Research Design & Methodologies: Types, Ethics, and Applications in Social Work

TAKEAWAYS:

• Research design is the blueprint linking research questions to data collection methods, with 5 core types (experimental, quasi-experimental, descriptive, exploratory, evaluative) each suited to different social work goals.
• Ontology (what exists) and epistemology (how we know it) shape methodology: positivist designs (quantitative) vs. constructivist designs (qualitative).
• Ethical challenges in social work research include power imbalances (e.g., interviewing vulnerable groups), informed consent (e.g., digital data collection via WhatsApp), and confidentiality (e.g., NTC’s customer complaint studies).
• Methodology selection depends on research purpose, resources, and participant context—e.g., a Daraz delivery delay study might use mixed methods (surveys + interviews with drivers).
• Real-world applications: eSewa’s evaluative design tests payment system trust; Pathao’s descriptive design maps rider safety patterns; NEPSE’s exploratory design probes investor sentiment.
• Research proposals must justify design choices, outline ethical safeguards, and align methods with social work values (e.g., anti-oppressive sampling for marginalized groups).

1. What Is Research Design?

Research design is the structured plan that guides how a study is conducted, ensuring validity, reliability, and ethical integrity. It bridges the research question and the data collection method, determining:

• What data to collect (quantitative/qualitative/mixed).
• How to collect it (surveys, interviews, observations).
• How to analyze it (statistics, thematic coding).
• How to generalize findings (sample representativeness).

Why it matters in social work: Social work research often deals with complex human behaviors, systemic issues (e.g., child trafficking), and ethically sensitive topics (e.g., domestic violence). A poor design can lead to biased results, unethical exploitation, or useless data. For example, a descriptive study on homelessness in Kathmandu might reveal patterns, but an experimental design could test whether cash transfers reduce recidivism.

2. Core Research Designs in Social Work

The 5 major types of research designs, with social work applications and Nepali examples:

MERMAID DIAGRAM:

flowchart TD
  A["Research Question"] --> B["Experimental"]
  A --> C["Quasi-Experimental"]
  A --> D["Descriptive"]
  A --> E["Exploratory"]
  A --> F["Evaluative"]
  B -->|"Example"| G["Does microfinance reduce domestic violence?"]
  C -->|"Example"| H["Impact of eSewa on financial inclusion in rural areas"]
  D -->|"Example"| I["Profile of street children in Kathmandu"]
  E -->|"Example"| J["Why do elderly avoid government healthcare?"]
  F -->|"Example"| K["Did Pathao’s safety training reduce accidents?"]
  B -->|"Strengths"| B1["High control"]
  B -->|"Weaknesses"| B2["Ethical concerns"]
  C -->|"Strengths"| C1["Real-world applicability"]
  C -->|"Weaknesses"| C2["Confounding variables"]
  D -->|"Strengths"| D1["Exploratory insights"]
  D -->|"Weaknesses"| D2["Lacks causality"]
  E -->|"Strengths"| E1["Flexible"]
  E -->|"Weaknesses"| E2["Low generalizability"]
  F -->|"Strengths"| F1["Actionable insights"]
  F -->|"Weaknesses"| F2["Political bias"]

3. Ontology and Epistemology: The Philosophical Foundation

Research designs are rooted in ontology (what is real?) and epistemology (how do we know it?).

Ontological Perspectives

Epistemological Approaches

4. Ethics in Research Design: Navigating the Politics

Social work research often faces ethical dilemmas due to:

1. Power imbalances (e.g., researchers vs. marginalized groups).
2. Sensitive topics (e.g., abuse, addiction).
3. Resource constraints (e.g., low-budget studies in rural areas).

Key Ethical Challenges

WORKED EXAMPLE: eSewa’s Trust Study Problem: eSewa wants to know why some users trust the platform for payments while others don’t. Design Choice:

• Mixed-methods evaluative design:
  • Quantitative: Survey 5,000 users on trust scales (e.g., "I feel safe using eSewa").
  • Qualitative: Focus groups with low-income women (who may face digital exclusion). Ethical Safeguards:
• Anonymized surveys (no IP tracking).
• Cash incentives for participants (to reduce coercion).
• Community advisory board (includes eSewa users). Findings:
• Quantitative: 60% trust eSewa, but only 30% of rural users do.
• Qualitative: Rural women cite lack of digital literacy and fear of scams. Policy Impact: eSewa launches SMS-based tutorials and partnered with banks for rural outreach.

5. Generating Research Questions and Objectives

Research questions must be:

• Clear (avoid vagueness like "What causes poverty?").
• Feasible (can you collect the data?).
• Ethical (does it harm participants?).
• Relevant (does it address a social work gap?).

How to Frame Questions

Worked Example: Pathao’s Safety Study

Research Question: "How does Pathao’s new rider safety training program affect accident rates among female drivers in Kathmandu?" Objectives:

1. Compare accident rates before/after training (using quasi-experimental design).
2. Identify common causes of accidents via driver interviews.
3. Assess drivers’ perceptions of safety via surveys. Methodology:

• Quantitative: Pre-test/post-test accident data from Pathao’s dashboard.
• Qualitative: Interviews with 50 trained drivers.

MERMAID DIAGRAM:

flowchart LR
  A["Research Question"] --> B["Descriptive"]
  A --> C["Exploratory"]
  A --> D["Evaluative"]
  B -->|"Example"| E["Profile of elderly abuse in nursing homes"]
  C -->|"Example"| F["Barriers to mental health services for LGBTQ+ youth"]
  D -->|"Example"| G["Effect of cash transfers on school dropout rates"]
  B -->|"Method"| B1["Quantitative: Pre-test/post-test accident data"]
  B -->|"Method"| B2["Qualitative: Interviews with 50 trained drivers"]
  C -->|"Method"| C1["Semi-structured interviews"]
  D -->|"Method"| D1["Controlled experiment"]

6. Quantitative vs. Qualitative Designs: When to Use Which?

7. Writing a Research Proposal: Step-by-Step

A well-structured proposal convinces funders/ethics boards and guides your study. Use this template:

A. Title

• Weak: "A Study on Poverty."
• Strong: "The Impact of Microfinance on Women’s Empowerment in Chitwan: A Mixed-Methods Study."

B. Background

• Why is this research needed?
  • Cite statistics (e.g., "60% of Chitwan women lack financial independence").
  • Link to social work theory (e.g., feminist economics).

C. Research Questions/Objectives

• Example:
  1. What is the current financial literacy level among Chitwan women?
  2. How does microfinance participation correlate with decision-making power in households?

D. Methodology

E. Timeline

Month 1: Ethics approval + pilot survey.
Month 3: Data collection (surveys + interviews).
Month 5: Data analysis + report writing.

F. Budget

WORKED EXAMPLE: NEPSE Investor Sentiment Study Proposal Title: "Exploring Investor Sentiment During Market Volatility: A Qualitative Study of NEPSE Traders."

Methodology:

• Design: Exploratory (to understand psychological factors behind trading decisions).
• Participants: 20 traders from Kathmandu Stock Exchange.
• Data Collection:
  • Semi-structured interviews (e.g., "How does news of political instability affect your trades?").
  • Trading journals (participants log decisions for 3 months).
• Ethics: Confidentiality agreements, no names in reports.

Expected Output:

• Policy brief for NEPSE on risk communication strategies.

8. Real-World Applications: Where Research Design Meets Social Work

Example 1: Khalti’s Fraud Detection System

• Design: Evaluative (testing whether AI fraud alerts reduce scams).
• Method:
  • Quantitative: Compare fraud rates before/after AI implementation.
  • Qualitative: User interviews with victims of attempted fraud.
• Ethical Issue: Bias in AI algorithms (e.g., flagging rural transactions as "suspicious").
• Outcome: Khalti adjusted thresholds and added manual review for flagged cases.

Example 2: Daraz’s Order Fulfillment Delays

• Design: Descriptive + Exploratory.
  • Descriptive: Map delivery times across districts (e.g., Kathmandu vs. Dhangadi).
  • Exploratory: Driver interviews to identify bottlenecks (e.g., traffic, warehouse inefficiencies).
• Method:
  • Quantitative: Order tracking data (10,000 orders).
  • Qualitative: Focus groups with 30 drivers.
• Finding: 70% of delays were due to last-mile traffic in urban areas.
• Solution: Dynamic routing algorithms + peak-hour delivery windows.

Example 3: NTC’s Customer Complaint Analysis

• Design: Evaluative (assessing whether complaint resolution training improves service).
• Method:
  • Pre-test: Survey 1,000 customers on satisfaction.
  • Training: Workshops for NTC staff on empathy and problem-solving.
  • Post-test: Repeat survey after 6 months.
• Ethical Challenge: Avoiding coercion (customers may feel pressured to give positive feedback).
• Result: 30% increase in resolved complaints within 24 hours.

MERMAID DIAGRAM:

mindmap
  root((Ethical Considerations in Mixed-Methods Research))
    NTC
      Problem: Low complaint resolution
      Design: Evaluative (Pre/Post)
      Output: Staff training
      Ethical Challenge: Avoiding coercion
      Result: 30% increase in resolved complaints
    eSewa
      Problem: User trust
      Design: Mixed-methods evaluative
      Output: SMS tutorials
      Ethical Challenge: Informed consent
    Daraz
      Problem: Delivery delays
      Design: Descriptive + exploratory
      Output: Dynamic routing
      Ethical Challenge: Data privacy

Exam Tip

1. Define clearly: Always start with precise definitions (e.g., "Research design is the framework that structures the research process, ensuring validity and reliability by specifying data collection and analysis methods.").
2. Link theory to practice: Examiners love Nepali examples. For instance:
   • "Like eSewa’s trust study, social work research must balance quantitative rigor (survey data) with qualitative depth (user narratives) to address complex issues like digital exclusion."
3. Compare designs: Use tables to contrast experimental vs. descriptive designs, highlighting strengths/weaknesses for social work.
4. Ethics is non-negotiable: Always discuss at least two ethical challenges in your answer (e.g., informed consent + confidentiality).
5. Proposal structure: For proposal questions, use the 6-step template (title, background, objectives, methodology, timeline, budget) and justify every choice (e.g., "A mixed-methods design was chosen because quantitative data would reveal trends, while qualitative data would explain why those trends exist.").
6. Avoid jargon: Replace "paradigm" with "approach" and "triangulation" with "using multiple methods to verify findings."
7. Diagrams save marks: Draw a simple flowchart (e.g., "Research Question → Design Selection → Data Collection → Analysis → Reporting") to visually structure your answer.

Final Visual Summary: