The literature review is where a research paper earns its right to exist. Before you present your own study, you must show what is already known, where the studies agree and disagree, and what remains unanswered — and then place your work in the gap that remains. A literature review that only lists previous studies, one after another, fails at this job. A good one builds an argument: it organises the field into themes, weighs the evidence, and ends with a clear reason for the new study.
This guide explains the purpose of a literature review, how to search the literature systematically, how to organise what you find into themes, how to write synthesis rather than summary, and how to structure the section. It applies to research papers across fields, including veterinary and animal science work submitted to journals such as the Global Journal of Veterinary and Animal Sciences.
A literature review has three jobs. First, it maps the field: it shows the reader the main lines of research on your topic, the key findings, and the methods that have been used. Second, it evaluates the field: it notes where evidence is strong, where it is thin or contradictory, and where methods differ in ways that affect the results. Third, it justifies the new study: it identifies the gap your paper addresses and explains why filling it matters. If the review ends and the reader cannot say what gap your study fills, the review has not done its third job, however thorough the first two were.
Note the distinction between a literature review section inside a research paper and a standalone review article. A standalone review surveys a whole field and is itself the contribution. A review section inside a research paper is shorter and selective: it covers only the literature that bears directly on the research question, and every paragraph should move the reader closer to understanding why the new study was needed. Length is not a virtue here; relevance is.
Start searching before you write, and search with a plan rather than by wandering. Define your question first, then list the key terms for each part of it, including synonyms and standard field terms. For a study of feeding practice and milk yield, for example, the terms would cover the population (dairy cattle, smallholder farms), the intervention or exposure (feeding practice, ration composition), and the outcome (milk yield, milk quality). Searching each term alone returns either too much or too little; combining them with AND and OR narrows the result to studies that actually match the question.
Use more than one source. General academic databases, subject-specific databases for your field, and the reference lists of the most relevant papers you find will each turn up studies the others miss. When a paper proves central to your topic, follow its citations both backwards, to the studies it builds on, and forwards, to the studies that later cited it. This citation-chasing step is often where the most relevant papers are found, because it follows the field's own connections rather than keyword matching alone.
Keep notes as you search. For each paper you may cite, record the full reference, the study design, the population and sample size, the main finding, and one line on how it relates to your question. This habit saves hours later: authors who skip it end up re-reading papers to recover details they forgot, and sometimes cite findings they can no longer verify. A simple table or reference manager entry per paper is enough, and it doubles as the basis for your reference list. Our guide on checking and avoiding plagiarism explains why careful note-taking also protects you against accidental copying of phrasing.
Once you have a set of relevant papers, do not write about them in the order you found them. Group them by theme: by the question each study answers, by the method it uses, by the population it studies, or by the position it takes on a debated point. The grouping you choose should reflect the argument your review will make. If studies disagree, group them by the side of the debate they support, so the disagreement becomes visible. If methods differ, group by method, so the reader can see how the method shapes the result.
A useful test for a grouping is whether each group can be given a heading that states a finding rather than a topic. A heading such as "Studies of feeding practice in smallholder systems" describes a topic; a heading such as "Feeding practice is associated with higher yield, but the effect varies by system" states a claim the paragraph must then support. Headings that state claims force you to synthesise, because a claim can only be supported by weighing several studies together, not by describing them one at a time.
Chronological organisation has its place — for example, when showing how understanding of a topic developed over time — but it should be a deliberate choice, not a default. A review that simply marches through the years usually reads as a list, and it hides the themes and disagreements that matter. In most research papers, thematic organisation serves the reader better, because the reader needs to understand the state of knowledge, not the history of the field.
Summary reports what each study did, one study at a time: "Author A found X. Author B found Y. Author C found Z." Synthesis compares studies and draws a conclusion from them together: "Studies using method P consistently report X, while studies using method Q report Y, which suggests the difference lies in the method rather than the population." The first tells the reader what exists; the second tells the reader what it means. Reviewers judge a literature review on the second.
Synthesis is built with comparison language: agreement, contradiction, consistency, variation, and explanation. When studies agree, say so and note how strong the agreement is. When they disagree, do not just report both sides — discuss why they might differ: different populations, different measures, different settings, or different methods. That discussion is often the most valuable part of a review, because it shows the author understands the field well enough to explain its disagreements, and it frequently points directly at the gap the new study will address.
A practical way to practise synthesis is to write each theme paragraph from your notes, not from the papers themselves. With the papers closed, state the theme's claim, then support it with the studies you remember as evidence, citing each one. If you cannot state the claim without reopening a paper, you do not yet understand the theme well enough to write about it — go back to your notes and clarify the grouping before drafting.
A literature review section in a research paper usually follows a funnel shape. It opens broadly, with the field and the problem, then narrows through the themes to the specific gap, and ends with the aim of the present study. This structure mirrors the introduction as a whole, and in many papers the literature review is the bulk of the introduction, with the aim stated at its end. Whatever the journal's structure, the narrowing logic should be visible: each theme should bring the reader closer to the gap.
Within each theme, use a consistent paragraph pattern: an opening sentence stating the theme's claim, evidence from two or more studies with citations, a note on agreement or disagreement, and a closing sentence linking the theme to the gap or to the next theme. This pattern keeps paragraphs focused and prevents the common drift where a paragraph starts on one theme and wanders into another. If a paragraph needs more than one claim, split it — one claim per paragraph is a reliable rule for this kind of writing.
End the review with an explicit gap statement and the study aim. The gap statement should name what is missing or unresolved, in one or two sentences, and the aim should follow directly from it. A reader who reaches the end of the review should be able to predict the study's purpose before reading the aim sentence. If the aim comes as a surprise, the review has not led the reader there, and the connection between the literature and the study needs strengthening.
A literature review is a snapshot of the field at the time of writing, and fields move. Before you submit, repeat your key searches and check whether important studies have appeared since your first search. A review that misses a recent, directly relevant paper looks dated to reviewers, and it can weaken the gap statement if the new paper already fills the gap you claim. This final search need not be exhaustive — rerunning the core search terms and scanning the most relevant journals is usually enough — but it should be done, and any new study that changes the argument must be worked into the review rather than appended as an afterthought.
Honesty in a review also means reporting the limits of your own coverage. If your search covered only certain databases, or only studies in a particular language, and the journal asks for search details, state them plainly. A review that acknowledges its boundaries is more credible than one that pretends to be complete, and it helps readers judge how far the conclusions reach. The same honesty applies to disagreements: when the evidence is mixed, say so and explain what would settle it, rather than smoothing over the disagreement to make the argument tidier. Readers and reviewers trust a review that shows its working, including its uncertainties.
The most common pitfall is the annotated bibliography: a review that describes studies one by one without comparing them. The second is cherry-picking — citing only studies that support the author's position and ignoring contradictory evidence. Reviewers know the field and will notice, and selective citation damages credibility more than an honest account of disagreement. The third is over-reliance on secondary sources: citing a review's summary of a study rather than the study itself, which risks repeating errors and shows shallow reading.
Other frequent problems include citing outdated work while missing recent studies, which suggests the search stopped early; citing a paper for a claim it does not make, which careful checking would catch; and padding the review with marginally relevant papers to reach a length target. Every citation should earn its place by bearing on the research question. If removing a citation changes nothing in the argument, the citation is padding and should go. For guidance on citing correctly, see our guide on formatting references and citations once it is published, and for the wider submission process see how to publish a research paper.
| Feature | Summary only | Synthesis |
|---|---|---|
| Organisation | Studies described one after another | Studies grouped by theme or position |
| Comparison | Little or none; each study stands alone | Agreement and disagreement discussed explicitly |
| Conclusion | No conclusion drawn from the studies together | A claim about the field, supported by several studies |
| Gap | Gap unclear or unstated | Gap follows naturally from the comparison |
| Reviewer impression | Author has read the field | Author understands the field |
These authoritative external resources explain the standards and guidance mentioned in this article:
How many sources should a literature review include?
There is no fixed number. Include the studies that bear directly on your research question, enough to map the field and support your argument, and no more. A short, relevant review is stronger than a long, padded one.
Should a literature review cover everything ever written on the topic?
No. A review section in a research paper is selective by design. Cover the studies that shape the question, the debate, and the gap — and explain your selection honestly where the journal asks for it.
What is the difference between summary and synthesis?
Summary reports what each study did, one at a time. Synthesis compares studies, discusses agreement and disagreement, and draws a conclusion from them together. Reviewers judge a literature review on synthesis.
How do I organise a literature review: by theme or by date?
By theme, in most cases. Thematic organisation shows the state of knowledge and the disagreements that matter. Chronological order works when the development of understanding over time is itself the point.
Can I cite a review article instead of the original studies?
Use review articles to find the field, but cite the original studies for the claims you rely on. Citing a review's summary risks repeating errors and shows shallow reading.
Where does the literature review end and my study begin?
With an explicit gap statement followed by your study's aim. The reader should be able to predict the aim from the review before reading it; if the aim surprises, the review has not led the reader there.