A student types a topic into a search box, opens the first result and starts copying. Ten minutes later, the notes are full but the research question has disappeared. Another student collects five websites that all repeat the same claim and calls that “five sources”. A third finds a statistic without recording who measured it, when it was measured or what the denominator was.
Research is not the act of gathering more text. It is a sequence of decisions about what you need to know, where to look, what to trust, how to compare and how to return what you found without pretending that the evidence is stronger than it is.
Secondary 2 research becomes powerful when students learn to build a question that can guide the search, then build an answer that remains traceable to its sources.
Why this belongs inside English
The current Singapore secondary English syllabus organises adolescent literacy across Listening and Viewing, Reading and Viewing, Speaking and Representing, Writing and Representing, Grammar and Vocabulary. It also treats digital networked sources of information as an important resource for language learning and emphasises creating meaning across different semiotic forms. Research therefore sits naturally between reading, evaluating, writing and representing rather than outside them.
Official reference: MOE English Language Syllabus 2020 Secondary.
International middle-school ELA systems also increasingly connect research to presentations, multimedia and discussion. This guide uses that broader direction as a comparator while keeping the actual tasks here original to eduKate.
This article belongs to the Secondary English Learning Hub. It extends Non-Fiction Reading: Claims, Evidence, Bias and Source Reliability by asking the learner to find and combine the evidence rather than only evaluate a supplied packet.
The research cycle: Question → Search → Select → Verify → Synthesize → Return
| Stage | Reader job | Typical failure |
|---|---|---|
| Question | Define what you actually need to know. | Topic remains too broad. |
| Search | Use terms that reveal different parts of the question. | One vague query repeated many times. |
| Select | Choose sources because they answer the question. | Famous or attractive pages selected regardless of relevance. |
| Verify | Check authority, evidence, date, scope and corroboration. | Claim accepted because several pages repeat it. |
| Synthesize | Combine sources by relationship, not by copy-and-paste order. | Paragraph becomes Source A, then B, then C. |
| Return | Present the answer with boundaries and traceability. | Opinion becomes stronger than the evidence. |
Begin with a question small enough to investigate
Topic: “Public transport.” Too broad.
Better question: “What makes public transport useful for students travelling to school?” Still broad, but workable.
More researchable version: “Which features of a public-transport journey most affect whether Secondary 2 students can arrive at school reliably: travel time, transfer count, waiting time or disruption recovery?”
The last version gives the search a structure. It also prevents the project from becoming a general essay about transport being “good for society”.
A useful question contains a decision boundary
Research questions improve when the learner can say what an answer would help them decide. “Is social media good or bad?” invites opinion before evidence. “What kinds of school communication are least likely to be missed by students?” directs attention to access, timing, reminders and message design.
Ask: What would I know after answering this that I do not know now? If the answer is “more facts”, narrow the question further.
Search is vocabulary applied to a knowledge problem
A single query often reflects only one way of naming the problem. Good research uses synonyms, related concepts, official terminology and narrower sub-questions.
For a question about student travel reliability, possible search terms might include journey reliability, public transport punctuality, service disruptions, school commute, transfer waiting time and the names of relevant agencies or datasets. Each term opens a different route.
This does not mean using every synonym. Search terms should be chosen because they reveal a missing part of the question.
Search results are leads, not evidence receipts
A search-result title or snippet is designed to help you decide whether to open a page. It is not enough to support a research claim. Open the source, locate the original statement, check its context and record the page itself.
If the result says “study finds 80% improvement”, ask: improvement in what, compared with what, among whom, measured how and over what period? The number becomes useful only after its relationship is recovered.
Source selection: relevance before prestige
A prestigious source can still be irrelevant to your exact question. A transport agency may be authoritative about service schedules and official operational data. A student interview may be the direct source for that student’s lived experience. A newspaper article may provide context or competing interpretations.
The useful question is not “Which source is best?” but “Best for which claim?”
The source-role table
| Source type | Often useful for | Common limitation |
|---|---|---|
| Official agency | Rules, schedules, definitions, administrative data. | May not describe every user experience. |
| Academic or research source | Methods, measured findings, reviewed evidence. | May study a different population or context. |
| News report | Events, interviews, context, competing views. | May summarise rather than provide primary data. |
| Organisation or campaign | Its own position, goals and evidence. | Has an interest in how the issue is framed. |
| First-hand account | Individual experience and direct observation. | Does not automatically establish prevalence. |
| Reference work | Definitions and background concepts. | May not answer a current or local question. |
Verify the source before borrowing its confidence
Check five things: authority, evidence, currentness, scope and independence.
Authority: who produced the information and why are they in a position to know it?
Evidence: what supports the claim—data, documents, observation, expert explanation or another source?
Currentness: does the date matter for this question? A grammar definition may remain useful for years; a transport schedule may not.
Scope: does the source concern the same population, place and problem?
Independence: are apparently separate articles actually repeating the same original source?
Three websites can still be one piece of evidence
Imagine three articles state that a school programme increased participation by 40%. Article B cites Article A. Article C cites the same school press release that Article A used. You do not have three independent measurements. You have one originating claim travelling through three pages.
This is why research should trace a claim towards its source when the claim matters. Repetition increases visibility; it does not automatically increase independence.
Original research case: Should a school keep printed notices?
Research question: “When is a printed school notice still useful if most routine communication is digital?”
Possible evidence categories: access to devices, need for information at a physical location, ability to update information, risk of outdated copies, urgency, accessibility and audience habits.
A weak search collects pages arguing “paper versus digital”. A stronger search assigns sub-questions: What problems do digital-only systems create? Where do physical signs serve a location-specific role? How are urgent updates handled? Which source could verify each issue?
Take notes in claims, not copied paragraphs
Instead of copying three paragraphs into a document, record:
- Claim: Printed signs can be useful at a physical point of decision.
- Evidence/source: Example or source showing users seek information at that location.
- Scope: Supports location-based use, not every printed notice.
- Question: Does the source show actual behaviour or only recommend a design?
This note format is slower at first and much faster later because the paragraph is already partly organised by reasoning.
Separate quotation from paraphrase from your own conclusion
Mark copied wording clearly in notes. Record the source immediately. Paraphrase only after understanding the idea. Then write your own conclusion in a separate line.
This prevents source language from becoming invisible in your draft. It also protects meaning because you can compare your paraphrase with the original before publishing or submitting.
Use the Summary, Paraphrasing and Meaning Preservation guide when note compression repeatedly changes scope or certainty.
Synthesis is not a queue of summaries
Weak structure: “Source A says… Source B says… Source C says…”
Stronger structure: “Several sources agree that fast updating is an advantage of digital notices, but location-based information creates a separate need that digital convenience does not remove. The disagreement is therefore not simply paper versus screen; it concerns which channel fits which reader job.”
The paragraph is organised by the relationship among sources. Individual sources still need attribution where appropriate, but they serve the argument instead of dictating the paragraph order.
Four synthesis relationships
| Relationship | What to write |
|---|---|
| Agreement | Sources support the same conclusion for similar or different reasons. |
| Qualification | One source narrows another source’s broader claim. |
| Conflict | Sources make incompatible claims about the same issue. |
| Gap | Available sources answer part of the question but leave a relevant issue unresolved. |
Do not force disagreement when sources address different questions
One source may measure how often students use a study room. Another may interview students about why they use it. Their findings can differ without contradicting one another because frequency and motivation are different questions.
Comparative reading helps here. Use Comparative Reading: Viewpoint, Agreement, Difference and Evaluation to align sources on the same axis before declaring conflict.
Numbers require a denominator and a definition
If a source says “participation rose by 50%”, ask whether it rose from 20 to 30 or from 2,000 to 3,000. Both are 50% increases, but the scale differs. Ask what counts as participation and whether the same definition was used before and after.
Numbers are not self-explanatory. Record the quantity, group, time, unit and measurement when they matter.
Original evidence ledger
| Research claim | Evidence | Source role | Boundary |
|---|---|---|---|
| Digital notices are easier to update. | Administrative example of changed venue. | Operational source. | Does not prove every student sees the update. |
| Printed notices can help at physical locations. | Observed need at an entrance or service point. | Location-based observation. | Does not justify printing every message. |
| Digital-only access can exclude some users. | Direct account of delayed access. | First-hand experience. | Does not establish prevalence. |
A ledger like this makes uncertainty visible before the conclusion becomes too strong.
Citation is part of reasoning
A citation does more than avoid plagiarism. It lets the reader inspect where a claim came from and how much authority the source should carry.
During school work, follow the citation format your teacher requires. The important habit is to preserve author or organisation, title, date when relevant and a stable route back to the source as you take notes, not after the draft is finished.
When the evidence is uncertain
Research does not always end with certainty. You may find two plausible explanations and insufficient evidence to choose between them. State that limit.
Useful wording includes: “The available evidence suggests…”, “The sources agree on X but do not establish Y”, “This conclusion applies to the surveyed group”, or “Further evidence would be needed to determine whether…”
This is stronger than pretending the gap does not exist.
Original mini-project: A better after-school study space
Question: “What conditions would make an after-school study space useful to more students?”
Build four sub-questions: access and timing; noise and environment; support available; transport home. Find one relevant source type for each. Record at least one limitation for each source. Then synthesize the findings into three design principles rather than four source summaries.
Possible synthesis: “A useful study space needs a workable time, a clear purpose and support that students understand correctly. Quietness may help concentration, but transport and teacher availability can limit access. A trial should therefore measure not only attendance but why students attend and why others do not.”
The model above is an original teaching answer, not a research conclusion about a real school.
A research error clinic
| Error | What failed | Repair |
|---|---|---|
| First search result becomes main source. | Selection. | Compare source roles before committing. |
| Three sites repeat same press release. | Independence. | Trace the originating source. |
| Statistic copied without denominator. | Scope. | Record group, measure and time. |
| Quote copied into notes without marks. | Source boundary. | Separate exact wording from paraphrase. |
| Paragraph organised Source A/B/C. | Synthesis. | Organise by agreement, conflict, qualification or gap. |
| Outdated page used for current rule. | Currentness. | Check publication or update date and official owner. |
Independent transfer challenge
Choose a school-life question that matters to you but is not answered by personal opinion alone. Examples: whether a club should change meeting time, what makes a notice easy to act on, or how students choose effective study spaces.
Write one researchable question and three sub-questions. Find three sources that play different roles. For each source, record one supported claim and one boundary. Then write a 250-word synthesis organised by ideas rather than by source order.
Finally, remove the source names from your notes and ask whether you can still explain why each claim is supported. If you cannot, you may have collected information without building a model.
For parents and teachers
Do not begin by asking students to find “five websites”. Ask them to find evidence for three specific sub-questions. This turns quantity into purpose.
Require a boundary with every strong claim. “What does this source not prove?” is one of the most productive research questions a learner can be taught.
When two sources disagree, resist solving the disagreement immediately. Ask whether they use the same definition, group, time period and question. Many apparent contradictions disappear when the axes are aligned.
Continue through the Secondary 2 gap set
Use Poetry and Spoken Text: Voice, Sound, Form, Image and Meaning when research gives way to close literary interpretation. Use Discussion and Seminar Skills when sources must be tested through conversation. Use Multimodal Representing when the findings need to become a presentation, infographic or spoken explanation.
Return to the Secondary English Learning Hub for the complete sequence.
Research begins before the search box. Decide what you need to know, find evidence that can answer it, preserve the source boundary, then build an answer that another reader can trace back to the world.
