How to Write a Literature Review: Full Structure Guide

What a Literature Review Actually Does

A literature review is not a summary of everything you've read. It's a critical synthesis of existing research that establishes what's already known about your topic, where the disagreements are, and — most importantly — where the gap is that your own study will fill.

Examiners read the literature review to judge two things: whether you understand the field well enough to place your own work within it, and whether your research is actually needed. A review that just lists source after source, one paragraph per study, fails on both counts even if every summary is accurate.

Step 1: Define a Search Strategy Before You Start Reading

Start with your research questions, not a general topic. Break each question into its core concepts, then build search terms for each — including synonyms and related terms a database might index differently than you'd expect.

Pick your databases deliberately. Scopus and Web of Science give broad, high-quality coverage across most fields; PubMed is essential for health and medical topics; field-specific databases (EBSCO's business or education indexes, for example) often surface sources the big general databases miss.

• Break research questions into core concepts and build search terms for each
• Include synonyms and alternate spellings ("e-learning" vs. "online learning")
• Use Boolean operators (AND, OR, NOT) to combine terms precisely
• Document your search strings — you'll need to report them if this is a systematic review

Step 2: Screen and Select Sources With Explicit Criteria

Set inclusion and exclusion criteria before you start screening results, not after you've already read them. Common criteria include publication date range, study type, geographic or population relevance, and language.

For a narrative review, this can stay relatively informal — documented but not necessarily reported in full. For a systematic review, this step needs to be rigorous and fully reported, typically visualized in a PRISMA flow diagram showing how many records were identified, screened, and ultimately included.

Step 3: Synthesize by Theme, Not by Source

This is where most weak literature reviews go wrong. Organizing a review study-by-study ("Smith (2019) found X. Jones (2020) found Y.") produces a list, not an argument. Organizing by theme — grouping what multiple studies say about a specific sub-question — produces synthesis.

A useful test: if you can't write a topic sentence for a paragraph that isn't just a study's name, you're probably still in list mode. A synthesized paragraph makes a claim about the field ("Evidence on X is mixed, with earlier UK-based studies finding a positive effect that more recent international studies have failed to replicate") and then supports it with sources.

Step 4: Write Critically, Not Just Descriptively

For every major source or cluster of sources, ask: what are its methodological strengths and limitations? Does it generalize to your context? Does it contradict or extend other work you've reviewed? A critical review evaluates evidence quality, not just what a study concluded.

End the review by explicitly stating the gap your study addresses. This should follow logically from everything before it — a reader who's just followed your synthesis should already sense where you're heading before you state it outright.