A literature review matrix is a table for comparing sources around a research question. Each row records a study; the columns capture the details you need to compare, such as the method, sample, findings and limitations. Its value comes from the patterns you can see across rows, not the number of papers you enter.
You can build one in a spreadsheet, a document or a notebook. Start with a few relevant papers and a small set of columns. Add detail only when it helps answer your review question. The blank structure and fictional worked example below show how to move from notes to synthesis. When the matrix is ready but the prose is not, use how to write a literature review to turn those comparisons into themes, claims and paragraphs.
What a literature review matrix helps you do
A reference manager helps you store and cite sources. A matrix helps you reason about their relationships. You might use both: keep the full bibliographic record in your reference manager and use the same short source identifier in your comparison table.
The University of Sheffield’s guidance on synthesising research explains how comparing themes, methods and disagreements helps build an argument across sources. A matrix supports that process; it does not replace reading the original studies or deciding which evidence is appropriate.
A copyable literature review matrix template
Copy these headings into a spreadsheet or document. Use one row per study, or clearly identify separate experiments within a paper if they need different entries. Keep direct quotations visibly distinct from your own paraphrases.
| Column | What to record |
|---|---|
| Source ID and citation | Author, year, title and a link or reference-manager key |
| Research question | What the study actually asks, in your own words |
| Context and sample | Who or what was studied, where and at what scale |
| Method | Design, data collection and analysis relevant to your comparison |
| Relevant finding | The result that addresses your review question, with page or section |
| Limitations | Constraints reported by the authors and your separately labelled observations |
| Theme or comparison | The issue this source helps you discuss across studies |
| Use in my review | What claim it may support, challenge or qualify; what to check next |
Blank row to copy
Source ID / citation:
Research question:
Context and sample:
Method:
Relevant finding and page/section:
Author-reported limitations:
My critical notes:
Theme or comparison:
Possible use in my review:
Unresolved question:For a wide spreadsheet, freeze the source column and wrap cell text. If the table becomes hard to scan, move extended notes into a linked document. Aim for enough detail to return to the evidence, not a second copy of every article.
Worked example: comparing study-planning approaches
The three studies below are invented teaching examples. Their labels, samples and findings are fictional and must not be cited as research. The example question is: “How do weekly planning activities relate to students’ completion of intended study tasks?”
| Fictional source | Method and context | Illustrative finding | Limitation and use |
|---|---|---|---|
| Example A | One-week survey of 40 volunteers in one seminar | Students who reported making plans also reported completing more intended tasks | Self-report and self-selection; cannot establish that planning caused completion |
| Example B | Interviews with 12 students who combine study with paid work | Participants described changing plans when work shifts changed | Explores experiences rather than estimating an average effect; adds context about flexibility |
| Example C | Four-week comparison of two existing tutorial groups using different planning activities | The group using weekly reviews recorded fewer unfinished tasks | Groups were not randomly assigned; baseline differences and recording practices could explain the result |
Notice that these entries do not all answer the question in the same way. A survey association, an interview account and a non-randomised comparison provide different kinds of evidence. The matrix should keep those differences visible rather than flattening them into “planning works.”
Turn the rows into a synthesis paragraph
First choose a theme that crosses more than one source. In this example, the theme might be the difference between making a plan and adapting it. Then identify where the sources agree, where their methods differ and what you cannot conclude.
Example synthesis using only the fictional studies: The examples suggest that planning and task completion may be related, but they do not establish a simple causal effect. Example A reports an association through student self-reports, while Example C compares existing groups whose differences may affect the result. Example B adds a possible explanation for why a fixed plan may be difficult to follow: participants describe changing work schedules. Taken together, the examples justify examining flexibility and weekly review, while leaving the effect of a particular planning method unresolved.
This paragraph makes a claim across sources and qualifies it. It does not simply repeat A, then B, then C. In a real review, replace the fictional labels with properly cited evidence and check every statement against the originals.
A practical workflow for your first five papers
- Write one provisional review question at the top of the matrix.
- Screen each paper for relevance before making detailed notes.
- Read closely enough to understand the method and the result you plan to use.
- Enter a concise row with page or section references.
- Compare the rows and write two sentences about agreement, differences or missing evidence.
The journal article reading guide helps with the reading stage. If your project is still taking shape, use the research proposal guide to connect the review to a question and feasible study.
When should you change the columns?
Let the research question decide. A methods review may need columns for datasets, measures and validation. A theoretical review may need concepts, assumptions and competing explanations. A qualitative review may need context and interpretive approach rather than a numerical sample comparison alone.
If you are conducting a formal systematic review, follow the required protocol and discipline-specific extraction and appraisal methods. This general study template is not a substitute for those requirements.
Common mistakes to avoid
- Recording only the abstract, then treating it as a full appraisal.
- Mixing an author’s conclusion with your own interpretation without labelling the difference.
- Calling a result “significant” without recording what was measured and how the term is being used.
- Treating a small or context-specific study as a universal answer.
- Creating so many columns that you stop comparing the evidence.
- Collecting papers indefinitely without writing any synthesis.
What to take to your supervisor
Bring a small, readable extract and a decision you need help with: whether the question is too broad, whether an important approach is missing or whether two bodies of work belong in the same review. Use the supervisor meeting agenda template to put that decision near the top of the discussion.
How many papers belong in the matrix?
There is no universal number. It depends on the assignment, scope and type of review. Start with a few relevant sources to test the structure, then expand according to your search and selection method. A large table is not evidence that a review is complete.
Is a synthesis matrix different from a literature review matrix?
The names are used in overlapping ways. Some synthesis matrices put themes in rows and sources in columns; others use one row per study. Choose the orientation that makes the comparisons you need easiest to see. The important step is turning those comparisons into a supported argument.
Featured photo: JESHOOTS.COM on Unsplash.