Bring two sources into one essay paragraph

Two sources belong together when their relationship helps answer your question. They might support the same point, disagree, examine different conditions or measure different parts of a problem. A synthesis paragraph makes that relationship visible.

Simply placing one source summary after another does not do this. The reader should be able to identify your paragraph’s point, what each source contributes and what follows from considering them together.

Begin with the question, not the authors

Before drafting, write the specific question the paragraph must answer. Then record each source’s relevant claim in relation to that question.

Purdue OWL’s synthesis guidance distinguishes synthesis from summary: synthesis looks for relationships among sources and uses them to reach a broader conclusion. UMGC likewise describes synthesis as communicating what a body of research suggests, rather than merely reporting each work in turn.

A paragraph organised as “Rao says … Meyer says …” may make the authors visible while leaving their relationship unclear. A question-led note is more useful:

Under which conditions might weekly planning reminders help learners progress?

This question gives you something to compare. It also stops a result about one population or outcome becoming a universal claim about “students.”

Use two clearly fictional source summaries

The studies, authors, institutions, samples, measures and results below are invented teaching material. They are not real publications and must not appear in an assignment or reference list.

Fictional Source A: Rao and Meyer (2026)

At the fictional Northbridge University, 120 first-year, full-time undergraduates in one introductory module were randomly assigned either to receive a weekly planning reminder or to receive the standard module email. Learning-platform records showed that 68% of the reminder group opened the next assessment page within 48 hours, compared with 55% of the standard-email group. Average final module marks did not differ meaningfully between the groups. Rao and Meyer conclude that the reminders may encourage earlier engagement with a task, but they do not establish improved academic performance.

Fictional Source B: Dlamini (2026)

At the fictional Eastport Professional Institute, 74 part-time certificate learners received the same weekly planning reminder during an eight-week module. Their module-completion rate was compared with the previous cohort’s rate and showed no clear increase. In an anonymous end-of-module questionnaire, 19 of 31 respondents said changing work shifts still prevented them from following their plans. Dlamini concludes that reminders alone may be insufficient where learners face unstable time access. The comparison used a previous cohort rather than concurrent random assignment, and the voluntary questionnaire cannot represent every participant.

Both sources concern weekly planning reminders, but they do not ask or answer exactly the same question. Source A measures early platform engagement and final marks among full-time first-year undergraduates. Source B measures module completion and reports voluntary questionnaire responses among part-time professional learners. Those differences are part of the synthesis, not details to discard.

Compare before you combine

On a narrow screen, scroll the table sideways to read every column.

What can and cannot be connected across the fictional sources
Feature	Rao and Meyer	Dlamini	Synthesis consequence
Population	120 first-year full-time undergraduates	74 part-time professional certificate learners	Do not generalise either result to all learners
Comparison	Concurrent random assignment	Previous cohort comparison	Give the first result more causal weight, but retain its narrow setting
Primary outcome	Assessment-page opening within 48 hours	Module completion	Earlier engagement and course completion are related but not interchangeable
Other evidence	No meaningful difference in final marks	19 of 31 questionnaire respondents reported shift barriers	Neither source establishes that reminders improve performance
Warranted claim	Reminder may prompt earlier task engagement in this module	Reminder alone may be insufficient under unstable schedules	Effect may depend on both the outcome measured and learners' practical constraints

Do not decide that the sources agree or disagree until you have compared their populations, designs, interventions, outcomes and claim strength. Similar topics do not guarantee comparable evidence.

Diagnose source stacking

Here is a source-stacked paragraph:

Rao and Meyer (2026) studied planning reminders with 120 first-year students. They found that 68% of the reminder group opened the assessment page within 48 hours compared with 55% of the standard-email group. Dlamini (2026) studied 74 part-time certificate learners. The study found no clear increase in module completion, and some questionnaire respondents mentioned changing work shifts.

Each sentence may accurately summarise its fictional source, but the paragraph does not say how the results relate or what they contribute to an answer. The transition from one author to the other is only chronological.

Adding “similarly” or “however” would not solve the problem by itself. A connector must name a defensible relationship. Here, “however” could falsely imply that failure to increase completion contradicts earlier opening of an assessment page. The sources measured different outcomes.

Build the relationship explicitly

Harvard’s source-integration guidance recommends beginning a paragraph with a point in the writer’s voice and explaining what source material does for the argument. Curtin’s integrating-sources resource similarly distinguishes one-source paraphrase or summary from synthesis across two or more sources.

Use four moves:

State the paragraph’s answer. Make a claim narrow enough for both sources to inform.
Present the first source’s relevant evidence. Include only what the comparison needs.
Name the relationship. Explain whether the second source supports, qualifies, extends or genuinely contradicts the first.
Draw the combined implication. Say what becomes clearer only after reading the sources together.

For the fictional material, a defensible synthesis is:

Weekly planning reminders may influence an immediate planning behaviour without being sufficient to improve a later academic outcome. In Rao and Meyer’s fictional randomised module study, reminders increased the proportion of full-time first-year students who opened the assessment page within 48 hours, although final marks did not meaningfully differ (2026). Dlamini’s fictional cohort comparison qualifies that result: completion did not clearly increase among part-time certificate learners, and 19 of 31 voluntary questionnaire respondents reported that changing work shifts disrupted their plans (2026). Considered together, the studies suggest that reminders may prompt early engagement, while progress towards completion can remain constrained by time access; the different populations, designs and outcomes prevent a broader causal claim.

The paragraph does not claim that Dlamini disproved Rao and Meyer. It uses the second source to qualify the conditions and limits of the first. The final sentence belongs to the writer: it states the relationship rather than introducing a third unsupported fact.

Choose the right relationship

University of Technology Sydney’s synthesis resource groups scholars by their position and uses reporting verbs and transition signals to make relationships explicit. Choose a relationship after checking the evidence:

Agreement: the sources support the same bounded claim with reasonably comparable evidence;
Extension: one source adds a mechanism, context, population or consequence to the other;
Qualification: one source shows conditions or limits on the other’s claim;
Tension: the results or interpretations differ in ways that remain meaningful after comparing methods and scope; or
No useful relationship: the sources share a topic but cannot jointly support this paragraph’s point.

Purdue advises writers not to force a relationship or ignore an outlier. “Both discuss reminders” is a topic overlap, not yet a synthesis. “One measures early engagement while the other tests completion under different scheduling conditions” identifies a relationship that affects interpretation.

Keep the writer’s reasoning visible

Source integration is not a contest to fit the most citations into one sentence. The reader needs to know which claim belongs to which source and which conclusion is yours.

Useful sentence patterns include:

“Both studies identify …, although their evidence applies to different …”
“Source B extends Source A by examining …”
“Source B qualifies this finding because …”
“The apparent disagreement narrows once the outcomes are distinguished: …”
“Together, the sources support … but do not establish …”

Treat these as reasoning prompts, not templates to paste without checking. A transition word cannot repair an inaccurate summary or make incomparable findings equivalent.

Check method and scope before claiming contradiction

When two findings differ, ask:

Did the studies examine the same population and setting?
Was the intervention or exposure actually the same?
Were the outcomes defined and measured in comparable ways?
Did one study examine short-term behaviour while the other examined a later result?
Did the designs permit the same strength of causal inference?
Are uncertainty and limitations reported differently?
Does one author directly challenge the other, or did the studies simply produce different observations?

Answering these questions may reveal genuine tension, a useful boundary condition or no valid comparison. Do not manufacture a debate merely to make the paragraph sound critical.

Keep this task smaller than a literature-review matrix

This guide addresses one paragraph built from two already selected sources. If you first need to document and evaluate one source, use the guide to writing an annotated bibliography that evaluates a source. A literature-review matrix has a broader job: it organises many sources across themes, methods or debates before a longer review is drafted. University of Wisconsin–Madison’s professional-writing guide describes organising multi-source notes by idea and looking for agreement, disagreement and nuance; that planning method remains useful, but it is not required for this two-source exercise.

If your assignment contains many studies, recurring themes and a research-gap claim, use the method prescribed for the full review. Do not turn one paragraph pattern into a substitute for systematic searching, appraisal or review methodology.

Run a two-source synthesis check

Before submitting, ask:

Does the paragraph begin with the point it must establish rather than an author’s name?
Does each source contribute to that same point?
Have I represented each source’s claim, population, outcome and certainty accurately?
Is the stated relationship—agreement, extension, qualification or tension—supported?
Have I mistaken different outcomes for contradictory findings?
Can the reader distinguish the sources’ claims from my combined implication?
Does removing either source materially weaken the reasoning?
Are both in-text citations accurate and connected to real reference-list entries?
Have I avoided a conclusion broader than both sources can support?
Would two paragraphs be clearer if each source needs substantial explanation?

If the reasoning is sound but the paragraph remains difficult to follow, our guide to academic editing and proofreading explains when the next problem is language-level revision.

If you want guided feedback on how sources interact in a draft you are developing, academic writing development & revision can help you test the comparison while you retain authorship of the assessed submission.