A water-filter poster can be eye-catching and still make a claim that its evidence does not support. G3 English Paper 2 tests whether you can read beyond the headline: identify a text’s purpose, notice small qualifications, distinguish statements from inferences and explain how language shapes a reader’s response.
This guide uses a fictional school exhibition notice, a short recount and an information report. It walks through literal answers, vocabulary, language for effect, evidence-based inference and a short summary. The scientific setting is simply the reading context. The assessed skill here is precise English comprehension.
- Section A: connect a headline, supporting detail and caution in a visual-text setting.
- Section B: track what characters know, say and change their minds about.
- Section C: distinguish claims and qualifications and compress relevant points accurately.
Follow the earlier learning route
Science named the limits of filtration; Mathematics showed how an output percentage differs from a removal claim. English now owns the reading questions that arise when those facts are written for an audience. These routes support, but do not replace, this guide.
- Volume 0965: Chemistry distinguishes filtered solids and dissolved substances
- Volume 0966: Mathematics distinguishes output, rate and percentage
- Volume 0964: English listening records numerical conditions
- Volume 0952: English Paper 2 mineral-supply reports
- G3 English owner for Sengkang readers
- PSLE English: who said what when speakers report others
- PSLE English: stop before extra unsupported claims
A caption can be true while the headline is too strong
Imagine a small school exhibition notice headed ‘Our New Filter Makes Water Safe’. Beneath it, a line in small type says ‘Visible sand grains were reduced in one trial; dissolved substances and germs were not measured’. The problem is not whether the photograph looks attractive. The problem is whether the headline makes a claim larger than the supporting statement.
A G3 English comprehension question might ask what the headline suggests, what the fine print limits, and why placing the two together could be misleading. The answer should not turn into a chemistry lecture. It should explain the relationship between language choices and the reader’s likely interpretation. You can say a headline makes an absolute promise while the body text provides only narrow observational evidence.
This guide treats scientific claims as reading material. Its subject owner is English Paper 2, not a manual for filtering water. Everything in the fictional school documents is invented for reading practice. The important transferable questions are: who speaks, what exactly is asserted, what evidence is given, what is only implied, and how does the form of the text influence its audience?
The actual 2027 G3 English Paper 2 format
The official 2027 SEC G3 school-candidate syllabus list identifies English Language K300, with 1184 as the 2026-and-earlier reference code. The 2027 K300 syllabus states that Paper 2 Comprehension is one hour 50 minutes, 50 raw marks and 35% of the English subject. Section A contains Texts 1 and 2, one involving visuals, and accounts for five marks. Section B uses a narrative or recount for 20 marks. Section C uses a non-narrative for 25 marks and includes a summary task of about 80 words excluding the provided introductory words.
The paper assesses literal understanding, inference, evaluation, vocabulary in context and language for effect. Section A may ask how layout, images and verbal wording work together. Section B often turns on chronology, motives, viewpoint and change. Section C calls for reading claims and qualifications precisely. Students are required to sit all four English papers, but this guide concentrates on reading and summary rather than writing, oral or listening techniques.
Our ‘visual-text’ material is represented here entirely in words to make it accessible without adding an image. This is a learning reconstruction, not a released SEAB passage, photograph, chart or paper. Never assume the fictional figures below are official examination information or evidence of real water safety.
Reading a visual text when you cannot trust the headline alone
A poster is designed to make readers notice some information before other information. A bold title, large numeral, arrow, icon or call-to-action may be more prominent than a technical qualification. In comprehension, ask both what the text says and how its features guide attention. ‘90% collected’ can be eye-catching; ‘90% impurities removed’ would be a different claim entirely.
The text’s purpose is often a useful first anchor. Is it recruiting volunteers, explaining an exhibit, selling a service, warning the public or sharing results? A school notice may combine several purposes: attract visitors and communicate a caution. A good answer names the actual purpose grounded in the provided text, rather than guessing a generic purpose such as ‘to educate people’.
Language-for-effect answers need evidence plus an explanation of effect. If a headline uses ‘guaranteed’, explain the certainty it conveys and why that may reassure readers. If a disclaimer uses ‘not tested’, explain how it limits that certainty. Merely saying ‘the words persuade’ does not identify the specific effect of this particular wording.
Inferences are licensed by details, not by your enthusiasm
An inference is a conclusion reasonably supported by the passage even if it is not stated word for word. It does not mean anything that might happen in real life. If a student hesitates before publishing an unverified claim, the text may support an inference that the student is concerned about accuracy or responsibility. It does not support a diagnosis of the student’s personality or a claim about what they will do ten years later.
To test an inference, point to the supporting clue and the link between the clue and the conclusion. If you cannot name a sentence, action or contrast that supports your answer, your interpretation may be too free. The same method helps with questions about a writer’s attitude: look for choice of verbs, comparisons, quoted speech, tone changes and the surrounding context.
Keep the source in charge. When a sentence says ‘the filter may have retained finer particles’, paraphrase the possibility as a possibility. Do not convert ‘may’ into ‘did’. When it says ‘only during the four-minute trial’, preserve that time boundary. Examiners often design questions around precisely these qualifiers.
Vocabulary in context: read the sentence before the dictionary
Words such as ‘retained’, ‘transparent’, ‘conclusive’, ‘qualified’, ‘representative’ and ‘tentative’ can have meanings shaped by their context. ‘Retained’ may mean kept in place; ‘tentative’ may mean not yet firm. The best paraphrase fits the actual sentence and grammar. A dictionary synonym can be wrong if it changes the tone or the relationship between ideas.
Consider ‘The team withdrew its sweeping claim’. Here ‘sweeping’ does not mean the physical act of brushing a floor. It means broad in scope, often too broad for the evidence. If the question asks for the effect of this description, explain that the claim covered more than the experiment could justify. Name the connotation rather than offering an unrelated synonym.
When a pronoun appears, track the nearest plausible referent and verify it against meaning. ‘This’ might refer to a test result, a caution or an entire preceding event. Write the noun phrase in place of the pronoun and reread the sentence. If it now makes sense without changing the passage’s point, you have likely found the referent.
How to write a concise, accurate short answer
Start with the command word. ‘Identify’ usually needs the relevant detail. ‘Explain’ needs detail plus reasoning. ‘How does the writer create the impression?’ requires a technique or language choice linked to a reader effect. ‘What can be inferred?’ needs an interpretation supported by a clue. Adding a paragraph of background chemistry will rarely rescue a failure to answer the English question.
Avoid repeating the entire passage. Select the minimum meaningful evidence and restate it clearly. If a question demands two ways, provide two distinct ways rather than two versions of the same observation. Number them mentally: one about the headline’s absoluteness; another about the disclaimer’s limitation. This prevents empty redundancy.
A particularly valuable proofreading habit is to remove unlicensed certainty. Replace ‘proves’ with ‘suggests’ where the text only gives partial grounds. Preserve ‘some’, ‘only’, ‘may’, ‘usually’, ‘not yet’ and ‘after’. These small words often carry the exact meaning that the examiner is testing.
A summary needs selection before compression
The 2027 syllabus specifies a summary of about 80 words, with introductory words supplied and excluded from the count. A good summary does not merely shorten every sentence proportionately. It selects the ideas relevant to the question, groups related points and expresses them in concise, accurate English. It does not include unrelated examples just because they are interesting.
First, mark the relevant portion and underline each distinct point required by the prompt. Next, remove repeated illustrations that express the same idea. Then group related points, such as preparation conditions, measured results, limitations and suggested next checks. Finally, rewrite in your own grammatical structure without changing the relationships. In the exam, the task’s particular scope determines what belongs in the answer.
Pay attention to nested qualifications. ‘The team found fewer visible grains after one trial but did not test dissolved substances’ contains both a finding and a limit. If the question asks for results and cautions, both matter. If it asks only for advantages of the procedure, that limit may still be relevant if it directly qualifies a claimed advantage. The prompt is your filter.
What makes the same passage more challenging
At first, a text may ask you to identify a straightforward number: 72 mL. A harder item asks what percentage the writer refers to and what the denominator must have been. Harder still, the passage may use that percentage as part of a persuasive argument. The best English response traces the exact wording and explains whether the argument follows from the stated data.
A second difficulty is source nesting: a school team writes a report quoting a visitor who quotes a poster. Do not attribute all three statements to the author. If the report says ‘A visitor called the filtrate pure, but the team rejected the label’, it is inaccurate to write ‘The team believed the filtrate was pure’. Attribution is not decorative: it changes the answer.
A third difficulty is time. A finding from a four-minute trial cannot automatically be rewritten as a permanent property of the filter. Listen for or notice ‘on this occasion’, ‘under these conditions’, ‘in the first trial’ and ‘further trials are needed’. Each is a useful boundary that protects the text from overreading.
A timed Paper 2 routine without rigid stopwatch rules
At the start, scan the overall structure so you know where each passage and question belongs. For a visual text, inspect the heading, body details, attribution and fine print together. For a narrative, follow the sequence of events and shifts in emotion. For a non-narrative, map the central claim, evidence, concessions and conclusion. Allocate time sensibly rather than treating every one-mark item as equally difficult.
When answering, locate evidence first, compose the sentence second and check whether your wording matches the requested quantity or relationship. For the summary, budget time to select points and rewrite them, then count approximately and tighten if necessary. The exam is not a test of literary flourish in every answer. A precise ordinary sentence is usually better than a grand one.
If you are stuck between two plausible interpretations, return to what the text actually allows. Ask whether one answer invents a motive, generalises from a single event or ignores a qualification. This often reveals the stronger option more quickly than rereading everything without a purpose.
Practise with feedback that names the error
Make a two-column record: my answer and what in the text supports it. If you cannot complete the second column, revise the first. Add a third column for the kind of error—unsupported inference, lost qualification, wrong speaker, vague effect, missed point or summary padding. A list of categories makes patterns visible over a week.
When a learner has a weak English answer, avoid the instruction ‘write more’. Ask ‘which word in the text changes the meaning?’ Sometimes the missing word is ‘some’; sometimes the error is an unsupported emotional interpretation. The fix may be one careful phrase, not two extra sentences.
The PSLE-to-secondary transition is an enlargement of reading responsibility. Primary learners already learn not to overstate a passage. G3 work asks them to handle longer arguments, multimodal features and summarising. The habits remain recognisable: read the question, locate evidence, answer precisely and stop.
Original text A: the exhibition notice as a text-only layout
Large headline: “A CLEARER RESULT IN FOUR MINUTES!”
Smaller subheading: “A school demonstration of separating visible suspended particles.”
Prominent figure: “72 mL collected from an 80 mL classroom sample.”
Visitor invitation: “Come and examine our two filter materials. Ask us what the numbers can—and cannot—show.”
Footer: “Our observation relates to one supervised classroom trial. We did not test dissolved chemicals, microorganisms or drinking-water safety. Do not drink samples.”
This constructed poster is a word-only exercise in visual-text hierarchy: the ordering and descriptions of type size stand in for a visual arrangement. It is fictional; no image, actual photograph or laboratory instruction is supplied.
Original text B: “The label nobody wanted to remove” — a fictional recount
Mira had been assigned the school exhibition’s final labels. By lunchtime, the others had brought the sample tables, checked the printed numbers and arranged the empty display bottles. Only the headline remained. On her screen stood three words: Pure Water Produced. She liked their confidence. They would look decisive above a neat table of results.
The problem began when Idris returned with the observation sheet. “We collected seventy-two millilitres from eighty,” he said. “And the liquid looked clearer than before.” Mira gestured towards the headline. “Exactly. It will attract visitors.” Idris did not correct her immediately. He placed the sheet next to the screen and asked whether anyone had measured substances that could not be seen.
Mira checked the worksheet again. It contained time, starting volume, collected volume and a note about visible grains. There were no results for dissolved material. For the first time, the empty cells looked more important than the filled ones. She remembered how readily she had admired a bright headline on another group’s poster the previous year, scarcely reading its notes.
“We could use ‘Clearer After Filtration’,” Idris suggested. Mira typed it slowly. The new headline looked less dramatic, but it did not make a promise the experiment had never tested. She added a small sentence explaining that the result came from one supervised demonstration. She had been reluctant to lose an impressive phrase; now she felt relieved that a visitor would not have to discover the limitation by asking awkward questions.
When a younger pupil stopped at the display, he pointed to the coloured numbers. “Does this mean we can drink it?” he asked. Mira did not rush to fill the silence. “No,” she said. “We measured collection volume and saw fewer visible particles. We did not test whether it was safe.” The pupil nodded, then asked how they could find out more. Instead of treating his question as a failure of the exhibit, Mira wrote it on a card marked Next Questions.
Original text C: “A result is not a guarantee” — a fictional information report
The Environmental Learning Club at Cedar School prepared a water-separation exhibit for its annual learning day. Its purpose was to demonstrate what a filter could make visible, and to help visitors distinguish observation from broader claims. The activity was not presented as a method for producing drinking water. The organisers used a prepared classroom mixture and displayed recorded quantities rather than inviting visitors to taste or handle samples.
In the first trial, the team began with 80 mL of a mixture containing visible suspended grains. Four minutes later, 72 mL had been collected below the filter. The collected liquid appeared clearer. The team described the difference as a reduction in visible suspended material and calculated the collected volume as 90% of the initial volume. This percentage was a volume relationship. It was not an estimate of how many dissolved particles or microorganisms had been removed.
The team repeated the demonstration under a different arrangement and recorded 66 mL after the same duration. The result did not by itself show that the second material was chemically better or worse. The setup differed, and the club had not measured all the conditions needed for a strict comparison. The organisers therefore removed an early sentence that had called the first filter ‘the best’. They replaced it with a small table that stated the starting volume, collection time and collected volume for each trial.
During a rehearsal, two volunteers interpreted the table differently. One said the first arrangement ‘processed more water’; the other said it ‘wasted less’. The report’s editor decided that both statements needed qualification. The table supported an observation about collection volume during the stated interval. It did not establish where the uncollected liquid had gone, whether the filtered liquid was pure or how either arrangement would perform over a long period.
The display’s revised notice described the observation in ordinary language. A bold line invited visitors to compare the two trials, while a caption under the table reminded them that the trials were not identical. A separate note said that water clarity and water safety are different matters. Some visitors found the revised version less exciting, but others used the caveat to ask better questions about what ought to be measured next.
The club did not want readers to confuse honesty with failure. A narrow conclusion can still be useful: visible suspended matter was reduced in a specified classroom demonstration. The next sensible step would be a properly controlled comparison and appropriate tests of the particular properties being investigated. Such work would require teacher oversight and established safe procedures. The club’s final message was not that every filter is unreliable, but that a result becomes more trustworthy when its limits are stated alongside it.
34 original comprehension questions with worked responses
These short questions are original practice linked to the three texts above, not copies of SEC paper questions. Answer each without rereading the model first. Then check whether your wording has preserved the detail, speaker, tone and qualification.
Question 01: Read the headline
Question. What does ‘A CLEARER RESULT IN FOUR MINUTES!’ draw attention to?
Suggested answer. An improvement in visible appearance during a short trial.
Reasoning check. The expression foregrounds a quick, appealing result. It does not state that the sample became pure. A strong answer links the wording to its immediate attention-grabbing effect.
Question 02: Check the footnote
Question. Why does the poster include the sentence about untested substances?
Suggested answer. To prevent readers from treating the visual change as proof of safety or complete purification.
Reasoning check. This identifies a precise purpose, not merely a claim that small print ‘adds information’. The caveat narrows what the displayed result supports.
Question 03: Read the 72 mL
Question. Which fact is stated in the prominent figure?
Suggested answer. 72 mL was collected from an 80 mL sample.
Reasoning check. Neither a time nor a health outcome is embedded in this single figure. Avoid replacing the text with the calculation it may prompt unless the question asks you to calculate.
Question 04: Find the audience
Question. What suggests that the notice is aimed at visitors rather than technicians?
Suggested answer. It invites people to come, compare and ask what the numbers mean.
Reasoning check. The imperative ‘Come’ and the ordinary-language invitation establish an accessible public-facing purpose. An answer about a specialist journal would not be supported.
Question 05: Avoid a headline inference
Question. Can the poster prove the recovered water was safe to drink?
Suggested answer. No; its footer expressly says drinking-water safety was not tested.
Reasoning check. The answer uses explicit text. It does not require independent scientific knowledge to contradict the claim.
Question 06: Recognise contrast
Question. Why is the confident headline paired with a cautious footer?
Suggested answer. The headline attracts attention while the footer limits interpretation.
Reasoning check. The contrast is a feature of how the document works. A complete response names both functions instead of calling the poster simply persuasive.
Question 07: Locate a reference
Question. In Text B, what does ‘their confidence’ refer to?
Suggested answer. The confident tone of the proposed words ‘Pure Water Produced’.
Reasoning check. The pronoun points to the headline’s bold certainty, not the students’ confidence in their social skills. Read the surrounding sentences.
Question 08: Identify the pause
Question. Why does Idris not immediately correct Mira?
Suggested answer. He instead invites her to inspect which results were actually measured.
Reasoning check. This is a text-supported description of his response. Do not invent a hidden motive, such as wanting to embarrass her, which is not established.
Question 09: Interpret the empty cells
Question. What does Mira realise when she sees the unfilled parts of the worksheet?
Suggested answer. That the team lacks measurements needed to justify its broad purity claim.
Reasoning check. Empty cells become significant because the proposed headline goes beyond the available evidence. This is a contextual inference anchored by the preceding dialogue.
Question 10: Describe the change
Question. How does Mira’s attitude to the new headline develop?
Suggested answer. She moves from reluctance at losing drama to relief that the wording is accurate.
Reasoning check. The story explicitly contrasts her earlier hesitation with later relief. Give both sides of the shift to demonstrate understanding of narrative development.
Question 11: Read the visitor question
Question. What is the function of the younger pupil’s question?
Suggested answer. It tests whether the display might encourage a false conclusion about drinkability.
Reasoning check. The moment brings the underlying risk into a conversation. It is not evidence that the pupil drank anything or was harmed.
Question 12: Read the final card
Question. Why does Mira write on a ‘Next Questions’ card?
Suggested answer. She treats unanswered questions as useful directions for further learning.
Reasoning check. The action illustrates inquiry instead of pretending the exhibit has explained everything. Avoid claiming she has already completed further tests.
Question 13: Interpret the title
Question. What is the significance of the title ‘The label nobody wanted to remove’?
Suggested answer. It highlights the tension between an attractive claim and a necessary correction.
Reasoning check. The title directs attention to the reluctance to change wording. It does not mean the students literally could not physically remove a label.
Question 14: Vocabulary in context
Question. What does ‘decisive’ suggest about the original headline?
Suggested answer. It suggests a firm, confident-sounding conclusion.
Reasoning check. The nuance matters: in the narrative the appearance of certainty is attractive before the group checks whether certainty is justified.
Question 15: Writer’s attitude
Question. What does Text B suggest about honest corrections?
Suggested answer. That accuracy may be more valuable than dramatic phrasing.
Reasoning check. This is supported by Mira’s relief and her later careful explanation. It is not a claim that every lively headline is dishonest.
Question 16: Purpose of Text C
Question. What is the club’s main purpose?
Suggested answer. To teach visitors to distinguish observations from larger unsupported conclusions.
Reasoning check. A complete answer captures the report’s central aim across paragraphs rather than repeating only a number from the first trial.
Question 17: Difference between trials
Question. Why can the second trial not be treated as a strictly fair comparison?
Suggested answer. The arrangement differed and important conditions were not all measured.
Reasoning check. A full inference names the missing comparability. Do not say the 66 mL result is automatically ‘worse’ without first identifying the question’s criterion.
Question 18: Correct the ‘best’ claim
Question. Why was the label ‘best’ removed?
Suggested answer. The trials did not support a general ranking of the filters.
Reasoning check. ‘Best’ requires a stated purpose and suitable evidence. The report instead gives data about collected volume under two different arrangements.
Question 19: The disputed word
Question. Why is ‘wasted less’ too strong?
Suggested answer. The trial did not establish where the uncollected liquid went.
Reasoning check. The word ‘wasted’ assumes a mechanism or value judgement not recorded. The safer statement reports the amount collected.
Question 20: A restricted conclusion
Question. What precisely did the first trial show?
Suggested answer. It recorded 72 mL collected in four minutes from 80 mL and clearer-looking liquid.
Reasoning check. The question asks for reported observation, not a new claim of purification. You may mention the 90% volume relationship if the wording requires it.
Question 21: What 90% is
Question. What does 90% describe in Text C?
Suggested answer. The collected volume as a proportion of initial volume.
Reasoning check. This detail is vital because the text explicitly rejects reading it as an impurity-removal percentage.
Question 22: Rephrase a caution
Question. Paraphrase ‘The table supported an observation … not a guarantee’.
Suggested answer. The measurements justify a limited finding about that trial, not broad claims about all conditions.
Reasoning check. Good paraphrase preserves the restriction while changing structure. Simply swapping ‘supported’ for ‘helped’ adds little.
Question 23: Identify concession
Question. What is the effect of ‘Some visitors found the revised version less exciting’?
Suggested answer. It acknowledges a cost to the display’s appeal before stressing the benefit of clearer questions.
Reasoning check. The report admits a drawback rather than pretending accuracy always feels entertaining. Recognising concession strengthens an evaluation answer.
Question 24: Evidence boundary
Question. Does the report establish how both filters would work for a whole day?
Suggested answer. No; it records brief demonstrations under specified conditions.
Reasoning check. This is a time-scope question. A short observation cannot be inflated into a long-term guarantee unless the passage provides data.
Question 25: Author’s conclusion
Question. What is the final message of Text C?
Suggested answer. Trustworthy reporting includes the limits of what was measured.
Reasoning check. Do not rewrite this as ‘filters never work’. The report expressly rejects that absolute conclusion.
Question 26: Cross-text connection
Question. How do Mira’s revision and the club report illustrate the same principle?
Suggested answer. Both replace broad promotional language with claims matching available observations.
Reasoning check. This answer identifies a shared principle without confusing a narrative event with a reported trial.
Question 27: Source attribution
Question. Who said the first arrangement ‘processed more water’?
Suggested answer. One of the volunteers during a rehearsal.
Reasoning check. The report records the volunteer’s interpretation; the editorial decision then qualifies it. It would be wrong to attribute the unqualified statement to the editor as a final conclusion.
Question 28: Tone of ‘guarantee’
Question. What effect would ‘guarantees pure water’ have if used on the poster?
Suggested answer. It would imply certainty about purity beyond the stated tests.
Reasoning check. The question is about wording and effect. You do not need to explain every scientific method that would be necessary in practice.
Question 29: Word choice: ‘narrow’
Question. What does ‘a narrow conclusion can still be useful’ mean?
Suggested answer. A carefully limited finding remains informative even if it does not answer every question.
Reasoning check. The adjective concerns scope, not physical width. Context determines the correct vocabulary interpretation.
Question 30: Summary selection
Question. The summary asks for limits of the team’s results. Should you include the invitation to come to the exhibit?
Suggested answer. No; it is unrelated to the specified limits.
Reasoning check. Relevant selection comes before elegant writing. A fact may be true yet not belong in the summary because it does not answer the task.
Question 31: Summary compression
Question. How can ‘the experiment did not test dissolved chemicals; it did not test microorganisms’ be compressed?
Suggested answer. ‘Neither dissolved substances nor microorganisms were tested.’
Reasoning check. One grammatical structure conveys two related limits without repeating the same subject and verb. Compression must not remove the word ‘not’.
Question 32: Avoid overclaiming
Question. Replace ‘the first filter was effective’ with a passage-grounded statement.
Suggested answer. ‘In the first trial, 72 mL was collected in four minutes and visible suspended particles appeared reduced.’
Reasoning check. This wording identifies the actual observations. ‘Effective’ lacks a stated criterion in the text, so it is too general.
Question 33: Conclude with purpose
Question. Why does the final line reject both blind enthusiasm and blanket scepticism?
Suggested answer. It distinguishes the usefulness of measured results from the need to state their limits.
Reasoning check. The answer captures the report’s balanced judgement. Avoid responding with a personal opinion on school exhibitions unless asked.
Question 34: Trace viewpoint
Question. Which character in Text B first insists on checking what was actually measured?
Suggested answer. Idris.
Reasoning check. This is a literal question; a direct answer is better than a lengthy discussion of his personality. Answer the level of question actually asked.
A complete summary-writing exercise
Practice task: Based on original Text C, summarise the limitations of the club’s observations and the steps taken to avoid misleading visitors. Write about 80 words. In a real K300 examination, follow the exact given starting words and the specified lines rather than this broad teaching prompt.
Original model response: The club recorded clearer-looking liquid and collected 72 mL from an 80 mL mixture in four minutes. This showed a volume relationship, not removal of dissolved substances or microorganisms. Another trial produced less collected liquid, but different arrangements prevented a fair comparison. The team therefore withdrew its claim that one filter was best. The trials could not explain uncollected liquid or predict long-term performance. Further controlled testing was needed, and the revised display stated these limits clearly.
Notice the selection: the summary retains the collection observation, its narrow meaning, the non-comparable second trial, the removed ‘best’ claim, the absent long-term evidence and the corrective display wording. It omits the visitors’ reactions and the exhibition’s promotional invitation because those details do not directly answer the task. The model is a practice demonstration, not an official SEAB marking scheme.
Improving a weak response in four passes
Pass one: remove claims absent from the source. ‘The experiment proved all water can be cleaned’ is not supported. Replace it with ‘The club observed fewer visible particles in its collected sample’. This correction alone repairs the central evidence error.
Pass two: restore a lost qualifier. If your answer says ‘The team tested purity’, reread the passage: the team expressly did not test dissolved substances or microorganisms. A negative qualification is not optional decoration; it often changes the entire claim.
Pass three: make speaker attribution exact. A volunteer said the first arrangement processed more water; another called it less wasteful. The editor decided both statements required conditions. Put the interpretation with the person who voiced it, not automatically with the narrator or the final report.
Pass four: tighten the language. Remove ‘in my personal opinion’ from a question that asks for textual evidence. Replace ‘because it is better and good’ with a detail and its consequence. Reread the question and confirm your final sentence answers the requested part of the text.
Frequently asked questions
Does an English answer need chemical calculations?
Only if the comprehension item supplies and asks you to interpret a quantity. The English goal is to read accurately, not to replace a textual question with a Chemistry explanation.
Is a summary exactly 80 words?
The K300 document specifies a summary of about 80 words, excluding the provided introductory words. Follow the exact instructions on the actual question and prefer relevant content over padding.
Can I use knowledge outside the passage?
Background knowledge may help understanding, but an answer about what the passage states or implies must be justified by the passage. External facts do not authorise invented motives or outcomes.
Where does this lead next?
The next EMS volume combines the scientific boundary, the numerical comparison and the careful language of reporting into one integrated evidence review. For the whole English route, see the G3 English owner.
