Using a Synthesis Matrix to Organise a Literature Review

A literature review should show relationships among studies rather than give each article its own disconnected paragraph. A synthesis matrix is a working table that places sources in rows and themes in columns. It helps you spot agreements, differences, limits and open questions before you draft. The table is a planning tool; the submitted review may be ordinary prose. Use your research question to choose themes, because a matrix built around vague labels such as “important” will not tell you what to write.

Create columns that reflect the question

Start with a brief source identifier, method and context, then add two to four substantive themes. For a review of peer feedback in writing courses, columns might include type of feedback, revision behaviour, learning outcome and setting. Leave space for caveats. A source can contribute to several themes, but record only what it actually studies. Keep your notes short enough to compare across rows without losing the nuance of the original paper.

Record findings and disagreement precisely

In each cell, paraphrase the relevant result and include a page or section pointer. Do not force agreement by writing “supports theme” for every paper. One study may describe students liking feedback while another measures whether revisions improve; those are different outcomes. When authors disagree, examine sample, intervention, timing and measurement. Mark cells where evidence is absent rather than treating silence as a negative finding.

Use the matrix to build paragraphs

Choose a theme and write a topic sentence that states the pattern you see. Bring two or more sources into the paragraph, explain their relationship and end by connecting the pattern to your research question. A paragraph can show that findings agree in one context but diverge in another. This moves your voice from reporting who said what to explaining what the evidence means. Cite each study at the point where its finding appears.

Identify a defensible gap

A gap is not automatically “no research exists.” It may be an understudied population, inconsistent definitions, limited follow-up or a method that has not answered a particular question. Search carefully before claiming novelty. If the literature is strong, your project may still have value by testing transfer to a different context. The matrix should reveal which claims are well supported and which remain tentative.

Keep the table alive during drafting

As you read new sources, update the matrix and your outline together. Remove a theme that no longer helps answer the question. Keep citation details in a reference manager or separate source list so abbreviations in the matrix do not become unverifiable. Before submission, compare each review section with the table to check that contrary evidence was not quietly omitted.

Worked example

If three papers report peer feedback benefits but measure satisfaction, writing quality and confidence separately, a matrix makes the difference obvious. A review paragraph can explain that students often value feedback, while the evidence about writing quality depends on how revisions were assessed. Calling all three studies proof of “better grades” would misrepresent them. The matrix protects against that shortcut.

A four-source matrix in practice

Draw columns for context, method, definition of feedback, revision outcome and caveat. Add four papers as rows. One may define feedback as comments from peers, another as automated suggestions; treating both simply as “feedback” hides the mechanism. Mark the two papers that measure revisions directly and the two that measure attitudes. Now try a paragraph beginning “Across these studies, feedback is associated with…” and finish it without making an unsupported outcome claim. If the paragraph cannot be written accurately, your theme may be too broad. Split it into feedback uptake and revision quality. This step converts a planning grid into a useful outline. Keep source pointers in cells so you can check exact findings before publication, and revisit the matrix whenever a new paper changes the pattern.

Checklist before you submit

Use question-led theme columns.
Record method, setting and actual outcome for each study.
Draft paragraphs around relationships across sources.
Check any claimed research gap against the wider search.

Questions students ask

How many sources go in one theme?

There is no fixed number. Include enough relevant studies to explain the pattern and important exceptions.

Should the matrix appear in my submission?

Usually it is a private planning tool unless the assignment asks for an appendix or evidence table.