Literature Reviews That Synthesise, Not List

The Purpose of a Review

A literature review sets out what is already known, how that knowledge was gained, where the evidence disagrees and what is still unresolved. It makes an argument about a body of work and is not an inventory of it.

Here is a simple test: should your review lose nothing when re-sorted alphabetically, it is a list and not a synthesis.

Types of Review

Planning the Search

1. Word the review question precisely so that readers can decide whether any particular paper is relevant.
2. Write down the concepts in your question along with the alternative terms for each one.
3. Select the databases in which your field really publishes, and search more than one.
4. Construct the search from your concept terms, joining synonyms with OR and concepts with AND.
5. Note down the precise search string and when you ran it.
6. Trace the reference lists of key papers both backwards and forwards.

Screening and Selection

Settle your inclusion and exclusion criteria prior to screening, not during it. They usually address population, method, setting, time period, language and type of publication.

Screen in two rounds: begin with titles and abstracts, then read the full text of those that remain. Note how many papers were dropped at each stage and the reason. These figures let readers assess whether your review is complete. A simple spreadsheet is usually enough to keep these records tidy.

Recording Your Method

A narrative review, too, gains from a brief statement of method covering the databases, search terms, date range and exclusions. Lacking it, readers cannot tell if you overlooked contrary evidence or left it out deliberately.

Start using a reference manager with your first search instead of afterwards. Remove duplicates early and note the reason each excluded paper was left out, because you will not recall it three months later.

Organising the Material

Arrange the review by theme, method, chronology or theoretical stance, whichever best serves your argument, but never paper by paper. A section heading ought to name an idea and not an author.

One easy technique is to create a matrix with papers as rows and the dimensions you care about (question, method, sample, finding, limitation) as columns. Patterns and disagreements show up at once, and the columns frequently turn into section headings.

Writing the Synthesis

Every paragraph should assert something about the literature and back it with several sources, for example: "Three studies with longitudinal designs find X, whereas cross-sectional research tends to find Y, a gap that may be explained by Z."

When studies disagree, state it and suggest why — different populations, measures or time periods. A review justifies itself by resolving, or at the very least describing, such disagreement.

Cite the original source rather than a secondhand account of it. Errors spread through a field when a paper is known only from another author's summary. When the original is unavailable, say plainly that you are relying on a secondary account.

Identifying the Gap

A gap is more than just something missing. "Nobody has studied this in my city" matters only when there is cause to expect a different result there. Explain why the gap is important and what closing it would change.

Pitfalls

• Describing one paper per paragraph with no argument linking them
• Quoting only studies that back your view
• Depending on abstracts and not full texts
• Overlooking earlier foundational studies, or overlooking the past two years
• Not stating the search method, so completeness cannot be judged
• Reusing wording from sources — refer to the plagiarism guidelines

At-a-Glance Checklist

1. Question setFocused enough to be answerable
2. Search loggedTerms, databases, dates and counts
3. Criteria documentedInclusion and exclusion, fixed before screening
4. Synthesis rather than a listCompare, contrast and link
5. Method statedSo the review can be replicated

Questions Researchers Often Ask

More Researcher Resources

This guide sits in the JMVRA Researcher Resources series and gives pointers that readers can verify on their own, without sales talk.

Browse all resources

Related Reading

Further reading on the same theme from the JMVRA blog.