Research Ethics: Principles, Cases & Compliance

TAKEAWAYS:

• Research ethics ensures truthfulness, fairness, and respect for participants/subjects, governed by codes like Belmont Report and Nepal’s National Health Research Council (NHRC) guidelines.
• Key principles: Autonomy (informed consent), Beneficence (minimize harm), Justice (fair selection), and Non-maleficence (avoid harm).
• Ethical dilemmas arise in human studies, animal testing, data privacy, and AI bias—solved via ethics review boards (ERBs) and institutional policies.
• Plagiarism and fabrication are severe violations; tools like Turnitin and reference managers (Zotero, Mendeley) help maintain integrity.
• Real-world ties: eSewa’s user data policies, Ncell’s telemedicine consent forms, and NEPSE’s insider-trading rules all reflect ethical research practices.
• Exam focus: Define principles, explain ERB roles, and analyze case studies (e.g., Stanford Prison Experiment, Facebook’s emotional contagion study).

1. Core Principles of Research Ethics

Research ethics is built on four foundational principles (derived from the Belmont Report, 1979) and Nepal’s NHRC guidelines for health/social research. These principles guide how researchers interact with participants, handle data, and design studies.

The Four Key Principles

mindmap
  root((Research Ethics Principles))
    Autonomy["Autonomy: Respect for Persons
    → Informed Consent
    → Voluntary Participation
    → Right to Withdraw"]
    Beneficence["Beneficence: Do Good
    → Maximize Benefits
    → Minimize Harm/Risk
    → Benefit-Risk Assessment"]
    Justice["Justice: Fair Treatment
    → Equitable Selection
    → Avoid Exploitation
    → Vulnerable Groups Protection"]
    Non-maleficence["Non-maleficence: Do No Harm
    → Avoid Physical/Psychological Harm
    → Confidentiality
    → Anonymity"]

How It Works in Practice

• Autonomy: Participants must understand the study (risks, benefits) and agree voluntarily. Example: A Khalti user survey must disclose how data will be used and allow opt-out.
• Beneficence: Researchers must weigh risks vs. benefits. Example: A Pathao driver study on stress levels must ensure drivers aren’t harmed by data collection.
• Justice: Participants should be fairly selected—not just convenient. Example: A NTC study on rural internet access can’t exclude villages without justification.
• Non-maleficence: Avoid harm. Example: A Nepalese hospital study on patient records must anonymize data to protect privacy.

Worked Example: eSewa’s Ethical Data Use eSewa collects user transaction data. Ethical concerns include:

1. Autonomy: Users must consent to data sharing (e.g., for fraud detection).
2. Beneficence: Data helps reduce financial crimes but must not expose users to identity theft.
3. Justice: Low-income users shouldn’t be excluded from benefits like loan approvals.
4. Non-maleficence: Data breaches (e.g., 2021 hack) violated confidentiality.

2. Ethical Review Boards (ERBs) and Compliance

Before starting research, proposals must be reviewed by an Ethics Review Board (ERB) or Institutional Review Board (IRB). In Nepal, bodies like:

• National Health Research Council (NHRC)
• Tribhuvan University Central Department of Research (TU-CDR)
• Pokhara University Ethics Committee

flowchart TD
  A["Research Proposal Submitted"] --> B["ERB Screening"]
  B -->|"High Risk?"| C["Full Review"]
  B -->|"Low Risk?"| D["Expedited Review"]
  C --> E["Modifications Required?"]
  D --> E
  E -->|"Yes"| F["Resubmit"]
  E -->|"No"| G["Approval"]
  G --> H["Commence Research"]

Key ERB Questions for Researchers

Real-World Case: Facebook’s Emotional Contagion Study (2014)

• Violation: Manipulated 689,000 users’ news feeds to study emotional contagion without full consent.
• Outcome: Criticized for lack of transparency and psychological harm risks.
• Lesson: ERBs must scrutinize digital/social media research for ethical lapses.

3. Special Ethical Considerations

A. Human Subjects Research

• Informed Consent: Must be clear, voluntary, and documented.
  • Example: A Daraz customer satisfaction survey must explain how responses will be used (e.g., improving services vs. selling data).
• Vulnerable Groups: Extra protections for children, prisoners, mentally ill, or economically disadvantaged.
  • Nepali Example: A school dropout study in rural Nepal must get parental consent and ensure no child feels pressured.

B. Animal Research

• 3Rs Principle: Replace, Reduce, Refine.
  • Replace: Use alternatives (e.g., computer models instead of lab animals).
  • Reduce: Minimize animal numbers.
  • Refine: Reduce suffering (e.g., better housing, pain management).
• Nepal’s CPCSEA (Committee for the Purpose of Control and Supervision of Experiments on Animals) oversees this.

C. Data and Privacy Ethics

• GDPR (Global) vs. Nepal’s Data Privacy Laws: Researchers must comply with local regulations.
  • Example: A Ncell telemedicine study must anonymize patient data and store it securely.
• AI and Bias: Algorithms can discriminate (e.g., hiring tools favoring certain demographics).
  • Solution: Audit AI models for fairness (e.g., Google’s What-If Tool).

D. Plagiarism and Fabrication

• Plagiarism: Using others’ work without credit.
  • Types:
    • Direct: Copy-pasting without quotes.
    • Paraphrasing: Changing words but keeping structure.
    • Self-Plagiarism: Reusing your own work without citation.
• Fabrication: Inventing data (e.g., fake survey results).
• Tools to Avoid It:
  • Turnitin (detects similarity).
  • Zotero/Mendeley (reference managers).
  • Grammarly (citation checks).

4. Ethical Dilemmas and Case Studies

Case 1: Stanford Prison Experiment (1971)

• Issue: Participants were psychologically harmed by role-playing guards/prisoners.
• Ethical Violation: Non-maleficence (harm caused) and autonomy (lack of exit options).
• Lesson: Debriefing and mental health support are mandatory post-study.

Case 2: Milgram’s Obedience Study (1963)

• Issue: Participants were tricked into believing they harmed others (fake shocks).
• Ethical Violation: Deception and psychological stress.
• Lesson: Full disclosure of risks is essential.

Case 3: Nepali Context – NEPSE Insider Trading

• Issue: Some traders manipulate stock prices using non-public info.
• Ethical Violation: Justice (unfair advantage) and integrity.
• Regulation: SEBON (Securities Board of Nepal) enforces penalties.

5. Writing Ethically: Avoiding Misconduct

A. Proper Citation

• APA/MLA/Harvard: Different styles for references, quotes, and paraphrases.
• Example (APA):

  "Research ethics ensures participant safety (NHRC, 2020, p. 15)." Reference: Nepal Health Research Council. (2020). Ethical Guidelines for Health Research. Kathmandu.

B. Handling Conflicts of Interest

• Disclose: Financial ties, personal relationships, or biases.
• Example: If a Ncell researcher studies its own app, they must declare conflict and use independent reviewers.

C. Open Science vs. Ethical Risks

• Open Access: Sharing data publicly risks privacy breaches.
• Solution: Anonymize data before publication.

In the Real World

1. eSewa’s Ethical Data Policies

   • Idea Used: Informed Consent & Data Privacy
   • How: eSewa’s terms of service explain how transaction data is used (e.g., fraud detection) and allow users to opt out of data sharing. Violations (like the 2021 hack) led to stricter ERB oversight for fintech research in Nepal.

2. Pathao’s Driver Well-being Studies

   • Idea Used: Beneficence & Non-maleficence
   • How: Pathao’s 2022 study on driver stress required:
     • Voluntary participation (drivers could skip surveys).
     • Anonymized data (no personal IDs linked to responses).
     • Mental health resources offered post-study.

3. NEPSE’s Insider Trading Regulations

   • Idea Used: Justice & Integrity
   • How: SEBON’s 2023 amendments mandate:
     • Whistleblower protections for employees reporting misconduct.
     • Real-time trading audits to prevent manipulation.
     • Public disclosures of conflicts of interest by board members.

Exam Tip

How to Score Full Marks in TU/PU Exams

1. Define Key Terms Precisely

   • Example: "Ethics Review Board (ERB) is an independent committee that evaluates research proposals to ensure compliance with autonomy, beneficence, justice, and non-maleficence principles."
   • Avoid: Vague answers like "ERB checks if research is ethical."

2. Use Real-World Examples

   • Link to Nepal: "Like Ncell’s telemedicine projects, researchers must obtain informed consent from patients and anonymize health records to comply with Nepal’s PDPA 2018."
   • Link to Global: "Facebook’s emotional contagion study violated autonomy by manipulating users without consent, leading to stricter IRB guidelines worldwide."

3. Compare Ethical Frameworks

   • Table Method: Use 2-column comparisons (e.g., GDPR vs. Nepal’s PDPA) to show understanding of jurisdictional differences.

4. Analyze Case Studies Critically

   • Structure:
     1. Identify the ethical violation (e.g., lack of consent).
     2. Explain the principle breached (e.g., autonomy).
     3. Suggest improvements (e.g., mandatory ERB review).

5. Memorize Key Acronyms

   • ERB/IRB: Ethics Review Board.
   • NHRC: Nepal Health Research Council.
   • PDPA: Personal Data Protection Act (Nepal).
   • GDPR: General Data Protection Regulation (EU).

6. Practical Tips for Short Notes

   • Plagiarism: "Types include direct, paraphrasing, mosaic, and self-plagiarism. Tools like Turnitin detect similarity."
   • Informed Consent: "Must be voluntary, informed, and documented—key for autonomy."

Common Mistakes to Avoid

• Overgeneralizing: Don’t say "All research is ethical"—always mention ERB oversight.
• Ignoring Local Laws: Nepal’s PDPA 2018 is different from GDPR; exams may test this.
• Skipping Examples: Always tie theory to real cases (e.g., eSewa, Ncell, NEPSE).
• Confusing Principles: Beneficence ≠ Justice—beneficence is about doing good, justice is about fairness.