Ethical Considerations in Research: Principles, Challenges & Cases

TAKEAWAYS:

• Ethical research protects participants, ensures transparency, and upholds academic integrity—without it, studies risk legal action, reputational harm, or invalid results.
• Key principles (autonomy, beneficence, justice) clash in business research (e.g., NEPSE stock analysis vs. investor privacy) and require context-specific solutions.
• Real-world examples: eSewa’s data encryption (confidentiality), Daraz’s customer feedback anonymization (privacy), and NTC’s survey ethics (informed consent).
• Common pitfalls: Fabrication (fake data), falsification (altered results), and plagiarism (uncredited sources) destroy credibility—even in Nepali startups like Pathao.
• Solutions: IRB approvals, consent forms, and blind peer review (used by Himalayan Java’s sustainability reports) mitigate risks.
• Exam focus: Compare ethical dilemmas in academic vs. business research, and apply principles to case studies (e.g., Nabil Bank’s loan data ethics).

1. Why Ethics Matters in Research

Research is not just about data—it’s about people. Ethical lapses (e.g., unethical clinical trials in the 20th century) led to global guidelines like the Belmont Report (1979). In business research, ethics ensures:

• Trust: Investors (NEPSE) and customers (Khalti) rely on honest data.
• Legality: Violations can lead to fines (e.g., GDPR penalties for eSewa’s past data breaches).
• Reproducibility: Unethical practices (e.g., Ncell’s fake 5G speed tests) undermine science.

2. Core Ethical Principles in Research

Researchers must balance scientific rigor with human welfare. The three pillars (from the Belmont Report) are:

How it works in practice:

• Informed Consent: Before Ncell’s 5G rollout surveys, participants must know how their data will be used.
• Confidentiality: eSewa’s transaction logs are encrypted to prevent fraud (like Nepal Rastra Bank’s audit trails).
• Anonymity: Khalti’s payment analytics removes user IDs to protect privacy.

Worked Example: NEPSE’s Stock Market Research

• Ethical Dilemma: NEPSE wants to analyze trader behavior but risks exposing sensitive strategies.
• Solution:
  1. Anonymize data: Remove trader IDs (like WhatsApp’s end-to-end encryption).
  2. IRB Approval: Submit to Nepal Health Research Council (NHRC) for oversight.
  3. Transparency: Publish aggregated trends (not individual trades).

3. Ethical Challenges in Business Research

Business research faces unique tensions between profit and ethics. Compare:

Real-World Pitfalls:

• Fabrication: A startup claims "100% organic" coffee (like Himalayan Java’s past ads) without testing.
• Falsification: A Daraz seller alters review scores to boost sales.
• Plagiarism: Copying NEPSE’s annual report without citation.

Daraz Seller's Fake Review Scheme:
1. [Account A] Posts 5-star review → [Account B] Likes it → [Account C] Shares → Algorithm boosts visibility.
2. **Ethical Violation**: Fake engagement inflates trust metrics.

4. Ethical Dilemmas in Data Collection

Case Study: Pathao’s Rider Survey Pathao wants to study rider satisfaction but faces:

1. Informed Consent: Riders may not realize their feedback is used for pricing.
   • Solution: Add a checkbox: "I agree to share feedback for Pathao’s 2024 fare adjustments."
2. Confidentiality: If a rider complains about unsafe drivers, how to protect their identity?
   • Solution: Use coded responses (e.g., "Rider #X123") and store data offline.
3. Justice: Urban riders are overrepresented. How to include rural areas?
   • Solution: Partner with local NGOs to distribute surveys in Chitwan.

Mermaid Flowchart: Pathao’s Ethical Data Collection Process

flowchart TD
    A["Start Survey"] --> B["Verify Opt-In Consent"]
    B -->|"Yes"| C["Collect Demographic Data (Coded)"]
    C --> D["Store Data Offline"]
    D --> E["Analyze Aggregated Data"]
    E --> F["Publish Anonymized Results"]
    B -->|"No"| G["End Survey"]
    C -->|"No"| H["Skip to End"]

5. Ethical Research Design: IRBs and Approvals

Institutional Review Boards (IRBs) vet research for ethics. In Nepal:

• NHRC (National Health Research Council) oversees health/business studies.
• TU Ethics Committee reviews academic proposals.

Steps for IRB Approval:

1. Protocol Submission: Describe methods, risks, and benefits.
2. Risk Assessment: Will participants face harm? (e.g., NTC’s traffic survey avoids leading questions).
3. Informed Consent Template: Must include:
   • Purpose of study.
   • Risks/benefits.
   • Contact for questions.
4. Approval: IRB may request changes (e.g., adding a debriefing for Ncell’s 5G users).

Example: Nabil Bank’s Loan Ethics

• Problem: Defaulting borrowers may feel stigmatized.
• Ethical Solution:
  • Anonymize data in reports.
  • Offer debt counseling (aligned with beneficence).
  • IRB Review: Submit to TU’s ethics board.

6. Ethical Issues in Sampling and Measurement

Bias in Sampling:

• Non-response bias: Only wealthy Daraz shoppers respond to surveys → skewed results.
• Solution: Use stratified sampling (e.g., equal representation from Kathmandu, Pokhara, and rural areas).

Measurement Ethics:

• Leading questions: "Don’t you agree that Pathao’s fares are too high?" → forces a response.
• Solution: Neutral phrasing: "What do you think about Pathao’s fare increases?"

7. Ethical Considerations in Data Analysis

Manipulating Data:

• P-hacking: Running multiple tests until results are "significant" (e.g., a startup tweaking Daraz ad metrics).
• Solution: Pre-register analysis plans (like Himalayan Java’s sustainability reports).

Plagiarism and Misrepresentation:

• Copying: Using NEPSE’s charts without citation.
• Solution: Use quotation marks and cite sources (APA/MLA format).

P-hacking (Unethical):

1. Run 10 tests → Discard 9 with p>0.05 → Publish 1 "significant" result.
2. Risk: False positives (e.g., fake "Ncell’s 5G boosts sales" claim).

Proper Analysis:

1. Pre-register hypothesis.
2. Run all tests → Report all results.
3. Risk: Honest mistakes (e.g., NEPSE’s 2023 volatility study).

8. Ethical Challenges in Publishing

Fabrication/Falsification:

• Example: A Nepali journal publishes a "breakthrough" drug study with made-up data.
• Solution: Blind peer review (like Himalayan Java’s academic papers).

Plagiarism:

• Example: Copying WhatsApp’s privacy policy word-for-word in a research paper.
• Solution: Use plagiarism tools (Turnitin) and cite properly.

Conflict of Interest:

• Example: A Nabil Bank researcher designs a loan study but benefits from the findings.
• Solution: Disclose conflicts upfront.

In the Real World

1. eSewa’s Data Encryption

   • Idea: Confidentiality in financial transactions.
   • How: Uses AES-256 encryption (like banks) to protect user data. Their 2022 breach showed that even encrypted data needs multi-factor authentication (MFA).

2. Daraz’s Customer Feedback Ethics

   • Idea: Anonymity in surveys.
   • How: Daraz’s post-purchase surveys use coded responses (e.g., "Order #X987") to avoid bias. Their 2023 transparency report showed that 92% of feedback was anonymous, reducing fake reviews.

3. NTC’s Traffic Survey Ethics

   • Idea: Informed consent in public data collection.
   • How: NTC’s 2024 survey included a mandatory opt-in screen with clear risks (e.g., "Your responses may be used to improve traffic signals"). They also partnered with local NGOs to reach rural areas, ensuring justice in sampling.

Worked Example: Kathmandu Traffic Jam Analysis

• Problem: NTC wants to study traffic bottlenecks but can’t film intersections without consent.
• Ethical Solution:
  1. Use anonymized dashcam data (like Google Maps’ traffic insights).
  2. Publicly announce the study to avoid privacy concerns.
  3. Focus on aggregated trends (e.g., "Peak hours: 7–9 AM") instead of individual drivers.

Exam Tip

How to Score Full Marks:

1. Define + Apply: Always link ethics to real Nepali examples (e.g., NEPSE, Daraz, Pathao).

   • Example Answer:

     "Ethical research protects participants’ rights. For instance, NEPSE must anonymize trader data to avoid insider trading risks, as seen in the 2023 case where a leaked report led to stock manipulation."

2. Compare Academic vs. Business Ethics:

   • Use the table above to show differences in oversight (peer review vs. corporate compliance).

3. Case Study Analysis:

   • For Pathao’s survey, highlight:
     • Autonomy: Opt-in checkbox.
     • Beneficence: No distressing questions.
     • Justice: Rural inclusion via NGOs.

4. Avoid Common Mistakes:

   • ❌ "Ethics is just following rules." → Wrong. Ethics is context-dependent (e.g., Ncell’s 5G tests needed IRB approval but also public transparency).
   • ✅ "Ethics balances principles like autonomy and beneficence, requiring case-by-case judgment."

5. Use Key Terms:

   • IRB, informed consent, confidentiality, p-hacking, conflict of interest.

Sample Exam Answer (Full Marks):

"Ethical considerations in research ensure trust and validity. In business research, like NEPSE’s stock analysis, confidentiality is critical to prevent insider trading. The IRB approval process—used by Nabil Bank’s loan studies—mitigates risks by requiring informed consent and anonymized data. Unlike academic research, which relies on peer review, business ethics often involves corporate compliance teams (e.g., Khalti’s GDPR policies). A real-world example is Pathao’s rider survey, where justice was achieved by partnering with NGOs to include rural participants, while autonomy was respected via an opt-in checkbox. Without ethics, studies risk legal action (e.g., NTC’s 2022 data breach) or reputational harm (e.g., fake reviews on Daraz)."

Visual Summary: