A folder containing thirty PDFs can still leave you staring at a blank literature review. You remember each paper separately, but cannot see what they say together. A literature review matrix solves that organisational problem by putting comparable evidence in one place.
This guide provides a copyable layout and an original worked example. It complements our full literature-review writing guide: the focus here is the working table between reading and drafting, not the whole review structure.
What is a literature review matrix?
A literature review matrix is a working table that records comparable information from several sources, such as their questions, methods, findings and limitations. It helps you identify patterns and disagreements across the reading. The matrix supports your analysis; it is not usually a replacement for the written review your assignment requires.
Monash University’s literature-review guidance distinguishes recording individual studies from organising evidence by theme. That distinction matters: a source-by-source table can help you collect notes, while a thematic view helps you develop an argument rather than a sequence of summaries.
Choose columns that answer your question
Begin with the question at the top of the sheet. Then create the columns below. This is our suggested template, not a mandatory university format. Use short notes in cells and retain a link to the original source for every entry.
| Field | What to record | Why it helps |
|---|---|---|
| Source and location | Reference, stable link, page or section | Lets you verify the note later |
| Question and context | What was investigated, where and with whom | Reveals whether studies are comparable |
| Design and measures | How the evidence was produced and what was measured | Stops unlike outcomes being treated as identical |
| Relevant finding | Your concise paraphrase of the result you need | Connects reading to the assignment |
| Limits for this review | A specific reason to qualify your use of the finding | Keeps claims proportionate to evidence |
| Theme and relationship | Supports, challenges or qualifies another source | Starts the synthesis before drafting |
Do not create a column simply because a template includes it. For a theoretical review, concepts and assumptions may matter more than sample size. For an empirical review, separate the study design from the outcome measure instead of hiding both under “method”.
Fill one row carefully before processing the pile
Test the layout on one relevant paper. Can you complete a row without copying its abstract? Can you explain why the source belongs in your review? If not, either the columns are too vague or you need to read more closely.
Use “not reported” when a detail is absent. Use “not yet checked” when you have not investigated it. Those labels mean different things. Do not silently infer a participant characteristic, research method or conclusion because it would make the table easier to complete.
The University of Leeds’ critical-reading guidance asks readers to test an argument’s evidence and reasoning. In the matrix, capture that judgement as a specific note: “measures intention, not observed behaviour” is more informative than “limited reliability”.
A worked example: compare before you conclude
Illustrative example only: the three records below are invented teaching scenarios, not published studies. Do not cite them as research. Imagine a review asking how remote onboarding supports informal learning among new employees.
- Example A: interviews describe uncertainty about when newcomers can interrupt colleagues. Useful theme: access to informal help. Limit: perceptions alone do not establish learning outcomes.
- Example B: a short survey records satisfaction with scheduled buddy meetings. Useful theme: structured support. Limit: satisfaction is not the same as demonstrated competence.
- Example C: observation notes show newcomers receiving quick answers in a team chat, but some questions remain unanswered. Useful theme: responsiveness. Limit: the observation captures one team context.
A summary-led draft would give A, B and C one paragraph each. A synthesis-led plan asks what links them. All three concern access to help, but they observe different parts of that process. B does not refute A merely because some respondents like buddy meetings; a scheduled meeting and spontaneous access are different conditions.
A defensible provisional interpretation would be: “These examples suggest that availability of support and access at the moment of need should be examined separately.” A claim that remote onboarding improves performance would go beyond these invented records. In a real review, replace every example with verified evidence and citations.
Turn the table into paragraph plans
Create a second view organised around emerging themes. For each theme, write four short lines: the question, sources that inform it, the relationship between their findings, and the conclusion you can currently defend. Keep a fifth line for a missing perspective that needs further searching.
For the example above, possible themes are “timing of help”, “formal versus informal access” and “how learning is measured”. These headings express analytical issues. Headings such as “Article A” and “Article B” only reproduce the filing system.
Use a provisional paragraph sequence: make a focused claim, compare relevant evidence, explain an important difference, and state the implication for the review question. Do not force every paragraph to contain the same number of papers. Some claims need a close examination of one influential study; others require a wider comparison.
Use gaps as questions, not automatic discoveries
An empty matrix column means you have not recorded evidence there. It does not prove that nobody has studied the issue. Your database choice, keywords or inclusion decisions may explain the absence.
Before describing a research gap, revisit the search. Sydney’s search-strategy guidance treats searching as a process of testing and adjustment. Record what you searched and qualify your conclusion: “The sources included in this review provide limited evidence about…” is narrower than claiming the entire field is silent.
Keep the matrix useful as the draft changes
Add a status field: read, needs checking, used in draft or excluded. Record an exclusion reason so you do not repeatedly reopen the same unsuitable article. Keep direct quotations visibly separate from paraphrases and retain page references even if your citation style does not always require them.
When your argument changes, update the “use in review” notes. Otherwise the matrix becomes a record of yesterday’s plan. Before submission, sample a few draft claims and trace each one back through the matrix to its original source. That check is especially useful after moving paragraphs between sections.
Frequently asked questions
Is a matrix the same as an annotated bibliography?
No. An annotated bibliography usually presents a separate summary or evaluation for each entry. A matrix places comparable features alongside one another to support cross-source analysis. Your course may require either or both.
Do I submit the matrix?
Only if the brief or supervisor requests it. It may be a planning document, appendix or assessed component. Check the requirements before spending time formatting a working sheet for submission.
Can AI fill the matrix for me?
Check your assessment rules first. Even where assistance is permitted, verify every extracted detail against the source and never invent missing findings. Keep a record of permitted assistance and follow the required AI-disclosure process.
Need help understanding the next step?
Ask about tutoring or feedback on your own draft. Tell us what your course permits; you remain responsible for the research, reasoning and submitted work.
Discuss study support