Synthesis Matrix: How to Build One for Your Literature Review

What a synthesis matrix is

A synthesis matrix is a grid that helps you compare sources by idea instead of reading them one at a time. Each row is a source, each column is a theme or question from your research, and each cell holds what that source says about that theme. It's one of the most practical tools for a literature review, a research paper or a thesis chapter, because it does the hard part of synthesis on one page.

The point is to change how you write. Without a matrix, literature reviews tend to sound like a list: Source A found this, Source B found that. With a matrix, you read down a column and see the pattern straight away: three studies agree, one disagrees, and nobody has looked at a particular group. That pattern is your paragraph.

You don't need special software. A spreadsheet, a table in a document, or a large sheet of paper all work. What matters is choosing good columns and filling the cells with short, specific notes.

Synthesis matrix vs. annotated bibliography

An annotated bibliography is organized by source. You write a paragraph about each paper: what it did, what it found, why it's useful. It's a good first step, and many courses ask for one, but it keeps every source in its own box.

A synthesis matrix is organized by theme. The same information gets broken up and placed next to what other sources say on the same point. That's what makes comparison possible. If you already have an annotated bibliography, it's a great starting point: go through each annotation and move its pieces into the right columns.

A useful rule of thumb: the annotated bibliography answers what does each source say, and the synthesis matrix answers what do the sources say together.

How to build a synthesis matrix, step by step

Start small and expect to revise. Your columns will change as you read, and that's a sign the matrix is working.

• Write your research question at the top. Every column should help answer it. If a theme doesn't, it doesn't get a column.
• Gather your first sources. Five to eight is enough to start; add more rows as you go.
• Add the basic columns: author and year, type of study or method, and sample or context. These help you spot why sources disagree.
• Skim your sources and list the themes that keep coming up. Pick three to six as your first theme columns.
• Fill the cells as you read. Write a short paraphrase in your own words with a page number, not a long quote. Keep each cell to a sentence or two.
• Leave cells empty when a source says nothing on a theme. Empty cells are information: a column with few entries may point to a gap in the research.
• Add a synthesis row at the bottom. For each column, write one or two sentences on the overall pattern: agreement, disagreement and gaps.
• Revise the columns. Merge themes that overlap, split ones that are too broad and add new ones as they appear.

Synthesis matrix example

This example is purely illustrative. The research question and all four sources are invented to show how the cells work. Research question: how does evening phone use affect sleep among university students? Columns: method and sample; effect on sleep duration; effect on sleep quality; role of notifications; limitations.

The four rows are below, written as one line per source with each column labeled. Read down the columns and the synthesis almost writes itself. On duration, the two surveys suggest later bedtimes, while the small sleep-lab study found a smaller effect. On quality, three sources point the same way. Only one source looked at notifications directly, which is a gap worth mentioning. And every source relies on student samples from a single country, which limits how far the findings travel.

• Source A (fictional survey, 2021). Method: online survey of 600 undergraduates. Duration: heavier evening use linked to later bedtimes. Quality: more self-reported restless nights. Notifications: not studied. Limitations: self-report only, one university.
• Source B (fictional sleep-lab study, 2022). Method: 40 students, phone versus book before bed. Duration: small difference in time to fall asleep. Quality: slightly lower in the phone condition. Notifications: phones were silenced. Limitations: small sample, lab setting.
• Source C (fictional diary study, 2023). Method: two-week sleep diaries from 120 students. Duration: no clear link once study workload was accounted for. Quality: worse on nights with late messaging. Notifications: night-time alerts linked to waking. Limitations: diaries can be incomplete.
• Source D (fictional review, 2024). Method: review of earlier studies. Duration: mixed findings overall. Quality: fairly consistent negative link. Notifications: noted as under-studied. Limitations: few studies outside one region.
• Synthesis row. Duration: mixed, depends on method. Quality: mostly negative. Notifications: a promising but thin area. Limitations: narrow samples across the board.

Synthesis matrix template

Copy this structure into a spreadsheet, with one column per item below. In a spreadsheet, freeze the first row and column so headings stay visible, and turn on text wrapping so cells stay readable.

• Research question: [one sentence at the very top]
• Column 1: Source (author, year, short title)
• Column 2: Method and sample (study type, who, how many, where)
• Column 3: Theme 1 [name it]
• Column 4: Theme 2 [name it]
• Column 5: Theme 3 [name it]
• Column 6: Theme 4, optional [name it]
• Column 7: Limitations (as stated by the authors, plus your own)
• Column 8: Notes and page numbers for quotes you might use
• Last row: Synthesis, with one or two sentences per theme on agreement, disagreement and gaps

Turning the matrix into your literature review

Each theme column becomes a paragraph or a section. Start the paragraph with the pattern, not with a source. Then support it with the sources, grouped by whether they agree or differ, and finish with what's unclear or missing.

Using the fictional example, a paragraph on sleep quality might start: Studies using different methods point to a negative link between evening phone use and sleep quality. You'd then cite the survey, the lab study and the diary study together, mention that the review found the same pattern, and end by noting that most samples come from a narrow group. Compare that with a paragraph that walks through Source A, then Source B, then Source C. Same sources, much stronger writing.

Watch for differences in method. When sources disagree, the method and sample column often explains why: a survey and a lab study may be measuring different things. Saying so shows you've thought about the evidence, not just collected it.

Tips for filling the matrix faster

The slowest part is usually reading each source closely enough to fill its row. A few habits help: read the abstract and conclusion first, then go back for the method; fill the basic columns before the theme columns; and keep the matrix open while you read so you fill cells as you go rather than afterward.

Avoid the common mistakes too. Cells stuffed with long quotes are hard to compare, so paraphrase and keep the quote for the notes column. Themes that are really just topics, like technology, are too broad to show agreement or disagreement; a column such as effect on sleep quality is far more useful. And don't wait until every source is read to start writing. Once a column has three or four entries, draft its synthesis sentence. You can always revise it when new sources arrive, and you'll find writing the review much less daunting.

If your sources are PDFs, ZensNote can help with the first pass. Put each paper into a notebook for the project as a PDF Document note, read its summary to see the main points, and use AI Chat on that note to ask focused questions, such as what the sample was or what limitations the authors mention. A notebook summary gives one overview across all the papers you've added. Check anything you put in the matrix against the paper itself, with page numbers, and write the cells in your own words. The thinking that connects the sources is still yours.