The literature review matrix: your secret organiser

Reading papers one after another feels productive, and each paper makes sense on its own. The trouble starts when you need to compare them, because your notes live inside seventy separate files, as highlights and margin comments that can never sit side by side.

Memory does not help much either. By the twentieth paper, the details of the fifth have blurred into the ninth. You remember that someone found a strong effect in Grade 2, but not who, with how many children, or in which state.

Your PDFs are not your notes. Your notes need a place where papers can sit next to each other.

A literature review is about comparison: what studies agree on, where they differ, and what nobody has looked at yet. So your notes need a shape that makes comparison easy.

Literature review matrixalso called a synthesis matrix

A table with one row for each source and one column for each question you ask of every source, such as its aim, method, sample, findings and limitations.

The idea is simple, and that is its strength. Reading across a row gives you one paper at a glance. Reading down a column shows how every paper answered the same question, and that second view is where synthesis happens, because patterns only appear when papers are placed next to each other.

In health research the same idea is often called an evidence table, and systematic reviews use a stricter version of it for data extraction. For a thesis chapter or a first review, a plain spreadsheet is enough. Our guide to writing a literature review shows where the matrix fits into the whole process.

Columns are questions. Choose them from your research question before you open the next paper, so that you read with a purpose. Six to ten columns are usually enough to start, and you can add one later if you find yourself asking the same new question of every paper.

Two practical columns are worth adding in any field. One records where you found the paper, such as a database search or another paper's reference list, which helps when your guide or an examiner asks how you searched. The other holds the file name or DOI, so you can reopen the right PDF in seconds instead of hunting through the folder.

The matrix only works if you fill each row while the paper is fresh, not at the end of the month. Meera set herself a routine for every paper.

1. Read the abstract and the conclusion first.Decide whether the paper belongs in your review at all. If it does not, note why in one line and move on.
2. Skim the method and the results.Look for the design, the sample and the main numbers. Our guide on how to read a research paper shows how to do this quickly.
3. Fill the row in your own words.Short phrases are fine. Writing in your own words forces you to understand the paper, and it protects you from copying by accident later.
4. Mark anything you copy.If you keep an exact phrase, put it in quotation marks and add the page number.
5. Add one line of your own.A doubt, a link to another paper, or a question for your guide.

When a paper does not report something, write "not reported" instead of leaving the cell blank. A blank cell looks like work you forgot. "Not reported" is information, and a column full of it is often the first sign of a gap.

Meera filled five rows a day. In two weeks all seventy papers were in the table.

Once the rows are filled, the real work begins. Reading across a row reminds you what one paper did. Reading down a column tells you what the whole field has done. When you read down a column, ask:

• Where do the studies agree?
• Where do they disagree, and is there a reason, such as different ages or different tests?
• Which methods keep repeating, and which are missing?
• Which places, groups or years appear again and again, and which never appear?
• How has the picture changed over time?

Colour helps here. Shade the cells of studies that agree in one colour and those that disagree in another, or sort the table by year to see how the findings shift over time.

When Meera read down her Sample column, she saw that almost every study came from urban schools. Rural classrooms, where many children speak a language at home that is different from the one used in school, barely appeared. That empty stretch of the column was a research gap, and it became the centre of her thesis.

A good literature review is organised by ideas, not by papers. The matrix makes this easy, because each theme or column can become a paragraph.

1. Pick a theme or a column.For example, everything your matrix says about reading gains in the early grades.
2. Write the pattern first.Say in your own words what the studies show together.
3. Support it with several studies.Cite two or three that agree, and say how strong their evidence is.
4. Add the contrast.Bring in the studies that disagree and explain why they might differ.
5. End with what it means.Point to the open question or gap that follows.

Compare that with a paragraph that summarises one paper after another. The first builds an argument; the second is a list. Meera wrote her first paragraph like this in twenty minutes, straight from two columns of her matrix.

You do not need special software. A spreadsheet in Excel or Google Sheets works well: freeze the top row, turn on text wrapping, and use filters to show one theme at a time. Some people keep the matrix in a notes app such as Notion or Obsidian, or use tags and notes inside a reference manager like Zotero.

Research tools can also build a first draft for you. Fiberarticle's Literature Reviewer reads each paper it selects and fills an evidence matrix with the contribution, methodology, models, dataset, tools, metrics, results, limitations, unresolved problems, assumptions, missing evaluations and opportunities, and it tells you whether it read the full text or only the abstract. Its Extract feature lets you set your own columns, up to twenty across up to fifty papers, with a short supporting quote from the paper for every cell. Both export as CSV, so the table opens straight in your spreadsheet.

Treat any automatic draft as a starting point. Check the cells against the papers, especially where only the abstract was available, and keep a column for your own comments, because the thinking in a review is still yours.

Meera's matrix now has eighty-four rows. At her next review meeting Professor Rathore asked about standardised reading tests again. She filtered one column and answered in ten seconds.