Quick answer. When two texts describe the same activity, the reader must work out what each text is doing. A procedure tells someone how to act in an authorised situation. A warning tells them what not to assume or do. A result describes what was observed. A recommendation proposes a future choice. A permission statement says who may proceed, under which conditions. The words ‘may’, ‘must’, ‘only after’, ‘if’ and ‘has been’ are therefore part of the evidence, not small details to skim.
Volume 0622 used a fictional sensor calibration to show how an exact threshold, range or missing premise changes a mathematical conclusion. This English lesson shifts the focus from numbers to language. It teaches how to read an instruction panel and a visitor notice side by side without turning a trial into an approved product, a numerical rule into a safety certification, or a possible future action into a finished action.
Where this skill belongs in the SEC
The first Singapore-Cambridge SEC is in 2027. Its G3 English Language code is K300, replacing the previous reference 1184. The official K300 Paper 2 Comprehension lasts one hour 50 minutes, carries 50 marks and contributes 35%. Section A tests texts including a visual text; Section B uses a narrative or recount; Section C uses a non-narrative text and includes an 80-word summary response. This guide concentrates on the language of instructions, conditions, visual-text features and source comparison. It does not claim that the fictional sensor texts are an actual SEAB paper.
For current subject structure consult SEAB’s 2027 K300 syllabus and the G3 school-candidate subject directory (updated 26 October 2026). The text-only mock panels below imitate information features, not official visual materials or copyrighted past-paper passages.
The original paired texts
Text A: Teacher trial card. Title: ‘Room Sensor Reading Practice: Trial 4’. Subheading: ‘For supervised learning only’. Body: ‘Students may inspect the table after a teacher has explained the output terminals. The amber label appears when the modelled voltage is below 3.00 V. It indicates that the example’s numerical rule has been met; it is not a declaration of danger. Values represent a prepared teaching dataset. Students must not change connections, handle mains wiring or treat the example readings as an approved room alarm. Any proposed practical activity requires teacher authorisation.’ Footer: ‘Trial card prepared for a classroom discussion; not a school-wide installation notice.’
Text B: Visitor proposal notice. Title: ‘A possible reading board for Science Week’. Subheading: ‘Planning stage — subject to review’. Body: ‘The team proposes a text-only display that shows how one sensor reading can support or limit an inference. If the school approves the display, visitors will be able to read the trial table and consider competing explanations. Until approval has been given, the board remains a proposal. The project team hopes that the display will help learners distinguish evidence from certainty. The school has not announced any new monitoring system.’ Footer: ‘Questions about approval belong with the teacher coordinating the activity.’
The two texts are deliberately similar enough to create traps. Both mention a sensor and a reading board. Their purposes differ, however. Text A governs a supervised teaching exercise; Text B describes a proposed public-facing learning display. The threshold occurs in the first text, whereas the approval condition controls the second. The stronger reader does not merge those distinct responsibilities.
Find purpose before extracting details
Ask why each text was made. Text A sets conditions for a supervised class trial and explains what the amber indicator means within a model. Text B makes a conditional proposal and limits public assumptions. A question about ‘what visitors can currently do’ cannot be answered simply by quoting the classroom permission in Text A.
Purpose is not merely a broad word such as ‘inform’. The precise answer identifies the reader and the action: to guide students within a supervised trial, or to explain to potential visitors that an idea remains under review.
Audience changes the force of a sentence
‘Students may inspect’ addresses participants under a teacher’s supervision. ‘Visitors will be able to read if the school approves’ addresses a different audience at a later stage. The two sentences do not contradict each other: they apply to different people and conditions.
An answer that omits audience often sounds plausible but becomes false. Learn to add a short subject phrase: ‘students in the lesson’, ‘visitors after approval’, or ‘the coordinating teacher’.
Distinguish ‘may’ from ‘must’
‘May inspect’ gives conditional permission. ‘Must not change connections’ states a prohibition. ‘Must receive approval’ states an obligation. ‘May’ does not mean ‘will’, and a proposed ‘will’ following an if-clause is not an unconditional prediction.
Read modal verbs as part of the main information. In a comprehension answer, it can be necessary to quote or paraphrase the modal rather than replace it with a stronger claim.
An if-clause carries the missing condition
Text B says visitors will be able to read the board if the school approves it. The approval is not reported as completed. Removing ‘if’ turns a possible future arrangement into a present fact. This changes the meaning even if the rest of the sentence is repeated word for word.
When reading quickly, circle the condition and ask whether the evidence establishes that it has been satisfied. A conditional statement cannot be reported as an achieved result without evidence.
Present perfect can distinguish completion from proposal
The phrase ‘has not announced’ concerns a completed announcement or its absence up to the stated point. ‘Proposes’ concerns an idea under consideration. ‘Will be able to’ concerns a possible later stage. Tense works with context; it is not a cosmetic grammatical choice.
Suppose a draft report says ‘the school installed the system’. This cannot be supported by the paired texts because neither says installation occurred.
Headings tell you how to read the body
‘For supervised learning only’ qualifies every permission on the trial card. ‘Planning stage — subject to review’ qualifies every future-facing sentence in the notice. A reader who focuses only on a large headline about a board might miss the smaller condition that controls the entire message.
Real visual texts use size, placement and contrasting design to direct attention. Even in text-only practice, read title, subheading, body and footer as a connected whole.
A footer can limit the scope of a claim
The footer ‘not a school-wide installation notice’ is a correction to an interpretation someone might otherwise make. It matters because the body uses words such as sensor, reading and indicator that could be mistaken for an operational device.
Do not treat text below a main paragraph as irrelevant. Depending on the question, a note may be the most direct evidence for an answer about scope or authority.
Distinguish observed data from interpretation
The sample reading table contains values; the amber label is a rule applied to them; ‘dangerous’ would be a further judgement requiring grounds not provided. Strong reading keeps these levels distinct. An author can accurately describe a measurement while still overstating what it proves.
This is the English counterpart to the Science mechanism and Mathematics threshold lessons. It concerns careful meaning and warranted inference, not the separate Science marking scheme.
An inference needs both a clue and a limit
It is reasonable to infer that the team is aware a reading could be misunderstood, because both texts explicitly restrict interpretations. It is not reasonable to infer that there has already been an accident or a school-wide failure. The texts contain no such evidence.
When a question asks what the writer suggests, give the relevant clue and the modest conclusion it supports. Do not invent an exciting background story.
Negative statements carry positive information about scope
‘The school has not announced any new monitoring system’ does not prove the school will never adopt one. It simply denies that the announcement has occurred. ‘Not approved yet’ differs from ‘rejected’. ‘Not safe to infer danger’ differs from ‘proved safe’.
Negatives are often where comprehension answers drift. Preserve their exact target and time frame so that a denial does not become a much broader claim.
Contrast a claim with its qualifying phrase
A draft headline might say, ‘Our board detects dangerous temperatures.’ Text A supplies only a modelled voltage condition, and Text B states the board is still a proposal. The responsible correction would remove both unsupported detection and premature completion claims.
The skill here is evaluation, not merely fact-finding. A learner should ask whether the evidence supports the whole sentence, including adjectives and implied certainty.
Use the most economical precise answer
A short answer to ‘Why can the public not use the board yet?’ could be: ‘It is still a proposal and requires school approval before visitors can use it.’ That sentence states both the condition and the consequence without copying the entire text.
Long answers sometimes add unsupported details. Choose the smallest amount of evidence that fully satisfies the question; then stop.
Explain tone through language choices
Text A has a controlled, cautionary instructional tone, shown by explicit prohibitions and the phrase ‘for supervised learning only’. Text B is tentative and informative, shown by ‘proposes’, ‘if’ and ‘subject to review’. Calling both ‘serious’ would be too vague to explain their distinct effects.
Connect language to audience and purpose: the first seeks compliant behaviour, while the second manages public expectations.
Separate fact from persuasive hope
The project team ‘hopes’ the display will help learners. That is its aim, not proof of an achieved learning outcome. If a question asks which parts are verified outcomes, do not count hopes or proposed benefits as measured successes.
This kind of distinction matters when students read advertisements, school notices, policy proposals and media reporting beyond examinations.
A quoted word needs interpretation
If asked why ‘possible’ is used in Text B’s title, a useful answer is that the display has not received approval and remains under consideration. Simply writing that ‘possible’ means ‘maybe’ does not explain the function of the word in this particular notice.
Examiners reward a contextual explanation, not a dictionary substitution. Show how the term shapes the reader’s expectations.
Read numbers as parts of sentences
‘Below 3.00 V’ is a strict numerical condition, and Text A explicitly says it is a modelled threshold. If an answer uses ‘3.00 V or below’, it changes the boundary. If it writes ’30 V’, it alters the amount. A number’s unit and comparison word can be just as significant as a noun or verb.
This reading skill is a natural bridge from Volume 0622’s inequality work without turning the English examination into a Mathematics test.
Avoid importing information from outside the given source
A candidate may know real thermistors and monitoring systems, but a comprehension question usually asks what the passage supports. Use relevant external knowledge only where the question invites it. Here an answer about school approval must come from the two texts, not from guesses about ordinary school procedures.
A confident reader may still be wrong by answering a different, more general question. Textual discipline is a learnable habit.
Build a reliable response under time pressure
Read the stem first, identify which text and which line of reasoning it requires, locate two or three possible evidence phrases and choose the one that answers the exact wording. For evaluative questions, name the clue, the conclusion and any limit.
This sequence is not a secret formula for marks. It is a way to prevent a plausible but unsupported answer from consuming examination time.
Original practice: thirty distinctions readers need to make
Question 1: Who may inspect the table?
Best-supported answer. Students in the supervised lesson, after teacher explanation. Guard against: Anyone who walks past the school display is not granted that permission.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 2: Who can authorise a practical activity?
Best-supported answer. The teacher must authorise the proposed practical activity under the stated trial conditions. Guard against: A student team’s interest is not the same as approval.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 3: What does the amber label mean?
Best-supported answer. The given modelled voltage rule has been met. Guard against: It does not by itself establish danger.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 4: What is the exact comparison?
Best-supported answer. The label appears when modelled voltage is below 3.00 V. Guard against: ‘At or below’ introduces an equality case the passage does not state.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 5: What stage is the visitor board at?
Best-supported answer. It remains a proposal subject to school review. Guard against: Do not say it is already installed.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 6: Can visitors read the board now?
Best-supported answer. The text promises access only if approval is granted; it does not establish present visitor access. Guard against: A future conditional must not become a present fact.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 7: Has the school announced monitoring?
Best-supported answer. No new monitoring system has been announced in the notice. Guard against: That does not mean one can never be announced.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 8: Why mention a coordinating teacher?
Best-supported answer. To identify where questions about approval should be directed. Guard against: The footer clarifies authority rather than announcing a final decision.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 9: Which text contains a prohibition?
Best-supported answer. Text A expressly says students must not change connections or handle mains wiring. Guard against: A ban is stronger than a suggestion.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 10: What is the writer’s aim in Text B?
Best-supported answer. To describe a proposed learning display and keep expectations conditional. Guard against: It is not evidence of a completed educational benefit.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 11: Why say ‘for supervised learning only’?
Best-supported answer. It restricts the context in which the trial card may be used. Guard against: The same data cannot be treated as a public product specification.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 12: Why say ‘possible’ in the title?
Best-supported answer. It signals that approval and implementation remain uncertain. Guard against: The adjective moderates the apparent certainty of the headline.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 13: What does ‘hopes’ communicate?
Best-supported answer. An intended benefit of the display, not a verified result. Guard against: The passage contains no measured learning outcome.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 14: What is the purpose of the numerical rule?
Best-supported answer. To classify a prepared reading within the fictional teaching model. Guard against: It is not presented as a real emergency criterion.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 15: Explain a contradiction claim
Best-supported answer. There is no contradiction between supervised student access and conditional visitor access: the audiences differ. Guard against: Two sentences about different audiences can both be true.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 16: Explain a tone difference
Best-supported answer. Text A is directive and restrictive; Text B is tentative and informative. Guard against: Support those descriptions with actual modal verbs.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 17: What is unsupported about ‘already approved’?
Best-supported answer. Text B says approval is still required and remains pending. Guard against: The precise distinction is between requested and granted approval.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 18: What is unsupported about ‘unsafe room’?
Best-supported answer. Text A limits the label to an illustrative voltage condition and denies it is a danger declaration. Guard against: A model output is not a safety judgement.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 19: Which phrase most strongly indicates prohibition?
Best-supported answer. ‘Must not change connections’ directly forbids an action. Guard against: This differs from ‘should consider’, which would merely advise.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 20: Which phrase gives qualified permission?
Best-supported answer. ‘May inspect the table after a teacher has explained the output terminals.’ Guard against: The timing and supervision condition belongs in the answer.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 21: Who is the pronoun ‘it’ referring to?
Best-supported answer. In ‘it is not a declaration of danger’, ‘it’ refers to the amber label or that label’s meaning in context. Guard against: Check the immediate sentence and the paragraph purpose.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 22: What does ‘prepared teaching dataset’ imply?
Best-supported answer. The values were supplied for learning, not identified as live building measurements. Guard against: A student’s assumption about real sensors is unsupported.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 23: Write a supported one-sentence summary
Best-supported answer. The team proposes a supervised learning display, but visitor access depends on school approval. Guard against: Avoid claiming the display has already produced results.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 24: Choose the safer headline
Best-supported answer. ‘How a sample voltage threshold changes a classroom reading’ is supported better than ‘New device detects danger’. Guard against: A modest title should still communicate the lesson accurately.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 25: Find a purpose mismatch
Best-supported answer. If asked to warn prospective visitors, do not copy classroom component-handling rules without explaining relevance. Guard against: Purpose and audience control which details are useful.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 26: Test a causal inference
Best-supported answer. Why is the notice cautious? Its repeated qualification suggests the proposal could be misunderstood as approved. Guard against: Do not speculate about a past accident.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 27: Rewrite a modal accurately
Best-supported answer. Change ‘Visitors may now use the board’ to ‘Visitors may use it if the school approves it’. Guard against: The conditional is the main correction.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 28: Explain the footer
Best-supported answer. It assigns approval enquiries to the coordinating teacher rather than to students or the audience. Guard against: Small print can govern large implications.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 29: Cross-check the two texts
Best-supported answer. Text A is a present supervised class resource; Text B is a proposed public display. Guard against: State the contrast rather than treating them as interchangeable.
Why it matters. To check this answer, locate the exact word, phrase or text feature that permits the claim. Then try replacing the key modal or time expression with a stronger one. If meaning changes, the original qualifier belongs in the answer. This is how evidence supports inference instead of merely sitting beside it. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 30: One detail you should not infer
Best-supported answer. The exact date of approval is not given and cannot be calculated from the notice. Guard against: Recognise when the correct response is that information is absent.
Why it matters. An effective response names the actual audience and purpose. Do not copy large amounts of source text when a short explanation suffices. Ask which detail answers the question and whether any other statement limits it. Include that limiting statement only when it changes the conclusion. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 31: Distinguish not approved and rejected
Best-supported answer. ‘Pending review’ means no completed approval is reported; it does not say the project was rejected. Guard against: Avoid converting a delayed decision into an adverse decision.
Why it matters. In examination conditions, keep three elements separate: the source detail, the inference, and the limit. A true source detail may support only a narrow inference. Extending it into a safety claim, an approval statement or an invented outcome turns a good beginning into an unsupported final answer. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Question 32: What would make a stronger answer?
Best-supported answer. A concise point supported by a word or phrase from the relevant text, with no invented facts. Guard against: Precision usually improves when unnecessary claims are removed.
Why it matters. Correct an error by editing the point where meaning changed, not by adding more adjectives. Pay special attention to ‘only’, ‘before’, ‘after’, ‘if’, ‘may’ and ‘must’. A sentence can be grammatical, fluent and completely wrong about who is allowed to do what. A useful self-test is to cover the model answer and explain your conclusion to a reader who has not seen either text.
Long-answer clinic: a claim review
Consider the following claim: ‘The school has introduced a sensor alarm that warns everyone when a classroom becomes dangerous.’ The school has not introduced a system according to Text B; a proposed display awaits approval. Text A contains a classroom-model label below a voltage threshold, and it explicitly says the label is not a declaration of danger. The claim is therefore unsupported in two different ways. A precise answer separates those two failures rather than writing only ‘the statement is false’.
A carefully written evaluation might say: ‘The statement is misleading because the visitor board is still a proposal subject to school approval, and the amber label in the trial card indicates only that a modelled voltage rule has been met. Neither text says a monitoring system has been introduced or that danger has been verified.’ The explanation quotes no long block and does not need technical voltage calculations. Its strength lies in connecting the source to each disputed part of the claim.
Students should also notice what the model answer does not do. It does not decide whether the proposal is good. It does not speculate about actual wiring, weather or classroom safety. It does not assume the team was dishonest. It reads the two texts closely and makes a proportionate judgement about what their words permit.
A seven-day reading practice plan
On the first day, read only headings and footers; predict audience and purpose. On the second, highlight modals and negation. On the third, rewrite five conditional sentences without altering meaning. On the fourth, read a numerical notice and preserve exact thresholds and units. On the fifth, compare the two texts and explain one apparent contradiction. On the sixth, produce short answers under a modest time limit. On the seventh, analyse wrong responses by scope, evidence, modal language, audience and over-inference.
Keep a small error notebook with an original wrong answer, the passage words that disprove it and a one-sentence correction. A student who repeatedly omits ‘only after approval’ needs a different intervention from one who misidentifies the writer’s tone. Repair the first failing reading operation, then practise on an unfamiliar notice rather than repeating the same memorised response.
Connected routes and next step
The local English Hub contains the wider reading, writing and language pathway. The narrative inference work in Volume 0615 explores a different kind of source; Volume 0611 remains the general owner for claim-and-evidence reading. For a younger learner, the PSLE mechanism bridge shows why naming the topic is not the same as explaining it. The next EMS volume will bring accurate English back into a timed Science–Mathematics–communication problem.
A final reader’s test
If you can tell another person who was being addressed, what has actually happened, what is merely proposed, which rule applies and what cannot be inferred, you have read the informational material rather than only its attractive headline. This is useful in the SEC examination and whenever a real-world notice offers a promise, a condition or a warning.
Sources: SEAB SEC overview; 2027 subject directory; K300 English syllabus. Paired texts and questions above are original illustrative material, not reproduced official examination texts.
