Literature Review Examples: Annotated Passages Showing Structure and Synthesis

A literature review is the part of a paper, chapter, or standalone article that shows what is already known about a topic, how that knowledge was produced, and where the gaps or disagreements are. Its job is not to list sources — it is to build an argument out of them. Searchers who look for “literature review example” are usually stuck on that second part: they know roughly what to write about, but they haven’t seen what a working literature review actually looks like on the page, sentence by sentence, versus a list of summarized sources with citations attached.

This guide is built around annotated examples rather than another abstract how-to. For the step-by-step process — search strategy, note-taking, drafting order — see CASRAI’s guide to writing a literature review. This page shows what the result should read like, at different scales and using different organizing patterns, and contrasts weak synthesis against strong synthesis using the same source material.

Literature Review Examples at Different Scales

“Literature review” describes several different objects depending on where it sits. The underlying moves — situating your work, grouping sources by idea rather than by author, and synthesizing rather than summarizing — are the same at every scale, but how much space each move gets changes a great deal.

1. A Short Review Embedded in an Article Introduction

In a journal article, the literature review is usually two to five paragraphs inside the introduction, ending in a clear statement of the gap the study fills. Here is an illustrative example of that compressed form:

Prior work on remote-work productivity has largely converged on two explanations. One line of research attributes productivity gains to reduced commuting and fewer office interruptions (Author A, Year; Author B, Year), while a second line attributes them instead to selection effects, arguing that employees who volunteer for remote arrangements were already higher performers (Author C, Year). These explanations have rarely been tested against each other directly, in part because most available datasets do not distinguish voluntary from mandated remote arrangements (Author D, Year). The present study addresses this gap using a natural experiment in which remote work was mandated rather than chosen, allowing the selection-effects explanation to be tested independently of self-selection.

Why this works: it groups four sources into two competing explanations rather than four separate sentences; it names the methodological reason the debate hasn’t been resolved; and the last sentence turns directly into the paper’s own contribution. Nothing here summarizes what any single source found in isolation — every sentence does comparative work.

2. A Full Dissertation Literature Review Chapter

A dissertation or thesis chapter is the largest-scale version, often running 20-40 pages and covering several sub-literatures. Rather than reproduce a full chapter here, it helps to see the structural skeleton such a chapter typically follows, since the shape matters as much as the sentence-level writing:

• Opening framing (1-2 pages). States the broad problem, why it matters, and previews the chapter’s organization — usually naming the thematic sections that follow.
• Thematic sections (bulk of the chapter). Each section covers one sub-question or construct, synthesizes the relevant sources, and closes with a short assessment of what is settled and what is contested within that theme.
• A theoretical framework section. Introduces the conceptual model or framework the dissertation will use, usually built directly out of the tensions surfaced in the thematic sections above it.
• A synthesis and gap section. Pulls the thematic sections together, states explicitly what existing research has not yet established, and states the research question(s) the dissertation will answer.

CASRAI’s companion guide, the literature review chapter of a dissertation, walks through this structure in full. The annotated thematic-section example in the next part of this page is representative of what one of those middle sections looks like at sentence level.

3. A Standalone Literature Review Article

Some journals publish literature reviews as their own article type, distinct from a systematic review because the search is purposive and representative rather than exhaustive and protocol-driven (see CASRAI’s comparison of literature review vs. systematic review vs. scoping review). A standalone review reads like the dissertation chapter above but compressed into 3,000-8,000 words, usually organized around the thematic or methodological patterns covered next, and ending with an explicit statement of open questions for the field rather than a single study’s research question. CASRAI’s critical literature review entry covers the evaluative sub-type of this article, where the goal is to appraise and argue rather than only map the field.

Three Organizational Patterns, Annotated

Almost every literature review is organized one of three ways. Which one fits depends on whether the field’s key divide is conceptual, historical, or methodological.

Thematic Organization

Sources are grouped by idea or construct, not by author or date. This is the most common pattern for a review with more than a handful of sources, because it forces synthesis: you cannot write a thematic paragraph well without comparing sources against each other.

Research on patient adherence to chronic-disease medication regimens clusters around three explanatory factors. Cost-related nonadherence is the most extensively documented, with multiple studies linking out-of-pocket expense directly to skipped doses among lower-income patients (Author A, Year; Author B, Year). A second, smaller body of work emphasizes regimen complexity, finding that adherence drops sharply once a patient is managing more than three concurrent medications, independent of cost (Author C, Year). A third strand focuses on the patient-provider relationship, arguing that adherence is better predicted by whether patients feel their concerns were heard during prescribing than by either cost or complexity alone (Author D, Year; Author E, Year). These three explanations are not mutually exclusive, but existing studies have tested them separately rather than in combination, leaving open how they interact.

Annotation: notice the paragraph never says “Author A found X, then Author B found Y.” Each sentence is a claim about the literature as a whole, with citations supporting the claim rather than driving the sentence structure. The closing sentence does synthesis work by naming a genuine gap — the factors have been studied separately, not together.

Chronological Organization

Sources are ordered by when the field’s thinking developed. This pattern fits topics where the current consensus only makes sense in light of how earlier positions were challenged or revised — it is a narrative of change, not just a timeline of publications.

Early accounts of organizational resilience treated it as a fixed trait: some organizations simply had more of it than others, largely as a function of financial reserves (Author A, Year). This static view was challenged over the following decade by research reframing resilience as a process rather than a trait, arguing that the same organization could be resilient to one type of disruption and fragile to another (Author B, Year; Author C, Year). More recent work has pushed further still, treating resilience as something built and rebuilt through repeated small disruptions rather than tested only in a single crisis (Author D, Year). The field has therefore moved from asking how much resilience an organization has to asking under what conditions resilience is built, a shift that reframes the research question this study takes up.

Annotation: each sentence marks a shift in thinking (“challenged,” “pushed further”), not just a change in date. A weak chronological review would list studies in publication order without explaining why the order matters; this one uses the order to tell a story about how understanding changed.

Methodological Organization

Sources are grouped by how the research was done — qualitative versus quantitative, lab versus field, cross-sectional versus longitudinal. This pattern fits fields where method strongly shapes findings, so that a reader needs to know how something was studied before they can weigh what was found.

Evidence on the link between green space and self-reported wellbeing comes from two largely separate methodological traditions. Cross-sectional survey studies consistently find a positive association between neighborhood green space and wellbeing scores, but cannot rule out that people who already report higher wellbeing simply choose to live near parks (Author A, Year; Author B, Year). A smaller number of longitudinal studies, which track the same individuals before and after a change in green-space access, report a similar but smaller effect, suggesting the cross-sectional estimates may be partly inflated by this selection issue (Author C, Year). Experimental and quasi-experimental designs remain rare in this literature, and the one study identified that randomized access to a new green space found an effect only for participants with low baseline access, not for the sample as a whole (Author D, Year).

Annotation: the grouping variable is method, not topic or date, and each paragraph explicitly states what that method can and cannot establish. This is what lets the review conclude something the individual studies cannot: that the size of the effect across the literature likely depends on which design produced it.

Weak vs. Strong Synthesis: The Same Three Sources, Two Ways

The single biggest difference between a literature review that reads as a list and one that reads as an argument is synthesis. Below are two illustrative treatments of the same three placeholder sources — one describing them, one synthesizing them.

Weak: Source-by-Source Summary

Author A (Year) studied the effect of team size on decision speed and found that larger teams took longer to reach consensus. Author B (Year) also looked at team size, but focused on decision quality rather than speed, finding no significant relationship. Author C (Year) examined a related question, the effect of team diversity on decision quality, and found a positive relationship between diversity and quality.

Why it’s weak: each sentence is a self-contained book report — subject, method, finding — with no sentence that relates one source to another. A reader has to do the comparative work themselves: is Author B’s null result on quality in tension with Author C’s positive result on quality? The paragraph doesn’t say, and doesn’t attempt to.

Strong: Synthesized Argument

Team size and decision quality appear to be only weakly related once diversity is accounted for. While larger teams take measurably longer to reach consensus (Author A, Year), that slower process does not translate into better decisions on its own (Author B, Year); instead, decision quality tracks more closely with the diversity of perspectives represented on the team (Author C, Year). Taken together, these findings suggest that team-size effects on decision speed and team-composition effects on decision quality operate through different mechanisms, a distinction that studies focused on size alone are not positioned to capture.

Why it’s strong: the same three findings now support a single claim stated in the first sentence. Every citation is doing argumentative work — showing agreement, tension, or a boundary condition — rather than just attribution. The closing sentence identifies a genuine limitation in how the existing research is framed, which is exactly the kind of observation that justifies further study.

Common Structural Signals That Mark a Literature Review as Working

• Topic sentences make claims about the literature, not about one study. “Researchers largely agree that…” or “Evidence on this point is mixed…” signals synthesis; “Author A found that…” as an opening sentence usually signals a summary.
• Citations cluster. Strong reviews frequently cite two or more sources together to support a single claim (Author A, Year; Author B, Year); reviews that cite one source per sentence are usually still in summary mode.
• Transitions do comparative work. Phrases like “in contrast,” “building on this,” “however, more recent work,” and “this raises the question of” signal that the writer is relating sources to each other, not just moving to the next one.
• Sections end with an assessment, not just a stop. A strong thematic section closes by stating what’s settled and what isn’t within that theme, which is what lets the next section or the final synthesis build on it.
• A synthesis matrix often sits behind the strongest paragraphs. Many writers track sources in a table — one row per theme, one column per source — before drafting prose; it’s a planning tool, not something that appears on the page, but its effects show up as the clustering and comparison described above. See the note-taking section of CASRAI’s literature review writing guide for how to build one.

Literature Review Example Checklist

Use this to check a draft section against the patterns shown above:

• Does at least one topic sentence per paragraph make a claim about the literature as a whole, rather than reporting a single study?
• Are two or more sources ever cited together in support of one sentence, or is every citation isolated to its own sentence?
• If two sources disagree, does the text say so directly, or are the disagreeing findings simply placed near each other and left unreconciled?
• Does each section end with an explicit statement of what is and isn’t established, rather than just trailing into the next topic?
• Is the organizing pattern (thematic, chronological, or methodological) consistent throughout, rather than drifting between them paragraph to paragraph?
• Does the review end by pointing at a specific, stated gap — not just “more research is needed,” but a named unanswered question the gap creates?

Frequently Asked Questions

What does a good literature review look like compared to a bad one?

A good literature review reads as a running argument: topic sentences make claims about the state of a field, citations are grouped to support or contest those claims, and sections close with an explicit assessment of what’s settled and what’s not. A weak one reads as a list: one source, one sentence, one finding, repeated, with the comparative work left for the reader to do. The side-by-side example above shows the same three sources treated both ways.

How long should a literature review example section be?

It depends entirely on scale. Inside a journal article introduction, two to five paragraphs is typical. A standalone review article commonly runs 3,000-8,000 words. A dissertation or thesis chapter is usually the longest, often 20-40 pages, because it has to cover several sub-literatures rather than motivate a single study. See CASRAI’s guide to writing a literature review for more on scoping length to your source’s purpose.

Is a literature review example the same as a systematic review example?

No. A (narrative) literature review uses a purposive, representative selection of sources chosen to support an argument; a systematic review commits in advance to an exhaustive, reproducible search and formal appraisal process, typically reported against the PRISMA reporting standard. See CASRAI’s comparison of literature review vs. systematic review vs. scoping review for the full distinction, including where a scoping review and narrative review fit relative to both.

What’s the difference between synthesis and summary in a literature review?

Summary reports what each source found, one at a time. Synthesis relates sources to each other — showing where they agree, disagree, or address different parts of a question — to build a claim that none of the individual sources makes on its own. The weak-vs-strong example above shows the same three findings written both ways.

Can I use AI tools to help draft a literature review?

Many researchers use AI-assisted discovery and summarization tools during the search and note-taking stage; CASRAI’s guide to AI literature review tools covers what these tools do well (surfacing and clustering related work) and where they still require human judgment (verifying every citation is real and accurately represents the source, and doing the comparative synthesis work itself).

Related CASRAI Resources

• How to Write a Literature Review — the full process: search strategy, note-taking, and drafting order.
• The Literature Review Chapter of a Dissertation — structuring the largest-scale version of a review.
• PRISMA and Systematic Review Methodology — reporting standards for the exhaustive-search version of a review.
• Literature Review vs. Systematic Review vs. Scoping Review — how the three review types differ in search strategy and output.
• Systematic Review vs. Meta-Analysis — how a quantitative synthesis differs from a narrative one.
• Critical Literature Review — the evaluative, argument-building sub-type of review.
• Narrative Review and Scoping Review — related review types and how they’re defined.
• AI Literature Review Tools — where AI assistance helps and where it doesn’t.