Research ethics and data protection: before you share a survey link
An apparently harmless online questionnaire can collect names, email addresses, workplaces and sensitive answers. Research ethics concerns how these data and the people behind them are treated from topic selection to the final report.
Ethics applies at every stage: design and access, collection, processing, storage, analysis and reporting. Obtain informed consent, honour promises about confidentiality and anonymity, and follow applicable data-protection requirements. Avoid collecting personal information that the research does not need.
1. What is research ethics?
Saunders et al. (2023, p. 253) describe research ethics as standards of behaviour guiding researchers in relation to the rights of participants and others affected. Social norms do not always produce an unambiguous answer, and ethical considerations can conflict.
A deontological position emphasises duties and rules; a teleological position evaluates actions through their consequences. Comparing benefits to one group with costs borne by another seldom resolves every dilemma. University and professional codes therefore offer principles rather than a formula for every situation (Saunders et al., 2023, p. 253).
Bryman and Bell (2019, pp. 50–52) discuss the Canadian TCPS2 principles of respect for persons, concern for welfare and justice. Avoiding harm takes priority over the pursuit of knowledge. Schindler (2022, p. 45) considers responsibilities of participants, sponsors and researchers: honest responses, transparent purposes, freedom from pressure to change results, and protection of consent, privacy, safety and confidentiality.
2. Ethical review
Find out which review procedures your institution requires and obtain the necessary approval before collecting data. Saunders et al. (2023, pp. 253–255, 277) distinguish lighter review for minimal-risk work from full committee review.
| Level | Typical application | Reviewers |
|---|---|---|
| Light-touch or fast-track | Uncontroversial proposals presenting minimal risk | Supervisor or academic staff using an approved process under committee oversight |
| Full review | Ethically sensitive or higher-risk projects | Faculty or university ethics committee |
Review considers participants' rights, dignity and welfare, ethical aspects of study quality, and researcher safety. Hospitals or other host organisations may require additional review, which should be allowed for in the timetable (Saunders et al., 2023, p. 255).
The questions and participant group, such as children, can matter more than the method. Do not assume that surveys are ethically harmless because interviews or observation seem more intrusive (Saunders et al., 2023, p. 255).
3. Ten ethical principles
Codes seek to prevent misconduct and harm while promoting integrity, respect, fairness and public benefit. Saunders et al. (2023, pp. 256–257) summarise common principles.
| Principle | Practical meaning |
|---|---|
| Integrity, fairness and openness | Be honest and accurate; avoid deception, misrepresentation and unrealistic promises; disclose conflicts |
| Respect for others | Recognise rights and dignity |
| Avoidance of harm | Anticipate emotional, physical and social harm, including embarrassment, pressure and identification |
| Privacy | Respect boundaries throughout participation and data management |
| Voluntary participation and withdrawal | Do not pressure people; explain withdrawal and any practical limits |
| Informed consent | Provide enough information for a free decision |
| Confidentiality and anonymity | Protect identities and honour explicit agreements |
| Responsible analysis and reporting | Do not fabricate or falsify; report findings that contradict expectations |
| Data-management compliance | Follow applicable rules on processing, storage and sharing |
| Researcher safety | Anticipate threats, distress and increased everyday risks |
Assess harm by both likelihood and severity (Saunders et al., 2023, p. 255). Ethical practice goes beyond an approval form: do not exaggerate benefits, and honour promises such as supplying a results summary (p. 258).
4. Informed consent
Agreement to a collection method is not agreement to every subsequent use of the evidence (Saunders et al., 2023, p. 264).
| Position | Participant | Researcher |
|---|---|---|
| No consent | Unaware of rights | Uses deception to obtain evidence |
| Inferred consent | Does not fully understand rights | Infers permission for use from access or a returned questionnaire |
| Informed consent | Agrees freely with adequate information | Explains participation, rights and uses; allows questions and time without pressure |
A participant information sheet should identify the researcher, institution, funder, invited participants, purpose, required data and time, possible consequences, rights, confidentiality, access, analysis, reporting, retention or destruction, and contact details (Saunders et al., 2023, pp. 265–266).
Returning a genuinely anonymous questionnaire may be treated as consent in some settings, but an explicit consent question is straightforward and the information sheet remains necessary. Interviews may require a signed form, including separate permission for recording (Saunders et al., 2023, p. 266).
Creswell and Creswell (2018, pp. 147–149), drawing on Sarantakos, similarly identify researcher and sponsor, purpose, benefits, participation, risks, confidentiality, withdrawal and contacts. Instructions must make clear that participation and signing are voluntary.
Failing to send an agreed summary can damage trust and future access. A promise of anonymity applies to the final report as well as collection (Saunders et al., 2023, p. 267). Explain in advance if withdrawing particular responses will become impossible after irreversible anonymisation.
5. Ethics throughout the research process
Saunders et al. (2023, pp. 262–274) discuss ethical concerns across the stages of a study.
Topic, design and access
Consider the purpose, evidence, collection method, participant consequences and possible uses from the outset. Much ethical risk is foreseeable. Avoid pressure when requesting access: insider researchers surveying colleagues or subordinates must not use their position to compel participation (pp. 262–263).
Collection
Participants retain the right to decline questions or withdraw. Stay within agreed uses or renegotiate consent. Record evidence accurately; falsification and fabrication are unacceptable. Covert collection after access has been refused is particularly problematic; deception requires appropriate review and debriefing. Protect researcher safety through safe meeting locations, contact arrangements and avoiding unnecessary disclosure of a home address (pp. 267–270).
Processing and storage
A data-management plan should specify what is collected, formats, backups, ownership, access, retention and destruction. Store identifying material separately and securely from de-identified working files. Encrypt portable storage and securely destroy paper. Some institutions restrict third-party cloud storage for sensitive information (pp. 271–272).
Analysis and reporting
Do not select only favourable evidence or misrepresent statistical precision. Indirect identifiers can reveal participants: a unique job title, an organisation's name in an appendix, real names in acknowledgements, recognisable initials or photographs can defeat a confidentiality promise (pp. 272–273).
Creswell and Creswell (2018, pp. 148, 152–153) also stress reporting unsupported hypotheses, separating respondent names from survey data during coding, citing paraphrased material and marking quotations. Agree authorship arrangements early when publication is planned.
6. Internet-mediated research
Saunders et al. (2023, pp. 258–261) identify particular concerns:
- Joining a community only to extract data without permission can involve deception and disrespect.
- Public accessibility does not by itself establish consent to research use. Consider participants' expectations and context.
- Verbatim quotations may be searchable back to the author; avoid traceable quotations without suitable permission.
- Transmission can fail through misdirected email or survey-system errors. Explain relevant risks during consent.
- Use institutional contact details and avoid unsolicited mass invitations.
7. Personal-data protection
Personal data can identify someone directly, through a name, image or IP address, or indirectly when combined with other information. Truly anonymous data are different (Saunders et al., 2023, p. 275). The book summarises seven GDPR principles:
| Principle | Application |
|---|---|
| Lawfulness, fairness and transparency | Explain what is collected and why |
| Purpose limitation | Do not reuse contact details for unrelated purposes |
| Data minimisation | Collect only variables necessary to the question |
| Accuracy | Correct or update information when needed |
| Storage limitation | Remove identifiers when no longer needed, subject to applicable safeguards and requirements |
| Security | Protect against improper processing, loss and damage |
| Accountability | Be able to demonstrate compliance |
Effective anonymisation differs from replacing names while retaining a re-identification key. Techniques include removing identifiers, using pseudonyms in reporting, aggregation, reducing precision and top- or bottom-coding distinctive values. The book's legal discussion is introductory; researchers must understand the law applicable to their own setting (Saunders et al., 2023, p. 276).
Vietnam's Personal Data Protection Law No. 91/2025/QH15 was adopted on 26 June 2025 and took effect on 1 January 2026. It recognises data-subject rights including being informed, consent and withdrawal, access, correction and requests for erasure under the applicable provisions. Decree No. 356/2025/NĐ-CP, issued on 31 December 2025, details implementation. Read the applicable texts and institutional rules when collecting names, email addresses, phone numbers or sensitive information. This overview is not legal advice.
8. Workflow, sample wording and checklist
Assess risks at topic selection
Identify participants, sensitive questions, possible harm and its likelihood and severity.
Decide whether identifiers are needed
Remove unnecessary fields. For linked waves, consider participant-generated or random codes and keep any lookup table separate.
Prepare information and consent
Place clear information and an explicit choice before the substantive questionnaire.
Obtain required review
Do not begin formal collection before the necessary approval.
Plan data management
Specify storage, backups, permissions, retention and destruction.
Review the final report
Check appendices, acknowledgements, quotations and descriptive tables for direct or indirect identification.
Illustrative introduction to an online questionnaire
I am [name], a student on [programme] at [institution]. This survey supports “[title]” and aims to [one-sentence purpose]. It takes approximately [n] minutes. Participation is voluntary; you may skip questions or stop at any time. The survey does not collect names or contact information [or specify the limited purpose and separate storage of email addresses]. Data will be used for the stated research, accessed only by [researcher/supervisor], reported in aggregate, and [securely retained/destroyed] by [date]. Please contact [institutional email] with questions.
[ ] I have read the information and agree to participate.
[ ] I do not agree to participate.
| Practice | Problem | Better response |
|---|---|---|
| Requiring email or phone “just in case” | Unnecessary personal information | Remove the field or make its purpose and optional status clear |
| Asking managers to distribute a staff survey | Possible pressure and doubts about voluntariness | Use a suitable invitation process and make refusal confidential |
| Attaching named raw data | Identification and broken promises | Include only appropriately anonymised evidence if necessary |
| Removing inconvenient observations without criteria | Selective reporting | Specify exclusions and report counts and reasons |
| Quoting scraped social posts verbatim | Searchable identification | Obtain suitable permission or use an ethically justified de-identified presentation |
Checklist
- Know the institutional code and required review level.
- Assess likelihood and severity of harm.
- Provide information and consent before collection.
- Explain refusal, skipped questions, withdrawal and applicable limits.
- Minimise personal information and separate identifiers.
- Document storage, access, retention and destruction.
- Check the entire report for identity disclosure.
- Understand applicable personal-data law.
Conclusion
Ethical failures can arise from small conveniences: an extra email field, a manager's invitation or an unedited data appendix. Collect only what is needed, explain what will happen to it, and honour that explanation throughout the project.
References
Main source
- Saunders, M. N. K., Lewis, P., & Thornhill, A. (2023). Research methods for business students (9th ed.). Pearson. Pearson
Additional sources
- Bryman, A., & Bell, E. (2019). Social research methods (5th Canadian ed.). Oxford University Press.
- Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). SAGE.
- Schindler, P. S. (2022). Business research methods (14th ed.). McGraw-Hill.
Vietnamese legislation
- Chính phủ. (2025). Nghị định số 356/2025/NĐ-CP quy định chi tiết một số điều và biện pháp thi hành Luật Bảo vệ dữ liệu cá nhân. https://thuvienphapluat.vn
- Quốc hội. (2025). Luật Bảo vệ dữ liệu cá nhân số 91/2025/QH15. https://thuvienphapluat.vn
This methodological overview is not legal advice. Textbooks are paraphrased with page references; authors quoted through those books are secondary citations. The questionnaire wording and the applications in Table 4 were written for this article.
