Literature Review: Sources, Analysis & Structuring for Research

TAKEAWAYS:

• A literature review is a critical synthesis of prior research that identifies gaps, not just a summary of sources.
• Primary vs. secondary sources differ in credibility and purpose: primary sources (e.g., original studies) are gold standards, while secondary sources (e.g., review articles) offer interpretations.
• Thematic analysis groups literature by trends, contradictions, or methodological approaches to reveal research patterns.
• Structuring a review follows a logical flow: from broad context to specific gaps, culminating in the research problem.
• Ethical considerations include avoiding plagiarism, citing fairly, and disclosing conflicts of interest—critical for academic integrity.
• Real-world tools like Google Scholar, ResearchGate, and university libraries streamline source discovery, but critical evaluation remains manual.

1. What Is a Literature Review?

A literature review is the foundation of any research project. It answers:

• What is already known about the topic?
• What gaps or contradictions exist in current research?
• How does your research fill those gaps?

Unlike a book review or a simple summary, a literature review is analytical and argumentative. It doesn’t just list sources; it evaluates, compares, and synthesizes them to justify the need for new research.

Key Characteristics

2. Types of Literature Reviews

Literature reviews vary by scope, purpose, and methodology. The two primary types are:

3. How to Conduct a Literature Review: Step-by-Step

Step 1: Define the Scope

Before searching, clarify:

• Research question: What problem are you addressing?
• Keywords: Use Boolean operators (AND, OR, NOT) for precise searches.
• Timeframe: How recent must sources be? (e.g., last 5–10 years for tech topics).

Worked Example: Daraz’s Delivery Optimization Problem: Daraz (Nepal’s largest e-commerce platform) wants to reduce delivery delays. Literature Review Scope:

• Keywords: "last-mile delivery optimization," "logistics in developing countries," "AI route planning."
• Sources: Studies on Pathao’s delivery algorithms, NTC’s traffic data, and Daraz’s internal reports.
• Gap: Most studies focus on urban areas; Kathmandu’s traffic congestion is understudied.

Step 2: Search for Sources

Use reliable databases:

• Academic: Google Scholar, IEEE Xplore, ScienceDirect, PubMed.
• Industry: ResearchGate, arXiv (for tech), NEPSE reports (for finance).
• Local: Tribhuvan University Central Library, Nepal Journals Online.

Step 3: Evaluate Sources

Not all sources are equal. Use the CRAAP test to assess credibility:

• Currency: Is the information up-to-date? (e.g., a 2010 study on mobile banking in Nepal may be outdated.)
• Relevance: Does it directly address your research question?
• Authority: Who published it? (Peer-reviewed journals > blogs > news articles.)
• Accuracy: Are claims supported by data?
• Purpose: Is it objective or biased? (e.g., a bank’s report on loan interest may favor their policies.)

Step 4: Organize and Analyze

Group sources by:

1. Themes: Trends, contradictions, or methodological approaches.
2. Chronology: How ideas evolved over time.
3. Methodology: Qualitative vs. quantitative studies.

Example: Kathmandu Traffic Study

• Theme 1: "Impact of traffic lights on congestion" (supported by NTC data).
• Theme 2: "Ride-sharing apps (Pathao, Yeti) reducing private vehicle use" (contradictory studies).
• Gap: No study combines NTC traffic data with app usage patterns.

Step 5: Synthesize and Write

Structure your review logically:

1. Introduction: Define the topic and its importance.
2. Body: Organize by themes, starting broad → specific.
   • Use signposting (e.g., "While Study X argues..., Study Y contradicts this by...").
3. Conclusion: Summarize gaps and justify your research.

4. Common Pitfalls and How to Avoid Them

5. Tools and Techniques for Efficient Reviewing

In the Real World

Literature reviews are everywhere—even in apps and companies you use daily. Here’s how:

1. eSewa and Khalti (Digital Payments)

   • Idea Used: Systematic review of fraud detection methods.
   • How: Before launching their fraud alert systems, eSewa and Khalti conducted literature reviews on:
     • Machine learning models for transaction anomaly detection (e.g., studies from banks like SBI or HSBC).
     • Behavioral biometrics (e.g., typing patterns, mouse movements) to identify fake logins.
   • Real Example: Khalti’s 2022 whitepaper on "Reducing Payment Fraud in Nepal" cites 15+ studies on AI in fintech, including a meta-analysis of global fraud detection tools.

2. Pathao (Ride-Hailing)

   • Idea Used: Scoping review of dynamic pricing algorithms.
   • How: Pathao’s surge pricing model was informed by:
     • Literature on Uber/Lyft’s pricing strategies (e.g., studies from MIT’s Transportation Lab).
     • Local data on Kathmandu’s traffic patterns (NTC reports) to adjust prices during festivals (Dashain, Tihar).
   • Real Example: Their 2021 blog post ["How We Optimized Delivery Routes in Nepal"] references a scoping review of "last-mile logistics in developing economies."

3. NEPSE (Nepal Stock Exchange)

   • Idea Used: Meta-analysis of stock market volatility predictors.
   • How: NEPSE’s risk assessment models rely on:
     • Systematic reviews of global financial crises (e.g., 2008 crash studies from Harvard Business Review).
     • Local data on Nepal’s GDP growth trends (published by the World Bank or Nepal Rastra Bank).
   • Real Example: NEPSE’s 2023 annual report includes a section titled "Literature on Emerging Market Volatility," citing 8 peer-reviewed studies.

4. NTC (Nepal Telecom)

   • Idea Used: Narrative review of 5G deployment challenges.
   • How: Before rolling out 5G, NTC reviewed:
     • Case studies from India (Jio’s 5G rollout) and South Korea (Samsung’s infrastructure).
     • Local challenges: terrain, rural connectivity gaps (studies from ITU or Nepal’s Ministry of Communications).
   • Real Example: NTC’s 2022 proposal for 5G trials in Kathmandu cites a narrative review of "5G in mountainous regions."

Worked Example: Bank Loan Interest Study

Scenario: A Nepalese bank wants to adjust loan interest rates based on inflation and GDP growth.

Literature Review Process:

1. Search: Keywords = "inflation loan interest correlation," "GDP growth and SME lending," "Nepal Rastra Bank policies."
2. Sources:
   • Primary: Nepal Rastra Bank’s annual reports (2018–2023).
   • Secondary: World Bank studies on "Interest rate policies in South Asia."
3. Analysis:
   • Theme 1: "High inflation → higher loan defaults" (supported by NRB data).
   • Theme 2: "SMEs benefit from lower rates during GDP growth" (contradicted by a 2020 study showing delayed repayments).
4. Gap: No study combines NRB data with real-time NEPSE market trends.
5. Conclusion: The bank’s new model should integrate NEPSE’s stock volatility data with NRB reports.

Visual:

flowchart TD
  A["Bank's Problem: Adjust Loan Rates"] --> B["Literature Search"]
  B --> C["Source 1: NRB Reports\n(Primary)"]
  B --> D["Source 2: World Bank Studies\n(Secondary)"]
  C --> E["Theme: Inflation → Defaults"]
  D --> F["Theme: SMEs & GDP Growth"]
  E & F --> G["Gap: Missing NEPSE Data"]
  G --> H["Proposed Solution: Hybrid Model"]

Exam Tip

How This Unit Is Tested

1. Definition Questions (5–10 marks):

   • "Define literature review and distinguish it from a book review."
   • Answer Tip: Use the mindmap above—highlight that it’s analytical, not descriptive.

2. Role and Organization (15–20 marks):

   • "Explain the role of literature review in research. How would you organize a review on ‘AI in Nepalese healthcare’?"
   • Answer Tip:
     • Role: Identifies gaps, builds theory, justifies research.
     • Organization: Use a thematic structure:
       1. Current AI tools in Nepal (e.g., eSewa’s chatbots).
       2. Challenges (e.g., data privacy laws).
       3. Global trends (e.g., WHO guidelines on AI in medicine).
       4. Gap: No study on "AI + traditional medicine in Nepal."

3. Worked Examples (10–15 marks):

   • "A Daraz manager wants to reduce delivery delays. Conduct a mini literature review."
   • Answer Tip:
     • Sources: Pathao’s route optimization studies, NTC traffic data.
     • Gap: No integration of real-time traffic (NTC) with app data (Daraz/Pathao).
     • Solution: Propose a hybrid model (like Google Maps + Pathao’s algorithms).

4. Pitfalls and Ethics (5–10 marks):

   • "What are two common mistakes in literature reviews? How would you avoid them?"
   • Answer Tip:
     • Mistake 1: Descriptive summaries.
       • Fix: Use comparison tables (like the one on review types).
     • Mistake 2: Plagiarism.
       • Fix: Cite using APA format (e.g., (Author, Year, p. X)).

Marks Distribution in Exams

Final Checklist for Full Marks

Before submitting:

1. Did I define terms clearly? (Use bold for key phrases.)
2. Did I use real-world examples? (eSewa, Daraz, NTC, NEPSE.)
3. Did I include visuals? (At least 3: flowchart, table, and one real-world diagram.)
4. Did I structure my answer logically? (Intro → Body → Conclusion.)
5. Did I avoid plagiarism? (Paraphrased all ideas and cited sources.)