Literature review matrix template (with example)
Short answer: A literature review matrix is a table with one row per source and one column per feature you want to compare, such as citation, year, setting, aim, design, sample, key variables, findings, limitations and relevance to your question. Filling it in as you read turns scattered notes into comparable data; reading it down the columns (not across the rows) shows the themes, agreements, contradictions and gaps that a synthesis-style literature review is built on. This guide gives a free CSV template, a worked example and the steps from matrix to paragraphs.
On this page
- Recommended columns
- A filled-in example
- Step 1: Fix the scope and columns before you read
- Step 2: Fill one row per source as you read
- Step 3: Sort and read down the columns
- Step 4: Turn themes into an outline
- Step 5: Write synthesis paragraphs from the groups
- Example paragraph built from the matrix
- Spreadsheet or reference manager?
- Common mistakes
- Quick checklist
The hardest part of a literature review is not finding sources; it is comparing them. Notes kept paper by paper lead to reviews that summarize one study after another. A review matrix fixes this by putting every source in the same structure, so you can compare them side by side and write about patterns instead of papers.
The approach is often called the Matrix Method, from Judith Garrard's textbook Health Sciences Literature Review Made Easy, but the idea works in any field.
Download: literature-review-matrix.csv (opens in Excel, Numbers or Google Sheets; one header row and three empty rows).
Recommended columns
Start with a small, fixed set of columns and add topic-specific ones as your question becomes clearer. University library guides based on Garrard's method typically begin with columns that identify the source, its year and its purpose, and leave the remaining columns to whatever you need to compare.
| Column | What to record | Why it helps |
|---|---|---|
| ID | A short number or code | Lets you cite rows in your notes and outline |
| Citation | Full reference in your citation style | Saves reformatting later |
| Year | Publication year | Shows how the evidence developed over time |
| Country / setting | Where and in what context | Reveals settings that have not been studied |
| Research question or aim | One sentence | Lets you group studies by what they asked |
| Theory / framework | Named theory, if any | Shows competing explanations |
| Design | Experiment, survey, panel, interviews, review… | Helps you weigh the strength of evidence |
| Sample | N and population | Shows who findings apply to |
| Key variables or concepts | How the main constructs were defined | Exposes inconsistent definitions |
| Measures / data source | Scales, records, datasets | Explains why results may differ |
| Main findings | Direction and size of key results, in your own words | The core of the synthesis |
| Limitations | Those the authors note and those you see | Feeds your gap statement |
| Relevance | How it relates to your question | Keeps the review focused |
| Theme(s) | One or more labels | Becomes your section structure |
A filled-in example
The rows below illustrate the format with invented, generic entries; your matrix should contain only sources you have actually read.
| ID | Citation | Design | Sample | Main findings | Limitations | Theme |
|---|---|---|---|---|---|---|
| 1 | Author A (2021) | Cross-sectional survey | 812 office workers, one country | Remote-work intensity positively associated with exhaustion | Self-report, single time point | Wellbeing costs |
| 2 | Author B (2022) | Two-wave panel | 430 software employees | Association weaker when employees reported clear work–home boundaries | Two waves only | Boundaries as mechanism |
| 3 | Author C (2023) | Interviews | 24 managers | Managers described boundary-setting as an individual responsibility | Small, purposive sample | Organizational role |
| 4 | Author D (2023) | Field experiment | 1,100 call-centre staff | Hybrid schedules did not change exhaustion over six months | One company | Wellbeing costs |
Reading down the "Main findings" and "Theme" columns already suggests a structure: evidence on wellbeing costs is mixed (rows 1 and 4), boundaries may explain the difference (row 2), and the organization's role is under-studied (row 3).
Step 1: Fix the scope and columns before you read
Write your question at the top of the sheet and choose 8 to 12 columns. Too many columns slow you down; too few leave out the features you will need to compare. You can add a column later, but then go back and fill it for earlier rows.
Step 2: Fill one row per source as you read
Complete the row while the paper is fresh, in your own words. Record findings with their direction and, where available, their size, not just "significant effect". Keep direct quotations in a separate notes column, in quotation marks with page numbers, so you never mistake them for your own wording later.
Step 3: Sort and read down the columns
Sort by year, design, setting or theme. Reading down a column answers synthesis questions: Which designs dominate? Do results differ by setting? Which populations are missing? Which definitions conflict? Mark agreements, contradictions and gaps directly in the sheet.
Step 4: Turn themes into an outline
Group rows by theme. Each theme becomes a section or a run of paragraphs, and each paragraph makes one claim about a group of studies. List the row IDs that support each claim under the heading in your outline.
Step 5: Write synthesis paragraphs from the groups
Write each paragraph from a group of rows, not a single row: a topic sentence that states the pattern, evidence from several studies, an explanation for any disagreement (often visible in the design, sample or measures columns), and a link to your question. Our literature review guide shows this paragraph structure in detail.
Example paragraph built from the matrix
Spreadsheet or reference manager?
A spreadsheet is simplest and easiest to sort. Reference managers such as Zotero let you tag and annotate PDFs, and some researchers export their library to a spreadsheet to build the matrix. Systematic reviews use more formal data-extraction forms, often based on PRISMA reporting guidance; the matrix here is suited to narrative reviews for theses, proposals and papers.
Common mistakes
- Copying abstracts into cells instead of extracting comparable details
- Leaving columns blank and filling them from memory later
- Recording "found an effect" without direction or size
- Writing the review row by row, which recreates a summary list
- Adding sources to the matrix that you have not read in full
Quick checklist
- Is your research question written at the top of the sheet?
- Do the first columns identify the source, year and aim?
- Are findings recorded with direction (and size where possible)?
- Have you sorted and read down the columns to find patterns?
- Does every theme in your outline list the rows that support it?
- Does each paragraph discuss several sources, not one?
- Have you noted the gaps that lead to your own question?
When the matrix is complete, use it to plan your sections with how to turn your notes into a first draft.