A literature review is a written argument about the research on a topic: it collects the key sources, puts them in conversation with each other, and shows what is established, what is contested, and what your study or paper will add. The Purdue OWL describes it as a document that "collects key sources on a topic and discusses those sources in conversation with each other" — a process called synthesis. This guide walks through the whole task: starting, searching, structuring, writing each section, and ending.
What is a literature review?
A literature review is both a process and a product. The process is the systematic reading and note-taking you do to master the research on a question. The product is the chapter or paper that presents that research as an organized argument — not as a list of summaries. The University of North Carolina at Chapel Hill Writing Center puts the key distinction plainly: a summary recaps each source, while a synthesis reorganizes the information so the reader can see relationships, debates and gaps.
Two implications follow. First, a literature review is not an annotated bibliography with connective tissue; every paragraph should be about the literature, not about one study. Second, a literature review is not a book report on everything you read. It is selective by design: you include the studies that shape your question, and you leave out the rest.
The name varies by context but the job does not. In the Philippines and much of Southeast Asia, the thesis chapter is called the Review of Related Literature, or RRL, and is usually Chapter 2; in journal articles the review is a section after the introduction; in some fields it is a standalone paper. Wherever it sits, the test is the same: synthesis, organized as an argument, ending in a gap.
How to start a literature review
Start from a question, not from a database. Write one to three review questions — the questions your review must answer for your study to make sense — and let them drive everything: which databases you search, which terms you combine, which papers clear your inclusion bar.
Turning a topic into questions is mostly a matter of adding a stake. "Teacher burnout" is a topic; "What working conditions are most consistently linked to teacher burnout, and in which school settings has this been studied?" is a review question — it names the variables, the comparison and the boundary, which is exactly what a search strategy needs. If your questions will not compress to a sentence each, the scope is probably still too broad; if they can be answered by a single study, it is too narrow.
Then set a scope you can defend: the years you will cover, the databases you will search, and the types of studies you will include. A scoped review you can finish beats a comprehensive one you cannot. Write the scope down; you will need it for the search-strategy paragraph, and it protects you from the two classic failure modes — reading forever, and missing a body of work your committee expected.
Before reading at full depth, skim widely. Titles and abstracts tell you which papers are actually about your question. Only the survivors deserve close reading — and close reading deserves a literature review matrix, so that each study's method, sample, findings and limitations land in comparable columns instead of scattered highlights. The filled matrix example shows how two real studies are recorded and compared.
How to conduct a literature review
Search
Combine two or three databases with different strengths (for education and social science, a common mix is one large interdisciplinary index plus one subject-specific one), and record the exact search terms and date filters you used. Free tools such as OpenAlex index hundreds of millions of works and make a useful first sweep. Track searches in one place: the same phrase searched three weeks apart should be findable.
Screen
Screen in two passes. First pass: title and abstract only — does this paper address your review question? Second pass: full text for the survivors — does it report something you can use (method, findings, framework)? Keep a record of what you excluded and why; systematic or not, your future self will need to explain the shape of your library.
Read
Read with your review questions in hand. For each study, extract: what it asked, how it was done, who the sample was, what it found, and what the authors said the findings mean. Note the page of every quote you might use — retrofitting page numbers later is slow and error-prone.
Extract
Move the extractions into your matrix as you go. The matrix is what turns reading into reviewable structure: sort by theme and the sections of your review start to appear; sort by year and you can see how the field moved. If typing citations is the bottleneck, the free literature review outline generator can draft the theme skeleton from real papers, and an outline template gives you the section prompts.
How to structure a literature review
Structure follows the argument, but three organizations cover most reviews:
- Thematic — sections are themes or claims about the literature. The default for dissertations and most course reviews; each section argues one point with several studies.
- Chronological — sections are periods, showing how the field developed. Useful when the evolution of an idea is the point.
- Methodological — sections are groups of methods or study designs, compared on what each can and cannot show. Common when method quality is part of your argument.
Most dissertation chapters mix these: a thematic core with a chronological opening or a methodological subsection. Whichever you choose, the literature review outline is the same — introduction, three to six theme sections, gaps, conclusion — and for a five-chapter dissertation, the Chapter 2 structure adds the search strategy and framework sections your committee expects.
A worked mini-outline makes the shape concrete. For a study on teacher burnout, a thematic outline might read:
1. Introduction: burnout as a driver of attrition; scope 2014–2024; three questions
2. How burnout is measured in school settings (ML Screener vs. self-report; why it matters)
3. Workload and working conditions (agreement: workload correlates; dispute: causal evidence)
4. Autonomy and school leadership (stronger links than workload in multi-site studies)
5. Gaps: few studies of early-career teachers in provincial schools; little longitudinal work
6. Conclusion: conditions over individuals; present study addresses gap 5Notice what the outline encodes before a single paragraph is drafted: sections 2–4 are claims with boundaries, not topics; each names its tension (dispute, stronger links); and section 5 points at the study. When your outline can do that, writing becomes translation.
How to write each section
Introduction. State the topic, the scope (years, databases, inclusion rules) and how the review is organized. Write it last if that is easier; it must match what follows.
Theme sections. Open each with a claim about that group of studies — not a topic label. "Teacher burnout is driven less by workload than by insufficient autonomy" is a claim; "Studies on teacher autonomy" is a label. Support the claim with several studies, compare their methods and findings, and close by noting what remains unresolved. If your paragraph pattern is "Author A found… Author B found… Author C found…", you are summarizing; rewrite around the relationship — agreement, contradiction, or gap.
A dependable paragraph shape inside a theme is claim → evidence → comparison → takeaway. First sentence states the claim. The next few sentences bring in the studies that support it, with citations woven into the sentence rather than bolted on ("classroom-level studies consistently report X (Aguilar, 2019; Brennan & Okada, 2021)"). Then one or two sentences compare: where the studies agree, where they diverge, and what explains the divergence — sample, setting, measure or decade. The last sentence says what the pattern means for your question. Writing this way keeps author names subordinate to the point, which is what separates a review from a roll call.
Search strategy. In dissertations, a short section reports how you searched: databases, terms, years, and how many studies made the cut. It reads dry; committees love it, because it makes your review reproducible.
Framework. If your program requires a theoretical or conceptual framework, present the theory, the studies that apply it, and why it fits your study.
Gaps. State what the reviewed literature does not cover — populations, settings, methods, questions — and point forward to your study. Lead with evidence from your matrix, not with conviction.
How to end a literature review
End by answering the question the review exists to answer. A strong conclusion does three things in a few paragraphs: it states what the body of evidence establishes; it names the main disagreements and their likely causes; and it states precisely what your study will add, in the terms the themes set up. Resist introducing new studies here — if a study belongs anywhere, it belongs in a theme.
For a dissertation, the last paragraph hands off to Chapter 3: name the gap, and state in one sentence how your design addresses it. The reader should close the chapter able to repeat your "so what" without notes.
How long should a literature review be?
Long enough to cover the themes that justify your study, and no longer — which is why the honest answer starts with your program's handbook rather than a number. Length scales with the document: a course-paper review may run a few pages, while a dissertation review chapter typically runs twenty or more. For scale, the University of Kent's study advice offers a rough rule of thumb of 8–10 significant works for an 8,000-word dissertation and up to 20 for 12,000–15,000 words, with more when the project mainly engages existing scholarship. Use numbers like these as planning aids, not as targets: a review is finished when its themes are argued, not when its pages are counted.
Literature review vs. annotated bibliography
An annotated bibliography is a list: one entry per source, each with a summary and perhaps an evaluation. A literature review is an argument: sources grouped by theme, compared with each other, and subordinated to claims. The practical test is the paragraph test — if most paragraphs are about a single source, you have written a bibliography; if most paragraphs need several sources to make their point, you have a review. Many programs assign a bibliography as a step toward a review, and that is exactly what it is good for: a way to force careful reading before synthesis begins.
Frequently asked questions
How long should a literature review be?
There is no universal length: it depends on your level, your field and the total size of the document. A course paper review may run a few pages; a thesis or dissertation review chapter usually runs 20+ pages. Start from the word limit in your program's handbook and let the themes that justify your study set the length.
How many sources should a literature review have?
It depends on the scope and level. As a rough rule of thumb, the University of Kent's study advice suggests 8–10 significant works for an 8,000-word dissertation and up to 20 for 12,000–15,000 words; dissertation review chapters typically draw on more. Coverage of the key lines of research matters more than the count.
What is the difference between a literature review and an annotated bibliography?
An annotated bibliography summarizes or evaluates one source at a time, entry by entry. A literature review organizes sources by theme and puts them in conversation with each other, building an argument about what is known, what is contested, and what your study adds.
Can a literature review include non-journal sources?
Yes, when they are appropriate evidence for your topic: books, reports, dissertations and policy documents all count as literature in this sense. Whatever the source type, the same rule applies: read it yourself, cite it accurately, and weigh it against the peer-reviewed work in the field.
