Combine multiple sources into cohesive arguments. Create original insights from diverse materials.
~10 min read
Synthesis is one of the most important and most misunderstood skills in academic writing. Many writers believe that gathering information from multiple sources and presenting it in sequence counts as synthesis. It does not. Synthesis requires you to combine ideas from different sources into a new, unified argument that is genuinely your own. It is the difference between being a messenger who delivers other people's ideas and being a thinker who uses those ideas to build something original.
This section explains what synthesis actually is, how it differs from summary and comparison, and why mastering it is essential for advanced academic and professional writing.
Synthesis is the process of integrating information, ideas, and arguments from multiple sources to develop a new understanding, a new perspective, or a new argument. The word comes from the Greek sunthesis, meaning "to place together." In academic writing, synthesis means placing ideas from different sources together in a way that produces insight that none of the individual sources provides on its own.
Think of it this way: if you have three sources about the effects of social media on mental health, a summary would tell the reader what each source says. A synthesis would identify the patterns, tensions, and gaps across all three sources and use them to build your own argument about the topic.
The key distinction is that synthesis creates something new. You are not simply reporting what others have said. You are using what others have said as building blocks for your own original contribution.
The most common mistake writers make is confusing synthesis with summary. Understanding the difference is critical.
A summary restates the main ideas of a single source in your own words. It is faithful to the original source and does not add interpretation or connect the source to other material. A summary answers the question: "What does this source say?"
For example, if you are summarizing a study by Dr. Martinez on screen time and adolescent sleep patterns, you might write:
Martinez (2023) found that adolescents who used screens for more than three hours per day experienced an average reduction of 45 minutes of sleep per night compared to those who limited screen time to under one hour.
This is a clear, accurate summary. It presents the source's findings without adding your own analysis or connecting it to anything else.
A synthesis combines ideas from multiple sources to develop a point that goes beyond what any single source says. It answers the question: "What do these sources, taken together, tell us?"
Using the same topic, a synthesis might look like this:
Research consistently links excessive screen time to reduced sleep quality among adolescents, with Martinez (2023) documenting a 45-minute sleep reduction for heavy users and Thompson (2022) identifying blue light exposure as the primary neurological mechanism. However, Park (2024) complicates this picture by demonstrating that the content consumed matters more than the duration of exposure, with passive scrolling producing significantly worse sleep outcomes than active, purposeful screen use. Together, these findings suggest that public health recommendations should shift from blanket screen time limits to more nuanced guidelines that distinguish between types of screen engagement.
Notice what happened. The writer did not simply summarize three sources in sequence. Instead, the writer identified a pattern (screen time reduces sleep), introduced a complication (content matters more than duration), and drew an original conclusion (guidelines should be more nuanced). That conclusion is the writer's own contribution, built from the synthesis of three sources.
Comparison identifies similarities and differences between sources. Synthesis goes further by explaining what those similarities and differences mean and using them to advance an argument.
Smith (2021) argues that remote work increases productivity, while Jones (2022) argues that it decreases productivity.
This tells the reader that two sources disagree, but it stops there. It does not explain why they disagree, what the disagreement reveals, or what the reader should take away from it.
The apparent contradiction between Smith's (2021) finding that remote work increases productivity and Jones's (2022) finding that it decreases productivity dissolves when we examine their methodologies: Smith measured output per hour among knowledge workers in technology companies, while Jones measured output per hour among customer service representatives in call centers. This discrepancy reveals that the impact of remote work on productivity depends heavily on the nature of the work itself, suggesting that organizations should adopt role-specific remote work policies rather than company-wide mandates.
The synthesis does not just note the disagreement. It explains it, draws an insight from it, and proposes an original conclusion. The writer has used the sources as evidence for their own argument, rather than simply reporting what the sources say.
Synthesis is not just an academic exercise. It is a fundamental thinking skill that separates competent writers from excellent ones. Here is why it matters:
Summary shows that you can read and understand a source. Synthesis shows that you can think about it. When you synthesize, you analyze, evaluate, and interpret information. You make judgments about which ideas are more convincing, which evidence is stronger, and which conclusions are best supported. These are the higher-order thinking skills that academic writing is designed to develop and assess.
Without synthesis, your writing is a patchwork of other people's ideas. With synthesis, your writing presents an argument that is genuinely yours. The sources support your argument; they do not replace it. This is the difference between a research paper that contributes to a conversation and one that merely echoes it.
Synthesis appears in standardized tests (IELTS, TOEFL, AP English), undergraduate essays, graduate theses, professional reports, and policy documents. The AP English Language synthesis essay explicitly requires students to combine ideas from multiple sources into a coherent argument. IELTS Academic Task 1 requires synthesizing data from charts and graphs. Business professionals synthesize market research, competitor analyses, and financial data to make strategic recommendations. The skill is universal.
A single source proves very little on its own. When you synthesize multiple sources that converge on the same conclusion, your argument becomes far more persuasive. Synthesis allows you to build a case from multiple pieces of evidence, showing the reader that your position is supported not by one study or one opinion but by a body of evidence.
Before you can write effective synthesis, you need to develop what might be called a "synthesis mindset." This means reading sources not in isolation but in conversation with each other. As you read each new source, you should be asking:
This active, questioning approach to reading is what makes synthesis possible. If you read each source passively and in isolation, you will produce summaries. If you read each source as part of a larger conversation, you will produce synthesis.
Synthesis operates at different levels of sophistication:
The simplest form of synthesis groups sources by theme or finding. You identify that several sources address the same topic and organize them together. This is a starting point, but on its own, it is not true synthesis because you are not yet creating new meaning from the grouping.
Example: "Several studies have examined the relationship between exercise and cognitive performance (Smith, 2020; Chen, 2021; Okafor, 2022)."
At this level, you identify specific relationships between sources: agreement, disagreement, cause and effect, or complementary perspectives. You begin to show the reader how the sources interact with each other.
Example: "While Smith (2020) and Chen (2021) both found that aerobic exercise improves memory retention, Okafor (2022) demonstrated that these benefits are significantly greater for adults over 50, suggesting an age-dependent mechanism."
At the highest level, you use the connections you have identified to generate a new argument, a new framework, or a new conclusion. The synthesis becomes a vehicle for original thought.
Example: "The convergence of Smith's (2020) findings on aerobic exercise and memory, Chen's (2021) neuroimaging data showing increased hippocampal volume in regular exercisers, and Okafor's (2022) evidence of age-dependent effects collectively point to exercise as a neuroprotective intervention that should be integrated into clinical dementia prevention protocols, not merely recommended as a lifestyle choice."
Strong academic writing operates primarily at Level 3. The sources are not the point; your argument is the point. The sources are the evidence you marshal in support of that argument.
Synthesis is the act of combining ideas from multiple sources to create an original argument. It goes beyond summary (restating what a source says) and comparison (noting similarities and differences) by producing new insight. Developing a synthesis mindset means reading sources in conversation with each other, looking for patterns, tensions, and gaps, and using what you find to build arguments that are genuinely your own. The sections that follow will teach you the specific techniques for doing this effectively.
~9 min read
Synthesis begins long before you start writing. It begins when you read. The ability to identify meaningful connections between sources is the foundation of all synthesis, and it is a skill that can be developed systematically. This section teaches you the four primary types of connections you should look for when working with multiple sources and provides strategies for finding them.
When you read multiple sources on the same topic, they will relate to each other in one or more of the following ways:
Learning to identify these connections is the single most important step toward writing effective synthesis.
When multiple sources reach the same conclusion, that convergence strengthens your argument significantly. However, simply noting that sources agree is not enough. Effective synthesis explores why they agree, how they agree, and what the agreement means.
A weak synthesis of agreeing sources simply lists them:
Smith (2021) found that sleep deprivation impairs cognitive function. Jones (2022) also found that sleep deprivation impairs cognitive function. Williams (2023) reached the same conclusion.
This tells the reader nothing beyond the fact that three people agree. It is a list of summaries, not a synthesis.
A strong synthesis of agreeing sources explores the convergence:
The link between sleep deprivation and cognitive impairment is now supported by converging evidence from three distinct methodological approaches: Smith's (2021) longitudinal study of shift workers, Jones's (2022) controlled laboratory experiments with university students, and Williams's (2023) meta-analysis of 47 clinical trials. The fact that researchers using different populations, different measures, and different study designs all reach the same conclusion substantially strengthens the case that this relationship is robust and generalizable, not an artifact of any single research method.
Notice how this synthesis does not just say "they agree." It explains how they agree (different methods), why the agreement matters (convergence across methods strengthens the finding), and what it means (the relationship is robust). This is synthesis that adds value.
Disagreement between sources is not a problem to avoid. It is an opportunity to think critically and develop a nuanced argument. When sources contradict each other, your job as a synthesis writer is to explore why they disagree and what the disagreement reveals.
Direct contradiction: Two sources make claims that cannot both be true.
Henderson (2020) argues that immigration reduces wages for native-born workers. Patel (2021) argues that immigration has no measurable effect on native wages.
Partial disagreement: Sources agree on some aspects but disagree on others.
Both Henderson (2020) and Patel (2021) agree that high-skilled immigration benefits the economy, but they disagree sharply about the effects of low-skilled immigration on native-born workers in the same sectors.
Qualified disagreement: One source challenges the scope or conditions of another's claim.
While Henderson (2020) found that immigration reduces wages, Patel (2021) demonstrated that this effect is limited to the first two years after a large influx and reverses within five years as the local economy adjusts.
When sources disagree, look for explanations in their methodology, scope, definitions, or assumptions:
The most sophisticated synthesis writing does not just report disagreements. It resolves them, or at least explains them. Here is a model:
The debate over whether homework improves student achievement has produced contradictory findings, with Cooper (2006) concluding that homework has a positive effect and Kohn (2006) arguing that it has none. However, this apparent contradiction reflects a failure to distinguish between types of homework. Subsequent research by Fan et al. (2017) demonstrated that meaningful practice homework (applying concepts learned in class) improves achievement, while busywork homework (repetitive worksheets) does not. The disagreement between Cooper and Kohn, then, is not about whether homework works but about what kinds of homework are measured. Both are partially right, and the real insight lies in understanding which homework practices are effective and which are counterproductive.
This synthesis resolves the disagreement by finding a more nuanced truth that neither original source captured on its own. This is synthesis at its best.
Not all sources directly agree or disagree. Many address different facets of the same topic. Identifying complementary perspectives allows you to build a more complete picture than any single source provides.
Imagine you are writing about the decline of local newspapers. Your sources might include:
None of these sources directly agrees or disagrees with the others because they are studying different dimensions of the same phenomenon. But together, they paint a comprehensive picture that is richer than any one source alone.
The decline of local newspapers is typically framed as a business problem, a result of advertising revenue migrating to digital platforms (Reynolds, 2020). But the consequences extend far beyond the media industry. In areas where local newspapers have closed, voter turnout in municipal elections has dropped by an average of 8 percentage points (Garcia, 2021), suggesting that local journalism plays an essential role in democratic participation. Meanwhile, Yamamoto (2022) found that communities without local news sources report lower levels of civic engagement, reduced trust in local institutions, and weakened social cohesion. Taken together, these findings reveal that the death of local newspapers is not merely an economic disruption but a civic crisis that undermines the social infrastructure of democratic communities.
The synthesis connects three different lenses (economic, political, sociological) into a unified argument about the significance of local journalism. No single source makes this argument. The writer creates it by combining complementary perspectives.
As you read, ask:
Sometimes the most valuable connection you can identify is the absence of one. Gaps in the literature represent opportunities for original thinking and are prized in academic writing at every level.
Missing populations: The research focuses on one demographic and ignores others.
Most studies of remote work satisfaction have focused on white-collar knowledge workers in North America and Europe. There is almost no research on how remote work affects blue-collar workers, part-time employees, or workers in the Global South, leaving significant gaps in our understanding of remote work's broader implications.
Missing perspectives: An important viewpoint is absent from the conversation.
The debate over school uniform policies is dominated by administrators and parents; student voices are conspicuously absent from the literature, despite students being the ones most directly affected by these policies.
Unexamined assumptions: All sources share an assumption that deserves questioning.
All three studies assume that economic growth is the appropriate measure of a policy's success, but none considers whether the growth is equitably distributed, raising the question of whether GDP growth is a meaningful indicator of societal well-being.
Unexplored interactions: Sources address individual factors but not how those factors interact.
While extensive research documents the individual effects of sleep deprivation and poor nutrition on academic performance, almost no studies examine the compound effect of experiencing both simultaneously, which is the reality for many low-income students.
Identifying a gap allows you to position your own argument as filling that gap:
Previous research has established the individual effects of sleep deprivation (Smith, 2021) and malnutrition (Park, 2022) on student performance, but neither addresses the compounding interaction between these factors. This essay argues that addressing either factor in isolation will be insufficient; effective interventions must target both sleep and nutrition simultaneously, as the negative effects of each amplify the other.
This is a powerful rhetorical move. You have shown that you understand the existing research and that your argument addresses something the existing research has missed.
Here is a practical process you can follow when working with multiple sources:
Before looking for connections, make sure you understand each source on its own terms. For each source, note:
After reading all sources, create a table that helps you see connections at a glance:
| Topic/Theme | Source A | Source B | Source C |
|---|---|---|---|
| Main claim | ... | ... | ... |
| Key evidence | ... | ... | ... |
| Method | ... | ... | ... |
| Limitations | ... | ... | ... |
| Population studied | ... | ... | ... |
Fill in the table and look for patterns across the rows.
Using your table, mark connections:
Based on the connections you have identified, ask yourself: "What do these sources, taken together, tell me that none of them says individually?" Your answer to that question is the seed of your synthesis argument.
When you write, organize your paragraphs around themes and connections, not around individual sources. Each paragraph should develop a point that draws on multiple sources, not summarize a single source. This is covered in detail in the next section.
~9 min read
Once you have identified the connections between your sources, you face a critical decision: how do you organize all of this material into a coherent, readable argument? The organizational structure you choose determines whether your synthesis reads as a unified argument or a disjointed collection of summaries. This section covers the two main approaches to organizing a synthesis and teaches you how to choose the right one.
There are two fundamental ways to organize a synthesis:
Thematic organization is almost always superior for synthesis writing, but understanding both approaches will help you see why.
In source-by-source organization, you devote one paragraph or section to each source, summarizing its main arguments and findings before moving on to the next source.
Smith (2021) studied the effects of class size on student achievement in elementary schools. She analyzed data from 200 schools across five states and found that reducing class size from 25 to 15 students improved test scores by an average of 12%. She concluded that smaller classes allow teachers to provide more individualized attention.
Jones (2022) examined the role of teacher quality in student achievement. His study followed 500 teachers over three years and found that teachers rated "highly effective" produced test score gains three times larger than those rated "needs improvement." He argued that teacher quality is the single most important in-school factor affecting student learning.
Williams (2023) investigated the impact of school funding on student outcomes. Her analysis of spending data from all 50 states revealed that a 10% increase in per-pupil spending was associated with a 7% improvement in graduation rates. She concluded that increased funding, when directed toward instruction rather than administration, meaningfully improves educational outcomes.
These three studies all suggest that investing in education improves student outcomes.
The source-by-source approach has several problems:
Source-by-source organization can be appropriate in a few limited situations:
Even in these cases, you should still include connecting commentary between source discussions.
In point-by-point organization, you organize your writing around themes, arguments, or findings rather than around individual sources. Each paragraph develops a point and draws on multiple sources as evidence.
The most robust finding in educational research is that the quality of instruction matters more than any other variable. Jones (2022) demonstrated that highly effective teachers produce test score gains three times larger than their less effective peers, while Smith (2021) showed that smaller class sizes improve achievement specifically because they allow teachers to provide more individualized instruction. The mechanism connecting both findings is the same: when teachers can interact meaningfully with each student, learning improves. This suggests that class size reductions are effective not because of smaller numbers per se but because they enhance the quality of teacher-student interaction.
However, improving instructional quality requires adequate funding. Williams (2023) found that targeted increases in per-pupil spending yield measurable improvements in graduation rates, and Smith's (2021) data confirms that the class size reductions she studied were only possible in districts that increased their education budgets. Jones (2022), while focusing on teacher quality, acknowledged that attracting and retaining highly effective teachers requires competitive salaries, which in turn requires funding. The evidence converges on a clear chain: funding enables smaller classes and better teachers, which together improve student outcomes.
Transitioning from source-by-source to thematic organization requires a deliberate process. Here is how to do it:
After reading all sources, list the major themes, findings, or arguments that recur across multiple sources. These become your organizational categories.
Using the education example:
Create a grid showing which sources contribute to which themes:
| Theme | Smith (2021) | Jones (2022) | Williams (2023) |
|---|---|---|---|
| Instructional quality | Class size enables it | Teacher quality is key | — |
| Role of funding | Budget needed for small classes | Salaries retain good teachers | Direct link to outcomes |
| Factor interactions | Class size + teacher attention | Quality + training | Funding + targeted spending |
For each theme, write a paragraph that:
Arrange your themes in an order that builds your argument progressively:
Sometimes a hybrid of source-by-source and point-by-point organization works well, particularly when one source is so central to your topic that it deserves extended treatment.
In this approach, you designate one source as your "anchor" and organize the rest thematically around it:
This works well when one source is significantly more important, comprehensive, or controversial than the others.
When the evolution of thinking over time is relevant, you can organize chronologically while maintaining thematic connections:
This preserves the story of how understanding evolved while keeping the focus on themes rather than individual sources.
This is the most common problem: a paragraph that moves from one source to the next without connecting them.
According to Smith (2021), class sizes should be reduced. Jones (2022) found that teacher quality matters more. Williams (2023) argued for increased funding.
Each sentence is about a different source with no connecting analysis. Fix this by identifying the shared theme and organizing around it.
A paragraph that discusses one theme but fails to connect it to the broader argument:
Several researchers have studied school funding. Williams (2023) found that more funding improves outcomes. Jackson et al. (2016) reached a similar conclusion. Some researchers disagree.
This paragraph floats in isolation. Every paragraph should connect to your thesis and to the paragraphs before and after it.
A paragraph that tries to address too many themes at once, jumping between topics without developing any of them:
Class size, teacher quality, and funding all affect student achievement. Smith found class size matters, Jones found teachers matter, Williams found money matters, and all of these factors are interconnected in complex ways that research is still trying to understand.
This covers everything superficially and nothing in depth. Break it into multiple paragraphs, each developing one theme thoroughly.
A paragraph that presents source material without any synthesis commentary:
Smith (2021) found that smaller classes improve test scores by 12%. Jones (2022) found that effective teachers produce three times the gains of ineffective ones. Williams (2023) showed that a 10% funding increase yields a 7% improvement in graduation rates.
The data is there, but the analysis is missing. The reader does not know what these findings mean together. Always include sentences that explain the significance of the combined evidence.
Thematic (point-by-point) organization is the foundation of effective synthesis. It forces you to organize around your own argument rather than around your sources, it eliminates redundancy, and it keeps your analytical voice at the center. Source-by-source organization has limited uses and should generally be avoided in synthesis writing. Whatever approach you choose, every paragraph should develop a clear theme, draw on multiple sources, and include your own analysis of what the combined evidence means.
~10 min read
You understand what synthesis is. You can find connections between sources. You know how to organize thematically. Now comes the hardest part: actually writing synthesis paragraphs that integrate multiple sources smoothly while maintaining your own voice and argument. This section covers the sentence-level techniques that make synthesis writing work.
A well-written synthesis paragraph follows a consistent structure:
The critical elements are the topic sentence and the analysis. These are where your own argument lives. The source evidence is supporting material, not the main event.
The evidence consistently shows that early childhood interventions yield returns that far exceed their costs. [Topic sentence — writer's voice] Heckman (2006) calculated that every dollar invested in high-quality early childhood programs returns seven to ten dollars in reduced social costs, including lower crime rates and higher tax revenue from better-employed adults. [First source] These economic projections align with longitudinal evidence from the Perry Preschool Project, which tracked participants for over 40 years and found that children who received early intervention were 44% more likely to graduate high school and 46% less likely to be arrested by age 40 (Schweinhart et al., 2005). [Connection + second source] More recently, Duncan and Magnuson (2013) demonstrated that these benefits are strongest for children from the lowest-income families, suggesting that early childhood investment is not just economically efficient but also the most effective tool for breaking intergenerational poverty cycles. [Third source + added nuance] Taken together, this body of evidence makes a compelling case that early childhood education is not a social program but an economic investment, one that governments cannot afford not to make. [Analysis — writer's voice]
Notice that the writer's voice bookends the paragraph. The sources provide evidence, but the interpretation is the writer's own.
One of the most common problems in synthesis writing is clunky source integration. When every sentence begins with "According to Smith..." or "Jones argues that...", the writing becomes monotonous and the writer's voice disappears. Here are techniques for integrating sources more smoothly.
Instead of foregrounding the source, foreground the idea and attach the source as evidence:
Source-first (clunky): According to Martinez (2023), screen time exceeding three hours per day reduces adolescent sleep by 45 minutes.
Idea-first (smooth): Adolescents who use screens for more than three hours per day lose an average of 45 minutes of sleep per night (Martinez, 2023).
The idea-first approach makes the finding the focus and keeps the citation in the background. Use this technique whenever the finding is more important than who found it.
Certain verbs naturally connect sources to each other. These "synthesis verbs" signal the relationship between sources:
Agreement verbs:
Chen's (2022) neuroimaging data corroborates Smith's (2021) behavioral findings, providing a biological mechanism for the observed cognitive improvements.
Disagreement verbs:
Park's (2024) findings complicate the conventional narrative by demonstrating that content type matters more than duration.
Extension verbs:
Okafor's (2022) age-stratified analysis extends Smith's work by revealing that the benefits are disproportionately concentrated among older adults.
These verbs do significant rhetorical work. They tell the reader how sources relate to each other, which is exactly what synthesis is about.
Certain phrases explicitly signal that you are synthesizing, not summarizing:
These phrases signal to the reader that you are not merely reporting what sources say but drawing conclusions from their combined evidence.
Rather than giving each source its own sentence, combine related sources into a single statement:
Separate (reads as a list): Smith (2021) found that exercise improves memory. Chen (2022) also found that exercise improves memory. Okafor (2023) reached a similar conclusion.
Combined (reads as synthesis): Multiple studies have demonstrated a robust link between regular exercise and improved memory function (Smith, 2021; Chen, 2022; Okafor, 2023), with the effect size increasing with exercise frequency and duration.
When sources agree on a point, grouping them in a single citation is cleaner and gives you space to focus on what their agreement means.
Concessive structures acknowledge one source's position before presenting another, creating a sense of dialogue between them:
These structures create intellectual tension, which is the engine of sophisticated academic writing.
The biggest challenge in synthesis writing is maintaining your own voice. When you are integrating three, five, or ten sources, it is easy to become invisible, merely shuttling the reader from one source to the next without asserting your own perspective. Here is how to keep your voice central.
Use clear language to signal when you are reporting a source and when you are offering your own analysis:
Source voice: Martinez (2023) found that... According to Park... As Thompson demonstrates...
Your voice: This finding suggests... The implication is... What these studies collectively reveal is... This pattern points to...
The shift from source voice to your voice should be deliberate and visible.
A reliable structure for maintaining your voice is the "analysis sandwich":
Reducing class sizes is effective only when combined with improvements in teacher quality. [Your point] Smith (2021) found that class size reductions from 25 to 15 improved test scores by 12%, but a follow-up study by Chen (2023) revealed that these gains occurred only in classrooms with experienced, well-trained teachers; in classrooms with inexperienced teachers, the smaller class size produced no measurable improvement. [Source evidence] This suggests that class size is a necessary but not sufficient condition for improved learning. Without a competent teacher to take advantage of the smaller group, reducing numbers alone is an expensive intervention that yields no return. [Your analysis]
The sources are contained within your argument. You set the agenda; the sources provide the evidence; you interpret the evidence.
Proper attribution is essential in synthesis writing because you are constantly moving between your ideas and others' ideas. The reader must always be able to tell which ideas are yours and which come from sources.
Synthesis writing relies heavily on paraphrasing rather than direct quotation. When you are combining ideas from multiple sources, quoting each one disrupts the flow and makes it harder to blend perspectives. Paraphrase by default and quote only when:
Compare:
Over-quoted: Smith (2021) states that "smaller class sizes allow teachers to provide more individualized attention to each student." Jones (2022) argues that "the single most important factor in student achievement is the quality of the teacher." Williams (2023) concludes that "targeted increases in per-pupil spending yield measurable improvements in graduation rates."
Paraphrased and synthesized: The evidence converges on a clear chain of causation: adequate funding enables smaller class sizes and better teachers (Williams, 2023), smaller classes allow teachers to engage meaningfully with individual students (Smith, 2021), and teacher quality is the factor that most directly drives student achievement (Jones, 2022).
The paraphrased version is smoother, more unified, and places the writer's analytical framework (the "chain of causation") at the center.
When you are synthesizing a large number of sources (10 or more), additional techniques help manage the complexity:
Instead of citing every source individually, group sources that share a finding and cite them collectively:
A growing body of research supports the link between green spaces and mental health (Thompson, 2019; Wu, 2020; Hernandez, 2021; Nielsen, 2022). These studies, conducted across four countries and using both survey and clinical data, consistently find that access to parks and natural areas is associated with lower rates of anxiety and depression.
Not all sources are equally important. Give detailed treatment to the two or three most significant sources and relegate the rest to supporting citations:
Heckman's (2006) landmark cost-benefit analysis remains the most influential study in this area, demonstrating a 7-to-10-dollar return on every dollar invested. His methodology has been replicated and confirmed by subsequent researchers (Garcia et al., 2016; Yoshikawa et al., 2019; Elango et al., 2020), establishing this finding as one of the most robust in education economics.
In longer papers, you can use footnotes or brief parenthetical references to acknowledge sources without disrupting the flow of your main argument.
Many studies have found that X (Smith, 2021; Jones, 2022; Williams, 2023; Chen, 2024; Park, 2024; Lee, 2024).
This tells the reader nothing about what these studies found or how they relate. Always include at least a sentence of analysis after a citation string.
Smith found X. Jones also found X. Williams also found X.
"Also" is not a synthesis word. It merely adds. Replace with synthesis verbs and analytical commentary.
Paragraphs that consist entirely of source reports with no analysis or interpretation. If you remove all the citations from a paragraph and nothing remains, your voice is missing.
Presenting your own analysis as though a source said it:
Smith (2021) shows that the government should invest more in class size reduction.
If Smith's study found that smaller classes improve test scores, that is what Smith showed. The policy recommendation is your interpretation, and it should be presented as such:
Smith's (2021) finding that smaller classes improve test scores by 12% provides strong evidence that government investment in class size reduction would yield measurable educational benefits.
~10 min read
This section provides concrete examples of synthesis writing at different levels of quality. Study the weak examples to understand common mistakes, then study the strong examples to see what effective synthesis looks like. Each example is followed by a detailed analysis explaining what works and what does not.
Imagine you are working with these three sources:
*Davis (2022) conducted a study on educational apps and math performance. He found that students who used math apps scored 15% higher on standardized tests than students who used traditional worksheets. He concluded that technology can improve student achievement.
Rivera (2023) studied the relationship between screen time and attention spans. She found that children aged 8-14 who spend excessive time on screens have reduced attention spans. Her study suggests that too much screen time is harmful for children.
Thompson (2023) researched technology integration in classrooms. She found that technology is most effective when teachers receive training and when screen time is structured. She concluded that teacher preparation is important for effective technology use.
These three studies show that technology in education has both benefits and drawbacks.*
Educational technology improves student achievement, but only under specific conditions. Davis (2022) demonstrated that students using math apps outperformed their peers on standardized tests by 15%, providing compelling evidence for technology's potential. However, this potential comes with a significant caveat: Rivera (2023) found that excessive, unstructured screen time reduces attention spans in the same age group that Davis studied, suggesting that the manner of technology use matters as much as the technology itself. Thompson (2023) resolves this apparent tension by demonstrating that structured, teacher-guided technology use produces the greatest gains, while unstructured screen time produces none. Together, these findings point to a clear conclusion: the question is not whether technology should be used in education but how. Schools that invest in digital pedagogy training for teachers and implement structured screen time protocols can capture the achievement benefits Davis documented while avoiding the attentional harm Rivera identified. Technology without pedagogy is just screen time; technology with skilled guidance is a powerful educational tool.
Herring (2019) found that companies with diverse management teams earn more revenue. This shows that diversity is good for business. Dobbin and Kalev (2022) found that diversity training does not work and can actually make things worse. Phillips (2023) found that diverse teams are more innovative but also have more conflict. All of these studies relate to diversity in the workplace.
*The relationship between workplace diversity and organizational performance is more nuanced than the simple "diversity is good for business" narrative suggests. Herring's (2019) large-scale analysis does confirm that companies with racially diverse management teams generate higher revenue, but this finding raises a critical question: if diversity is so profitable, why do many organizations still struggle to achieve it? Dobbin and Kalev (2022) provide part of the answer, demonstrating that the most common tool for increasing diversity, mandatory training programs, is largely ineffective and sometimes counterproductive, triggering resentment rather than behavioral change. Phillips (2023) adds another layer of complexity by showing that even when diverse teams are assembled, they experience greater interpersonal conflict than homogeneous teams, and the innovation benefits of diversity materialize only when skilled facilitators help the team navigate that conflict productively.
What emerges from these three studies is a picture of diversity as a high-potential, high-maintenance organizational strategy. The benefits are real and substantial (Herring, 2019), but achieving them requires abandoning simplistic interventions like mandatory training (Dobbin & Kalev, 2022) in favor of sustained investments in facilitation, conflict resolution, and inclusive leadership practices (Phillips, 2023). Organizations that treat diversity as a checkbox will capture its costs (conflict, resentment) without its benefits (innovation, revenue). Those that treat it as an ongoing organizational capability to be developed will outperform their competitors.*
Nordhaus (2021) says carbon pricing is the best way to reduce emissions. Mildenberger (2020) says it is hard to implement because of political opposition. Roberts (2022) looks at the EU's carbon pricing system and finds it has reduced emissions somewhat but has also had problems. Overall, carbon pricing has both strengths and weaknesses.
*Carbon pricing occupies a paradoxical position in climate policy: it is simultaneously the approach most strongly supported by economic theory and the one most consistently undermined by political reality. Nordhaus (2021) builds a rigorous economic case that carbon pricing, whether through taxation or cap-and-trade, is the most cost-efficient pathway to emissions reduction because it harnesses market incentives rather than relying on regulatory mandates. The logic is compelling in theory. In practice, however, Mildenberger (2020) demonstrates that the political coalitions needed to implement effective carbon pricing rarely form, because the costs are concentrated and immediate (higher energy prices) while the benefits are diffuse and distant (avoided climate damage decades later). This political dynamic explains why, as Roberts (2022) documents, even the EU's Emissions Trading System, the world's most ambitious carbon pricing experiment, has been diluted by industry lobbying and permit over-allocation to the point where the carbon price has often been too low to drive meaningful behavioral change.
The synthesis of these three perspectives suggests that the debate between carbon pricing advocates and their critics is, in an important sense, beside the point. The question is not whether carbon pricing would work if implemented properly, as Nordhaus convincingly demonstrates it would, but whether the political conditions for proper implementation can ever be created. Roberts's analysis of the EU ETS shows that even in a political environment relatively favorable to climate action, industry pressure systematically weakens carbon pricing mechanisms. Effective climate policy may therefore require not just the right economic instrument but a parallel political strategy to overcome the opposition that Mildenberger identifies, one that perhaps combines carbon pricing with direct regulations and public investment to build political constituencies that benefit from the transition to clean energy.*
Across all three strong examples, notice these recurring features:
The writer has an argument. Every strong example is driven by a thesis or central insight, not by the sources themselves.
Sources interact, not just coexist. One source's finding is qualified, extended, complicated, or resolved by another's. The sources are in conversation.
The writer's voice frames and interprets. Topic sentences and concluding analysis belong to the writer. Sources provide evidence within that framework.
Organization is thematic. Paragraphs are built around points, not around sources.
The synthesis produces original insight. The conclusion of each example says something that none of the individual sources says. That new insight is the whole point of synthesis.
Tensions are embraced, not avoided. Disagreements between sources are not problems to paper over but opportunities for deeper analysis.
After writing a synthesis paragraph, ask yourself these questions:
If you answer "no" to any of these questions, revise before moving on.
~8 min read
This section provides practical reference materials for synthesis writing: templates for common synthesis structures, signal phrases for connecting sources, a catalog of common mistakes with fixes, and a comprehensive self-assessment checklist. Bookmark this section and return to it whenever you are working on a synthesis assignment.
These templates provide structural scaffolding for common synthesis patterns. Use them as starting points, then adapt the language to fit your specific argument.
Use when multiple sources support the same point.
[Your topic sentence stating the point]. [Source A] found/demonstrated/showed [specific finding]. [Connection phrase], [Source B] [synthesis verb] this finding by [specific finding that supports or extends A's]. [Optional: Source C adds further evidence that...]. [Your analysis: what the combined evidence means, implies, or suggests].
Filled example:
Early intervention programs produce long-term economic returns that dwarf their upfront costs. Heckman (2006) calculated a return of seven to ten dollars for every dollar invested, based on reduced incarceration, higher employment, and increased tax revenue. Garcia et al. (2016) corroborated these projections with updated longitudinal data, finding an even higher return of 13 dollars per dollar when accounting for improvements in health outcomes that Heckman's original model did not include. The convergence of these analyses, using different datasets and different cost-benefit models, makes the economic case for early intervention one of the most robust findings in education policy.
Use when sources conflict and you want to explore or resolve the tension.
[Your topic sentence framing the disagreement]. [Source A] argues/found [position X], based on [evidence]. However, [Source B] [challenges/contradicts/complicates] this by [showing/demonstrating/arguing] [position Y], based on [different evidence]. [Your analysis: why they disagree, what the disagreement reveals, or which position is better supported].
Filled example:
The effect of minimum wage increases on employment remains contested. Card and Krueger (1994) found no significant job losses following minimum wage increases in New Jersey, challenging the classical economic prediction that price floors reduce demand for labor. However, Neumark and Wascher (2014) reanalyzed the data with a different control group and reached the opposite conclusion, finding measurable job losses among the youngest and least skilled workers. The discrepancy likely stems from methodological differences in how the two studies defined their comparison groups, suggesting that the employment effects of minimum wage increases are real but smaller and more narrowly concentrated than critics claim.
Use when sources address different dimensions of the same topic.
[Your topic sentence identifying the broader theme]. From [a/an X perspective], [Source A] shows [finding]. From [a/an Y perspective], [Source B] adds [finding about a different dimension]. [Optional: Source C contributes a Z perspective by...]. [Your analysis: what picture emerges when these perspectives are combined].
Filled example:
The opioid crisis is a multi-dimensional problem that cannot be understood through any single lens. From an economic perspective, Case and Deaton (2020) trace the epidemic to the collapse of stable, well-paying employment in rural communities, arguing that "deaths of despair" are fundamentally a consequence of economic dislocation. From a medical perspective, Kolodny et al. (2015) document how pharmaceutical companies' aggressive marketing of OxyContin in the 1990s created a generation of patients physically dependent on opioids. From a policy perspective, Dasgupta et al. (2018) show that states with stricter prescribing regulations have lower rates of opioid abuse but higher rates of heroin use, as dependent individuals turn to illicit alternatives. Together, these perspectives reveal that the opioid crisis sits at the intersection of economics, medicine, and policy, and that interventions targeting only one dimension will inevitably fail.
Use when you want to position your own argument as filling a gap in the existing research.
[Your topic sentence identifying what existing research covers]. [Source A] has examined [aspect 1], and [Source B] has explored [aspect 2]. However, [gap: what has not been studied/considered/examined]. [Your analysis: why the gap matters and what filling it would reveal].
Filled example:
Existing research on food insecurity has focused primarily on its nutritional consequences. Gundersen and Ziliak (2015) documented the caloric and micronutrient deficiencies associated with food insecurity, while Seligman et al. (2010) established its connection to chronic diseases like diabetes and hypertension. However, neither study examines the cognitive and psychological effects of food insecurity, particularly the mental burden of chronic uncertainty about where the next meal will come from. This gap is significant because emerging evidence in behavioral economics suggests that scarcity of any kind imposes a "cognitive tax" that impairs decision-making, which may explain why food-insecure individuals sometimes make choices that appear irrational from the outside but are entirely predictable responses to the psychological stress of deprivation.
Use when you acknowledge a counterargument before advancing your own position.
[Acknowledging the opposing view]. [Source A] argues [counterargument], supported by [evidence]. This is a legitimate concern. However, [Source B] and [Source C] demonstrate that [your position], because [reasons/evidence]. [Your analysis: why the balance of evidence favors your position].
Filled example:
Critics of universal basic income raise legitimate concerns about work incentives. Mulligan (2012) argues that unconditional cash transfers reduce the motivation to seek employment, citing economic models predicting a 5-7% decline in labor force participation. This concern is understandable in theory. However, evidence from real-world experiments tells a different story. Forget (2011) found that during Canada's Mincome experiment, only two groups reduced their working hours: new mothers, who spent more time with infants, and teenagers, who stayed in school longer, both arguably positive outcomes. Similarly, Marinescu (2018) found no significant reduction in employment rates in Alaska following the introduction of the Permanent Fund Dividend. The weight of empirical evidence suggests that the theoretical work disincentive, while logically plausible, does not materialize at the levels of cash transfer that have been tested.
The following phrases help you connect sources and signal the type of relationship between them. Organize your writing by choosing signal phrases that match the relationship you are describing.
| Phrase | Example |
|---|---|
| consistent with | These results are consistent with earlier findings by... |
| in line with | Chen's data is in line with what Martinez reported... |
| corroborates | The longitudinal evidence corroborates the cross-sectional data... |
| reinforces | This pattern reinforces the conclusion that... |
| mirrors | The European experience mirrors the American findings in that... |
| parallels | Rivera's conclusions parallel those of earlier researchers who... |
| Phrase | Example |
|---|---|
| contradicts | This finding contradicts the conventional assumption that... |
| challenges | Park's data challenges the narrative established by... |
| calls into question | The failure to replicate calls into question whether... |
| complicates | This evidence complicates the straightforward conclusion that... |
| undermines | Subsequent research undermines the original claim by showing... |
| qualifies | Thompson's work qualifies Henderson's conclusion by demonstrating that it applies only when... |
| Phrase | Example |
|---|---|
| builds on | Okafor's analysis builds on Smith's foundational work by adding... |
| extends | This research extends previous findings to a new population... |
| adds nuance to | The qualitative data adds nuance to the quantitative picture by revealing... |
| deepens | These interviews deepen our understanding of the mechanisms behind... |
| refines | Garcia's updated model refines Heckman's original estimate by accounting for... |
| expands | The cross-cultural comparison expands the evidence base beyond Western contexts... |
| Phrase | Example |
|---|---|
| Taken together, these findings suggest... | Taken together, these findings suggest that screen time guidelines should distinguish between active and passive use. |
| What emerges from this body of evidence is... | What emerges from this body of evidence is a pattern of declining civic engagement. |
| The convergence of... points to... | The convergence of economic, sociological, and epidemiological data points to housing as the central determinant of health outcomes. |
| This pattern reveals... | This pattern reveals a fundamental tension between efficiency and equity in healthcare delivery. |
| The collective weight of evidence indicates... | The collective weight of evidence indicates that early intervention is the most cost-effective strategy. |
| Reading these studies alongside each other... | Reading these studies alongside each other reveals that the disagreement is not about the data but about the values underlying the interpretation. |
Symptom: Each paragraph summarizes one source. Sources never interact.
Diagnosis: You organized by source, not by theme.
Fix: Identify 3-4 themes that cut across your sources. Reorganize each paragraph around a theme, drawing on multiple sources.
Symptom: Smith found X. Jones also found X. Williams also found X.
Diagnosis: You are listing agreements without analyzing them.
Fix: Ask: Why did they all find the same thing? Do they use different methods? Different populations? What does the convergence mean? Write that analysis.
Symptom: Every sentence attributes an idea to a source. No sentence presents your own analysis.
Diagnosis: Your voice has been crowded out by source voices.
Fix: Ensure every paragraph begins with your topic sentence and ends with your analysis. Sources belong in the middle.
Symptom: A paragraph that places sources near each other but does not explain how they relate.
Martinez (2023) studied screen time. Thompson (2022) studied blue light. Park (2024) studied content types. All three studied aspects of technology.
Diagnosis: Proximity is not synthesis. The sources need to interact.
Fix: Add explicit connection language. How does Thompson's finding about blue light relate to Martinez's finding about sleep? What does Park's finding add to or change about what Martinez found?
Symptom: You present only sources that agree with your position and ignore those that do not.
Diagnosis: You are building a one-sided argument, not a synthesis.
Fix: Acknowledge opposing evidence using concessive structures. This actually strengthens your argument by showing you have considered multiple perspectives.
Symptom: Multiple direct quotes strung together, with minimal connecting text.
Diagnosis: You are outsourcing your argument to your sources.
Fix: Paraphrase most sources and reserve direct quotes for language that is especially precise or memorable. Use your own words to build the connections between ideas.
Symptom: A source is mentioned once and never referenced again.
Diagnosis: You have not integrated the source into your argument.
Fix: If a source is worth including, it should connect to at least one other source and to your thesis. If it does not, either find the connection or remove it.
Use this checklist before submitting any synthesis assignment.
15 questions on exactly what you just read — beginner to advanced, unlimited retakes. A free account also tracks your reading and scores your essays against your goal’s own rubric.
14 days of full access, no credit card required.
