Quick answer: read the claim, then test its limits
A scientific report can contain accurate observations and still make an overconfident conclusion. The key to advanced G3 English comprehension is to separate what the text states, what the figures support, what the writer assumes and what remains unknown. This guide uses an original root-tip research scenario to train non-narrative reading, visual evidence, language effect, summary selection and careful phrasing. No special scientific qualification is needed to answer a reading question; the aim is to understand the source on its own terms.
Start with five questions. Who is speaking? What was actually observed? Which comparison is being made? Which word limits the claim? What did the question ask you to do? A good answer is usually shorter and more precise than a confident guess. The following examples are invented classroom material, not previous or specimen SEC questions.
Where this lesson belongs in the 2027 SEC
The Singapore-Cambridge Secondary Education Certificate begins in 2027. The current G3 English Language syllabus is K300, replacing the earlier reference code 1184. Paper 2 Comprehension is 1 hour 50 minutes, 50 marks and 35% of the subject grade. Section A uses two texts including a visual text (5 marks); Section B is based on a narrative or recount (20 marks); Section C is based on a non-narrative text and includes an approximately 80-word summary response (25 marks). This is a strategy guide for those assessed skills, not a prediction of the exact theme of a real examination. The official syllabus also values literal, inferential and evaluative reading, vocabulary in context, and language for effect.
The original research text: A school club counts cells
A science club investigated what could be learned from prepared onion root-tip slides. Its first report drew attention to apparently dividing cells, which some members recognised by the condensed material visible in the nucleus. Volunteers examined fields of view and recorded the number of cells they judged to show clear division features. Their supervisor reminded them that a static image captures a moment: the image does not show a complete cell cycle or the rate at which a whole root lengthens. When a cell boundary was unclear, the club marked that observation as uncertain rather than silently forcing it into a category.
The first group counted 120 cells and marked 24 as showing visible division features. A second group counted 40 cells and marked 12. At an early meeting, one pupil pointed to the first tally and announced that the first group had more dividing cells. A classmate replied that the second group had a larger observed proportion. Both pupils could point to a number in the record; only the second had considered a fair proportional comparison. The supervisor added that even correct percentages do not remove differences in field selection, slide quality or classification rules.
The club revised its method. It decided in advance how to choose fields, recorded the total number of cells inspected, and agreed to flag unclear cases. Its next note did not state that any treatment had caused faster growth. Instead, it reported the observed proportions within the inspected material. A newsletter editor wanted a headline about dramatic results. The club proposed a more modest heading: ‘What a small slide sample can—and cannot—tell us.’ It explained that repeated sampling and a fair comparison would be necessary before drawing a stronger conclusion.
The scientific language matters, but the reading skill matters more. A reader must keep track of a claim’s speaker and scope. Is a percentage about one field, one slide or all plants? Is the table a method, an observation or a conclusion? Does ‘may suggest’ mean the same thing as ‘proves’? Do two figures use the same denominator? A careful reader can answer each of these questions without inventing facts outside the text. This distinction makes comprehension more reliable when the passage concerns science, finance, health, transport or any other unfamiliar subject.
Build a disciplined response before attempting harder questions
On a first reading, identify the writer’s central message in one sentence: the club improved the fairness of its counting and remained cautious about what its images proved. On the second reading, underline contrast words such as ‘however’, ‘although’, ‘instead’ and ‘only’. Circle the precise units and denominators. Use a simple evidence ladder: quotation or observed figure, accurate paraphrase, justified inference, and limitation. An answer that jumps from quotation to a much stronger claim has missed a rung.
For a literal question, locate the statement and answer it directly. For an inference question, combine details that lead to a reasonable interpretation while avoiding certainty not in the passage. For language effect, name the choice, explain the impression and connect it to the writer’s purpose. For evaluation, check whether the claim survives a change in denominator, context or missing information. For summary, choose relevant ideas and recast them economically. These moves are related but not interchangeable.
Thirty-two worked reading decisions
Each decision below gives a short extract or situation, a question, a model interpretation, the common overreach, and a small transfer task. Read the task before looking at the response. Explain your answer aloud if possible, then write a concise version under timed conditions.
Decision 01: A percentage that seems decisive
Text and question. A student report calls 24 dividing-looking cells out of 120 ‘one in five cells dividing’ and another sample 12 out of 40 ‘only twelve cells’. What conclusion is justified, and how should a careful reader answer?
Model response. Compare like with like: 24/120=20% and 12/40=30%, so the smaller count has the higher observed proportion. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. The denominators are different. Comparing 24 with 12 alone substitutes a count for a rate. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Rewrite the author’s claim with both sample sizes and percentages, then state what sampling information is still missing. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 02: The title promises too much
Text and question. A notice says ‘Microscope proves new fertiliser makes plants grow twice as fast’, but the only evidence is a single root-tip photograph. What conclusion is justified, and how should a careful reader answer?
Model response. The photograph may illustrate cell appearance; it does not measure whole-plant growth rate or establish a causal effect. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A photograph is not a time series, a controlled experiment or a guarantee of representative sampling. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Replace ‘proves’ with a fair statement and name one additional type of observation needed. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 03: The qualifier quietly changes the result
Text and question. A teacher writes ‘The prepared slides may suggest a higher proportion of visibly dividing cells in this field’. What conclusion is justified, and how should a careful reader answer?
Model response. The claim is expressly tentative and field-specific. It does not assert a population-wide difference. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Words such as ‘may’, ‘in this field’ and ‘visibly’ are part of the scientific meaning, not verbal clutter. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Underline all three qualifiers and explain how removing each would widen the claim. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 04: Observation versus explanation
Text and question. A report records dark, condensed structures in some cells and then states that mitosis causes root growth. What conclusion is justified, and how should a careful reader answer?
Model response. The first statement describes an observable feature; the second links a biological process to growth. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. You can accept a plausible biological explanation while still recognising that the image alone does not measure its effect. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Write one sentence of observation and one sentence of explanation without using them interchangeably. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 05: One field was chosen after inspection
Text and question. The writer admits that the clearest field was selected from eight possible fields for the published image. What conclusion is justified, and how should a careful reader answer?
Model response. The displayed field is selected, not necessarily representative. The choice may increase the apparent frequency of clear figures. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A reader should not infer the overall slide distribution from a field chosen precisely because it looked exceptional. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Propose a transparent field-selection method and one qualification to add to the report. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 06: A denominator is hidden
Text and question. A poster reports that ’18 cells were dividing’ but omits how many cells were examined. What conclusion is justified, and how should a careful reader answer?
Model response. The number of observed cases is given, but the proportion cannot be calculated without the total examined. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A strong answer does not invent an unstated denominator or treat eighteen as a percentage. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Draft a question to the author that would make the quantitative claim testable. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 07: Two different times
Text and question. The passage compares a morning slide photographed yesterday with an afternoon slide prepared a week earlier. What conclusion is justified, and how should a careful reader answer?
Model response. Time of observation and time of preparation differ. Without controls, a claimed treatment effect is uncertain. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Chronology is evidence; read temporal phrases rather than assuming that the two slides were handled alike. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Construct a two-row timeline and flag one confounding difference. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 08: A caption and paragraph disagree
Text and question. The paragraph says Sample B has 15 dividing-looking cells in 100; the image caption says 15 in 150. What conclusion is justified, and how should a careful reader answer?
Model response. The text contains an internal inconsistency. The percentages would be 15% and 10% respectively. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Do not silently repair an inconsistency. Identify exactly which source fields disagree and request verification. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Compose an answer beginning ‘The sources disagree about…’ followed by the numerical consequence. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 09: The author is quoting a pupil
Text and question. A science club article says one pupil called the experiment ‘certain proof’ before its supervisor warned that no control group was used. What conclusion is justified, and how should a careful reader answer?
Model response. The exaggerated assertion belongs to the pupil; the qualifying judgement belongs to the supervisor. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Speaker attribution is essential to a fair inference question. You cannot assign a quotation to the whole article. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Name who made each claim, and state whose judgement the published author ultimately endorses. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 10: A cause is replaced with an association
Text and question. Slides from two schools differ in observed cell patterns and the writer concludes that one school’s water caused the change. What conclusion is justified, and how should a careful reader answer?
Model response. The observation reports a difference between groups; it does not isolate water quality as the cause. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Other variables, including sampling, slide preparation and root age, could account for the difference. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Rewrite the conclusion as an association and propose a fair controlled comparison. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 11: A graph uses two scales
Text and question. A column chart shows counts for Group A and percentages for Group B without saying that the vertical axes differ. What conclusion is justified, and how should a careful reader answer?
Model response. The bars cannot be compared as though they measured the same quantity. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Graphic features are part of multimodal reading. Check units, baselines and axis labels before accepting visual prominence. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Explain precisely why the taller bar does not necessarily mean a higher rate. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 12: A missing label looks harmless
Text and question. A microscope table contains columns headed ‘Field’, ‘Total’, and ‘Dividing’, but no definition of ‘Dividing’. What conclusion is justified, and how should a careful reader answer?
Model response. The reader knows the counts but not the decision rule used to classify cells. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Operational definitions matter because an ambiguous borderline cell can change the tally. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Propose a short classification note without pretending every stage can be confidently identified. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 13: The passive voice conceals an actor
Text and question. A report says ‘Twenty difficult cells were excluded’ and gives no reason. What conclusion is justified, and how should a careful reader answer?
Model response. A methodological choice has been made, but the actor, criterion and consequence remain unclear. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Passive structures can be appropriate, yet they can also obscure responsibility for selection. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Write two questions that would clarify who excluded the cells and why. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 14: The photo is illustrative, not exhaustive
Text and question. The report states ‘A representative image is shown’, yet gives no information about how representation was judged. What conclusion is justified, and how should a careful reader answer?
Model response. The word ‘representative’ is a claim requiring evidence of a selection procedure. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A single attractive image may support description but not automatically justify a general numerical estimate. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Suggest one sentence that would substantiate the selection process. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 15: The numbers refer to different tissues
Text and question. A passage compares dividing-looking cells in root tips with mature leaf cells. What conclusion is justified, and how should a careful reader answer?
Model response. The observations may concern different cell populations and purposes; a numerical comparison is not automatically fair. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Before explaining a difference biologically, inspect whether the comparison groups match the question. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. State the population for each denominator and explain the limits of direct comparison. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 16: A conditional conclusion
Text and question. A writer says, ‘If the counting rules were applied equally, the difference could be meaningful’. What conclusion is justified, and how should a careful reader answer?
Model response. The conclusion depends on a condition that the passage has not yet established. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Conditional language separates a hypothetical possibility from verified evidence. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Restate the conclusion without turning the condition into a fact. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 17: A misleading average
Text and question. The writer averages 10%, 20% and 40% from fields containing 10, 20 and 100 cells respectively. What conclusion is justified, and how should a careful reader answer?
Model response. The simple average, about 23.3%, is not the pooled fraction; the pooled count is 1+4+40=45 out of 130, about 34.6%. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. The English task is to interpret what the author claims. Maths checks help expose the ambiguity without dominating the reading response. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Explain in prose why ‘average percentage per field’ and ‘percentage of all observed cells’ answer different questions. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 18: A footnote reverses confidence
Text and question. Bold copy says ‘Every trial improved’, while the footnote states that two of five trials were excluded as unreadable. What conclusion is justified, and how should a careful reader answer?
Model response. The text supplies only three assessed trials; it cannot support a claim about all five. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Small print is still part of the source and may limit the headline’s reach. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Write a faithful one-sentence replacement for the headline. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 19: The sample has been renamed
Text and question. A paragraph begins with ‘cells in these slides’ but ends with ‘all onion roots’. What conclusion is justified, and how should a careful reader answer?
Model response. The conclusion shifts from observed material to an unexamined population. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Track the nouns. A move from sample to universal class is often the exact reasoning error being tested. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Underline the narrowest and broadest population labels and correct the final sentence. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 20: A correlation is treated as a timetable
Text and question. A report finds more visibly dividing cells in one prepared field and claims that growth will finish tomorrow. What conclusion is justified, and how should a careful reader answer?
Model response. A static count does not predict completion time unless further valid temporal evidence supports such a model. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Scientific language may sound technical while skipping a required inferential step. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Identify the missing time information and write a measured replacement. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 21: An adjective asks the reader to decide
Text and question. The article calls a difference ‘dramatic’ without giving the size of the groups. What conclusion is justified, and how should a careful reader answer?
Model response. ‘Dramatic’ is evaluation, not measurement. Assess the adjective against reported data. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A useful comprehension answer identifies both the language choice and its effect on the reader. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Complete: ‘The adjective encourages readers to…, although the evidence…’. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 22: A method is described as a result
Text and question. The passage says students ‘counted cells along five lines’ and then calls that statement a finding. What conclusion is justified, and how should a careful reader answer?
Model response. Counting along lines is a procedure. The resulting observations or figures are the findings. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. In summary writing, methods and results have different roles; include the one the question requests. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Sort four statements into method, observation, inference and recommendation. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 23: The source inserts a concession
Text and question. After presenting a clear pattern, the writer notes that one of the groups was small. What conclusion is justified, and how should a careful reader answer?
Model response. The concession limits certainty while acknowledging that a pattern appears in the available observations. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A nuanced answer can identify both the main claim and the concession without dismissing either. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Write a sentence using ‘although’ that keeps both ideas accurate. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 24: A result is marked provisional
Text and question. A school newsletter calls its table ‘preliminary classroom data’. What conclusion is justified, and how should a careful reader answer?
Model response. Readers should treat the table as an early result, not a final scientific conclusion. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. The label changes what can responsibly be inferred from the same numbers. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Explain why the word ‘preliminary’ matters when the report is shared publicly. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 25: A picture lacks a scale
Text and question. Two enlarged pictures show cells at different magnifications, but the caption labels both simply ‘close-up’. What conclusion is justified, and how should a careful reader answer?
Model response. Apparent cell size in the photographs cannot be compared without magnification or scale information. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A clear comparative answer must distinguish physical size from displayed image size. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Propose the caption information needed and a cautious visual conclusion. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 26: A negative result is ignored
Text and question. In four repeat trials, two suggest a higher dividing-cell share and two do not. What conclusion is justified, and how should a careful reader answer?
Model response. Reporting only the two positive trials would give an incomplete account of the evidence. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Evaluation means attending to evidence that weakens a preferred story as well as evidence that supports it. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Draft a balanced two-sentence conclusion that reports the split honestly. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 27: The question asks for purpose
Text and question. An author includes a method diagram before offering percentages and a short caution about sampling. What conclusion is justified, and how should a careful reader answer?
Model response. The diagram helps readers understand how counts were collected and why the conclusion may be limited. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Purpose questions ask what a feature accomplishes for its intended audience, not simply what it contains. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Answer using the pattern ‘By showing…, the diagram helps readers…’. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 28: The question asks for effect
Text and question. A report ends with ‘Look again before you claim certainty’. What conclusion is justified, and how should a careful reader answer?
Model response. The direct imperative asks readers to reconsider their assumptions and reinforces the cautionary conclusion. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Language-effect answers should connect the chosen words to reader response and wider argument. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Explain the effect without claiming the line makes the findings scientifically true. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 29: A summary prompt narrows the task
Text and question. The task asks for ways the club improved the reliability of its counting, not for the history of microscopes. What conclusion is justified, and how should a careful reader answer?
Model response. Select only measures such as consistent criteria, planned fields, repeat counts and complete records. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Summary quality depends on selection as well as paraphrase; attractive irrelevant details consume words. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. List the relevant measures, group similar ones, then produce one accurate sentence in your own words. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 30: The source leaves a decision open
Text and question. The final paragraph recommends a larger, blinded repeat before deciding whether two treatments differ. What conclusion is justified, and how should a careful reader answer?
Model response. The recommendation acknowledges unresolved uncertainty while proposing a way to test the comparison. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Avoid summarising this as ‘the treatments did not differ’; the passage says the present work cannot settle the question. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Write a final sentence preserving that distinction. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 31: A table combines observation and interpretation
Text and question. The heading reads ‘Healthy cells’, but the column records only cells with a visible nucleus. What conclusion is justified, and how should a careful reader answer?
Model response. A visible nucleus is an observation, not a sufficient test of overall cell health. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. Word choice in headings can silently smuggle an interpretation into apparently factual data. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Replace the heading with the observable feature and explain the advantage. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
Decision 32: A score is rounded too early
Text and question. A display rounds 17/64 to 27% and later calls the difference from 16/64 ‘2 percentage points’. What conclusion is justified, and how should a careful reader answer?
Model response. 17/64=26.5625% and 16/64=25%, so the difference is 1.5625 percentage points before rounding. This directly answers the evidence question rather than guessing what the writer intended beyond the passage.
Why a tempting answer fails. A reading answer should notice when rounding affects the apparent size of a comparison. The distinction is important because comprehension marks reward a warranted inference, not the most impressive-sounding statement.
Try it. Explain why raw fractions should be checked before narrating a small difference. Then check the words ‘all’, ‘only’, ‘may’, ‘because’ and ‘therefore’ in your own answer. These often reveal a hidden change of scope.
An original Paper 2-style language-effect clinic
Consider the sentence: ‘The photograph offered a window, not a verdict.’ The contrast between a window and a verdict works because a window gives a view while a verdict settles a judgement. The image invites inspection but does not finish the scientific argument. A weak answer says only that the phrase is a metaphor. A stronger answer names the contrast, explains the limited view and connects it to the writer’s caution about evidence. That explanation is the work the language is doing.
Consider next: ‘The bold headline outran the quiet footnote.’ The verb ‘outran’ personifies the headline as something moving faster than the caution beneath it. Readers may take in the exciting claim before noticing its limits. The sentence criticises careless presentation, not the use of headlines in itself. Avoid writing that the author wants to stop all science reporting; that would go beyond the text.
A worked 80-word summary exercise
Practice prompt: Summarise the steps the club took to make cell counting more dependable and its conclusions more responsible. Use the material in the original passage and your own words. In an official Paper 2 Section C summary, candidates write about 80 words excluding any provided introductory words; follow the wording and boundaries of the actual question.
Model summary (81 words): The club agreed on classification rules before counting and chose fields using a planned method rather than selecting attractive images. Members recorded the total number of cells examined and marked uncertain observations separately. They compared proportions instead of raw counts and checked whether samples differed in size. Their revised report limited its statements to the material observed. Rather than claiming a treatment caused faster growth, the club recognised the need for repeated sampling and fair comparisons before reaching a stronger conclusion.
Why it works: each item is tied to the requested improvements, not to general microscope history. The actions are grouped into method, calculation and conclusion. Wording has been recast rather than copied mechanically. The summary does not add a control group as though one was already used, and does not turn a proposed repeat into a completed trial.
Two timed practice routines
For a 12-minute precision drill, spend two minutes reading a 300–450-word informational passage, six minutes answering three questions that require different operations, two minutes writing a compact summary of the central improvement, and two minutes checking whether each answer overstates the source. Use passages from school notes, public information notices or science museum material only when permission allows. The aim is not to memorise specialised facts. It is to practise transfer across subjects.
For a 25-minute deeper drill, write a 500-word original informational passage about a measured classroom change, place a small two-column table alongside it, and ask a partner to find two ambiguous terms, one unjustified causal conclusion, one point of fair comparison and one sentence whose language effect can be explained. Revise the passage after feedback. Becoming a careful writer is also a way to become a careful reader.
How to correct your answer using evidence
Do not change your answer simply because a model uses different words. Compare the claim, the supporting source detail and the degree of certainty. Ask whether your answer conveys the same relationship. If a question asks why a writer included a diagram, a numerical description alone is not enough: you must explain its function for a reader. If it asks what can be inferred, an exact quotation alone may not perform the required reasoning.
When the source includes scientific data, read the table before interpreting the prose. Check the unit, denominator, sampling period and any note that changes the meaning. A good English response can name a mathematical weakness without turning into a long mathematical solution. If calculations are needed, show only the arithmetic that establishes the inference, then return to the author’s argument. The goal is to use every tool in the right place.
Your next route
This guide follows Volume 0653 on root-tip biology and Volume 0654 on pooled proportions. For visual-source inference with another scientific context, use Volume 0647 on chromatogram claims. For the transition from primary-school practice, use PSLE English revision without full papers; for related reasoning, the PSLE Mathematics word-problem guide and PSLE Science explanation guide are useful.
Official syllabus and exam reference
Check the current 2027 SEAB G3 syllabus index and K300 G3 English Language syllabus. SEAB explains the transition to SEC and its examination schedule at SEC official overview. This article is independent educational guidance; its fictional research example and exercises are not official examination questions.
