Literature Review — Search & Protocol
Session at a Glance
Purpose & types of literature review; SLR protocol; academic databases; Boolean search; citation chaining; reference management
SLR scoping exercise; database search execution; search string refinement; Zotero/Mendeley setup
2 hrs Lecture + 12 hrs Lab/Project
—
Learning Objectives
By the end of this session, you will be able to:
- Distinguish between narrative, systematic, scoping, and meta-analytic literature reviews and select the appropriate type for your capstone
- Execute a systematic literature search using Boolean operators, controlled vocabulary, and database-specific syntax across discipline-appropriate databases
- Construct a PRISMA-style search protocol with inclusion/exclusion criteria, quality assessment, and data extraction plan
- Apply citation chaining (forward and backward) to expand and validate your search results
- Set up a reference management workflow using Zotero, Mendeley, or EndNote
Session Planner
Suggested breakdown of the 4-hour contact session.
| Time | Segment | Activity | Mode |
|---|---|---|---|
| 0:00–0:08 | Opening | Recap Week 3 problem statements; why lit review is the foundation of your capstone | Whole class |
| 0:08–0:28 | Lecture 1 | Purpose of lit review; types (narrative, systematic, scoping, meta-analysis); the SLR protocol step-by-step | Lecture |
| 0:28–0:50 | Lecture 2 | Academic databases (BBA + BCA); Boolean search construction; citation chaining; reference management tools | Lecture + Demo |
| 0:50–1:10 | Activity | Search string construction exercise; Boolean operator practice | Pairs |
| 1:10–1:25 | Discussion | Share search strings; troubleshoot common construction errors | Whole class |
| 1:25–1:40 | Break | — | — |
| 1:40–2:00 | Lab Briefing | SLR scoping exercise walkthrough; Zotero/Mendeley setup demo; database access check | Demo |
| 2:00–3:30 | Lab Work | Individual SLR scoping: define search string, run on 2+ databases, document hits, apply inclusion/exclusion | Individual |
| 3:30–3:50 | Discussion | Share search results; compare database coverage; discuss refinement strategies | Whole class |
| 3:50–4:00 | Exit Ticket | Submit preliminary search protocol; self-assess search quality | Individual |
1. Why Literature Review Matters
A literature review is a systematic, critical synthesis of published research on a topic. It is NOT a summary of papers one by one. It IS a coherent argument that maps the intellectual landscape, identifies what we know, exposes what we don't know, and positions your study within that landscape.
1.1 Five Purposes of a Literature Review
Understand what research exists, who the key authors are, which journals publish on this topic, and how the field has evolved over time.
Find what's missing: a relationship not yet tested, a population not yet studied, a method not yet applied, findings that conflict and need resolution.
Ensure your capstone hasn't already been done. Nothing is more embarrassing than discovering in Week 20 that someone published your exact study three years ago.
Situate your RQ within the existing conversation. "Prior research has established X and Y, but we still don't know Z — this study addresses Z."
Identify which theories explain your phenomenon. Your conceptual framework (Week 3) should emerge from your literature review, not precede it.
A well-written literature review chapter has a funnel structure: start broad (the domain and its significance), progressively narrow (specific streams of relevant research), and funnel down to the precise gap your study fills. The final paragraph of your lit review should make the reader think: "Yes — this study needs to be done."
2. Types of Literature Review
| Type | Purpose | Method | When to Use | Typical Size |
|---|---|---|---|---|
| Narrative / Traditional | Provide a broad overview and critical synthesis of a topic | Flexible, author-driven selection and interpretation of literature | When you need to establish context broadly; common in dissertation Chapter 2 | 30–80 papers |
| Systematic Literature Review (SLR) | Rigorously identify, evaluate, and synthesize ALL available evidence on a specific question | Protocol-driven: predefined search strategy, inclusion/exclusion, quality assessment, data extraction, synthesis | When the question is narrow and you need comprehensive, reproducible coverage; recommended for capstones | 15–50 papers after filtering |
| Scoping Review | Map the breadth and boundaries of a field; identify what evidence exists | Systematic search but broader question; doesn't assess quality in depth | When the field is broad or emerging; often a precursor to an SLR; good for initial capstone exploration | 50–200+ papers mapped |
| Meta-Analysis | Statistically combine results from multiple quantitative studies to produce a pooled effect size | Statistical synthesis of effect sizes from homogenous studies | When multiple studies have tested the same relationship and you want a precise estimate; rare at UG level | 10–60 studies (statistically combined) |
For most undergraduate capstones, a hybrid approach works best: an SLR for the core of your topic (ensuring you don't miss key papers) nested within a broader narrative review (providing context). Pure SLRs are excellent but time-intensive; pure narrative reviews risk missing important literature. Your supervisor can help calibrate the right balance.
3. The Systematic Literature Review Protocol
An SLR follows a predefined, transparent protocol. The protocol is your research plan for the literature — it ensures rigour, reproducibility, and completeness. The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework provides the standard structure.
3.1 The Six-Step SLR Process
3.2 Key Protocol Elements — Explained
Inclusion Criteria — Rules that a paper must satisfy to enter your review. Be specific and justified:
Example: "Peer-reviewed journal articles and conference papers published 2015–2025 in English; empirical studies (quantitative, qualitative, or mixed methods) that examine influencer marketing effectiveness on consumer purchase behaviour; studies involving social media platforms (Instagram, YouTube, TikTok)."
Exclusion Criteria — Rules that disqualify a paper. These prevent irrelevant material from cluttering your review:
Example: "Editorials, opinion pieces, book reviews; studies focused exclusively on traditional celebrity endorsements (not social media influencers); studies where the dependent variable is not purchase-related; duplicate publications (keep the most complete version)."
Quality Assessment — A structured evaluation of each included study's methodological rigour. For quantitative studies: sampling adequacy, measurement validity, appropriate analysis. For qualitative studies: credibility, transferability, dependability, confirmability. Low-quality studies may be excluded or flagged with caveats.
Data Extraction Form — A standardized template for recording information from each included paper: author(s), year, country/context, research design, sample size, key variables, main findings, limitations. This transforms reading into structured data that you can synthesize.
A standard element of any SLR is the PRISMA flow diagram — a visual record of how many papers were found, screened, excluded (with reasons), and included at each stage. It makes your search process transparent and auditable. Records identified → Duplicates removed → Records screened → Full-text assessed → Studies included in synthesis.
4. Academic Databases & Search Techniques
4.1 Database Selection — BBA vs. BCA
| Database | Discipline | Coverage | Best For | Special Features |
|---|---|---|---|---|
| Scopus | Both | 90M+ records, 25K+ journals | Broad multidisciplinary search; citation analysis; author profiles | Citation counts, h-index, journal metrics (CiteScore, SJR) |
| Web of Science | Both | 170M+ records, 21K+ journals | High-impact literature; backward/forward citation tracking | Impact Factor, citation reports, cited reference search |
| Google Scholar | Both | Largest coverage; includes grey literature | Initial exploration; cited-by searches; finding obscure references | Free; "Cited by" chain; alerts; BUT no quality filter, limited advanced search |
| IEEE Xplore | BCA | 5M+ records | CS, EE, software engineering, AI/ML, networks, security | Premier CS/engineering venue; conference + journal papers |
| ACM Digital Library | BCA | Full ACM publication history | Computing research — HCI, software engineering, algorithms, systems | ACM Computing Classification System; SIG newsletters |
| arXiv | BCA | 2M+ preprints | Cutting-edge CS, ML, AI research (pre-publication) | Free; rapid dissemination; BUT not peer-reviewed — use with caution |
| DBLP | BCA | 6M+ CS publications | Author-centric CS bibliography; finding an author's complete work | Free; clean author disambiguation; conference tracking |
| EBSCO / ProQuest | BBA | Business, management, economics, social sciences | BBA dissertations; business news; industry reports | Company profiles, industry reports, market research |
| JSTOR / SSRN | BBA | Social sciences, business, economics | Classic papers; working papers; pre-publication business research | SSRN: early access to research; JSTOR: historical depth |
4.2 Boolean Search — The Mechanics
Boolean operators are the grammar of database searching. They allow you to combine search terms with precision. Most academic databases support Boolean logic.
| Operator | Function | Example | Result |
|---|---|---|---|
| AND | Narrows — both terms must appear | influencer AND "purchase intention" | Papers containing BOTH terms |
| OR | Broadens — either term can appear | influencer OR "content creator" OR "KOL" | Papers containing ANY of these terms |
| NOT | Excludes — removes a term | influencer NOT celebrity | Papers about influencers but NOT about celebrities (use cautiously — may exclude relevant papers) |
| Quotes "" | Exact phrase matching | "social media influencer" | Only papers with this exact phrase — not "social" and "media" and "influencer" separately |
| Parentheses () | Grouping — controls logic order | (influencer OR "content creator") AND (Instagram OR TikTok) | Papers about influencers or content creators on Instagram or TikTok |
| Asterisk * | Truncation/wildcard — matches word variants | influenc* | influencer, influencers, influencing, influence — any word starting with "influenc" |
4.3 Building a Search String — A Worked Example
Start with your RQ: "How does influencer tier (micro vs. macro) affect purchase conversion among Gen Z consumers of Indian D2C brands on Instagram?"
Step 1: Extract key concepts
Concept 1: Influencer (micro, macro, tier, type)
Concept 2: Purchase (conversion, intention, behaviour)
Concept 3: Social media platform (Instagram, social media)
Concept 4: Consumer (Gen Z, young, demographic)
Step 2: Generate synonyms for each concept
C1: "social media influencer" OR "digital influencer" OR "micro-influencer" OR "macro-influencer" OR "content creator"
C2: purchas* OR "purchase intention" OR "buying behaviour" OR "conversion rate"
C3: Instagram OR "social media" OR "social networking"
C4: "Gen Z" OR "young consumer*" OR "youth" OR "millennial"
Step 3: Combine with Boolean operators
("social media influencer" OR "digital influencer" OR "micro-influencer" OR "macro-influencer" OR "content creator") AND (purchas* OR "purchase intention" OR "buying behaviour" OR "conversion rate") AND (Instagram OR "social media") AND ("Gen Z" OR "young consumer*" OR "youth")
4.4 Citation Chaining
Citation chaining is a powerful technique for finding papers that database searches might miss. There are two directions:
- Backward chaining (snowballing): Look at the reference list of a key paper. Which older papers did it cite that are foundational to your topic? This moves you backward in time to the roots of the literature.
- Forward chaining: Use Google Scholar, Scopus, or Web of Science to find papers that have cited a key paper. Who built on this work? This moves you forward in time to the cutting edge.
Pro tip: Find 3–5 "seed papers" — the most relevant, highly cited papers in your area. Backward-chain to find the classics they cite. Forward-chain to find the latest studies that cite them. This alone can capture 70%+ of the relevant literature.
5. Reference Management — Your Second Brain
A reference manager is not optional for a year-long capstone. It saves you from the nightmare of manually formatting 60+ references and the panic of "I know I read this somewhere but I can't find it."
| Tool | Cost | Best For | Strengths | Limitations |
|---|---|---|---|---|
| Zotero | Free (open source) | Both BBA & BCA | Browser connector (one-click save); automatic metadata extraction; groups for sharing; 300MB free storage; Word/Google Docs plugin; best PDF handling | Storage fills up with large PDFs; sync can be slow |
| Mendeley | Free (2GB) | Both BBA & BCA | PDF annotation; social network for researchers; citation plugin for Word; article recommendations based on your library | Elsevier-owned (privacy concerns for some); less flexible than Zotero |
| EndNote | Paid (institutional access often available) | BBA (APA heavy users) | Most comprehensive citation style support; powerful deduplication; advanced search within library | Expensive; steep learning curve; less intuitive interface |
| BibTeX (with Zotero) | Free | BCA (LaTeX users) | Native LaTeX integration; Zotero can export .bib files; version-controllable (plain text) | Requires LaTeX; manual editing for complex entries |
Zotero is the recommended tool for this course — it's free, open-source, works for both BBA (Word/Google Docs + APA) and BCA (LaTeX/BibTeX + IEEE), and its browser connector makes saving papers effortless. Install it this week, create a folder for your capstone, and save EVERY paper you read. Future you will thank present you.
6. From Reading to Writing — Synthesis Not Summary
The #1 mistake in literature reviews is writing an annotated bibliography instead of a synthesis. An annotated bibliography says: "Smith (2020) found X. Jones (2021) found Y. Patel (2022) found Z." A synthesis says: "Studies consistently show X (Smith 2020; Jones 2021; Patel 2022), though they disagree about the mechanism — Smith argues A, while Jones and Patel find evidence for B."
6.1 Summary vs. Synthesis — A Side-by-Side Comparison
| Summary (Wrong — Annotated Bibliography) | Synthesis (Right — Literature Review) |
|---|---|
| "Smith (2020) conducted a survey of 200 Instagram users and found that 68% had made a purchase based on an influencer recommendation. Jones (2021) studied YouTube influencers and reported that authenticity was the strongest predictor of follower trust. Patel (2022) examined beauty influencers on Instagram and found that micro-influencers generated higher engagement rates than macro-influencers." | "Research consistently demonstrates that social media influencers affect consumer purchase behaviour, with studies reporting that 60–70% of users have made influencer-inspired purchases across platforms (Smith 2020; Gupta 2021). However, the mechanism of influence remains contested. One stream of research emphasizes source characteristics — authenticity (Jones 2021), expertise (Kumar 2020), and similarity (Chen 2021) — while a second emerging stream focuses on platform affordances and content format. Most notably, studies converge on the finding that micro-influencers (10k–100k followers) generate 40–60% higher engagement than macro-influencers (Patel 2022; Singh 2023), yet these studies exclusively examine Western markets — leaving a significant gap regarding Indian consumer contexts." |
6.2 The Funnel Structure
Establish the domain's significance. Define key terms. Orient the reader. "Social media influencer marketing has grown from a fringe tactic to a $21 billion industry..."
Organize by themes (not by author or chronology). Synthesize within each theme. Show agreements, disagreements, and trends. "Three streams of literature inform this study: (a) source credibility in digital contexts, (b) the micro-influencer advantage, and (c) Gen Z consumer behaviour..."
Identify the precise gap. State it explicitly. Transition to your study. "While prior research establishes the micro-influencer advantage in Western markets, no study has examined... This gap is particularly significant because... This study addresses this gap by..."
6.3 Common Pitfalls
Paper-by-paper summary. No thematic grouping. No comparison across studies. The reader gets a list, not an argument.
Omitting seminal or highly relevant studies because the search wasn't systematic. An examiner who knows the field will notice immediately.
Relying on textbooks and papers from 15+ years ago. A literature review should be current — aim for 70%+ of references from the last 5–7 years.
Citing papers you haven't actually read — or worse, citing based on another paper's summary. Always read the original (or at minimum the abstract and key sections).
Treating every published paper as equally valid. Good literature reviews critique methodology, note limitations, and weigh conflicting evidence — not just report it.
Claiming a gap exists without showing evidence. The gap must be visible in the literature you've reviewed — the reader should see it, not just be told it's there.
Think Deeper — Cross Questions
Discuss in pairs before sharing with the class.
Your database search returns 2,347 papers. Your SLR protocol says you'll review everything. Realistically, what should you do? What does this tell you about the importance of a well-constructed search string before you start searching?
A BCA student plans to use only Google Scholar for her literature search on "federated learning privacy." What specific papers, venues, or types of content might she miss compared to also searching IEEE Xplore and arXiv? Is Google Scholar sufficient for a capstone?
You find a perfect paper — it addresses your exact RQ, uses your exact method, and was published in 2024 at a top journal. Is this good news or bad news for your capstone? What does this mean for the "N" (Novelty) in FINER?
Look at the "Summary vs. Synthesis" comparison in Section 6.1 again. The synthesis paragraph says things the summary paragraph cannot — what are they? Identify three specific functions the synthesis performs that the summary does not.
Quick Check — Search String Construction
For each scenario, select the best Boolean search string.
1. You want papers about "blockchain" in the context of "supply chain management." You don't care about cryptocurrency. Which search string is best?
2. You're studying sentiment analysis for Hindi-English code-mixed text. You want to capture papers that use any variant of "sentiment analysis" and any variant of "code-mixed" or "code-switched," focused on Hindi and English.
3. You want to find papers about "employee retention" or "turnover intention" specifically in the "IT sector" or "technology industry."
Knowledge Check — Interactive Quiz
Test your understanding of literature review concepts.
Q1. Which of the following BEST describes the difference between a systematic literature review (SLR) and a narrative literature review?
Q2. A BCA student is researching "lightweight encryption algorithms for IoT devices." Which database combination provides the MOST comprehensive coverage for this topic?
Q3. What is the primary purpose of backward citation chaining (snowballing)?
Q4. In the PRISMA flow diagram, 500 records are identified from databases. After removing 120 duplicates, 380 records remain. After screening titles and abstracts, 290 are excluded. Of the 90 full-text articles assessed, 45 are excluded for specific reasons. How many studies are included in the final synthesis?
Q5. Which of the following is an example of literature SYNTHESIS (as opposed to summary)?
Lab Activity — SLR Scoping Exercise
Part A: Define Your Search Protocol (30 min)
Complete the following for YOUR capstone topic. This will become part of your methodology chapter.
- Review RQ(s): State your research question(s) that guide the literature search.
- Key concepts: Extract the 2–4 main concepts from your RQ.
- Search string: Build a Boolean search string with synonyms for each concept, using OR within concepts and AND between concepts. Use quotes for phrases and * for truncation.
- Databases: List 3–4 databases you'll search. Justify each choice based on your discipline.
- Inclusion criteria: Define 4–5 criteria (publication type, date range, language, methodology, relevance to RQ).
- Exclusion criteria: Define 4–5 criteria (what specifically will you reject and why?).
- Quality assessment: How will you evaluate the quality of included studies? List 3–5 questions you'll ask of each paper.
Part B: Execute and Document (90 min)
- Run your search on at least 2 databases. Record the number of hits for each.
- Export results (titles + abstracts) to Zotero/Mendeley. Remove duplicates.
- Screen titles and abstracts: Apply your inclusion/exclusion criteria. Mark papers as "include," "exclude," or "maybe." Record reasons for exclusion.
- Document the process: Start your PRISMA flow diagram with the numbers at each stage.
- Identify 5–10 key papers that appear most relevant. These will be your starting point for critical reading in Week 5.
Part C: Reference Manager Setup
- Install Zotero (zotero.org) + browser connector.
- Create a folder for your capstone project.
- Import the papers from Part B into Zotero.
- Install the Zotero plugin for Word or Google Docs.
- Practice: insert 3 citations and generate a reference list in your discipline's style (APA 7th for BBA, IEEE for BCA).
Exit Ticket
Submit with your search protocol draft.
- Paste your Boolean search string (the one you used in your database search).
- How many hits did each database return? Was the volume manageable? If not, how will you refine?
- Name 2 databases you searched and explain why each is appropriate for your topic.
- How many papers remain after duplicate removal and title/abstract screening?
- One challenge you encountered during the SLR scoping exercise:
Key Takeaways — Week 4
A literature review is a thematic synthesis that builds an argument toward a gap. It is NOT an annotated bibliography. Organize by themes, not by authors.
A predefined protocol (search string, inclusion/exclusion criteria, quality assessment, PRISMA diagram) makes your review reproducible and defensible in your viva voce.
Google Scholar alone is insufficient. Search at least 3–4 databases spanning your discipline. BBA: Scopus, EBSCO, ProQuest. BCA: IEEE Xplore, ACM DL, Scopus, arXiv.
Set up your reference manager NOW. Save every paper you read. The 60+ references in your final dissertation need to be formatted perfectly — Zotero makes this a button press, not a weekend of pain.
Facilitator Notes
Preparation Checklist
- Verify institutional database access before the session — students will hit paywalls if not on campus network or VPN. Prepare proxy/remote access instructions.
- Prepare a live demo of search string construction in Scopus or IEEE Xplore — show hits with a poor string vs. a refined string.
- Have Zotero installation instructions ready. If possible, pre-install on lab machines.
- Prepare a sample PRISMA flow diagram as a visual reference (or show one from a published SLR).
- For BCA cohorts: emphasize arXiv's non-peer-reviewed status and when it's appropriate to cite preprints.
- Share a completed Data Extraction Form template (Excel/Google Sheets) for students to adapt.
Common Student Difficulties
- "I searched and found nothing": The search string is too narrow or uses the wrong terminology. Teach them to start broad and narrow down — it's easier to filter out irrelevant papers than to find papers that were never captured.
- "I found 5,000 papers": The search string is too broad or missing AND operators. Add more specific terms, limit to title/abstract/keyword fields, or add date/language filters.
- Confusing Google Scholar with a database: Google Scholar is a search engine, not a curated database. It has no quality control, includes non-academic sources, and its advanced search is limited. Frame it as a supplementary tool, not a primary one.
- Not recording search details: Students run searches, find papers, and can't remember which database or string produced them. The search protocol document forces systematic record-keeping.
- Analysis paralysis — reading instead of screening: At the screening stage, students start reading full papers. Teach the discipline: title + abstract only for screening. Full-text comes later.
Pacing Tips
- The search string construction demo is the highest-value 10 minutes of the lecture — don't rush it. Use a student's own topic (volunteer) to make it real.
- The SLR scoping exercise is lab-intensive. If lab time is insufficient, have students complete the database searches during the week and bring results to Week 5.
- Zotero setup should be done in the first 15 minutes of lab — if students leave without it working, most won't set it up independently.
- For slow database access (common on shared institutional connections): have students work in pairs, one searching while the other documents.