eduKate Singapore

VIEW THIS AS

Auto mode follows the Route Engine until you choose a viewpoint.

YOU ARE HERE

ROUTE CHECK

CONNECTED TO

WHAT NEXT

Use the canonical route for this room, or HELP if you are unsure.

Next useful routeCanonical hubHELP

Everything about Essay Writing for Year 2027 | How to Write a Literature Review Essay with Themes, Evidence and Research Gaps

—

by

eduKate Asia
Illustration of university students comparing printed pages and grouping note cards at a library table.

A literature review essay helps readers understand what a body of published research says about one defined question, how different studies relate, which findings deserve confidence and what remains unresolved. It is not a row of abstracts with connecting phrases, and it is not a competition to cite the largest number of papers. A strong review explains how its sources were chosen, compares like with like, evaluates methodological limitations and builds a coherent map of themes and disagreements.

This 2027 guide develops the next skill after research-proposal writing: turning existing scholarship into an evidence-informed account that another reader can inspect. Its running example concerns how public instructions are understood by first-time readers. Any sample “studies” used as source cards are explicitly fictional demonstrations of how to organise notes, not real papers or findings. For your actual review, only cite publications you have located and read.

Quick answer: literature review checklist

  • Define the review question, audience, period and scope before opening dozens of tabs.
  • Decide whether the assignment calls for a narrative, scoping, systematic or other type of review.
  • Search with relevant terms and synonyms, then document what was actually searched.
  • Read the underlying paper rather than trusting a snippet, abstract or AI summary alone.
  • Use inclusion and exclusion rules that fit the question.
  • Record each study’s method, sample, outcome, findings and limitations in a source matrix.
  • Organise the body by themes, mechanisms or methodological differences—not by one author per paragraph.
  • Distinguish agreement from repeated citation of the same evidence.
  • Identify research gaps carefully instead of claiming a subject has never been studied.
  • Conclude with an appropriately bounded account of knowledge and a useful next question.

Prepared in November 2026 for the 2027 writing series. Sample research cards are fictional illustrations of organisation, not evidence of real study findings. A completed review must cite authentic publications and disclose its actual search scope.

Essay Writing 2027: four connected advanced guides

Follow this advanced sequence: design a research study, organise existing research, analyse a bounded case and evaluate policy alternatives. Each chapter is also a self-contained reference for the writing task in its title.

Chapter-by-chapter contents

1. What a literature review essay does

A literature review explains and evaluates existing scholarship on a defined question. It does more than summarize each paper. It identifies patterns of inquiry, compares results, explains disagreement and shows the reader how the available research bears on a problem. Purdue OWL emphasises summary, synthesis, critical evaluation and connections to the research question as central functions.

Imagine an organisation asking what research says about whether layout affects understanding of short public notices. Reading one paper about visual headings does not answer the whole question. Other studies may consider plain language, prior knowledge, task design or accessibility. A literature review can show which aspects have been examined and how closely their measures match the practical question.

The review’s result is not necessarily a recommendation. It might reveal that several studies measure preference while fewer measure correct interpretation, suggesting that the outcomes should not be treated as interchangeable. That kind of conceptual clarification is useful even when no simple verdict emerges. The writer’s job is to make the evidence landscape intelligible, not manufacture consensus.

2. Choose the review type deliberately

A narrative review offers a reasoned synthesis of a chosen body of scholarship and should be transparent about its scope. A scoping review commonly maps the range of literature and concepts on a topic. A systematic review follows a structured, prespecified approach to identifying and evaluating evidence for a defined question, with reporting expectations that are much more demanding than a typical classroom essay. Other fields have additional traditions.

Do not label a document a systematic review merely because you searched a few databases. Claims about completeness and method require appropriate protocols, screening and documentation. Likewise, do not treat a focused narrative essay as a failure because it does not include every article ever published. The appropriate method depends on the assignment, resources and intended conclusion.

State your approach early. For example: “This is a focused narrative review of selected studies about comprehension of public instructions, not an exhaustive systematic review.” That sentence gives readers a fair expectation. If the task requires a systematic or scoping method, follow relevant discipline-specific guidance rather than a simplified general writing template.

3. Define one answerable review question

A broad subject such as “communication” can produce thousands of unrelated results. A review question should identify which aspect of communication matters and which outcomes would help answer it. A possible example is, “How have studies evaluated whether presentation and wording of short public instructions influence first-time readers’ task comprehension?” It is narrow enough to guide source selection but still allows comparison across methods.

Define the population, format, context and outcome as needed. Some reviews may use established frameworks for intervention or exposure questions, while others are conceptual or qualitative. Do not force an experimental template onto work asking about meanings or lived experience. The useful principle is being clear about what belongs in the review.

Keep the question visible while searching. A paper about general visual attractiveness may not be central if the review concerns correct interpretation of instructions. It could supply context, but it should not carry the same evidentiary role as a comprehension study. A strong review does not confuse topical similarity with direct relevance.

4. State what is inside and outside the scope

Scope decisions affect what readers can infer from a literature review. You might include studies about adult newcomers reading brief public notices, while excluding papers about long technical manuals or advertising preference. The boundary must be justified by the question, not silently chosen to favour a thesis. Date, language, publication type and method restrictions may also influence the literature you find.

Write inclusion and exclusion notes before examining every result. The rules may be refined as terminology becomes clearer, but meaningful changes should be documented rather than hidden. If the review is selective because the assignment is short, say so. Do not claim the absence of evidence from material you did not search.

Be particularly careful when the literature crosses disciplines. Terms such as “comprehension,” “readability” and “usability” may be defined differently. A source can be relevant to one dimension without measuring the others. Stating the scope helps readers see why some apparently related studies are discussed and others are not.

5. Translate the question into search concepts

Search terms should reflect the concepts in the question, not only the phrase used in the proposed title. For public instruction comprehension, possible concept families include notice, public information, instructional text, reading comprehension, information design and user interpretation. The relevant vocabulary may differ across disciplines. A strong search uses synonyms thoughtfully and records what was tried.

Use suitable databases and catalogues for the field. Search engines can help with orientation, but structured scholarly sources may offer better control over publication type and metadata. Follow citation chains from relevant papers, including both references they cite and later work that cites them when appropriate. Always check the actual paper rather than treating a search hit as included evidence.

Do not fill the review with irrelevant results just because they match one keyword. Screen for question, setting and outcome. Save the search strings and date for your records, particularly when the review is intended to be updated. That documentation makes the selection more transparent than a vague statement that “many sources were consulted.”

6. Document the search without pretending it was exhaustive

A reader may need to know where and when you looked, which terms were used and how sources entered the review. In a formal systematic review this requires structured reporting, but a focused narrative review can still supply a concise description of its approach. The goal is honest scope. Do not copy a flow diagram or invent a count of records to imply a search process that did not happen.

Distinguish a planned search from a completed one. A student writing a proposal might state which databases they would search. A completed literature review should describe what they actually did. If access limitations prevented reading certain papers, record that. A citation from a secondary summary does not mean you inspected the original article.

Transparent search documentation also helps future updates. A new publication may change the picture, and a clearly dated review can be revisited. The year in the page title is not evidence that it covers all studies published in that year; readers should see exactly when the review was conducted and which material it included.

7. Screen titles and abstracts cautiously

Titles and abstracts help decide which sources to investigate, but they are compressed descriptions. A study may sound relevant until the full paper reveals a different population or outcome. Conversely, an unfamiliar title may conceal a useful method. Screening should therefore use the actual inclusion criteria and move to full text when needed.

Record why a source is excluded when the task requires transparency. “Different outcome measure” is more informative than “not useful.” If a paper examines how much readers liked a notice rather than whether they understood it, it may provide supporting context but not direct evidence for comprehension. That distinction helps avoid a review whose conclusions are built on mismatched measures.

Do not assume the abstract establishes methodological quality. Locate the methods, results, definitions and limitations before quoting a strong finding. If the paper is inaccessible, do not invent its contents from a search snippet. You may describe the limited information you found, but should not present it as a fully appraised study.

8. Read full papers for their actual questions

Before comparing studies, write a short record of each paper’s research question, design, sample or materials, measurements, results and limits. The same label may cover different work. Two papers discussing “clear instructions” could measure subjective ease, task success or memory after a delay. Their findings are not automatically contradictory or complementary without checking what was observed.

A literature matrix makes these distinctions visible. Include the author’s actual conclusion but also note what the design can establish. An experimental comparison may support a narrow causal inference under its tested conditions; a qualitative interview study may illuminate how participants interpret experiences. Neither is interchangeable with the other merely because the topic overlaps.

Read the sources as arguments grounded in methods. A paper’s discussion may contain hypotheses beyond its measured results. Those suggestions can help develop new research questions, but they must not be quoted as established findings. A careful review separates observation, author interpretation and your own synthesis.

9. Record source metadata while reading

Accurate notes prevent a last-minute citation scramble. Record author names, article title, publication venue, year, stable identifiers, page or section locators and version status where relevant. Papers can have corrections or retractions, so inspect current publication records when a claim matters. Do not rely on a reference generated by a language model unless you verify it against an actual source.

Separate direct quotations from paraphrase and your own comments. Mark which claims belong to which paper. It is easy to accidentally attach a finding to the author of a different study when several tabs are open. A source matrix helps prevent that error by keeping claims and locators together.

A review’s bibliography should be a reliable map of the material used. It should not list papers you never read as if you evaluated them. If you rely on a secondary source for a historical claim, make that chain clear. Source integrity is not a cosmetic matter; it underpins the reader’s ability to inspect your synthesis.

10. Build a matrix for comparison, not decoration

A good source matrix has one row per study and columns for question, setting, method, evidence, reported findings, limitations and relevance. Add a column asking what the study cannot establish. This is especially helpful when several studies use the same words for different outcomes. The table is a private analytical tool; you do not need to publish its full contents when a concise narrative communicates the result better.

For the fictional example, suppose the matrix contains placeholder cards for a study of visual headings, a study of simplified wording, an interview project on reader expectations and a field observation of notice use. These cards are not actual research. They illustrate different evidentiary roles. The real matrix must be populated only from genuine, traceable papers.

The writer’s next job is to find relationships among rows. Which methods address the same outcome? Which examine complementary mechanisms? Which limitations recur? A matrix succeeds when it helps build the review’s argument, not when it becomes a large catalogue readers have to decode without explanation.

11. Find themes that arise from reading

Themes are meaningful patterns across the included literature. They should emerge from the review question and the sources, not simply from the writer’s preferred outline. In a review about instructions, possible themes might include wording, visual hierarchy, prior knowledge, testing methods and accessibility. Each theme needs enough relevant material to justify its place.

Group studies conceptually and examine whether their findings or approaches relate. A paper on the position of headings and another on the sequence of steps may both concern information organisation, but their mechanisms differ. The review can explain the shared question and then the difference. Avoid forcing every paper into one neat theme when a result cuts across several categories.

Keep the hierarchy clear. “Comprehension,” “older readers” and “mobile phones” are not three parallel types of findings; they are outcome, population and delivery context. A useful outline may give each dimension its own role rather than mix them on the same level. Thematic structure should help readers interpret evidence, not simply group keywords.

12. Compare studies with compatible measures

Two studies can be compared directly only when their research questions, outcomes and settings are sufficiently compatible. One may measure whether readers recall a date; another whether they complete a form correctly. Both concern interpretation, but their measured outcomes are different. A review should name the difference instead of averaging them in an informal statement about “better comprehension.”

Consider time frames and participant characteristics. Immediate recognition of a heading may not establish long-term recall. A small study of expert readers may not describe first-time users. A visual design tested on a desktop display may function differently on a small screen. These conditions can explain why findings seem inconsistent without any study being wrong.

When direct comparison is not justified, write about the relationship more cautiously: the studies illuminate different dimensions of usability. A well-written review can still gain insight from complementary evidence without pretending it forms a single homogeneous result. Such distinctions are among the main contributions of synthesis.

13. Explain genuine disagreement among studies

A literature review should not hide studies that complicate its emerging story. If comparable studies reach different conclusions, examine possible reasons: design, sample, context, operational definitions, measurement and random variation. Do not choose the paper with the most striking finding simply because it supports the preferred thesis.

Sometimes disagreement reflects the theoretical account rather than a measured outcome. Researchers may interpret similar patterns differently because they emphasise prior knowledge, attention or layout. The review can compare those mechanisms and identify observations that would distinguish them. That is more valuable than announcing that “experts disagree” without explaining what they disagree about.

Give contrary evidence a fair place. A conclusion may become conditional: one approach is promising under certain conditions but less clearly supported in others. The review does not fail when it reaches a nuanced result. It succeeds by revealing the circumstances and uncertainties that a one-paper summary would miss.

14. Recognise shared sources and dependent evidence

Several publications may repeat one widely cited finding. Their agreement is not equivalent to several independent tests. A review should trace whether papers use the same dataset, cite the same foundational experiment or draw on one institutional report. Counting citations as votes can overstate certainty.

A review might find four articles recommending a particular layout, yet discover that three rely on one small trial. The correct conclusion concerns the strength and independence of the original evidence, not the popularity of its repetition. In our teaching example, this is a hypothetical caution, not a report that such a cluster exists in the actual field.

Use careful language such as “several sources cite the same study” rather than claiming independent replication. If subsequent investigations genuinely test the finding under new conditions, explain how their methods compare. The credibility of a literature review depends on seeing the evidence network beneath the publication list.

15. Evaluate methods rather than author reputations

A prestigious institution may produce useful work, but affiliation does not by itself settle whether a method answers a particular question. Examine recruitment, outcome definitions, comparison conditions, analysis and stated limitations. A small qualitative study might be strong for exploring experiences while unsuitable for estimating prevalence. A large survey may describe reported patterns but not prove a causal mechanism.

Do not use a one-size-fits-all hierarchy that automatically dismisses valuable approaches. The right method depends on the question. Instead, ask what the design was capable of establishing, how clearly the researchers reported it and what uncertainty remains. When the statistical or technical approach is beyond your expertise, state the limitation instead of guessing.

A good evaluation does not need hostile language. You can say that a study offers useful insight but cannot support a broad generalisation. Explain the exact boundary and its relevance to the review. Readers learn more from precise methodological comparison than from a list of papers called good or bad.

16. Distinguish a research gap from a convenient claim of novelty

A research gap is something important that existing studies have not adequately answered within the scope of your search and question. It does not always mean no paper exists. A field may have many studies of user preference but fewer studies of correct interpretation under realistic conditions. A particular population or task may be underexamined. A method may have a limitation that a new design could address.

Do not claim “there is no research” after a quick web search. Describe what you searched, what you found and which question remains unresolved. Your proposed gap should arise from actual evidence rather than from the desire to make a research proposal sound original. A gap may also concern conflicting findings or inadequate comparison, not simple absence.

Purdue OWL’s literature review guidance encourages identifying patterns, debates and areas needing further research. The responsible wording is calibrated: “In the selected literature, few included studies directly examined…” is narrower than a universal declaration about the entire field. Choose the wording that the documented search can support.

17. Write a controlling thesis for the review

A literature review needs a coherent central insight even when it does not advocate a policy. A weak thesis says, “Many researchers have studied information design.” A stronger one identifies a relationship: “The reviewed work examines several different meanings of clarity, making comparisons difficult unless reader preference, comprehension accuracy and task completion are distinguished.” This is a model thesis illustrating structure, not a finding established about real papers in this article.

The body must supply evidence for every clause from the actual included literature. Each section should explain how studies contribute to the theme, where their methods diverge and what limits follow. The review’s thesis may change as more material is read. A revised conclusion is a normal result of responsible synthesis.

Avoid the temptation to sound conclusive when the evidence is mixed. The strongest thesis may be about the state of knowledge: what is comparatively well examined, where measurement differs and which next question would be most informative. That provides readers with a useful map rather than a manufactured yes-or-no result.

18. Write paragraphs around relationships

A theme-based paragraph begins with a question or pattern, introduces the studies that address it and explains their relationship. For instance, a paragraph on visual organisation might compare studies that vary heading placement with studies that examine sequence of steps, then clarify which outcomes were measured. The writer should not simply list author A, author B and author C without drawing a conclusion about why they belong together.

Use a claim-evidence-explanation structure, but allow the order to vary. Cite material near the claim it supports and identify relevant limitations. If a study reports one result and another reports something different, show whether the methods and populations make them comparable. The relationship is the intellectual work of the review.

Do not force multiple sources into every sentence to satisfy a citation quota. A complex method may need a paragraph of explanation before it can be compared fairly. The goal is not maximum citation density. It is a sequence of meaningful relationships that help the reader answer the review question.

19. Use synthesis rather than citation stacking

Citation stacking occurs when a paragraph collects many sources after a broad claim without explaining their distinct contributions. The result may look researched but leaves the reader unable to assess what the sources establish. A good review gives each source a role: direct evidence, contrasting method, conceptual definition, replication, contextual limitation or counterexample.

Try writing a one-line annotation beside every citation in your draft: “Why is this source here?” If the answer is merely “It mentions the topic,” the citation may be unnecessary. If the source supports the central claim, explain exactly how. If it contradicts another, identify the point of disagreement. If it provides a definition, state whether other papers use that definition.

A review earns trust when its synthesis is transparent. A reader should understand not only that several papers exist but what they collectively allow the writer to conclude. This is particularly important when published work influences a real decision or research proposal. Evidence needs explanation before it becomes guidance.

20. Model a source matrix with fictional study cards

The following cards are invented for a writing exercise. Card A represents a hypothetical study comparing two heading layouts while recording correct identification of a deadline. Card B represents a hypothetical interview project asking participants which words confused them. Card C represents a hypothetical field observation of how visitors consult printed notices. Card D represents a hypothetical study of reader preferences between short and long instructions. None corresponds to an actual publication or result.

Place each card in a matrix with question, method, outcome and limitation. A potentially informs deadline comprehension, B reveals experiences of confusion, C concerns observed information-seeking behaviour and D concerns preference. The cards overlap in topic but do not measure the same thing. A responsible synthesis might therefore distinguish comprehension, lived interpretation and preference as separate dimensions.

The exercise demonstrates how a reviewer should think when authentic sources are available. It cannot establish whether headings improve comprehension in real life. To make that claim, locate and read actual research, record its methods and compare its findings. The fictional cards are instruments for learning synthesis, not evidence to cite.

21. Model thematic synthesis paragraph

Here is an original hypothetical model passage: “The imagined studies approach public notice clarity from different directions. A task-based experiment asks whether readers identify the correct deadline, while an interview explores which wording participants interpret ambiguously. These questions are related but not equivalent: a person may report that instructions feel clear while still selecting the wrong answer. The two approaches could therefore complement one another if their participants, materials and contexts were described carefully.”

“A separate field-observation study might offer insight into whether people consult the notice at all, while a preference study could reveal which layout readers like. Neither would substitute for a comprehension measure. Their value lies in illuminating different stages of an information-use process, from encounter to perception to interpretation.”

Because the sources are fictional, this passage is a structural demonstration only. A genuine literature review must replace these hypothetical claims with precise references to actual papers and report what those papers found. The lesson is how to relate methods and outcomes without flattening them into one vague idea of clarity.

22. Describe limitations without erasing useful findings

A literature review should identify the important limits of its included evidence. Studies may use narrow samples, different outcome definitions or settings that do not match the intended application. The review itself may be limited by access, search scope, language or time. These factors affect the confidence and transferability of its synthesis.

A limitation is not a decorative disclaimer placed at the end. Explain how it changes a particular conclusion. If included studies examine computer-screen notices, the review should be cautious about claims concerning physical signs in a noisy public space. If many studies measure self-reported confidence, their results cannot directly prove task accuracy. The relationship should be made explicit.

Do not dismiss all research because no study is perfect. Well-designed work can provide useful narrow knowledge. A careful review preserves the contribution while marking what remains unresolved. Readers benefit from learning which conclusions are supported under which conditions rather than receiving either an exaggerated consensus or a blanket statement of uncertainty.

23. Connect the review to a new research question

A literature review often supports a future study by showing what remains unexamined or poorly distinguished. The connection should be logical. If the review finds that studies use different measures of comprehension, a next project might compare outcomes with clearly defined tasks. If the main gap concerns a population or setting, a proposal could investigate whether findings transfer there. The proposed study must be feasible and ethical.

Do not manufacture a research gap simply to justify a project already chosen. The review may show that the question has been answered sufficiently for the immediate purpose. In that case, a replication, application review or different problem might be more valuable than another superficial study. Intellectual honesty sometimes changes the research plan.

Explain what the proposed investigation could add, not what it is guaranteed to discover. The literature review supports the question and design, but it cannot predict the observed outcome. This distinction connects the review with research-proposal writing while preserving the different purposes of the two genres.

24. Avoid overclaiming systematic coverage

Some assignments expect a rigorous systematic or scoping review with formal reporting protocols. Others ask for a narrative synthesis of selected readings. Use the correct label. A short set of papers gathered from a general search is not automatically a systematic review, even if the essay includes a colourful flowchart. False claims of methodological completeness undermine trust.

Be honest about the work actually done. If you searched two databases on particular dates, record that. If you excluded publications by language or access, state the boundary where it matters. Do not invent the number of records screened, the number of reviewers or a preregistered protocol. A reviewer should be able to distinguish a planned search from a completed one.

For a public article, concise disclosure may be enough; for a formal review, more detailed reporting may be required. Follow the standards of the field and assignment. Transparent scope is preferable to borrowed terminology that creates the appearance of rigour without the underlying method.

25. Revise the literature review for an evidence trail

Create a reverse outline. Does each section develop a meaningful theme, contradiction or gap? If the headings are author names, the draft may still be a serial summary. Replace unnecessary repetition with relationships across sources. Check whether each paragraph’s conclusion is supported by the specific papers cited. An interesting generalisation without a suitable source should be narrowed or removed.

Next audit the references. Verify names, dates, titles, page numbers, DOIs, quoted passages and corrections. Look for accidental citation of a paper through a secondary summary when you imply that you read the original. Check that the review’s claims about agreement and disagreement reflect what the methods actually measured. Avoid letting AI-generated references enter the bibliography unverified.

Finally ask an informed reader which questions the literature answers, where the findings differ and what research gap the essay has established. If they cannot distinguish these, revise the structure and definitions. A successful review is a map of knowledge that can be inspected, not an impressive-looking stack of titles.

26. Build mastery through repeated source comparison

Start with three real, credible papers on a narrow question. Summarise each accurately, noting the method and the limits. Then close the individual notes and write a paragraph that compares their questions or outcomes. Add another source that complicates the pattern. Revise the thesis to reflect the relationship rather than defend your first impression.

Use a source matrix and maintain a dated search record. Each revision should make the review more accurate, not merely longer. The practice transfers to policy analysis, professional research, historical essays and evidence-based public guidance. Readers repeatedly need someone to distinguish what is known, what is merely asserted and where different kinds of evidence do not align.

For 2027 and beyond, the goal is intellectual reliability. A literature review becomes useful when it explains relationships among authentic sources with enough clarity that another reader can retrace the argument and identify the unresolved question. That is a stronger service than claiming the last word on a field.

Four-week literature review practice

Week 1 — Scope and find sources

Write one review question and inclusion boundaries. Search appropriate scholarly sources, record the terms and date, and read several original papers. Distinguish actual findings from abstracts and opinion.

Week 2 — Appraise and group

Create a source matrix with design, population, outcome, findings and limitations. Develop themes from what the literature actually contains. Identify one apparent disagreement and check whether the studies measure the same thing.

Week 3 — Draft synthesis

Write an introduction that states scope and purpose. Develop thematic paragraphs integrating relevant sources, explain relationships and distinguish any research gap from a claim of universal novelty.

Week 4 — Audit and refine

Reverse-outline the review, verify every citation and ensure the conclusion matches the included evidence. Ask a reader to explain the themes, methods and remaining gap. Remove any unsupported claim of exhaustive coverage.

Frequently asked questions

What is a literature review essay?

A structured explanation and critical synthesis of existing research on a defined question, showing what studies collectively establish and what remains uncertain.

How is it different from a synthesis essay?

Synthesis is the reasoning process; a literature review is a research-focused genre with particular attention to scholarship, methods, search scope and gaps.

Must a literature review be systematic?

No. Narrative, scoping and systematic approaches have different aims and standards. Use the method required by the task.

How do I choose sources?

Define relevance criteria, search appropriate literature, inspect original papers and evaluate their methods rather than choosing results that merely support your opinion.

What is a source matrix?

A table recording questions, methods, outcomes, findings and limitations so sources can be compared accurately.

Can I organise by author?

Sometimes chronology or author groups matter, but thematic or methodological organisation often reveals more than one-paper-per-paragraph summaries.

How do I describe disagreement?

Check whether studies address comparable outcomes and contexts before calling their findings contradictory.

What is a research gap?

An important question the examined literature has not adequately answered, not necessarily a topic no one has ever mentioned.

Can I cite only abstracts?

Abstracts can guide screening, but significant claims should be checked against the relevant full paper whenever possible.

Should I include negative or conflicting studies?

Yes when relevant. Omitting inconvenient evidence can distort the review’s conclusion.

How should I end the review?

State the themes, the strength and limits of the evidence, and the next useful question or application.

Can AI build the bibliography?

Use it only where permitted and verify every reference against actual publications. Never publish invented citations.

Original and authoritative writing resources

Continue the eduKateSG essay-writing pathway

A well-written literature review neither invents consensus nor claims novelty without checking. It makes the included research intelligible through themes, methods, differences and honest limits, so readers can recognise the next worthwhile question.