Search website

Quick links

↑ ↓ to select · Enter to open · Esc to close

Blog

Literature Synthesis Matrix: How to Build One That Works

Updated · 14 min read

Literature Synthesis Matrix: How to Build One That Works

Listen to this post

0:000:00

Share this post

XLinkedInReddit
Contents

A literature synthesis matrix is a table that puts your sources in rows and your themes in columns, so you can compare what every author says about the same idea side by side. You fill each cell with what that source says about that theme, then read down the columns to see agreement, disagreement and gaps. This guide shows you how to set one up, what to put in the cells, and how to turn the finished grid into paragraphs.

TL;DR

  • A synthesis matrix has one row per source and one column per theme or analytical question. The insight comes from reading down a column, not across a row.
  • Draft the columns from your research question before you read everything, so you only extract what your argument needs.
  • Paraphrase instead of quoting, and keep a citation, page number or DOI in every row so you can trace each cell back to the source.
  • The matrix is scaffolding for your writing, not a replacement for it.
  • Conflicts between sources are the most valuable cells. Mark them and build paragraphs around them.

Table of Contents

What a literature synthesis matrix is

Academic librarians describe the matrix as a chart for recording the main points of each source and documenting how sources relate to each other. Johns Hopkins University's library guide frames it that way, and most university guides agree on the basic layout: sources down the side, ideas across the top.

You may also see it called a literature review matrix, a literature synthesis table, or the matrix method. They are the same idea with slightly different column choices. The matrix method of literature review is documented in the academic literature and is used in coursework as a way to organize a body of reading before writing.

It is usually built in a spreadsheet such as Excel or Google Sheets. A Word table works for small projects. What matters is that you can sort, filter and scan it.

What a literature synthesis matrix is — overview diagram

Matrix versus reading notes

Reading notes follow the order in which you read. A matrix follows the order of your argument. Notes tell you what Author A said. A matrix tells you what Authors A, B and C said about the same question, and where they split.

That difference is why instructors push the matrix so hard. Stanford's TeachingWriting page describes it as an activity that prepares students to write a literature review by relating sources to each other according to themes, topics, keywords or codes. It is a cure for the common habit of summarizing one article after another, the so-called patchwork review.

Before you build: decide what you are asking

The step most people skip is the first one. Start from your research question, then define the column headers that question needs. The Oklahoma State open textbook on writing for publication recommends exactly this order: question first, columns second, rows filled as you read.

Why it matters: if you build columns after reading forty papers, you will have forty summaries and no structure. If you sketch a draft matrix first, you read with a purpose. A guide to scoping reviews from Birmingham City University's open-access repository makes a related point for professional work: plan your data extraction form in advance so it is unambiguous and structured, and you will not waste time pulling out information that never supports your final argument.

Your first draft of the columns will be wrong, and that is fine. Read three or four sources, see which columns stay empty or keep needing a note, and revise.

Question-first sequence for drafting a synthesis matrix

Choosing your columns

Think in two groups: descriptive columns that identify the source, and thematic columns that carry the synthesis.

Descriptive columns

These are the columns most guides agree on:

  • Citation (author, year, title, and a DOI or other stable identifier)
  • Aim or purpose, in one or two sentences
  • Methodology and sample or population
  • Key findings
  • Relevance to your own study

The Toronto Metropolitan University graduate guide suggests starting with a citation column and a one-to-two-sentence purpose column, then adding what your review needs, such as methodology, population, geography, intervention or outcomes. It also suggests ordering sources from oldest to newest so you can see how research on the topic has changed over time.

Thematic columns

These are the heart of the matrix. Each one is a concept, debate or variable that appears across several sources. If you are reviewing remote-learning research, for example, your themes might be student engagement, teacher training, and access to technology. Add a last column for gaps, problems, questions and notes. That is where your own thinking goes.

Aim for three to six themes at first. Fewer than three and you have a list. More than six and you probably have two reviews hiding in one.

A worked example

Here is a small matrix for a made-up review question: How does peer feedback affect writing quality in first-year college courses? The sources below are placeholders to show the layout, not real studies.

Source Method and sample Theme 1: Quality of feedback Theme 2: Student uptake Theme 3: Instructor role Gaps and notes
Author A (2019) Survey, 210 students Students rate peer comments as vague Low revision rates Not addressed Single institution
Author B (2021) Experiment, 2 sections Trained reviewers give specific comments Higher revision rates after training Instructor models feedback Small sample
Author C (2023) Interviews, 18 students Quality varies by relationship Students revise when they trust the reviewer Instructor sets tone Conflicts with A on uptake

Now read down the "Student uptake" column. A reports low revision, B reports higher revision after training, and C ties it to trust. That is already a paragraph: uptake seems to depend on training and trust, and A may have measured an untrained group. You could not see that by reading three separate summaries.

The "Gaps and notes" column flagged the conflict between A and C. Do not smooth that over. Academic librarians stress that a literature review is a conversation between authors, and the University of Georgia lecture on the synthesis matrix treats the matrix as the visual aid for that dialogue. Disagreement is often the most useful thing in the grid, because it points to the debate you are meant to address.

How to build the matrix, step by step

  1. Write your research question. One sentence. If you cannot, your columns will drift.
  2. Draft the columns. Use the descriptive set above plus three to five themes you expect to matter.
  3. Test it on three sources. Read them closely and fill the rows. Adjust any column that is vague or always empty.
  4. Fill rows as you read. Paraphrase rather than quote. Paraphrasing forces you to understand the point, and it keeps your later draft from sounding like a collage of quotations.
  5. Record where each claim lives. Add page numbers or a DOI so any cell can be traced back.
  6. Read down each column. Ask which sources agree, which disagree, which method each used, and what nobody has studied.
  7. Sort and group. Reorder rows by theme, date or method. Merge columns that overlap and split columns that hide a disagreement.
  8. Outline from the columns. Each thematic column becomes a section or a cluster of paragraphs.

From matrix to paragraphs

Many students believe the finished matrix is the review. It is not. It is the blueprint. The Stanford guide describes the goal as moving from source-by-source summary to theme-by-theme analysis, and that move happens in your prose.

A reliable paragraph pattern for one thematic column:

  1. State the claim the column supports in your own words.
  2. Cite the sources that agree, grouped in one sentence.
  3. Introduce the source that disagrees or qualifies, and say why they may differ (method, sample, date).
  4. Close with what remains unknown, drawn from your gaps column.

If you want a checklist for the analysis itself, Texas Tech's Graduate Writing Center offers the 5 Cs: cite, compare, contrast, critique and connect. Each one maps to a matrix move. Cite is your citation column. Compare and contrast are what you do when reading down a column. Critique lives in your methodology and limitations columns. Connect is your relevance column, where you link findings to your own project.

Professional and systematic work: a stricter matrix

If you are doing a scoping review, a systematic review, or research for a client, the matrix turns into a data extraction form, and the standards go up.

  • Plan it like a protocol. Decide the fields before extraction starts and define each one so that two people would fill it the same way.
  • Expect duplicate reports. The same study can appear in several publications, sometimes with conflicting figures. A Cochrane-style chapter on collecting data notes that multiple reports of a study need to be identified and handled together. Decide in advance which report takes precedence.
  • Check a sample. Single-reviewer extraction is common for quick projects, but for high-stakes work have a second person verify at least a random sample, or extract in duplicate.
  • Keep an audit trail. Include a DOI, PMID or equivalent in every row so the review stays transparent and reproducible.

The audit-trail habit carries over to other kinds of evidence. If your sources are corporate disclosures rather than journal articles, the same discipline applies: every claim in a cell should point back to a document. A walkthrough of traceable textual analysis of SEC filings shows how that looks when the source material is public company records.

Common mistakes

Building one row of long summaries. If a cell holds a paragraph, you are summarizing, not synthesizing. Keep cells to a phrase or two.

Too many columns. Every column is a promise to extract that information from every source. Cut any column you would not use in your argument.

Ignoring the empty cells. A blank cell can be a finding. If nobody addresses a theme, that is a gap you can name.

Quoting instead of paraphrasing. Direct quotes are fine for a few key phrases, but long quoted cells defeat the purpose and tempt you toward the patchwork review.

Treating the matrix as finished. It should change as your understanding does. Rename themes, split columns, and re-sort rows.

Where the matrix breaks down: the reading itself

The matrix is only as good as your reading. The slow part is rarely the table. It is understanding a dense paper well enough to state its finding in one line, and remembering where in the paper you saw it. Skimming to fill cells quickly gives you cells you cannot trust.

Our perspective

We built Omphalis around a simple belief: a table of summaries is no substitute for understanding. A matrix works when each cell reflects a source you actually grasped, and it fails when the cells are filled with fluent-sounding text you cannot defend.

That is why we keep you with the full source. In Omphalis you can read a paper in a cleaned, structured view, get an explanation of a difficult passage without leaving the page, and attach your own note to the exact original passage. When you later fill a matrix cell, the note and the text it came from sit together, so you can check your paraphrase against the author's words.

We also think questions should come with evidence. When you ask about a source in Omphalis, the answers come with citations from that source, so a claim in your "Key findings" column has somewhere to point. Our approach favors comprehension over speed, because a fast summary that you do not understand costs you more time at the writing stage.

Our practical advice: build the matrix in the tool you already like, whether that is a spreadsheet or a table in your writing software, and use your reading environment to make sure each row is something you could explain to a classmate.

— Omphalis Team

Keeping your reading and your matrix connected

One workflow that holds up: save each source, read it once for the argument, then return with your matrix open and annotate the passages that answer each column. Our research papers page shows how reading and annotating papers works, and the ask page explains questions answered with citations from the source. If you want a structure for the notes that feed your table, our guide to research notebook templates walks from reading notes to reproducible work, and the reading journal templates help with long reads.

Omphalis accepts PDFs, EPUBs, Word documents and feeds, and offers export and notebook organization, so your notes are not locked in one place.

Try it with your own sources

There is a free plan, $0, with 100 comprehension units a month, and new accounts start with a trial. 30-day trial, no card required. Includes 600 comprehension units, 200 Ask voice minutes and 15 assistant answers a day. Paid plans on our pricing page add capacity and features: Pro is 9 USD per month (300 comprehension units) or 90 USD per year, Premium is 19 USD per month (600 units) or 190 USD per year, and Scholar is 39 USD per month (1,200 units) or 390 USD per year. Unit packs start at 10 USD for 200 units. PDFs are read on Pro and above and during the trial; on Free a PDF is kept with its name and page count but not read. If you are managing a stack of papers for a thesis or a review, start at omphalis.ai and bring one source to see whether the annotated reading changes how you fill your first row.

FAQ

What is the difference between a literature matrix and a synthesis matrix?

In practice, the terms overlap. A literature matrix often leans toward recording details about each study, such as method, sample and findings. A synthesis matrix emphasizes themes across sources and how they relate. Most useful matrices combine both: descriptive columns first, thematic columns after.

How many sources do I need for a synthesis matrix?

There is no fixed number. A short course assignment might use eight to fifteen sources, while a thesis or systematic review can involve far more. What matters is that your columns come from your research question and that you can compare at least a few sources on each theme.

Should I use Excel, Google Sheets, or a Word table?

Use whatever lets you sort and filter comfortably. Spreadsheets are the common choice because you can reorder rows and hide columns. A Word table is fine for a small matrix you will paste into an appendix.

What are the 5 Cs of a literature review?

They are cite, compare, contrast, critique and connect, as set out in Texas Tech's graduate writing guide. Use them as prompts while you read down your matrix columns.

Sources