A scientific report might say that one group of seedlings recorded an 85% survival rate while another recorded 80%. A publicity poster might turn the same figures into ‘Our method guarantees healthier plants.’ In English Paper 2, the central reading skill is not solving the experiment; it is identifying what each text actually says, the effect of its wording and what can reasonably be inferred. This guide teaches those distinctions through an original fictional habitat crop-trial story, with a visual-text description, narrative passage, informational passage, practice questions and a model summary.
The previous Biology guide on osmosis owns the scientific explanation, while Mathematics guide on pooled rates owns the calculations. The present article owns English comprehension: literal detail, inference, writer’s attitude, language for effect, text comparison and concise synthesis. This distinction prevents comprehension from becoming a second mathematics lesson. Readers moving from PSLE English revision can begin with evidence sentences before attempting the more complex task of evaluating a claim.
- Paper 2 K300: visual texts, narrative comprehension and non-narrative summary.
- Separate literal answers, inference and evaluation.
- Compare promotional and factual claims using evidence from each.
- Explain language effect rather than naming techniques alone.
- Build an accurate approximately 80-word summary with source limits preserved.
What the 2027 K300 paper really requires
The Singapore-Cambridge SEC G3 English Language syllabus is K300 in 2027, corresponding to the earlier 1184 code. English Paper 2, Comprehension, lasts 1 hour 50 minutes, carries 50 marks and contributes 35% of the subject’s overall weighting. Section A carries five marks and uses two texts, one with a visual element; Section B carries 20 marks and uses a narrative or recount; Section C carries 25 marks and uses a non-narrative passage, including a summary of about 80 words excluding provided introductory words. This is the official examination scheme, not a guess from a practice book. The tasks below are original illustrations and do not reproduce an actual SEC question.
A reader’s first three passes
On the first pass, work out the subject and purpose of the passage. On the second, underline the words that limit or qualify claims: ‘in this trial’, ‘may’, ‘not yet’, ‘although’ and ‘under identical conditions’. On the third, match each question to the sentence or detail that answers it. Do not highlight every line. The point of annotating is to find evidence efficiently. In an examination, excessive marking makes the page look busy without making reasoning sharper. A short margin note such as ‘observed rate’ or ‘writer’s caution’ is more useful than an elaborate colour code.
Literal, inferred and evaluated information
A literal answer is directly supported by what the text states. An inference combines details from the text into a reasonable conclusion. Evaluation considers the purpose, stance or effect of the writer’s choices. For example, ‘the notice was revised on Friday’ is literal if Friday is printed. ‘The editor became more cautious’ may be inferred if the revision removes a guarantee and replaces it with a limited claim. ‘The revision makes the message more trustworthy’ is a judgement that needs a reason. Examiners can ask for these separately, so the verbs in a question matter.
How visual language does work
Visual texts use placement, relative size, colours and short wording to direct attention. In a text-only lesson, we can describe a poster layout without inserting an image: a large heading ‘SUCCESS?’ occupies the top; a prominent ‘85%’ appears in the centre; small print below says ’51 of 60 seedlings survived in one classroom trial’. If the examiner asks about the impact of the large number, discuss its visibility and persuasive emphasis. If asked what the smaller line does, explain that it provides the specific scope and evidence behind the percentage. Do not confuse visual emphasis with factual proof.
The danger of reading only one number
A percentage is not an argument on its own. The reader needs the denominator, the comparison and the definition of success. But an English comprehension answer should not spend a page recalculating statistics unless the question asks about the meaning of the data. Explain how the writer presents a number to create confidence or caution. Compare ‘85% survived in our trial’ with ‘85% of all seedlings will survive’: the former limits itself to an observation, while the latter predicts a broad outcome. One changed phrase can turn a defensible report into a misleading promise.
Identify hedges and their purpose
Words such as ‘appears’, ‘may’, ‘approximately’, ‘so far’ and ‘under these conditions’ are sometimes called hedges or qualifiers. They limit the strength or scope of a claim. That does not automatically make the writer weak or uncertain about everything. A well-chosen qualifier may show precisely what the evidence supports. When a question asks about effect, explain the particular limitation: ‘so far’ signals that future results may differ; ‘under these conditions’ restricts the conclusion to the experiment; ‘approximately’ warns against false precision.
Paraphrase without changing the idea
A good paraphrase changes language while preserving meaning. ‘The revised report restricted its conclusion to the six trays tested’ can become ‘The new report made clear that its finding applied only to those six trays.’ It cannot become ‘The new report proved the method was harmless’ because the source did not say that. Precision matters especially with negatives and conditions. Try replacing a single key word at a time, checking that the relationship between evidence and conclusion is unchanged. Paraphrasing is not a search for impressive synonyms.
Language for effect: what to write
If a writer calls a confident claim ‘a promise built on a handful of seedlings’, the image suggests a weak foundation beneath an expansive assertion. A strong response identifies the expression and explains its impact: it makes the promise seem unreliable because the evidence base is small. Merely calling the phrase ‘effective’ or ‘dramatic’ does not answer how it works. Conversely, claiming that every image is a metaphor may overread ordinary description. Always connect the selected words with a specific effect on the reader.
Compare texts, not isolated phrases
In a comparison question, place the sources side by side. What does Text A emphasise? What does Text B emphasise? How do those emphases differ in purpose or tone? One might attract visitors to a greenhouse display; another might tell staff which data remain incomplete. Quote or refer to one detail from each if requested. A response that analyses only one text is not a comparison. Nor should you say ‘both say plants grow’ when one text is actually discussing the uncertainty of future growth.
Narrative inference needs sequence
A narrative may reveal a character’s motive through action, silence or a change of plan. Track the order of events. If a student notices that Mara deletes a sentence, rereads the table and invites a colleague to check the wording, these actions suggest caution and accountability. They do not prove that Mara was afraid of punishment or that the trial was dishonest. Inference should be plausible and proportionate to the evidence supplied. Avoid supplying an invented private thought to make the answer sound more elaborate.
Summary begins with the question, not the passage
The K300 Section C task includes a summary of about 80 words, excluding the printed introductory words. Identify the requested focus before selecting points. If asked to summarise reasons why a trial report should be revised, do not include decorative descriptions of the greenhouse, the characters’ feelings or the exact colour of the poster unless those details are relevant. Gather the required points, combine related ideas, use your own precise language and check the word count. The aim is accurate synthesis, not writing an attractive miniature essay.
How to avoid summary distortion
Compressing two claims may alter their relationship. ‘One method produced a higher recorded survival rate, but its trays began with stronger seedlings’ must not become ‘The method succeeded because it used stronger seedlings’, which asserts causation. Likewise ‘some measurements were missing’ is not the same as ‘all the data were fabricated’. Keep contrasts, conditions and negatives intact. A summary that is shorter but changes the source is not better. Re-reading the summary against the source takes less time than defending a confidently wrong sentence.
Use time with intention
Paper 2 lasts 1 hour 50 minutes and covers several kinds of comprehension. The syllabus does not prescribe a single minute-by-minute strategy for every learner. Students should practise a personal allocation that leaves time for all sections and a final evidence check. If a question is temporarily difficult, mark the evidence location and return after answering the more accessible items. A single inference should not consume the whole paper. The last review should check whether each answer uses the requested number of reasons and whether the summary stays near its specified length.
The examiner tests reading, not plant science
Even where a passage uses laboratory vocabulary, the comprehension questions may be answerable from the text itself. Background knowledge can help you understand a scene, but it should not override explicit information. If a passage says ‘the group used a different source of light’, do not replace it with an assumption about temperatures. Likewise, a student who knows osmosis should not insert its definition into a question about the writer’s tone. The best answer is the narrowest complete answer to the question asked.
How this differs from earlier English volumes
Vol 1240 focused on a blackout notice and the reading of urgent public communication. This volume examines comparison across a promotional display, a narrative decision and a trial report, then uses the information for a Section C-style summary. The common habit is evidence-led reading, but each owner addresses a distinct question type and setting. Studying them in order increases transfer instead of repeating the same notice analysis with a new title.
Original practice Texts 1 and 2: the two crop-trial notices
Text 1, a described promotional poster (fictional). At the top, large letters say ‘THE GROWTH-BED BREAKTHROUGH’. The centre displays ‘85%’ in very large type above the small line ’51 of 60 seedlings in Method B survived a fourteen-day classroom trial’. A green circle says ‘Come see what is possible’. At the bottom, in small type: ‘Results describe one comparison; further trials are planned’.
Text 2, an internal notice (fictional). ‘The display must identify the starting numbers for both methods. Method A recorded 48 survivors from 60 starts; Method B recorded 51 from 60. Visitors may see both results. The wording must not promise outcomes in future beds, because the sample is small and the growing conditions have not been independently reproduced.’
Read the two notices twice. Which is promotional? Which is procedural? Both include numbers, but they serve different audiences. One seeks interest; the other seeks accuracy of public claims. The difference is not that one must be false. It is that their word choices and visual presentation aim to produce different responses.
Original practice Text 3: Mara and the poster
Mara had promised the students that their greenhouse exhibition would look ready by Thursday. The printer had already produced thirty leaflets, each carrying the line ‘A method that works every time.’ When she checked the figures on Wednesday night, the words began to trouble her. There were six trays in the data record, and not one observation from outside the classroom. The higher survival rate was real within the table, but she could not find a note showing that every tray began with plants of comparable size.
At first, she reached for a red pen to change ‘every time’ to ‘usually’. It would have been an easy amendment. But the new word still spoke about trials that nobody had run. She put the pen down and asked Amin, who had designed the chart, to show her the original counts. He found a spreadsheet with every observation, including a missing height measurement from one plant that had not survived.
‘That doesn’t change the survival count,’ Amin said. ‘But it changes which height average we can describe.’ Mara nodded. She had thought the poster needed a more impressive number. Now she realised that it needed a more precise sentence. She crossed out the guarantee and wrote: ‘In this fourteen-day classroom comparison, Method B recorded a higher survival proportion. The result requires confirmation in further trials.’
By morning, the new text was less triumphant. The students had to move the percentage down from the centre of the leaflet to make room for the explanation. Mara expected complaints, but a visitor stopped at the revised chart and asked how the team had measured success. For the first time that week, she was pleased to answer the question in full.
The story does not tell us exactly what Mara privately feared. It shows her reasoning through action: questioning a promise, checking a source, separating two outcomes and accepting a less dramatic design. Those observable choices support a defensible inference about responsibility without allowing us to invent a hidden motive.
Original practice Text 4: why the report needed revision
A growing-bed trial can produce useful measurements without settling every question about the method. The classroom team began with sixty seedlings in each of two groups. Method A recorded forty-eight survivors, while Method B recorded fifty-one. The corresponding rates were 80% and 85%. These figures support a narrow comparison of survival in the recorded fourteen-day period. They do not establish which method would work better in every greenhouse.
The report originally used the phrase ‘guaranteed improvement’. The phrase exceeded the evidence for several reasons. The sample was limited to a small number of trays. The initial size and condition of individual seedlings were not fully recorded. Light, temperature and watering were described, but no independent team had repeated the work under the same protocol. Finally, the trial counted survival and collected some height measurements, which are different outcomes: an apparently successful survival record does not mean that every surviving plant grew to the same size.
To improve the record, the team preserved the original tray counts and identified which measurements were missing. It changed the headline to describe an observed difference rather than a universal promise. A new table displayed denominators beside the percentages, and a note explained that further comparable trials would be needed. The revised report remained positive about the students’ effort, but it invited readers to inspect the evidence and its limits.
None of these revisions erased the observation that Method B had three more surviving seedlings in the particular trial. Rather, the revisions changed the scope of the conclusion. A reader can value an interesting preliminary finding while also recognising that preliminary does not mean final. Careful reporting makes future experiments easier to plan because it shows precisely which questions have already been answered and which remain open.
Try summarising only the reasons for revision, not the chronology of the classroom project. The strongest points are the difference between an observation and a guarantee, the small and potentially unequal sample, incomplete height records, limited replication and the need to show denominators. The last paragraph gives the logical relationship: the narrower report keeps what was measured and removes what was not proven.
A model 80-word practice summary
Practice instruction: Summarise why the original trial report required revision, using approximately 80 words in your own phrasing.
The initial report made an unjustified promise from a limited classroom trial. Only a few trays were observed, and the starting condition of individual seedlings was not fully recorded. Some measurements were missing, while survival and growth were different outcomes that could not be treated as identical. The work had not been independently repeated under the same protocol. These weaknesses meant the reported percentage difference could not establish a universal improvement, so the report needed narrower claims and clearer denominators.
This is a teaching model for selecting relevant points; different correct paraphrases are possible. A summary is not evaluated by matching a particular sentence word for word. Check each assertion against Text 4, count the words, and notice how the paragraph omits the story’s timing and personal emotions.
Twenty-eight original comprehension exercises with explanations
These are practice examples, not examination questions issued by SEAB. Attempt each question and explain the supporting words from the passages before reading the answer.
Question 1: Poster headline
What response does ‘BREAKTHROUGH’ seek? It encourages excitement and the impression of a major discovery. To complete the answer, link the forceful headline to the purpose of attracting visitors. The word itself does not establish that the trial was scientifically conclusive; that is a separate question about evidence.
Question 2: Large percentage
Why is ‘85%’ displayed in large type? It makes the favourable rate the most immediately noticeable detail and can create confidence before readers inspect the sample size. A high-scoring explanation identifies the design choice and its intended effect instead of saying only that it ‘looks nice’.
Question 3: Small print
What does ‘one classroom trial’ limit? It restricts the observation to the specific recorded context, rather than promising all future growing beds will show the same rate. Do not paraphrase it as ‘only one seedling was tested’; the text gives a group of sixty.
Question 4: Audience contrast
Which notice is more likely written for staff preparing a display? Text 2. It gives procedural instructions about stating denominators and avoiding promises. Text 1 addresses visitors and invites them to come see the exhibition. Name evidence from both texts when the question asks for a contrast.
Question 5: Purpose of caveat
Why include ‘further trials are planned’? The phrase acknowledges that current evidence is not final and signals a next step. It may make the publicity sound more responsible. It does not guarantee that those trials will confirm the existing result. The future-oriented wording matters.
Question 6: What the reporter claims
Does Text 2 deny that Method B recorded a higher proportion? No. It explicitly lists the figures. It cautions that the recorded advantage cannot justify an unconditional prediction. An accurate comprehension answer separates a reported observation from the strength of an inference drawn from it.
Question 7: Narrative opening
What does Mara’s promise that the exhibition would be ready by Thursday establish? It creates time pressure and explains why revising leaflets is inconvenient. It does not prove that Mara intends deception. The story introduces a practical conflict between attractive presentation and careful reporting.
Question 8: A change of plan
Why does Mara abandon ‘usually’? She sees that the replacement word would still generalise beyond the trials conducted. The clue is the sentence saying nobody had run those additional trials. Good inference joins that clue to the decision rather than inventing a private fear.
Question 9: Amin’s spreadsheet
Why is the original data file significant? It permits Mara to inspect what was measured and discover the missing height record. The file improves transparency, but its existence does not automatically verify every measurement. The story’s point is that checking primary records improves the wording of a conclusion.
Question 10: Two outcomes
Why does Amin distinguish survival count from height average? An unmeasured height can affect the set used for an average without necessarily altering the number known to have survived. The reader should explain the different denominators or populations in plain language, not launch into unrelated calculations.
Question 11: Less triumphant
What does ‘less triumphant’ suggest about the revised leaflet? It is less celebratory or boastful because the text now includes qualifications. Do not state that the whole exhibition was a failure. The wording describes tone, not the value of the experiment.
Question 12: Visitor response
Why is Mara pleased when someone asks about measurement? The question lets her explain the evidence openly, which now matters more to her than displaying an impressive headline. The inference is supported by her earlier changes and the final sentence. Avoid claiming she is happy because the visitor agrees with the method.
Question 13: Writer’s stance
Does the writer condemn the students’ project? No. The narrative treats their work as useful while criticising the overconfident wording. A nuanced answer recognises both respect for the effort and concern about unsupported claims. Quoting only a critical phrase could distort the overall attitude.
Question 14: ‘A promise built on…’
If an examiner supplies the expression ‘a promise built on a handful of observations’, explain the implied weak foundation: a broad guarantee is unsupported by a limited evidence base. Identify the image and its effect. Do not label it a metaphor without explaining what the reader is encouraged to think.
Question 15: Chronology
Which happens first: the printer completes leaflets, Mara checks the figures, or the visitor asks a question? The leaflets are printed first, the data check comes next, and the visitor responds to the revised display later. Chronological questions are often simpler than inference questions, but only if the reader tracks the sequence.
Question 16: Meaning of preliminary
In Text 4, ‘preliminary’ refers to an early finding that may be useful but still needs confirmation. It does not mean the data are necessarily false. Context is crucial: the following sentences explain why further investigations would strengthen conclusions. An isolated dictionary meaning may be too broad.
Question 17: Reason one for revision
The report needed revision because ‘guaranteed improvement’ claimed more than the recorded fourteen-day comparison could establish. A summary point should capture the mismatch between scope and evidence. Merely stating that the headline was changed describes an action rather than the reason the action was needed.
Question 18: Reason two
The number of trays was small and the starting condition of individual seedlings had not been fully documented. The answer should treat these as limitations of generalisation and comparability. Do not convert the point into ‘the seedlings were all unhealthy’; the passage supplies no such claim.
Question 19: Reason three
A single class collected the observations, and no independent team had repeated the protocol. That limits confirmation across different settings. The passage does not say repetition never occurred within the class; it refers specifically to independent reproduction. The distinction between repeated measurements and independent testing should remain intact.
Question 20: Reason four
The passage distinguishes survival from height. Some height information was incomplete, so a claim about all plants growing better would be unsupported. A precise summary may say ‘the outcomes measured did not establish uniform growth’. It should not imply that the missing height value changed the known survival total.
Question 21: How the report improved
The revised text included denominators, preserved observations and identified the need for further trials. These steps help readers evaluate the finding. If asked for improvements rather than reasons for concern, select these positive actions. Summary scope matters: reasons and solutions are related but not interchangeable.
Question 22: Paraphrase safely
Original: ‘The report described an observed difference rather than a universal promise.’ A good paraphrase is ‘The new report limited its claim to the difference actually recorded.’ A bad paraphrase is ‘The new report proved the method failed everywhere.’ The latter reverses and expands the meaning. Accuracy outranks unusual vocabulary.
Question 23: Negative wording
Text 4 says the revisions did not erase the recorded difference. This double structure means the observation remained valid within its limits. A student who reads only ‘did not’ may answer that Method B had no advantage. Work through negation in full and identify what is being denied.
Question 24: Comparative answer structure
To compare Text 1 and Text 4, begin with the difference in emphasis, then support each side: Text 1 foregrounds a large favourable figure to attract attention; Text 4 foregrounds the limits of the trial to inform responsible judgement. End with one sentence explaining how that shapes reader response. Avoid two unrelated mini-summaries.
Question 25: Summary filter
A question asks for reasons to revise the report. Do not include the leaflet colour, Thursday deadline or Mara’s final pleasure: those details concern the narrative, not the requested informational passage. Include small sample, unequal starting conditions, incomplete measurements, missing independent replication and misleading guarantee. Filtering is as important as shortening.
Question 26: Word economy
The phrase ‘the results of the experiment which had been carried out by a group of students’ can often become ‘the students’ trial results’. Saving words allows the summary to include more essential points. Compression should remove empty structure, not qualifiers. Never discard a word like ‘not’ or ‘only’ just to save space.
Question 27: Inference without speculation
Mara asks a colleague to check the spreadsheet. A supported inference is that she wants to verify the evidence before making a public claim. An unsupported one is that she fears disciplinary action. State what her action suggests and cite the action itself. Inference should stay one step beyond the text, not several imaginary steps.
Question 28: Integrated final response
If asked whether the passage endorses the project, answer that it values the preliminary observation while urging caution about general claims. Support the first part with the retained numerical result and the second with the limitations and planned repetition. A balanced two-part answer is usually stronger than an absolute claim that the author is either wholly for or against the project.
FAQ: Paper 2 preparation
Must I know osmosis to answer the English question?
No. A comprehension question should be answered from the passage’s information and language. The linked Biology guide is enrichment, not a substitute for textual evidence.
Is the summary exactly 80 words?
The K300 syllabus describes a summary of about 80 words, excluding the supplied introductory words. Follow the exact wording of the actual examination question and maintain accurate content.
Do I need to recalculate 85% in Paper 2?
Only if the question requires it. Most comprehension questions here ask what the number means and how it is used by the writer.
Why practise across multiple kinds of text?
The paper assesses comprehension of visual, narrative and non-narrative material. Each calls for a slightly different reading habit.
Can I copy entire sentences into a summary?
The published assessment guidance rewards meaningful use of one’s own words and structures. Paraphrase accurately, preserve the key points and avoid inserting unsupported information.
Next reading
For the authoritative format, consult SEAB’s 2027 G3 English K300 PDF and the official 2027 G3 list. The wider Sengkang G3 EMS route connects the subject owners. Use the forthcoming cross-subject case review to practise communicating a decision grounded in all three skills; EMS is a study grouping, not a single SEAB subject code.
