive research practices.
5. Describe the relationship among ethics review, scientific review, data governance and site governance.

Analytical / Application Questions

1. A hospital manager asks a researcher to interview employees whom the manager directly supervises. Analyse the risks and propose a recruitment and consent strategy.
2. A study team wants to upload identifiable interviews to a free online AI transcription service. Evaluate the ethical and governance issues and recommend an alternative process.
3. A community objects to publication of a finding that it believes may increase stigma, although the finding is methodologically well supported. Develop a response that protects participants while preserving research integrity.

Multiple Choice Questions with Answers

Question 1

Which feature is essential to valid informed consent?

A. A signed form in every circumstance

B. Voluntary agreement based on adequate understanding

C. Approval by the study funder

D. A guarantee that no distress will occur

Correct Answer: B. Voluntary agreement based on adequate understanding

Question 2

What is the most accurate distinction between anonymity and confidentiality?

A. Anonymity protects data during storage, whereas confidentiality protects data during collection

B. Anonymity means the participant cannot withdraw, whereas confidentiality means withdrawal is possible

C. Anonymity prevents linking data to identity, whereas confidentiality limits unauthorized disclosure

D. They are identical terms in qualitative research

Correct Answer: C. Anonymity prevents linking data to identity, whereas confidentiality limits unauthorized disclosure

Question 3

Which practice best reflects trauma-informed interviewing?

A. Requiring participants to describe events in chronological detail

B. Avoiding all discussion of difficult experiences

C. Allowing participants to skip questions, pause or stop the interview

D. Promising professional therapy after every interview

Correct Answer: C. Allowing participants to skip questions, pause or stop the interview

Question 4

Why can removing a participant’s name fail to ensure anonymity?

A. Names are never needed for analysis

B. Distinctive combinations of details may identify the participant

C. Pseudonyms automatically create legal liability

D. Anonymous data cannot be securely stored

Correct Answer: B. Distinctive combinations of details may identify the participant

Question 5

Which principle is most closely associated with decolonizing research?

A. Maximizing extraction of data for external institutions

B. Allowing only academic researchers to interpret findings

C. Redistributing authority and supporting community control of knowledge

D. Avoiding all community involvement to preserve objectivity

Correct Answer: C. Redistributing authority and supporting community control of knowledge

Question 6

What is data minimization?

A. Deleting all data immediately after collection

B. Collecting only data necessary for the specified purpose

C. Sharing data with as many researchers as possible

D. Replacing qualitative data with numerical data

Correct Answer: B. Collecting only data necessary for the specified purpose

Question 7

A researcher evaluates a programme they personally designed. What is the primary ethical issue?

A. Anonymity

B. Conflict of interest

C. Translation validity

D. Data saturation

Correct Answer: B. Conflict of interest

Question 8

Which action is most appropriate when a protocol changes substantially during fieldwork?

A. Implement the change secretly to avoid delay

B. Ask participants to approve the entire institutional protocol

C. Seek the required ethics amendment or governance approval

D. Destroy all data already collected

Correct Answer: C. Seek the required ethics amendment or governance approval

Question 9

Which is an appropriate concern when using a commercial AI transcription service?

A. AI tools cannot process spoken language

B. The service may retain or transfer identifiable data outside approved systems

C. Transcription eliminates the need for confidentiality

D. Human verification is prohibited

Correct Answer: B. The service may retain or transfer identifiable data outside approved systems

Question 10

What does equity require in research recruitment?

A. Identical support for every participant regardless of barriers

B. Selection of only participants who are easiest to reach

C. Adjusting procedures to address differing barriers and needs

D. Excluding all groups considered vulnerable

Correct Answer: C. Adjusting procedures to address differing barriers and needs