A marker can tell within a page whether a literature review compares studies or merely lists them. This guide to how to write a literature review, published by EdCitation, covers the patterns a review can follow, its introduction, body and conclusion, synthesis rather than summary, and the gap. It serves a stand-alone review and a dissertation chapter alike, and is the literature review chapter guide for how to write a dissertation.
Every rule below was read on 26 September 2026 at six universities' writing centres and libraries: the University of North Carolina at Chapel Hill (UNC), Purdue and the University of Southern California (USC) in the US, Kent and Southampton in the UK, and Toronto in Canada. Where they disagree, we say so.
EdCitation never writes any part of a review. It finds the reading with Find sources, keeps each reference, with a free account, in a list you can tag by theme, and checks the finished list with Verify references.
How to write a literature review, step by step
Read before you draft, as Kent and Southampton both advise; we would add that the introduction is easiest written last.
- Read the brief for length, source count, recency and style. Check your paper reads those requirements into a checklist, free.
- Fix the scope and the research question the review serves.
- Map the field. Find sources searches about 300 million published works by topic or by the claim a sentence needs, filtered by kind of source, years and open access.
- Take notes that set sources side by side on the same points.
- Choose the organising pattern and tag every reference with its theme.
- Draft the body section by section, each paragraph led by your own point.
- Write the conclusion, ending on the gap, then the introduction.
- Check the list with Verify references, or the whole chapter with References from a file.
- Check the draft against the brief. Theoretics QA (Max) judges it with quotations from your text.
What is a literature review for?
A literature review shows what is known about a question, where research disagrees, and where there is room for new work. University of Southampton Library (2026) gives it three jobs: showing how your field has developed, identifying the gap your project fills, and justifying the research. Labaree (2026), for USC, adds resolving apparent conflicts between studies.
Taylor (n.d.), for Toronto's Health Sciences Writing Centre, says the review is organised around a guiding concept, such as a research objective or a thesis, and is not a descriptive list. That separates a literature review from an annotated bibliography, which takes sources one at a time; Southampton traces some disconnected reviews to confusing the two.
A stand-alone review or a dissertation chapter
The parts are the same; their size changes. Purdue Online Writing Lab (n.d.) says a review inside a larger paper may need only a few sentences of introduction and conclusion, while a stand-alone review gives both more space.
In a dissertation it becomes the literature review chapter, though Southampton adds that some subjects have no separate chapter and work the reading into the introduction and themed chapters. A systematic review, with a protocol and a reproducible search, is a different method from the narrative review this guide covers.
How should I organise a literature review?
By theme, in most cases, though a review can also run by time, method, theory or debate, and many combine two.
| Pattern | The sections follow | What the guides say |
|---|---|---|
| Thematic | Subtopics or issues | Most reviews use a version of it (Southampton); most common in the social sciences (USC); harder to keep coherent (Kent) |
| Chronological | When work appeared, or how the field changed | Strict date order only if the order shows a trend (UNC); purely by date reads as fragments (Southampton) |
| Methodological | How the studies were done | Compares results from different approaches, such as qualitative and quantitative (Purdue) |
| Theoretical | Theories, models and definitions | The foundation for a theoretical framework (Purdue) |
| By debate | The positions taken on a question | Works for a position, against it, and offering alternatives (USC); include studies against your view (Toronto, Southampton) |
Where the guides disagree
They split on the source-by-source review. University of Kent (2025) presents working through the chosen books and articles in turn as the conventional plan, and calls the thematic approach more creative but riskier, since coherence is harder to establish. Southampton calls the same shape the "islands" structure and says it suits very few projects, and Taylor treats paragraphs that each open with a researcher's name as a bad sign.
All six guides ask for synthesis, and a source-by-source review makes synthesis hardest, so in our view organise by theme unless your brief asks otherwise. On chronology they differ only in degree, from Purdue's full pattern to Southampton's caution against a purely chronological review.
Combining patterns: the funnel
A common combination, in Southampton's and USC's guides, is themes as headings with time inside each. Southampton plans the themes as a funnel, broadest first, narrowing to the one nearest your question and giving that one the most space. Where themes do not fall in one line, several funnels meet at the gap: its ADHD example narrows a clinical section to gender, a classroom section to teacher-led and family-led interventions, and joins them in the last.
How do I write the introduction of a literature review?
Say what the review covers, what it leaves out and in what order it proceeds. Purdue suggests stating the topic or thesis, previewing the key texts and, where useful, how the sources were found; Kent asks for the strategy used to select them. Set the scope concretely: years, population, kinds of source.
An introduction in three sentences
Our own example, for a chapter on screen use and adolescent sleep:
This chapter reviews research published since 2010 on screen use and sleep among adolescents aged 11 to 18, drawn from systematic reviews and cohort studies. It is organised in three themes: sleep duration, sleep timing, and how screen use has been measured. It argues that the size of the association depends on the measure used, the gap this study's diary design addresses.
How do I write synthesis rather than summary?
Put several sources in each paragraph and say how they relate: where they agree, where they differ, and why. University of North Carolina at Chapel Hill Writing Center (n.d.) describes a summary as a recap of a source and synthesis as a reorganisation that can produce a new reading. Purdue suggests picturing the sources in one room, talking to each other, and offers a test: a paragraph citing one source is usually summary.
UNC's model paragraph opens and closes with the writers' own point, sources in between. Southampton's exercise for a stuck writer: one sentence from one source, then a sentence from a second that supports, extends or contradicts it, then a third. The notes behind this are in the synthesis matrix; your own words are in paraphrase, quote or summarise.
A worked example: the same three sources, twice
This example is ours, written from three real papers; every figure comes from their abstracts on PubMed. First, summary:
Hale and Guan (2015) reviewed 67 studies of screen time and sleep in school-aged children and adolescents. Carter et al. (2016) carried out a meta-analysis of portable devices and sleep. Orben and Przybylski (2020) analysed data on 11,884 adolescents from the UK Millennium Cohort Study.
Then synthesis:
Research agrees that adolescents who use screens more tend to sleep less, but not on how much less, and the answer seems to depend on how screen use is measured. Hale and Guan (2015) found an adverse association, mostly shorter or later sleep, in 90% of the 67 studies they reviewed, and Carter et al. (2016), pooling 20 studies of children aged 6 to 19, found that bedtime use of a portable device went with more than double the odds of inadequate sleep (odds ratio 2.17). All 20 were cross-sectional, though, and Hale and Guan list unconfirmed causality and measurement error among the field's limits. When Orben and Przybylski (2020) compared time-use diaries with questionnaire recall, the associations were very small, about nine minutes less weekday sleep per extra hour of screen use, and came mainly from recall and whole-day measures. The link is consistent; its size is not settled, and a study recording screen use as it happens would test it.
The second version opens on our claim, groups two studies by what they share, sets the third against them on design and measurement, and closes on what is open, where a gap statement begins. Whether that gap is open today would need its own search: the paragraph shows a shape, not a finding.
How do I identify the gap in the literature?
The gap is what the review shows to be missing; the body should lead to it, not announce it at the end. Southampton says a gap need not be an empty space: it lists gaps of population or place, of theoretical framework, of method, of scarcity (too few studies to rely on) and of elapsed time (findings that changed circumstances may have overtaken). The worked example points to a gap of method.
When no gap appears
Change the scope. Southampton advises zooming in when a field is crowded, to a narrower lens, object or group, and zooming out when a topic is so new there is little to review. Our glossary's entry on the research gap adds that a gap is worth filling only if the answer would matter.
State it no more strongly than your search supports; Kent warns that sweeping generalisations let many dissertations down. "No study has examined" is a claim about everything published; "none of the studies reviewed here" is one you can defend. Finding sources for a literature review fast covers the search behind it.
How do I write the conclusion of a literature review?
Draw the themes together into what is known, what is disputed and what is missing, then look forward. UNC asks where the discussion might go next, Purdue asks you to connect the findings to your research question, and Labaree (2026) adds which works are most convincing. In a dissertation the last paragraph hands over to the methodology, where the gap becomes your question; a stand-alone review ends on questions for further research.
How long is a literature review, and how many sources?
As long as your brief or handbook says. University figures are rough guides, and they disagree. University of Kent (2025) adds that a dissertation built mainly on existing scholarship needs a longer review than one leading into new empirical research.
| University guide | Length of the review | Number of sources |
|---|---|---|
| Kent (UK) | Undergraduate no more than about 2,000 words; master's 2,000 to 3,000; PhD one chapter of perhaps 8,000 to 10,000 | 8 to 10 significant pieces for an 8,000-word dissertation; up to 20 for 12,000 to 15,000 words |
| Southampton (UK) | Ask your supervisor | 30 to 40 sources in the review of a dissertation of about 10,000 words, as a ballpark |
| UNC and USC (US) | No figure | No figure; ask your instructor |
| Purdue (US), Toronto (Canada) | No figure | No figure |
The two counts measure different things. Kent counts significant pieces, the major works a review analyses; Southampton counts every source cited, several sharing one parenthesis, and says one dissertation can carry 50 more references than another and be as rigorous. How many references you need covers a whole paper.
How EdCitation helps with a literature review
EdCitation is the best tool for a review's source work: every reference it gives comes from the publisher's record and is never written, which no chatbot can promise, and it checks a whole list at once. We tested both on 26 September 2026.
Mapping the field. In Find sources, screen time adolescent sleep, filtered to review articles and sorted by most cited, returned 9,114 results. First was Hale and Guan's systematic review in Sleep Medicine Reviews, cited 1,467 times, then Sampasa-Kanyinga and colleagues' 2020 review in the International Journal of Behavioral Nutrition and Physical Activity, cited 451 times. The first eight split by age group (under-3s, under-5s, school-age children) and by outcome (sleep, depression, mental health): two candidate sets of themes. The list dates Hale and Guan 2014, when the review appeared online; Cite a source built the reference from Crossref's record, with the year of its June 2015 issue.
Building the structure. A free account keeps every reference in a list you can tag and filter by tag. Tag each source by theme and each filtered list is one section's reading; organising references with tags sets out a scheme.
Checking the list. We gave Verify references three entries: Carter and colleagues' meta-analysis, correct; Orben and Przybylski's study with 2018 in place of 2020; and one we invented. The first came back verified ("The DOI resolves to this record and the title matches."), the second doubtful, shown as "check this" ("A record with this title exists, but it was published in 2020, not 2018."), and the third not found ("No publisher's or registry's record matches this reference."). A reference that could not be checked is shown as such, never as not found. Read each suggested entry against the article: Carter and colleagues' record holds only the first page, so its entry ends at 1202, not 1208. Whether a real paper says what the review claims is a reading job, set out in detecting fake citations in a literature review.
For the whole chapter, References from a file checks every reference and matches every in-text citation to the list: three checks a month with a free account, no limit with Pro at $8 a month. Check your paper turns the brief into a checklist, free; Theoretics QA, part of Max at $24 a month, judges whether the review does what that brief asks, quoting your text and rewriting nothing. None of these writes a word of the review, and none can promise a mark.
Quick questions
What are the main ways to structure a literature review?
Thematic, chronological, methodological, theoretical and by debate. Southampton and USC describe thematic organisation as the most common, and many reviews combine two, with themes as headings and time inside each.
What is the difference between summary and synthesis in a literature review?
A summary restates one source; synthesis relates several, saying where they agree, differ and why. A paragraph citing only one source is usually summary.
How long should a literature review be in a dissertation?
As long as your handbook says. Kent suggests up to about 2,000 words for an undergraduate dissertation, 2,000 to 3,000 for a master's and 8,000 to 10,000 for a PhD, as rough guides.
How many sources should a literature review have?
No university sets a universal number. Kent suggests 8 to 10 significant pieces for an 8,000-word dissertation; Southampton suggests 30 to 40 sources for the review in a 10,000-word dissertation, counting every source cited.
Can EdCitation check my literature review against the brief?
Yes. Check your paper reads the brief's requirements into a checklist, free, and Theoretics QA, part of Max, judges the review against them with quotations from your text. It never rewrites anything.
References
- Carter, B., Rees, P., Hale, L., Bhattacharjee, D., & Paradkar, M. S. (2016). Association between portable screen-based media device access or use and sleep outcomes: A systematic review and meta-analysis. JAMA Pediatrics, 170(12), 1202–1208. https://doi.org/10.1001/jamapediatrics.2016.2341
- Hale, L., & Guan, S. (2015). Screen time and sleep among school-aged children and adolescents: A systematic literature review. Sleep Medicine Reviews, 21, 50–58. https://doi.org/10.1016/j.smrv.2014.07.007
- Labaree, R. V. (2026, September 8). Organizing your social sciences research paper: The literature review. USC Libraries, University of Southern California. https://libguides.usc.edu/writingguide/literaturereview
- Orben, A., & Przybylski, A. K. (2020). Teenage sleep and technology engagement across the week. PeerJ, 8, Article e8427. https://doi.org/10.7717/peerj.8427
- Purdue Online Writing Lab. (n.d.). Writing a literature review. Purdue University. https://owl.purdue.edu/owl/research_and_citation/conducting_research/writing_a_literature_review.html
- Taylor, D. (n.d.). The literature review: A few tips on conducting it. Writing Advice, University of Toronto. https://advice.writing.utoronto.ca/types-of-writing/literature-review/
- University of Kent. (2025, September 29). Writing a literature review. Skills for Academic Success. https://student.kent.ac.uk/studies/literature-reviews/writing-a-literature-review
- University of North Carolina at Chapel Hill Writing Center. (n.d.). Literature reviews. https://writingcenter.unc.edu/tips-and-tools/literature-reviews/
- University of Southampton Library. (2026, June 18). Literature review. Writing the dissertation: Guides for success. https://library.soton.ac.uk/writing_the_dissertation/literature_review