How to write a literature review

A literature review shows what is already known about your topic, how researchers have studied it, and where the gaps are that your own work will fill. It is one of the longest and most important parts of a project, dissertation or thesis, and also one of the most commonly misunderstood. The most frequent mistake is writing a list of summaries. A good review synthesises: it groups and compares sources to tell the story of the research.

Step 1: Define your scope

Start with your research question or topic, then decide what is in and out:

• Key concepts: the two or three ideas your topic connects, for example mobile money, financial inclusion and rural households.
• Time span: how far back you will go, apart from seminal papers.
• Context: a country, region, population or sector, if your topic has one.
• Types of sources: usually peer-reviewed journal articles first, then books, book chapters and reputable reports.

Write your scope down. It stops the review growing without end.

Step 2: Search systematically

Turn your concepts into search terms, including synonyms (mobile money OR M-Pesa OR digital financial services). Search more than one place:

• your university library’s databases (Scopus, Web of Science, JSTOR, EBSCO and others),
• Google Scholar,
• the CiteSwap finder, which searches more than 150 million articles, books and conference papers and gives you a ready-made citation for each.

Keep a simple log of where you searched and which terms you used. Supervisors often ask, and in some disciplines you report it in your methods.

Step 3: Find the seminal papers first

Every field has a handful of papers that everyone cites. They give your review its backbone and show your supervisor that you know the field. The quickest way to find them is to rank a topic search by citations: the most cited papers finder does exactly this, and you can limit it to recent years when you need the current state of research.

Then snowball from the key papers:

• Backward: read their reference lists to find the earlier work they build on.
• Forward: look at newer papers that cite them, to see how the field has developed.

In CiteSwap, open References in this paper and Papers citing this under any search result.

Step 4: Read and take structured notes

You don’t need to read every paper word for word. Read the abstract, introduction and conclusion first, and read closely only what is relevant. For each source, note:

• the research question and context,
• the method and sample,
• the main findings,
• strengths, limitations and how it relates to other papers.

A synthesis matrix, a table with one row per source and one column per theme or question, makes the next step far easier, because you can see at a glance which papers say what about each theme.

Step 5: Choose a structure

Organise the review around ideas, not around individual papers. Common structures:

• Thematic: one section per theme or debate (the most common and usually the strongest).
• Chronological: how thinking on the topic has developed over time.
• Methodological: how different research approaches have produced different findings.
• Theoretical: the theories or models used to explain the topic.

Many reviews combine these, for example themes within a broadly chronological story.

Step 6: Write it

A literature review usually has:

1. An introduction that states the topic, why it matters, the scope of the review and how it is organised.
2. A body of sections that synthesise the research theme by theme. Compare and contrast sources, explain agreements and disagreements, and evaluate the quality of the evidence.
3. A conclusion that summarises what is known, identifies the gaps, and leads into your research question or methodology.

The difference between summarising and synthesising looks like this:

• Summary: Suri and Jack (2016) found that mobile money reduced poverty in Kenya. Aker and Mbiti (2010) studied mobile phones in Africa.
• Synthesis: Evidence from Kenya suggests that mobile money reduces poverty, particularly for female-headed households (Suri and Jack, 2016), building on earlier findings that mobile phones lower the cost of information and trade across Africa (Aker and Mbiti, 2010).

Step 7: Cite everything correctly

Every idea you take from a source needs an in-text citation, and every citation needs a matching entry in your reference list. Before you submit:

• run your list through the citation checker to confirm every reference exists and its details are right,
• use the in-text citation checker to make sure every citation matches the list, and nothing in the list is uncited.

Common mistakes

• Listing summaries instead of comparing sources.
• Relying on old sources with nothing from the last few years.
• Leaving out the seminal papers of your field.
• Too many direct quotations. Paraphrase and cite; quote only when the exact words matter.
• Ignoring evidence that contradicts your argument. Discussing it makes your review stronger.
• No clear link to your own research. The review should end by showing the gap your work addresses.