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 | Research, Evidence, Fact-Checking and Citations

—

by

eduKate Asia
Illustration of a university student checking research-report pages with a pencil at a campus desk.

A well-researched essay is not an essay with the longest reference list. It is an essay in which each important claim can be traced to suitable evidence, the writer explains what that evidence means, and the reader can see where uncertainty remains. To research an essay for 2027, begin with a focused question, find the primary or most authoritative available sources, record exactly what each source supports, and cite ideas as well as quotations when the relevant style requires it. Never treat a fluent AI summary or a confident search snippet as a substitute for checking the underlying document.

This guide is useful for a school essay, university assignment, professional commentary, scholarship application, community proposal or independent longform article. It explains how to choose and compare sources, read statistics, identify misleading arguments, paraphrase accurately, build a source log, handle conflicting evidence and give credit. You will also learn when personal experience is appropriate evidence and when it cannot justify a general conclusion. Examples are explicitly illustrative; they do not pretend to report real studies, budgets or policies.

  • Start from the question and list the claims that need evidence before searching.
  • Prefer original documents and directly relevant, authoritative sources where feasible.
  • Record who made each claim, when, under which conditions and with what method.
  • Separate observation, interpretation, inference, prediction and recommendation.
  • Check units, dates, denominators, definitions and uncertainty before using statistics.
  • Use short quotations only when the exact words matter; otherwise paraphrase with attribution.
  • Keep reference details while researching rather than reconstructing them at the deadline.
  • Verify every cited page and every AI-suggested reference; never invent sources.
  • Follow the assignment’s required citation system, and disclose AI assistance where required.

This is a 2027-oriented skill guide written from sources available by November 2026, not a prediction of an examination, legal or citation requirement.

Contents: from a question to reliable evidence

1. Research is a chain of accountability

Research for an essay begins with a promise: a reasonable reader should be able to ask “How do you know?” and receive an honest answer. A citation alone does not fulfill that promise. A source may exist but be irrelevant, outdated, misinterpreted or quoted without the qualification that changes its meaning. The writer’s task is to build a traceable chain from question to claim, from claim to appropriate evidence, and from evidence to a proportionate conclusion.

Suppose you are writing about whether a town should add shade to a public walkway. A personal observation that the route feels hot can explain why the question arose. It cannot, by itself, establish how much temperature would change after planting trees. A design drawing may show an intended layout but not actual performance. An evaluation of another project may offer useful comparison, but its climate, materials and users may differ. Each source answers a different part of the question.

The first research skill is therefore classification. Ask what the source actually knows, how it knows it, and which part of your argument it can support. An essay becomes more trustworthy when the writer openly distinguishes what is established, what is plausible and what remains to be tested. Confidence should be the product of evidence, not the tone you adopt before researching.

2. Build a claim–evidence map before opening twenty tabs

Write the question at the top of a page. Beneath it, place each claim your proposed essay would need to make. Alongside every claim, leave a column for a suitable kind of evidence and another for uncertainties. You may discover that one paragraph needs a primary document, another needs an example, and a third needs clear reasoning more than another quotation.

For a hypothetical essay on improving a local crossing, your map might include the physical problem, who uses the crossing, what alternatives could exist, the constraints that would affect implementation and the criteria for judging success. You should not fill the cells with fictional numbers or unsupported assumptions. Mark them “information needed” until you find a relevant source. If the source cannot be obtained, revise the scope or qualify the proposal.

This map prevents the endless-search habit. Without it, writers collect interesting articles that barely address their actual question. With it, each search has a purpose. You can stop researching a minor claim once it has sufficient support and redirect effort to the weakest part of the argument. A good research process is not maximal collecting; it is targeted reduction of important uncertainty.

3. Know the difference between sources and evidence

A source is where information comes from. Evidence is material from that source that is relevant to a claim. A long official report may be a credible source but offer no evidence for the particular assertion you have made. A short, first-hand record may provide excellent evidence for a narrow historical question while offering little basis for a broad national claim. Credibility is not a magical property that transfers to every sentence you write after a citation.

Sources may include official regulations, published research, books, archived records, interviews, direct observation, datasets, technical manuals and reputable reporting. Each has strengths and limitations. An interview is valuable for the interviewee’s experience or statement; it does not automatically represent a whole community. A law explains what a rule says; it may not tell you how people experience its implementation. A dataset can show recorded patterns; it may omit relevant context.

Ask two separate questions: “Is this source worth taking seriously?” and “Does this particular passage or figure support the sentence I intend to write?” Both must be answered. Treat the source as a witness with a limited field of vision, not as a stamp that makes unrelated prose true.

4. Search strategically instead of randomly

Convert your question into several search terms. Begin with the central subject, then add a location, organisation, time period, method or alternative term. If you are researching a policy, search the responsible authority and the actual rule. If you are researching a scientific claim, look for the study, methods and the original publication rather than only summaries of summaries.

For example, “tree shade urban walkway” is broad. A targeted search might include the city, the relevant agency or a technical term for outdoor thermal comfort. Do not assume that a search result near the top is the most reliable or the most relevant. Search engines rank pages for many reasons; placement is not verification of the claim. Open the document, inspect what it covers, and distinguish a webpage’s publication date from the date of the event or study it discusses.

Keep a search note recording which terms worked and which did not. Useful negative findings are part of research. If the original data are not publicly available, write that down. Do not silently replace them with a claim made by a promotional page. A research log is a memory aid and a check against accidental source substitution.

5. Evaluate authority in context

Authority is specific to the question. A transport agency may be the right place to verify a published service notice. A public health agency may be appropriate for its own current guidance. A researcher may have expertise in a defined method but not every topic adjacent to it. A personal account may be the strongest evidence of what that person experienced, even if it cannot establish a population rate.

Use a practical credibility audit: who authored or issued the material; what competence or access did they have; what is the date; what evidence or method is shown; what interests may be involved; and can an independent source confirm the central point? Consider whether the document is a draft, consultation, final rule, opinion, advertisement or correction. Labels matter because they change what the text can legitimately prove.

Avoid an easy mistake: equating an attractive design with reliability. A polished webpage may still omit methods. A plain-looking archive may preserve an essential primary document. Equally, do not dismiss information solely because it is unfamiliar. Investigate the provenance and relevance. The result should be a reasoned judgement about each claim rather than a reflex based on branding or appearance.

6. Read a source for its actual claim

Skimming can help decide whether a document deserves attention, but close reading is needed before citing it. Find the question the author answered, the population or cases studied, the time period, the definitions and the caveats. Read enough surrounding text to know whether a sentence is the author’s conclusion, an example, a hypothesis or a position the author is criticising.

Imagine a report that discusses traffic at one junction during a particular survey period. It may support a statement about observations at that junction and time. It does not automatically support a claim about all roads in a city. If a study reports an association, do not convert that into proof of causation. If a document says a proposal is under review, do not write that it has been approved. The verb in your essay must match the strength of the source.

Make a margin note in your own words: “This source directly supports X; it does not establish Y.” This small discipline prevents many large errors. It also makes later writing faster because you have already identified the source’s proper role. Serious research often consists of learning what you cannot responsibly say.

7. Use primary sources carefully

Primary sources are materials directly produced in the event, study or decision you are examining: a statute, an original dataset, meeting minutes, a recorded interview or an author’s original text. They can be valuable because they reduce the number of interpretive layers between you and the evidence. But “primary” does not mean automatically correct. A first-hand witness can misremember, an official document can have a narrow purpose, and a dataset can reflect imperfect collection.

Secondary sources interpret, analyse or synthesise primary material. Good secondary analysis can reveal context, disagreements and methodological problems a beginner would miss. For a historical question, the useful combination may be an original document plus scholarship explaining the conditions in which it was produced. For a current public issue, it may be an official notice plus careful independent reporting.

Cite the source you actually used. If you learned about a result from a review but have not read the underlying study, do not pretend you inspected that study yourself. When the original source is accessible and central to your argument, read it. When it is not, be transparent about the intermediary. Traceability matters more than the appearance of an impressive bibliography.

8. Handle numbers, units and denominators

Numbers often make weak claims appear strong. The first question is “Out of what?” A percentage without a denominator can hide the scale of an observation. A change without the starting point can be difficult to interpret. An average may conceal wide variation. An absolute number may reflect population size rather than individual risk. These distinctions matter whether the essay concerns transport, business, health, technology or community life.

Suppose a hypothetical questionnaire receives responses from twenty volunteers, of whom fifteen favour a proposal. You can accurately write that fifteen of those twenty respondents expressed support. You cannot safely write that three-quarters of an entire neighbourhood supports the plan unless the sampling design justifies that inference. The example is invented to illustrate denominator discipline, not a report of a real survey.

Always copy the unit and period with the figure. Is it people, trips, dollars, hours, a rate per population or an index? Was it measured over a day, month or year? Has the definition changed? Keep enough context beside the number so that future editing cannot detach it from its meaning. If you cannot reconstruct how a figure was obtained, do not build a confident conclusion on it.

9. Distinguish correlation, cause and mechanism

When two things occur together, one does not automatically cause the other. People may visit a park more often during a period of improved weather, but that observation alone does not prove a change in programming created the increase. The reverse could be true, a third factor could affect both, or the pattern could be coincidental. Causal claims demand an explanation of the mechanism and, where possible, suitable research design.

In your essay, choose verbs that match the evidence. “Was associated with,” “coincided with,” “may have contributed to” and “caused” are not interchangeable. A careful essay can describe a plausible mechanism while acknowledging that it has not been established in the specific case. That is better than making the prose stronger than the research.

A useful test is to ask what other explanation could produce the same observation. Then consider what evidence would distinguish the explanations. This habit also improves personal reflection: noticing that you felt more confident after a particular event does not prove that one event alone changed everything. Clear reasoning keeps stories meaningful without turning them into experiments they were never designed to be.

10. Watch the time dimension of every claim

Many sources become misleading when their dates are stripped away. A webpage updated yesterday may describe a survey conducted several years earlier. A news item may report an announcement that has since changed. A legal or administrative policy may have a consultation phase, a commencement date and later amendments. The writer should record the date of the underlying information, not merely the date displayed beside a search result.

For 2027-oriented essays, distinguish three categories explicitly: rules already in effect, official future measures that have been announced but are not yet in effect, and possibilities that remain uncertain. Do not write future developments as accomplished facts. Verify official text for the relevant jurisdiction and date, particularly where a claim affects financial decisions, eligibility, safety or assessment requirements.

If a source is undated, identify it as such under the relevant citation style rather than inventing a year to make your reference look complete. When a page may change, record the date you accessed it if the style or task requires that information. Temporal precision is part of intellectual honesty, not mere administrative detail.

11. Compare sources that disagree

Disagreement between sources can reveal a real uncertainty, a difference in definitions, incompatible methods or a genuine dispute about values. Do not resolve it by choosing whichever source supports your preferred conclusion. Put the claims side by side and ask what each was trying to measure or explain. Did they use the same population, period and threshold? Are they studying the same phenomenon? Does one report an initial result and another a later revision?

For example, two hypothetical reports about noise may use different measures: one counts complaints, while another measures sound levels at particular locations. The results need not be contradictory because they answer different questions. A responsible essay explains the difference before using either as evidence for a broader claim. If sources truly disagree on the same proposition, state the disagreement and evaluate the strength and limits of each.

You may conclude that the evidence is inconclusive. That is a legitimate finding when it follows a careful comparison. Readers often need to know what is unsettled and why. Do not conceal uncertainty behind “experts say,” a phrase that removes names, methods and disagreement from view. Let the reader see the reasoning that led to your judgement.

12. Quote only when the exact wording matters

A quotation is appropriate when the wording itself is evidence: a legal definition, a distinctive argument, a revealing phrase in a literary text or a statement whose precise nuance you will analyse. Use the smallest passage needed. Introduce its speaker, source and context, then explain what it shows. A quotation that is merely dropped between your sentences without interpretation may interrupt rather than strengthen the argument.

Copy accurately, including important qualification. If you omit words, follow the relevant conventions for ellipses and ensure the omission does not change meaning. Do not create a composite sentence from separate passages and present it as one continuous quote. Avoid quoting a person from a secondary summary when the primary statement can be inspected directly.

Copyright and fairness also matter. Quoting large passages is not the same as explaining an idea in your own prose. Paraphrase when the exact wording is not necessary, and attribute the underlying thought. In a public article, a source link provides an accessible route for readers, though academic tasks may require additional formal citation details. The writer’s own analysis should remain the central material.

13. Paraphrase without disguising copying

Paraphrasing means expressing another source’s meaning accurately in your own language and structure, with appropriate credit. It is not replacing a few nouns with synonyms while preserving every sentence turn of the original. A good paraphrase begins with understanding: read the passage, look away, write what it means, compare your result with the source and correct any distortion.

Suppose a document describes a process with three necessary conditions. You should preserve those conditions and the relationships between them. You may use a clearer explanation or a different example if it remains faithful. Do not remove an exception that weakens the point simply because your thesis would sound stronger without it. A good paraphrase can be shorter than the source, but it must not be more certain than the source.

Purdue OWL explains that even ideas expressed in new words need attribution when they come from another source. Keep the reference beside your notes as soon as you paraphrase. Waiting until the essay is finished increases the risk that borrowed material will be mistaken for your own observation. Research ethics becomes easier when it is built into the note-taking process.

14. Create a source log that survives drafting

A useful source log contains the author or organisation, full title, publication date, stable link or identifier, date accessed when relevant, the exact page or section used, a short account of the claim, and any limits. Add a status such as “read in full,” “checked relevant section” or “secondary summary only.” This prevents accidental overstatement about what you actually consulted.

You can maintain the log in a simple table, a note-taking document or reference-management software. The tool matters less than the discipline. Separate the source’s exact words from your own notes using a clear convention. When copying a quotation, record the page at the same time. When summarising, mark that the wording is yours. When a source contradicts your thesis, keep it visible rather than discarding it.

Include a column for the sentence or section of the essay that the source will support. If an entry has no clear role, ask whether you need it. Conversely, if a major paragraph has no evidence where evidence is necessary, the log reveals the gap before publication. The log is a map of accountability, not a competition to gather as many citations as possible.

15. Understand citation styles without inventing rules

Citation systems such as APA, MLA and Chicago answer related questions in different formats: whose idea or text is being used, which work contains it, and how a reader can locate the material. The appropriate style depends on the assignment, discipline, publication or editor. Follow the current official manual or institutional guidance for that context, because the details of punctuation, order, dates and nonstandard sources vary.

Do not mix examples from several systems into a reference list that looks formal but is inconsistent. Check how the required style handles webpages, organisational authors, missing dates, multiple editions, page numbers, audiovisual material and works with no individual author. A citation generator can save time, but inspect its output. It may misread a webpage’s date, confuse a site name with an author, or produce a reference missing essential information.

The purpose is traceability and credit, not decorative punctuation. A reader should be able to find the relevant material and distinguish what comes from someone else. For research submitted in 2027, check the rules actually required then; the year in this guide does not establish a universal new citation standard.

16. Integrate sources into your own reasoning

A researched essay fails when it becomes one paragraph from Source A, followed by Source B, then Source C, with little explanation of what the materials jointly show. Synthesis means organising material around the question rather than around the order in which you discovered the references. Two sources may support different parts of the same paragraph; one source may provide a general framework, while another provides a concrete case.

Start a paragraph with the claim or distinction the reader needs. Then bring in the relevant evidence and explain the relationship. For example, if your argument concerns the challenge of maintaining a public facility, one document might define maintenance responsibilities and another describe user needs. The writer’s synthesis asks whether the responsibilities are suited to those needs. Neither source alone answers the whole question.

Remember to present the limits of your synthesis. Combining two sources does not make them automatically comparable. Explain differences in location, period and purpose. If you offer a new inference, signal it as your interpretation. The reader should be able to see when the essay is reporting a source and when it is reasoning beyond it.

17. Make graphs and images accountable

Visual material can strengthen an essay when it communicates a relationship more clearly than prose, but it can also mislead. Check axes, scales, missing values, units, date ranges and whether the plotted data answer the question. A cropped graph can hide the starting point. A visually dramatic change may be small in the original units. A map can omit categories, boundaries or groups whose absence alters the interpretation.

For images, verify provenance and context. A photograph of a crowded platform might show a moment, not the usual condition at that site. An illustration generated for the essay should be labelled as an illustration and not passed off as documentary evidence. A diagram can represent a proposed process, but its arrows do not make the process an observed fact.

Use meaningful captions and alternative text where publishing tools allow it. Credit the creator and source in the way your licence and citation requirements demand. Do not assume that because an image appears in search results it is free to reproduce. Sometimes the strongest choice is to describe the relevant point in plain words and link to the source instead.

18. Verify changing information and corrections

A source can be legitimate and still be superseded. Look for corrected versions, errata, updated tables, amended regulations, retractions or notices of withdrawal. A report’s first public version may differ from its final edition. If the information matters to a live decision, checking the latest authoritative version is part of the writing task.

Keep an audit trail when you update a draft. Record which sentence changed and why. If a public essay includes a material correction after publication, explain the correction clearly rather than silently leaving readers with conflicting versions. An update note can separate a new development from a mistake in the original article. Readers deserve to know the difference.

Use special care with figures copied from charts or PDFs. Verify the table headings and footnotes; a number may refer to a subgroup, a projected scenario or a particular confidence interval rather than a whole population. When you cannot establish the status of a claim, state the uncertainty. Refusing to overclaim is a professional strength.

19. Treat AI-assisted research as a lead, not a source

Generative AI can help suggest search terms, clarify an unfamiliar concept or offer questions to investigate, subject to the rules and privacy requirements of your task. It can also fabricate titles, merge two real studies into one nonexistent citation, misstate an author’s conclusion and write plausible quotations that do not appear in the source. A polished paragraph is not evidence that its references exist.

Verify each substantial claim through original material or a trusted and clearly identified source. Open the cited page. Check that the author, title, publication date and relevant passage match the assertion. If the source is unavailable, do not preserve the claim merely because the surrounding prose is attractive. Keep confidential documents and personal information out of tools unless you are authorised to share them.

The Modern Language Association updated its guidance on citing generative AI in August 2025, including attention to the model version and a stable shareable link when possible. The exact disclosure requirement for your essay depends on its rules. Using AI for permitted brainstorming is different from presenting generated prose as independently researched, personally experienced or originally authored where that is prohibited. Responsibility remains with the human writer.

20. Worked example: a claim about shade

Here is an entirely hypothetical research case. Question: “Should a walkway near a busy public facility receive additional shade?” The initial claim might be “Shade would make the walkway more comfortable for some users.” The writer should first ask what users experience, which parts of the route lack shade, how the proposed intervention would work, and what constraints might apply. Those are research questions, not facts already established.

A careful evidence plan separates categories. Site observation can help document the present layout when collected responsibly. Relevant technical guidance may explain how particular shade structures or planting designs operate under defined conditions. Public consultation may reveal preferences and accessibility concerns but should not be generalised beyond its participants without appropriate sampling. Official project documents, if any, would be needed before describing a specific approved plan or budget.

The eventual essay could present a conditional argument: examine the site’s exposure and users’ needs, compare options, assess maintenance and accessibility, then recommend a decision process. It should not invent a temperature reduction percentage, claim an unannounced authority decision or cite a fabricated study. The point of research is to determine which conclusion is justified, even when that conclusion is narrower than the one that first seemed appealing.

21. Protect people in interviews and personal evidence

Some essays involve interviews, lived experience or observations about other people. Obtain appropriate permission, respect privacy and follow the relevant ethical or institutional rules. A person sharing a sensitive story for background may not have agreed to public attribution. Anonymising a name may not be sufficient if the surrounding details identify the person. Consider what information is necessary for the reader and what exposure it creates for someone else.

Separate direct quotations from your own reconstruction. Keep notes about when and where an interview occurred, but do not publish private contact details. Do not present a colleague’s perspective as a company position unless authorised. In a reflective essay, distinguish what you saw from your interpretation of another person’s motives. “They left early” and “They did not care” are different claims.

For research involving human participants, additional review or consent procedures may apply. Do not assume a generic online guide replaces those requirements. If you cannot use an interview ethically, look for less intrusive evidence or narrow the essay. A good argument is not worth building by exposing a person unnecessarily.

22. Write with confidence proportional to the proof

Academic and public writing often rely on qualifiers: may, appears, in this sample, under these conditions, suggests, according to the report, or remains uncertain. These are not always signs of weakness. They tell the reader how far the evidence extends. Yet constant hedging can become evasive if the source actually establishes something directly. The challenge is accurate calibration.

Choose strong verbs for documented facts and appropriately cautious verbs for uncertain inferences. If a source states the wording of a published rule, you can report that wording confidently while checking its effective date. If two small examples suggest a broader possibility, label it as a hypothesis or illustration rather than a demonstrated pattern. When different readings remain reasonable, explain why you prefer one without pretending the alternative has disappeared.

Before publication, scan the draft for “always,” “never,” “proves,” “everyone,” “the best” and unsupported numerical claims. Ask what evidence each requires. Sometimes a strong phrase should remain. Often it should be narrowed. Your essay earns authority when a reader can see that certainty has been considered instead of manufactured.

23. Audit the finished essay sentence by sentence

Do one final factual pass separate from language editing. Highlight every empirical claim, quotation, date, named policy, technical specification and reference. Open the corresponding source and confirm the wording. Trace reported figures to their units, denominators and time periods. Check that a paraphrase preserves the source’s qualification. Verify that the working links lead to the intended document and not a similarly named page.

Then look for a different class of problem: claims that are logically unsupported even when the sources are real. Does the cited statistic address the population you are discussing? Does the source document an outcome or merely propose one? Does your conclusion depend on several steps that have not been explained? Remove, revise or qualify the sentence as necessary. Do not conceal the remaining limits.

Finally read the essay as a new reader. Can someone separate verified material, your interpretation and your recommendation? Can they find the cited source? Do they know which information may have changed since you wrote it? If the answer is yes, the essay has achieved a more important quality than an impressive reference count: it has become checkable. That is the standard to carry into 2027 and beyond.

A four-week research practice plan

Week 1 — Build a claim map

Take one essay question and identify six claims you might make. Classify each as a definition, observation, explanation, causal claim or value judgement. Write what type of support it needs and mark what you cannot presently verify.

Week 2 — Compare source quality

Find three relevant sources and record their authors, dates, methods and scope. Explain which sentence each can genuinely support and which sentence it cannot. Identify whether any source is secondary reporting of material you could read directly.

Week 3 — Practise quotation and paraphrase

Choose a short, permitted passage. Write a careful paraphrase with attribution, then compare it against the source for altered meaning. Write a separate sentence explaining how the material supports your own question.

Week 4 — Perform a verification audit

Draft a short research essay. Highlight every factual assertion, date and quotation. Check each against source material. Remove unsupported numbers, repair overconfident language, and keep a source log that another reader could follow.

Frequently asked questions

How many sources does an essay need?

Enough relevant, reliable material to support the claims made, while obeying the task’s minimum or maximum requirements. A large bibliography is not evidence that the argument is well supported.

Is Wikipedia allowed as a source?

It can be a starting point for orientation, but check the task rules and inspect its underlying references. For central claims, use primary or authoritative material where practical rather than stopping at a summary page.

How can I tell whether a website is reliable?

Check the author, publisher, purpose, date, evidence, methodology and possible interests. Also ask whether its specific content supports the sentence you intend to write.

Must every sentence have a citation?

Not every sentence. Cite borrowed ideas, evidence and quotations according to the required style. Your own reasoning may not need a source, but its factual premises often do.

Is paraphrasing enough to avoid plagiarism?

No. A good paraphrase uses your own structure and wording while preserving meaning, and it still credits the original source when the ideas come from elsewhere.

Can I cite an AI summary instead of reading the source?

Do not claim to have checked a source you have not inspected. If AI was used and the task permits it, follow the disclosure and citation rules, but verify the underlying evidence before relying on the summary.

What if two studies disagree?

Check whether they asked the same question and used comparable definitions, populations and periods. Explain the disagreement and the limits of each, rather than selecting the preferred conclusion automatically.

Which citation style should I use in 2027?

Use the style and edition required by your assignment, discipline or publication. Check current official instructions because there is no universal 2027 essay citation format.

What should I do with an undated webpage?

Do not invent a publication year. Follow the relevant style’s convention for missing dates, record identifying information and add an access date if required.

Can I use a statistic without the full report?

It is risky when the figure is central to your argument. Try to inspect the original table, definitions and time period. If you cannot, either qualify the claim and cite the intermediary honestly or choose a better-supported point.

How do I cite a photograph or chart?

Follow the relevant style and licence requirements, giving appropriate creator, title, source and date details. Do not treat a search result as proof of permission to reproduce an image.

What if I cannot find evidence for my thesis?

Change or narrow the thesis rather than inventing support. A more modest but honest conclusion is preferable to a confident conclusion without a traceable basis.

Sources and further reading

Keep exploring on eduKateSG

Research finishes when you can tell the reader what the evidence supports and what it cannot. The final guide in this series focuses on revision, feedback, style and responsible use of AI tools.

Complete 2027 Essay Writing Series

Follow the whole process in order or open the guide that matches your current problem: