How to write a literature review that actually synthesizes the field

A literature review is not a sequence of paper summaries. Its job is to show what is known, how it is known, where findings disagree, what methods dominate and what gap or unresolved question remains.

1. Define the review question

Start with a scope narrow enough to search and compare meaningfully. Define the phenomenon, population or system, variables or concepts, relevant methods, time range and disciplinary boundaries where needed.

2. Build a reproducible search strategy

List the main concepts and synonyms, then combine them into search terms suitable for the databases or indexes used in your field. Record which databases, date ranges and filters you used so the search can be explained later.

3. Screen sources against explicit criteria

Decide what counts as relevant before reading everything in depth. Inclusion and exclusion criteria may involve publication type, methodology, population, language, date, theoretical framework or direct relevance to the research question.

4. Extract comparable information

For each source, capture more than the conclusion. Record the question, data, method, assumptions, sample or system, key result, limitations and how the paper relates to the rest of the literature.

5. Group by theme, claim or method

Do not organize the final review as “Paper A says…, Paper B says…”. Organize around the structure of the field: competing explanations, methodological families, chronological shifts, theoretical frameworks or areas of agreement and disagreement.

6. Synthesize instead of stacking summaries

Synthesis asks questions such as: Do several studies converge despite different methods? Are conflicting findings explained by different samples or assumptions? Is a popular conclusion supported by direct evidence or repeated citation? Where is uncertainty concentrated?

7. End with the gap that follows from the evidence

A research gap should not be invented merely because “few studies exist.” Explain why the unresolved issue matters, what current work cannot determine and what kind of evidence or method would improve the situation.

Using AI in a literature review

AI can help create a taxonomy of themes, compare notes, challenge a synthesis or detect contradictions in material you provide. It can also hallucinate citations or misrepresent papers. Verify references, quotations, methods and claims directly against the source material before relying on them.

How XYSHIMA Uni can support the process

XYSHIMA Uni can keep papers, PDFs, notes, questions, claims and writing progress connected to the relevant course or project. Literature reviews remain part of the normal academic project workflow.