What is a literature review – purpose, common types, and what examiners expect
Struggling to turn a pile of papers into a clear, defensible argument? A literature review does more than summarize studies: it evaluates evidence, synthesizes patterns and gaps, and shows why your research matters. This how-to guide walks you through planning, searching, appraising, synthesizing and writing a review that examiners or journal reviewers can follow and trust.
A literature review surveys, critiques and integrates existing research on a focused question so you can justify, position and design your study. It typically follows the introduction and precedes methods in articles and dissertations. Examiners expect an explicit scope, transparent selection and critical appraisal that leads to a clear research gap or hypothesis.
- Narrative (traditional) – broad context and theory development; use when interpretation and conceptual framing matter more than exhaustive coverage.
- Systematic review – structured, reproducible search and appraisal for precise questions; use in clinical, policy or tightly defined effect questions.
- Scoping review – maps the extent and nature of literature for emerging or heterogeneous topics.
- Integrative review – combines theoretical and empirical work across methods to build new frameworks.
- Meta-analysis – statistical synthesis of quantitative results when many comparable studies exist; usually a component of a systematic review.
Key distinctions: an annotated bibliography lists items one-by-one; a literature review synthesizes across studies. A systematic review documents a protocol and emphasizes reproducibility; a meta-analysis pools effect sizes to estimate average effects.
Plan first: define your research question, scope and protocol
Start by turning a broad topic into a sharply focused review question. Use PICO for clinical topics (Population, Intervention, Comparison, Outcome) or PEO for observational questions (Population, Exposure, Outcome). For theory or methods, create keyword clusters (concept A + mechanism B + context C). A focused question guides search terms, inclusion rules and extraction fields.
- Example: Broad “social media and mental health” → Focused “How does daily social media use affect anxiety symptoms in undergraduate students?” (P: undergraduates; E: daily use; O: anxiety symptoms).
- Theory example: “mechanisms linking remote work to productivity” → Keywords: remote work OR telecommuting AND productivity AND mediators (autonomy OR distractions).
Set inclusion/exclusion criteria up front-study designs, date range, languages, populations and outcome measures-and tailor them to your discipline. Decide the review type before searching: if you need reproducibility and low bias, prepare a systematic review protocol; for concept-building choose narrative or integrative approaches and document your method in a clear methods paragraph. Keep practical records: a search log (databases, dates, search strings), a selection log (reasons for exclusion) and a PRISMA-style flow summary when applicable. Schedule realistic time for scoping, screening, extraction and iterative writing.
Search efficiently: databases, search strings and grey literature
Use multiple, discipline-appropriate databases and add Google Scholar and targeted repositories for grey literature. Don’t rely on a single source-each database indexes different journals and conference proceedings. Build search strings from concept clusters: list keywords and synonyms, combine with Boolean operators (AND, OR, NOT), use truncation and phrase quotes, and pilot strings until precision and recall are acceptable.
- Database choices by discipline: PubMed/Medline for biomedical topics, PsycINFO for psychology, JSTOR and Web of Science for humanities and social sciences, EconLit for economics, IEEE Xplore for engineering. Add institutional repositories and conference proceedings for recent or unpublished work.
- Sample search strings (adapt to your syntax and fields):
- “(social media OR Facebook OR Instagram) AND (anxiety OR anxious OR ‘mental health’) AND (undergraduate OR ‘college students’)”
- “(remote work OR telecommut*) AND (productiv* OR performance) AND (mediator OR mechanism OR autonomy)”
- “(diabetes) AND (self-management OR ‘self care’) AND (randomized OR RCT) AND (‘quality of life’ OR ‘health outcomes’)”
Use snowballing and citation chasing-backward (reference lists) and forward (citations to a paper)-and set alerts for key authors and terms to capture ongoing work. Manage results with a reference manager and a screening spreadsheet; deduplicate early and track inclusion decisions and reasons to make selection defensible.
Evaluate and extract: critical appraisal and data extraction
Before deep extraction, screen studies for basic quality and relevance. A short appraisal checklist speeds triage and helps you weight evidence during synthesis:
- Authority: peer-reviewed journal or reputable institution?
- Methodology: is the design appropriate and clearly described?
- Sample and power: adequate for the claims?
- Bias and conflicts: funding, conflicts of interest disclosed?
- Relevance: does it directly inform your question?
Record methodological red flags (unclear randomization, high attrition, absent control) and note limitations for synthesis. Design a structured extraction table with fields such as citation, design, sample, measures, key findings, effect sizes (if any), limitations and theme tags. For speed, keep a one-line summary, one-line critique and 1-2 tags per paper-these synthesis-friendly notes reduce rewriting later.
Decision framework: choose the right review type and structure
Match your aim, the shape of the evidence and your audience when choosing structure. Common organization options are thematic (by topic), chronological (development over time), methodological (compare methods), and theoretical (competing frameworks). Heterogeneous evidence often suits thematic or scoping structures; comparable trials lend themselves to systematic review and meta-analysis.
- Step 1: Define the primary aim-context/theory building or estimating an effect?
- Step 2: Assess evidence shape-are studies comparable in design, measures and outcomes?
- Step 3: Consider the audience-specialists expect detailed methods; thesis examiners want comprehensive critique.
- Step 4: Combine aim + evidence + audience to select structure (for example, mechanisms + varied methods → thematic with methodological subsections).
Practical signals: choose a systematic review when many comparable studies exist and you can define clear eligibility and outcomes; prefer narrative or integrative approaches when methods and outcomes vary or when developing conceptual frameworks. When transparency and reproducibility matter, register a protocol and report your planned eligibility, search and appraisal methods to reduce selective reporting.
Organize and synthesize: structures, synthesis techniques and writing templates
Adopt a clear overall structure: introduction (scope and question), body (organized by chosen scheme), and conclusion (synthesis, gaps, and how your study fits). Within body sections, use a consistent paragraph pattern so readers can follow comparisons and weight of evidence.
- Paragraph template: topic sentence → synthesis of 2-4 studies → methodological note → critical evaluation → transition.
- Synthesis techniques: thematic matrices, concept maps, vote-counting (with caution), narrative synthesis and integrating conflicting findings by explaining methodological or contextual differences.
- How to present gaps and tensions: state what evidence shows, where it is thin or inconsistent, and specific next steps (types of study, methods or contexts needed).
Two short examples applying the paragraph template: Social media & anxiety: several cross-sectional surveys report a positive association between platform time and self-reported anxiety, while experience-sampling studies show the link weakens after controlling for baseline social comparison; the predominance of cross-sectional designs limits causal claims and longitudinal evidence is sparse. Intervention effectiveness: small RCTs report brief telecoaching reduced symptom scores by 0.3-0.5 SD compared with control, but outcome heterogeneity and short follow-up constrain generalizability and identified biases should temper clinical recommendations.
Weave citations smoothly with comparative phrasing (“X found…, Y suggests…”), use past tense for study findings and present tense for your synthesis, and prefer paraphrase over quotation. Use matrices or concept maps to convert extraction tables into a readable, evidence-weighted narrative that highlights consensus, divergence and gaps.
Revise, common mistakes to avoid and final checklist before submission
Common problems are predictable and fixable. Use deliberate revision steps to improve balance and rigor before submission.
- Oversummarizing: Fix by grouping related findings into synthesis sentences that reveal patterns rather than listing studies.
- Poor or drifting scope: Revisit inclusion criteria and remove off-topic material; test your scope with a focused pilot search.
- Cherry-picking studies: Document search and selection, report contradictory findings and explain how you weighed evidence.
- Weak synthesis: Use matrices, maps or a reverse outline to force comparison across studies and draw higher-level conclusions.
- Sloppy citations: Check every reference against the source; ensure consistent citation style and accurate quotation or paraphrase.
- Unclear structure: Make the organizational logic explicit in the introduction and signpost section transitions.
- Ignoring contradictions: Dedicate a paragraph to reconcile inconsistent findings by design, context or measurement differences.
- Inadequate coverage: Perform backward/forward citation searches and consult key authors to reduce missing relevant studies.
Quick revision steps: ask a peer or supervisor to review for coverage and balance; create a reverse outline mapping paragraphs to claims and evidence; map each main claim to 1-3 citations; proofread citations and formatting.
- Scope is explicit and justified (dates, populations, designs).
- Search strategy and selection documented (search strings, databases, PRISMA flow where relevant).
- Coverage checks completed (key authors, backward/forward searches).
- Quality appraisal applied and limitations reported.
- Synthesis is organized and more than a series of summaries.
- Citations are accurate and consistent with required style.
- Ethics and originality: no plagiarism; proper attribution of ideas.
- All figures and tables (matrices, maps) are clear and referenced in the text.
If you plan to publish, adapt the review to the target journal’s expectations (systematic vs narrative), condense methods where required, and consider a registered protocol or meta-analysis when evidence volume and comparability allow.
Conclusion: final practical steps to finish and publish your review
A robust literature review begins with a clear question, explicit scope and documented method. Use systematic search, critical appraisal and structured extraction to make your synthesis defensible. Choose an organization that matches your aim and audience, and use practical tools-matrices, concept maps and narrative synthesis-to turn extraction notes into a persuasive, readable argument.
Final practical step: build your extraction table and short protocol summary, run a focused pilot search to test feasibility, adjust scope if needed, then proceed to full screening. That front-loaded clarity saves time and improves the credibility and publishability of the final review.