How to Write a Literature Review: Structure, Steps and Examples

What a literature review is — and what it is for

A literature review is a structured, critical discussion of the published research that already exists on your topic. It sits near the start of a dissertation, thesis or research article, and its job is to show your reader where your work fits into an ongoing conversation. It is not a reading diary, and it is not a string of book reports. It is an argument about the state of knowledge in your field.

Think of it this way. Before you can claim your research adds something, you have to demonstrate that you understand what is already known, how it was found out, and where the loose ends are. A strong review does four things at once: it maps the territory, evaluates the quality and relevance of what others have found, groups those findings into a coherent picture, and points to the space your own study will occupy. Everything else — the searching, the note-taking, the referencing — exists to serve those four aims.

The purpose shifts slightly depending on the assignment. In a standalone review essay, the review is the whole piece of work. In a dissertation or dissertation project, it is the chapter that justifies your research questions. Either way, the reader should come away understanding not just what has been written, but what it all adds up to.

Types: narrative, systematic and everything between

Not all literature reviews follow the same rules, and picking the wrong style for your assignment is a common early error. The two you will meet most often are the narrative review and the systematic review.

Narrative (or traditional) reviews

A narrative review gives a broad, interpretive overview of a topic. You select the most relevant and influential work, organise it around themes or debates, and build an argument about what the field shows. There is judgement involved in what you include, and that is expected. Most undergraduate and taught-masters literature reviews are narrative reviews, even when the module handbook does not use the word.

Systematic reviews

A systematic review answers a tightly defined question using an explicit, reproducible method. You set inclusion and exclusion criteria in advance, search named databases with recorded search terms, screen every result against your criteria, and report exactly how many studies you kept and discarded (often in a flow diagram). The aim is to remove selection bias so that another researcher following your method would reach the same set of studies. Systematic reviews are common in health, psychology and the sciences, and they take considerably longer.

Other variants

You may also encounter scoping reviews (mapping the breadth of a topic without full appraisal), integrative reviews (combining different research designs), and meta-analyses (statistically pooling results from comparable quantitative studies). For most coursework, the decision is simply narrative versus systematic — so check your brief, and if it is unclear, ask your supervisor before you start searching. The method you choose shapes everything that follows.

Searching for and selecting your sources

A review is only as good as the reading behind it, so treat searching as a deliberate stage rather than something you do the night before writing. Start by turning your topic into search terms. Break the question into its key concepts, then list synonyms and alternative spellings for each — British and American, technical and lay. Combining these with the operators AND, OR and NOT lets you widen or narrow results precisely.

Search where the scholarship actually lives. Your university library discovery tool and subject-specific databases (such as Scopus, Web of Science, PsycINFO, PubMed or your field’s equivalent) will give you far better results than a plain web search. Google Scholar is useful for chasing citations and finding who has cited a key paper since, but it should supplement, not replace, the databases your library pays for.

Deciding what to keep

You cannot read everything, so you need selection criteria. Prioritise peer-reviewed journal articles, then scholarly books and reputable reports. Weigh each source on relevance to your question, recency (with allowance for foundational older work), the quality of the outlet, and the strength of the method. A quick way to expand a promising cluster is to look at the reference list of one excellent recent paper (backward searching) and at the papers that have since cited it (forward searching). When a search stops turning up anything new, you have reached a reasonable stopping point for a taught assignment.

Keep a simple log of what you searched, where, and what you kept. Even for a narrative review this saves hours later and makes your process defensible if a marker asks how you found your sources. If you would rather work from a proven starting layout, our literature review template gives you a search log and a note-taking grid ready to fill in.

Reading critically, not passively

The single biggest difference between a weak review and a strong one is how the sources were read. Passive reading — highlighting sentences and hoping the meaning sticks — produces a review that just repeats what each author said. Critical reading interrogates every source and produces a review that evaluates it.

As you read each study, ask a consistent set of questions. What exactly was the research question? What method was used, and was it appropriate? Who or what was studied, and how large or representative was the sample? What did the authors actually find, as opposed to what they claim it means? What are the limitations, and do the conclusions overreach the evidence? How does this study speak to the others you have read — does it agree, complicate or contradict them?

Record the answers as you go, ideally in a matrix with one row per source and columns for aim, method, sample, findings, limitations and “link to my question”. This grid becomes the raw material for synthesis, because it lets you read across studies rather than down a single one. Do not skip this into a pile of unread PDFs; a smaller set of sources you have genuinely digested beats a huge pile you have merely collected.

Organising: thematic, chronological or methodological

Once you have read and logged your sources, you need a structure to arrange them. There are three main options, and choosing well is what stops a review reading like a list.

Thematic

You group sources by the ideas, debates or concepts they address, with each theme becoming a section. This is the default for most reviews because it forces you to compare studies rather than march through them one by one. A theme might be a particular explanation, a recurring finding, or a point of disagreement in the field.

Chronological

You trace how thinking on the topic has developed over time. This works well when the story of the field is genuinely about change — a theory being revised, a method maturing, a consensus forming or breaking. The risk is that it slides into “in 2015 X said…, in 2018 Y said…”, which is a timeline, not an analysis. Use it only when the sequence itself carries meaning.

Methodological

You organise by the approach researchers took — for instance, grouping qualitative studies separately from quantitative ones, or comparing experimental and observational designs. This is useful when the method is central to the debate, such as when different designs produce conflicting results on the same question.

These are not mutually exclusive. A strong dissertation review is often thematic overall, with a methodological comparison inside one theme where it matters. Whatever you choose, the organising principle should be visible to the reader in your section headings and topic sentences.

“If your reader can tell which author you are discussing before they can tell which idea you are discussing, your review is organised by source rather than by argument — and that is the fault to fix first.”

Synthesising versus summarising

This is the heart of the whole task, so it is worth being precise. Summarising restates what one source says. Synthesising combines several sources to make a point that none of them makes alone. A review built from summaries answers the question “what did each author say?” A synthesised review answers “what does the field, taken together, tell us — and where does it disagree?”

The practical test is your paragraph structure. In a summarised review, most paragraphs begin with an author’s name and cover one study. In a synthesised review, most paragraphs begin with an idea or a finding, and several studies are cited within it as evidence, agreement or dissent. You should routinely be writing sentences such as “three studies report the same effect, although two used student samples that limit how far the finding travels”. That single sentence does what a summary cannot: it groups, compares and evaluates.

Synthesis also means putting sources into conversation. When two studies contradict each other, say so and offer a reason — a difference in sample, method, context or definition. When a cluster agrees, name the pattern and then probe its limits. If you find yourself paraphrasing one paragraph after another, our paraphrasing tool and summarising tool can help you compress individual sources into your own words quickly — but remember the real work is joining them together, which no tool does for you.

Identifying the gap

Every review should end up somewhere, and that somewhere is usually a gap: the thing the existing literature has not settled, tested or even asked. Finding it is not a separate step bolted on at the end — it emerges naturally from good synthesis, because comparing studies exposes what none of them covers.

Gaps come in recognisable shapes. There may be a population that has never been studied (research done only on adults, never on adolescents). There may be a context gap (findings established in one country but never tested in another). There may be a methodological gap (a question only ever approached quantitatively, never through lived experience). There may be a contradiction that nobody has resolved, or a theory that has never been tested against real data. Naming which kind of gap you have found makes your justification sharper.

The gap is the bridge from your review to your own research questions, which is why it belongs at the end of the chapter. State it plainly: “while the effect is well established in [context], no study has examined it in [your context], and it is this gap the present study addresses.” That sentence turns a review into a rationale. If you are writing a proposal rather than a full dissertation, our guide on how to write a research proposal shows how the gap you identify becomes the spine of the whole document.

The writing structure that works

With reading, organisation and synthesis in hand, the writing structure is straightforward. A literature review has three parts, whatever its length.

Introduction

Set the scene in a few paragraphs. State the topic and why it matters, define the scope (what you are covering and, just as usefully, what you are not), explain how you searched and selected sources if the brief expects it, and tell the reader how the review is organised. A reader should finish your introduction knowing exactly what is coming and why.

Body

This is the themed core, arranged by whichever principle you chose. Each section handles one theme, opens with a topic sentence stating the point of that section, then marshals the relevant studies as evidence, comparison and contrast. Signpost between sections so the argument flows — each theme should feel like it earns its place and connects to the next. Keep your own analytical voice present: after presenting what studies show, tell the reader what it means.

Conclusion

Draw the threads together. Summarise what the literature collectively establishes, highlight the tensions and limits you found, and then state the gap that leads into your own research. The conclusion should not introduce new sources; it should distil everything the body has built. For a dissertation, this is where you hand the baton to your methodology chapter.

If you want to see the shape on the page before you draft, browse a worked literature review sample to see how introduction, themed body and conclusion fit together in practice.

Referencing and keeping track

A literature review is dense with citations, so referencing discipline is not optional. Use whichever style your department requires — Harvard, APA, MLA, Vancouver or another — and use it consistently for every in-text citation and every entry in the reference list. Mismatched or missing references are among the easiest marks to lose and the easiest to protect.

Set up a reference manager early. Tools such as Zotero, Mendeley or EndNote store your sources, generate citations in your chosen style, and save you from rebuilding a bibliography by hand at 2am. Add each source the moment you decide to keep it, along with a line or two of notes, so your reading log and your reference list grow together.

Two integrity points matter throughout. First, every claim that is not your own or common knowledge needs a citation — synthesis does not dilute this obligation, it increases it, because you are drawing on many sources at once. Second, paraphrase properly: rewriting a sentence by swapping a few words is still too close to the original. Put the idea fully into your own words and cite it, and use quotation marks on the rare occasions you quote directly.

Common mistakes to avoid

Most weak reviews fail in a handful of predictable ways. Watch for these:

• The annotated bibliography in disguise. One paragraph per source, no comparison. The commonest fault, and the one thematic organisation cures.
• Description without evaluation. Reporting what studies found but never judging their quality, relevance or limitations.
• No argument. A review that lists everything but builds towards nothing, leaving the reader unsure why any of it was included.
• Ignoring disagreement. Cherry-picking studies that agree and quietly dropping the ones that complicate the picture. A good review handles conflict, it does not hide it.
• Over-reliance on quotation. Long quotes stitched together are a sign the ideas have not been digested. Paraphrase and synthesise instead.
• Scope drift. Wandering into loosely related topics because a source was interesting. Stay inside the boundaries you set.
• No gap. Finishing without saying what the literature leaves unanswered, so the review does not justify any further research.

If several of these feel familiar, the fix is almost always the same: go back to your source matrix and rebuild the review around ideas rather than authors.

Example: a synthesised paragraph versus a list

The difference between summarising and synthesising is easiest to see side by side. Both passages below use the same three fictional studies; only the handling differs.

Notice that the strong version is barely longer. Synthesis is not about writing more; it is about organising the same evidence around an argument instead of around a reading order.

Tools and templates that speed it up

You do not have to build every part from scratch. A few resources take the repetitive load off so you can spend your energy on the analysis that earns marks.

A ready-made literature review template gives you the introduction-body-conclusion skeleton plus a source matrix to fill in as you read. Worked literature review samples and full dissertation samples show the standard you are aiming for. When you are compressing individual sources into your own words, the summarising tool and paraphrasing tool speed up the drafting — just keep the synthesis and the citing in your own hands.

And if the review is part of a larger project you are struggling to hold together, structured guidance on writing a research proposal keeps the whole thing pointing in one direction.

Related guides

• Literature review writing services
• Literature review template
• Literature review sample
• Dissertation sample
• How to write a research proposal
• Paraphrasing tool
• Summarising tool

Frequently asked questions