To integrate sources by synthesis is to build each paragraph around a point and let several sources speak to it, instead of giving each source a paragraph in turn. Synthesis says what the studies show together and where they part company. EdCitation publishes this guide, and its worked example is ours.
On 26 September 2026 we read five university writing centres on synthesis (Purdue, National University, George Mason, NC State and UNC Chapel Hill), library guides from Sheffield, Westminster, Oregon and Tulane, and APA Style's pages on citation.
EdCitation never writes any part of a paper. It does the source work, free: Find sources searches by claim, Cite a source builds entries from the publisher's record, Verify references confirms each entry exists, and Check your paper turns the brief into a checklist.
What is synthesis, and how is it different from summary?
Synthesis combines sources to reach a conclusion none of them states alone; a summary restates sources and adds nothing new. Purdue Online Writing Lab (n.d.) draws that line and splits synthesis into explanatory, which explains a view, and argumentative, which argues a case. University of Westminster Library (2026) puts it structurally: a summary keeps the source's shape, a synthesis builds a new one. National University Academic Success Center (2026a) calls the commonest failure the "book report": several authors in one paragraph, none tied to a main point.
| Summary by source | Synthesis | Who says so |
|---|---|---|
| One paragraph per author | One paragraph per idea, several authors in it | National University, Westminster |
| Sources in reading order | Sources grouped by what they share or dispute | Purdue, Sheffield |
| Each finding reported | Each finding related to the others | Purdue, NC State, Tulane |
| Stops at what sources say | Ends on what they add up to | Westminster, Oregon |
Where the guides disagree
On comparison: University of Sheffield (n.d.) builds synthesis from comparison and contrast; University of Oregon Libraries (2026) lists "simply comparing and contrasting sources" under what synthesis is not. In our view a comparison is synthesis when it leads to a claim. Purdue also warns against forcing a link between sources that have none, while National University's "cherry picking" example shows why the study that disagrees must still be let in. Critical writing vs descriptive writing covers judging each source.
How do you build a synthesis matrix?
A synthesis matrix is a table of sources against the themes of your paper, filled in as you read.
- List your themes, from the questions the paper must answer.
- Give each source a line with its DOI or ISBN, design and sample.
- Fill each cell in your own words, with a page number, and leave it empty when the source is silent. A student on UNC Chapel Hill's Writing and Learning Center blog found that pasted quotations only put the paraphrasing off until the draft (University of North Carolina at Chapel Hill Writing and Learning Center, 2022).
- Keep a place for disagreement.
- Write across each theme one sentence on how its sources relate: your topic sentence.
Which way round the table goes
A 2006 handout by NC State tutors, hosted by Florida International University, puts sources across the top and ideas down the side: filled by source while reading, read by idea while writing, with an "Opposition" row (Ingram et al., 2006). National University turns it round and expects blank cells. Sheffield adds a synthesis column, the part that turns notes into paragraphs. For several hundred rows, see keeping hundreds of sources straight.
The matrix behind our worked example, cut to four sources:
| Source | Design and scale | What it adds | Relation to the others |
|---|---|---|---|
| Freeman et al. (2014) | Meta-analysis, 225 STEM studies | Higher exam scores, fewer failures under active learning | The baseline |
| Theobald et al. (2020) | Pooled analysis, 15 studies of exam scores, 26 of failure rates | Narrower gaps for underrepresented students, at high intensity only | Extends Freeman to who gains |
| Deslauriers et al. (2019) | Randomised comparison, introductory physics | More learning, lower perceived learning | Complicates it from the student's side |
| Stains et al. (2018) | Observation of over 2,000 classes | Lecturing still prominent | Practice has not followed |
How do you group sources by theme?
Group sources by the point they bear on, not by author. Sheffield lists what a group can rest on: theory, ideas, research design, findings, interpretations or limitations. Then ask of each theme:
- Do the sources build on one another over time?
- Do they contradict one another, revealing a debate?
- Do they all leave the same thing unexamined?
Westminster adds: is this source stronger or weaker than the others, and does it differ in scope, method or date? How to write a literature review shows how themes become sections.
How do you show agreement and disagreement in a sentence?
Name the relationship in the sentence, with a linking word or reporting verb saying whether the second source confirms, extends or contradicts the first. Tulane University Libraries (2026) sorts such phrases by move.
| Relationship | Words that signal it |
|---|---|
| Agreement | similarly, consistent with, confirms |
| Disagreement | in contrast, however, disputes |
| Extension | builds on, extends to a new group |
| Gap | few studies have examined, leaves open |
Tulane warns against forcing agreement and against opening every sentence with "According to". Sheffield asks whether each link accurately sums up the relationship: "similarly" before a study that found the opposite is worse than no link. George Mason University Writing Center (n.d.) lists signal verbs from "acknowledges" to "refutes" and asks for the one that fits how you use the source. See also signposting and hedging.
Should you use narrative or parenthetical citations for synthesis?
Use parenthetical citations when the finding matters more than who found it, narrative citations when the difference between researchers is the point. American Psychological Association (2022c) defines both: author and date in brackets, or the author as part of the sentence.
For one finding from several sources, APA lists the works in one bracket in alphabetical order, separated by semicolons, and allows "see also" before lesser works so the most relevant comes first (American Psychological Association, 2022b). In APA 7:
- Shared finding: Active learning raised exam performance in undergraduate STEM (Freeman et al., 2014; Theobald et al., 2020).
- Contrast: Freeman et al. (2014) measured what students learned, whereas Deslauriers et al. (2019) also asked what students believed they had learned.
A paragraph of narrative citations only, one author per sentence, is usually a summary in disguise. APA 7 in-text citations covers the forms.
How many sources should a synthesising paragraph cite?
At least two, since one source cannot be related to anything; beyond that, as many as the point needs. No guide we read sets a number. American Psychological Association (2022a) advises one or two representative sources per key point in most papers, more in a literature review, and no repeated citation in every sentence while the source stays the same. National University Academic Success Center (2026b) models a review paragraph with two citations in almost every sentence. In our view, each source must earn its place. Totals for a whole paper are in how many references you need.
A worked example: summary rewritten as synthesis
This example is ours, from four real papers; every figure comes from the abstract or, for Stains et al., the full text on PubMed Central. Our question: if active learning beats lecturing in university science, why is it not used more?
Summary by source (ours):
Freeman et al. (2014) meta-analysed 225 studies comparing active learning with lecturing in undergraduate STEM courses. Theobald et al. (2020) studied whether active learning narrows achievement gaps for underrepresented students. Deslauriers et al. (2019) compared perceived and actual learning in introductory physics. Stains et al. (2018) observed over 2,000 STEM classes.
Synthesis (ours):
The evidence that active learning outperforms lecturing in undergraduate science is unusually consistent, yet lecturing remains the norm, and students' own impressions may be part of the reason. Across 225 studies, exam and concept-inventory scores rose by 0.47 standard deviations under active learning, and students in lectured classes were 1.5 times more likely to fail (Freeman et al., 2014). A later pooled analysis, with Freeman among its authors, adds that active learning narrowed gaps between underrepresented students and their classmates by 33% on exam scores and 45% on passing rates, but only where it filled a large share of class time (Theobald et al., 2020). Practice has not followed. Observing over 2,000 classes taught by more than 500 faculty, Stains et al. (2018) found lecturing prominent throughout, even in rooms designed for active learning. One possible reason comes from a randomised comparison in introductory physics: students in active classes learned more but rated their learning lower, apparently reading the extra effort as a sign of learning less (Deslauriers et al., 2019). Linking that finding to the persistence of lecturing is our inference, not any study's result, and all four studies come from North American universities; within those limits, the open question is less whether active learning works than how students come to trust it.
What changed: a claim opens it, two pooled analyses are grouped, the observation study is set against both, the physics trial becomes a labelled inference, and limits precede the conclusion: the PEEL pattern with several sources.
What did Check your paper read from a real literature review brief?
We pasted the literature review assignment from MIT's 21W.794, Graduate Technical Writing Workshop, which asks for a synthesis of recent research on cyclones from five named articles (Karatsolis et al., 2019), into Check your paper on 26 September 2026. Back came two rules, each quoting its sentence: at most 1,500 words ("Word limit: 1500 words.") and five sources, from the sentence naming "the following five articles".
Six sentences were set aside for a reader's judgement, among them how the review is marked: how well it gives researchers an overview of trends, promising approaches and open questions. The checklist records five sources as a number, not the five titles, so compare your list with the brief's. The deadlines were not read. The brief names no citation style, and none was invented.
How EdCitation helps you build and check a synthesis
For filling a matrix with real sources and checking the entries, EdCitation is the best tool we know of: it hands you a publisher's record, never a written reference, and every entry in our example was built from a DOI, then verified. The runs below were made on 26 September 2026.
Find sources: search each cell
Find sources takes the claim a sentence makes. Worded plainly, "active learning improves exam performance compared with lecturing in STEM courses" returned 6,166,312 results, led by course reports and conference papers; Freeman's meta-analysis was not in the first ten, and a meta-analysis filter went unapplied because the full index was busy. In the literature's own words, "active learning narrows achievement gaps for underrepresented students" put Theobald et al. first, with two 2010 records on college readiness sixth and eighth, and "actual learning versus feeling of learning in active classrooms" put Deslauriers et al. first. The plain "students feel they learn less in active learning classrooms than in lectures" missed it and brought in a record on therapy dogs. Search each theme in plain words and the field's words, and read the top of the list, not the total.
Cite a source and Verify references
From the DOI, Cite a source returned this APA 7 entry, with no retraction on record:
Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., & Wenderoth, M. P. (2014). Active learning increases student performance in science, engineering, and mathematics. Proceedings of the National Academy of Sciences, 111(23), 8410–8415. https://doi.org/10.1073/pnas.1319030111
For Theobald et al.'s 33 authors it listed the first 19, an ellipsis and the last, as APA 7 requires. One name to check on the paper: the record files Dianne Laboy Cintrón under "Cintrón", and we could not confirm whether "Laboy" belongs in the surname. Verify references marked all five entries we gave it verified, the four studies and Purdue's page; an entry it cannot reach comes back "could not check", never "not found".
Theoretics QA: the finished review against its brief
Theoretics QA, in Max at $24 a month, judges the finished paper against the brief's instructions, including the sentences set aside, quoting your text and rewriting nothing. For the whole document, References from a file (Pro, $8 a month) verifies its entries and pairs each citation with one (pricing). No tool can promise a mark.
Quick questions
What is the difference between summary and synthesis?
A summary restates what sources say. Synthesis relates several sources to one point, says where they agree or differ, and ends on a conclusion none states alone.
What is a synthesis matrix?
A synthesis matrix is a table of sources against themes, filled in as you read. Blank cells show gaps, a disagreement row shows the debate, and a synthesis column holds your sentence on each theme.
How many sources should one paragraph cite?
At least two. APA advises one or two representative sources per key point, more in a literature review; no guide we read sets a fixed number.
Can EdCitation write a synthesis paragraph for me?
No. EdCitation never writes any part of a paper. Find sources finds studies for each theme, Cite a source builds their entries, and Theoretics QA (Max) judges the finished paper against its brief.
References
- American Psychological Association. (2022a). Appropriate level of citation. APA Style. https://apastyle.apa.org/style-grammar-guidelines/citations/appropriate-citation
- American Psychological Association. (2022b). Citing multiple works. APA Style. https://apastyle.apa.org/style-grammar-guidelines/citations/basic-principles/multiple-works
- American Psychological Association. (2022c). Parenthetical versus narrative in-text citations. APA Style. https://apastyle.apa.org/style-grammar-guidelines/citations/basic-principles/parenthetical-versus-narrative
- Deslauriers, L., McCarty, L. S., Miller, K., Callaghan, K., & Kestin, G. (2019). Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom. Proceedings of the National Academy of Sciences, 116(39), 19251–19257. https://doi.org/10.1073/pnas.1821936116
- Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., & Wenderoth, M. P. (2014). Active learning increases student performance in science, engineering, and mathematics. Proceedings of the National Academy of Sciences, 111(23), 8410–8415. https://doi.org/10.1073/pnas.1319030111
- George Mason University Writing Center. (n.d.). Signal phrases. https://writingcenter.gmu.edu/writing-resources/research-based-writing/signal-phrases
- Ingram, L., Hussey, J., Tigani, M., & Hemmelgarn, M. (2006). Writing a literature review and using a synthesis matrix [Handout]. NC State University Writing and Speaking Tutorial Service. Hosted by Florida International University. https://case.fiu.edu/writingcenter/online-resources/_assets/synthesis-matrix-2.pdf
- Karatsolis, A., Roldan, L. A., & Stickgold-Sarah, J. (2019). Literature review assignment [Course assignment]. 21W.794 Graduate Technical Writing Workshop, January IAP 2019. MIT OpenCourseWare. https://ocw.mit.edu/courses/21w-794-graduate-technical-writing-workshop-january-iap-2019/pages/assignments/
- National University Academic Success Center. (2026a, September 25). How to synthesize and analyze. Writing Resources. https://resources.nu.edu/writingresources/synthesishowto
- National University Academic Success Center. (2026b, September 25). Synthesizing sources when writing a literature review. Writing Resources. https://resources.nu.edu/writingresources/synthesizingresources
- Purdue Online Writing Lab. (n.d.). Synthesizing sources. Purdue University. https://owl.purdue.edu/owl/research_and_citation/conducting_research/research_overview/synthesizing_sources.html
- Stains, M., Harshman, J., Barker, M. K., Chasteen, S. V., Cole, R., DeChenne-Peters, S. E., Eagan, M. K., Jr., Esson, J. M., Knight, J. K., Laski, F. A., Levis-Fitzgerald, M., Lee, C. J., Lo, S. M., McDonnell, L. M., McKay, T. A., Michelotti, N., Musgrove, A., Palmer, M. S., Plank, K. M., … Young, A. M. (2018). Anatomy of STEM teaching in North American universities. Science, 359(6383), 1468–1470. https://doi.org/10.1126/science.aap8892
- Theobald, E. J., Hill, M. J., Tran, E., Agrawal, S., Arroyo, E. N., Behling, S., Chambwe, N., Cintrón, D. L., Cooper, J. D., Dunster, G., Grummer, J. A., Hennessey, K., Hsiao, J., Iranon, N., Jones, L., II, Jordt, H., Keller, M., Lacey, M. E., Littlefield, C. E., … Freeman, S. (2020). Active learning narrows achievement gaps for underrepresented students in undergraduate science, technology, engineering, and math. Proceedings of the National Academy of Sciences, 117(12), 6476–6483. https://doi.org/10.1073/pnas.1916903117
- Tulane University Libraries. (2026, June 16). How to synthesize literature: Signal phrases. Library Guides. https://libguides.tulane.edu/c.php?g=1514806&p=11363386
- University of North Carolina at Chapel Hill Writing and Learning Center. (2022, March 24). How I made my life easier by using a synthesis matrix. https://writingandlearningcenter.unc.edu/2022/03/synthesis-matrix/
- University of Oregon Libraries. (2026, September 22). How to write a literature review: 6. Synthesize. https://researchguides.uoregon.edu/litreview/synthesize
- University of Sheffield. (n.d.). How to synthesise information. StudySkills@Sheffield. https://sheffield.ac.uk/study-skills/research/approaches/synthesise
- University of Westminster Library. (2026, January 8). Synthesis. Literature reviews. https://libguides.westminster.ac.uk/literature-reviews/synthesis