How long should a literature review be?

It depends on the type of work. In a thesis the review is often one chapter, in a journal article a section of a few pages, and a stand-alone review article can be much longer. Your department or the journal's author guidelines are the binding reference.

What is the difference between a narrative and a systematic review?

A narrative review selects and discusses literature to give an informed overview. A systematic review follows a predefined, documented protocol for searching, selecting and appraising studies so that others can reproduce it. Most theses use a narrative or structured review unless the method itself is systematic.

How many sources should a literature review include?

There is no universal number. Cover the key studies, the major positions in the debate and recent work. If new searches only return papers you already have, you have probably covered enough.

Can I write the literature review before my research question is final?

Yes, and many people do. Writing a first draft often sharpens the question. Expect to revise the review once your question and method are settled.

How to write a literature review: structure and steps

A literature review is a structured, critical account of what is already known about your topic. It does not just summarise papers one after another; it groups them, compares their findings, points out strengths and weaknesses and ends by showing the gap your own work addresses. Think of it as an argument built from other people’s evidence.

What a literature review has to do

Whether it is a chapter in a thesis or a section of a paper, a good review does four things:

Maps the field: the main concepts, theories and lines of research.
Synthesises: shows how studies relate to each other, where they agree and where they conflict.
Evaluates: comments on methods, samples, contexts and limitations.
Positions your work: leads to the open question or research gap your study responds to.

If a paragraph only reports what one author did, it is probably still a summary. Reviewers and examiners look for the comparison and the judgement.

Step-by-step

Define the scope. Write down your working question and the boundaries of the review: time span, disciplines, types of study, languages.
Search systematically. Use the method in our guide to literature search and keep a search log.
Read with a purpose. For each source, note the question, method, main findings, limitations and how it relates to your topic.
Build a synthesis matrix. A table with sources as rows and themes or variables as columns quickly shows clusters, contradictions and empty cells.
Choose an organising principle (see below) and draft an outline with one point per paragraph.
Write the first draft, leading each paragraph with your claim and supporting it with several sources.
Revise for argument. Check that every section moves towards your gap and that the conclusion of the review is clear.

A simple synthesis matrix

Source	Context	Method	Key finding	Limitation
Study A	Secondary schools	Survey	Positive effect	Self-reported data
Study B	Universities	Experiment	No effect	Small sample
Study C	Universities	Interviews	Effect depends on support	Single institution

Seen together, these three rows already suggest a paragraph: findings differ by context and by method, and nobody has yet tested the role of support with a larger sample.

Ways to structure a review

Structure	Works well when	Watch out for
Thematic	The field has several clear sub-topics	Themes overlapping without clear boundaries
Chronological	The development of an idea is part of your argument	Turning into a history lesson
Methodological	Results depend strongly on how studies were done	Losing sight of the content
Theoretical	Competing theories frame your question	Abstract discussion without evidence

Many reviews combine these: a thematic structure with a short methodological comparison inside each theme.

Writing techniques that help

Lead with the point, not the author. “Several studies find that feedback timing matters more than frequency (A, 2021; B, 2023)” is stronger than “A (2021) studied feedback”.
Use connecting phrases that show relationships: in contrast, building on this, a different explanation is offered by.
Be precise about strength of evidence. Distinguish between one small study and a consistent body of research.
Paraphrase and cite correctly. Choose one citation style and apply it consistently; our guide to citation styles compares APA, Harvard and IEEE.
Keep your voice present. Short evaluative sentences of your own tie the evidence together.

Note: requirements for the review chapter vary. Some departments expect a separate “state of research” chapter, others integrate it into the introduction or theory chapter. Check your module handbook and the thesis structure your department uses.

Checklist before you hand it in

Each section has a clear point stated in its first sentence.
Most paragraphs cite more than one source.
Conflicting findings are discussed, not ignored.
Limitations of key studies are named.
Recent research is included, as well as foundational work.
The final paragraph leads directly to your research question or gap.
Every in-text citation appears in the reference list and vice versa.

Typical problems and quick fixes

“My review reads like a list.” Rebuild the outline around themes from your synthesis matrix and rewrite topic sentences as claims.

“I keep finding new papers.” Set a cut-off date for the search, note it in your methods, and only add later papers if they change your argument.

“My supervisor says it lacks critical analysis.” Add one sentence per key study about what its design can and cannot show, and compare studies directly.

For more guides on searching and framing your topic, see the literature section.