Online Tool Store
· 4 min read
How to Organise Sources for a Literature Review
Heshan Fernando
Co-founder & COO
The most common failure in a literature review isn’t missing sources. It’s a review that reads: “Ahmed (2019) found X. Berg (2020) found Y. Chen (2021) found Z.” Twenty paragraphs, one per paper, no argument.
That structure comes directly from how the notes were taken — one document per paper — and it’s very hard to escape at the writing stage if the reading stage produced it.
What a synthesis matrix changes
A matrix flips the organisation. Sources go down the side, themes go across the top, and each cell holds what that source says about that theme.
The consequence is that you write from columns rather than rows. A column is one theme with eight sources in it, which forces a sentence like “measurement approaches vary considerably: six studies used self-report, while two used observational coding” — a synthesis, with several citations, making a point.
You can’t easily write that from per-paper notes. You can barely avoid writing it from a matrix.
The other thing the grid makes visible is absence. A theme column with two entries out of fourteen sources is a gap, and gaps are findings. Most “further research is needed” sentences in published reviews came from someone noticing an empty column.
Why people get stuck here
- Notes organised by paper. Which produces a review organised by paper.
- Themes chosen too early. Themes set from your proposal rather than from the reading tend not to match what the literature actually discusses.
- Too many themes. Fifteen columns each with two entries isn’t a matrix, it’s a spreadsheet of quotes.
- Re-reading. Without a structured record, you’ll open the same paper four times looking for the thing you know it said.
What a good matrix looks like
Themes drawn from the reading
Start with a rough set, then revise them after ten papers. The themes that actually recur are rarely exactly the ones you predicted.
Method recorded alongside findings
A finding from a sample of 30 and a finding from a sample of 3,000 aren’t equivalent. Keeping the method in the row means the synthesis can say so.
Cells with your words, plus the quote
Paraphrase for synthesis; keep the exact quote for when you need to cite precisely. Both in the cell saves going back to the PDF.
| Organised By | You Write | Reads As |
|---|---|---|
| Source (rows) | One paragraph per paper | An annotated bibliography |
| Theme (columns) | One paragraph per theme | A synthesis |
| Neither | From memory | Vague, hard to cite |
Common mistakes to avoid
- Building the matrix after reading everything, from memory.
- Filling cells with quotes only, leaving the interpretation to be done later under deadline.
- Adding a theme for every interesting idea, until no column has enough entries to say anything.
- Ignoring an empty column instead of naming it as a gap.
- Recording findings without sample sizes or methods, so contradictions can’t be explained.
How to do it with Literature Matrix Builder
The Literature Matrix Builder lays out the grid so the theme columns are what you read.
- List your sources with author, year and method.
- Name the themes you keep noticing across them — from the reading, not the proposal.
- Fill the cells with your paraphrase plus any quote you’ll need.
- Read down each column. Full columns are where you can synthesise; empty ones are gaps worth naming.
- Write one paragraph per column, not one per source.
Everything stays in your browser. Other research and writing tools are in the tools directory.
Frequently asked questions
Why a matrix rather than notes per paper?
Because per-paper notes produce a review that lists summaries. Reading down a theme column forces you to write about the theme citing several papers at once, which is what synthesis means.
What if a theme has only one source?
Either it’s a genuine gap worth naming, or it isn’t really a theme. Both are useful conclusions, and both belong in the discussion rather than being quietly dropped.
How many sources before a matrix is worth it?
Around ten. Below that you can hold the comparisons in your head. Above it, the matrix is what stops you re-reading the same paper repeatedly.
Final thought
Build the matrix while you read, not after. The gap between “I’ve read forty papers” and “I can say what they collectively found” is exactly the work the grid does for you.