G3 SEC English K300 | Learner’s Guide Vol 0189. Two statements can look contradictory while both are true. One may describe a particular group; the other may describe everyone. One reports an event in June; the other reports the position in September. One names what was observed; the other predicts what might happen. Skilled comprehension begins when a reader stops treating similar words as identical claims.
This original learner guide teaches a narrow, useful skill: compare claims across written and visual material without inventing a disagreement—or erasing a real one. It follows earlier work on evidence verbs and summary accuracy. It is for classroom practice, parents and tutors, not an official examination paper or an assurance that a particular question type will appear.
1. The current 2027 SEC English examination context
The SEAB 2027 G3 school-candidate directory, last updated 26 October 2026, lists English Language as K300. Under the official K300 syllabus, the four papers are Writing, Comprehension, Listening and Oral Communication. Writing and Comprehension each contribute 35 per cent; Listening contributes 10 per cent and Oral Communication 20 per cent. Do not mistake this cross-text practice for a fifth paper.
Paper 2 Comprehension lasts 1 hour 50 minutes, carries 50 marks and has three sections. Section A uses Texts 1 and 2, one including visuals; Section B uses a narrative or recount; Section C uses a non-narrative text and includes a summary response of about 80 words. The assessment objectives cover literal, inferential and evaluative comprehension, including multimodal texts and language for effect. These published details describe where careful claim-reading helps, not a promise that candidates must produce a particular comparative essay.
2. What looks like disagreement?
Suppose a notice says, “The school library opened for lunchtime study on Mondays.” A later announcement says, “The library is now open every lunchtime.” The sentences are not necessarily opposites. The first may describe an earlier trial; the second may describe an expanded timetable. Their shared subject hides a change of date and range.
A genuine contradiction would require the same relevant conditions and incompatible propositions—for instance, two records asserting that the same library was both open and closed at exactly the same time, in the same room and under the same meaning of “open”. Even there, verify whether one record has been corrected. Different wording alone is not enough to prove that sources conflict.
3. Reduce every claim to a testable core
Before comparing sources, identify five parts: who or what is discussed, what is asserted, when the claim applies, which group or setting it covers and how certain the wording is. You can call these subject, predicate, time, scope and strength. A sixth part—source or evidence—often tells you whether the statement is observation, explanation or prediction.
Write a compact paraphrase without changing its conditions. “Most participating pupils preferred the new schedule during the pilot” is not “Most pupils preferred the new schedule.” The word “participating” narrows the group, while “during the pilot” narrows the time. A student who strips those conditions has already changed the evidence before reaching the question.
4. The scope test: all, some, most and none
Quantifiers are small words with large consequences. “Some students completed the form” asserts that at least one did. “Most students completed the form” usually indicates more than half of the relevant group. “All students completed the form” allows no exceptions in that group. “No students completed the form” is a completely different assertion.
Two sources saying “some” and “most” are not automatically contradictory: if most participated, some participated as well. Conversely, “all” and “some did not” cannot both apply to the same group at the same time with the same meaning of completion. Never rewrite “not all” as “none”; the first permits partial completion, while the second denies any completion.
5. The group test: different denominators
A report may say “Eight in ten pupils who joined the club enjoyed the workshop,” while a caption says “Only four in ten pupils at the school attended it.” Both might be true. One is a satisfaction proportion among attendees; the other is an attendance proportion across the school. The percentages describe different denominators.
Students often compare the size of two percentages before finding the population each describes. This produces false conflict and false agreement. Ask “Eight in ten of whom?” and “Four in ten of whom?” before drawing a conclusion. A result about participants may be informative, but it cannot automatically represent non-participants who never answered the survey.
6. The time test: before, after and still
Imagine a January article says, “The footbridge is closed for repairs.” An April update says, “The footbridge reopened after inspection.” These reports can be mutually consistent: the bridge’s status changed. A sentence beginning “As of January” must not silently be presented as a current statement in April.
Look for dates, stages, phrases such as “during the initial trial”, and words such as “previously”, “now”, “subsequently” and “remains”. “The device was unreliable in early tests” does not conflict with “The redesigned device passed later tests” if the design and testing conditions changed. Chronology is evidence, not decoration.
7. The measurement test: the same thing?
“Average waiting time fell” may coexist with “The longest queue became longer.” One compares a mean; the other compares an extreme value. Similarly, a sentence about total electricity used need not disagree with a statement about electricity used per device. Different metrics can change in opposite directions.
Name the measurable variable before comparing outcomes. A graph caption may show the percentage of respondents who completed an activity, while the accompanying prose describes how many minutes an activity took. Do not convert one into the other. If the measure is not stated clearly, say what is missing rather than manufacturing the connection.
8. The certainty test: might, suggests, demonstrates
“The results suggest a benefit” is weaker than “The results demonstrate a benefit under controlled conditions”, and both differ from “The benefit will occur for everyone”. The first leaves room for alternative explanations; the second claims stronger support in a particular setting; the last makes a universal prediction. The verb changes the evidential strength.
This chapter extends Volume 0185 on evidence verbs. We now compare the full propositions, not merely the strength of individual verbs. A sentence may use a cautious verb yet still be misleading if its subject, group or date is wrong.
9. The source test: observation or interpretation?
A table may record twenty people arriving before 8 a.m. A writer may call this “excellent punctuality”. The count is an observation; “excellent” is an evaluation requiring a standard or comparison. A second writer who calls the same result “insufficient” might be using a different benchmark. Their evaluations differ even though they agree on the count.
A strong response separates the shared data from competing interpretations: “Both sources report twenty early arrivals, but the writers assess that number differently.” This is better than saying “The sources disagree about the attendance figure” when the numerical record actually matches.
10. Primary evidence, repeated reporting and independent proof
Three websites copying the same press release are three presentations, not three independent observations. Their agreement might show that the quotation circulated consistently, but it does not provide three separately measured confirmations. If a question asks whether several sources strengthen the claim, identify whether they rely on the same underlying data.
The earlier source-independence workshop, Volume 0133, addresses this trap. In Paper 2, the immediate task may simply require an inference or comparison; bring in source independence only when the texts make that issue relevant. Do not turn a straightforward comprehension answer into an unsupported investigation.
11. Work through a complete two-source example
Fictional text A — pilot note: “During a four-week trial, 72 of the 90 pupils who tried the after-school reading room rated it useful. Some said that the seats were noisy after 4 p.m. The trial team recommended continuing the room while improving quiet study areas.” Every date, group and figure in this paragraph is invented for practice.
Fictional text B — school bulletin: “At the end of term, 120 pupils used the reading room at least once. The school expanded seating after complaints about noise and will review demand next term.” These statements do not refute the first survey result. Text B refers to unique users by the end of term, whereas Text A reports usefulness ratings among 90 trial participants and notes earlier noise concerns.
A defensible synthesis is: “The trial survey found strong approval among participating pupils, although some reported noise. The later bulletin describes increased usage and a seating response, but it does not report a new usefulness rating.” The answer preserves what is known and what is absent. It does not claim that the expansion caused all 120 pupils to participate.
12. How a visual text can silently change the comparison
Suppose a bar chart displays counts for weekdays, while a photograph shows a crowded room on one Saturday. A writer might use the photograph to imply that the room is crowded every day. Yet one image records a particular view at one time; a chart may summarise many observations over an interval. Check the caption, axes and sampling period before combining them.
The 2027 K300 Paper 2 Section A includes a visual text. Visual details can carry meaning through layout, labels, images and scale. The reader should ask what the visual actually shows, not what its design encourages us to feel. If a bar has a shortened axis, its apparent height difference may exaggerate the numerical change without altering the values printed beside it.
13. Different aims can explain different emphases
A fundraising appeal may highlight the number of pupils who benefited from a programme, while an evaluation report notes its cost and limits. Both can use truthful facts yet select those most relevant to their aims. Understanding purpose helps explain why certain details appear; it does not automatically license you to call the writer dishonest.
To support a claim about persuasive purpose, identify actual choices in language, evidence or presentation. “The appeal foregrounds positive experiences to encourage support” is a reasoned interpretation when the text uses testimonials and a donation request. “The author deliberately deceives the reader” demands stronger evidence of intention, which a short passage rarely supplies.
14. Compare quotations with their surrounding context
A quotation can look categorical when removed from a conditional sentence. “The programme is suitable for everyone” and “The programme may be suitable for everyone who meets the entry requirements” are different claims. If a text includes the qualifying phrase, keep it attached when paraphrasing. An answer that drops a condition may create a contradiction not present in the source.
Search for nearby phrases indicating exceptions, reservations or stages. The sentence before a quotation may define the group being discussed; the sentence after it may limit the conclusion. This is why scanning a single keyword is a weak substitute for reading the whole paragraph.
15. Decide whether there is a genuine conflict
Use a three-part decision. First, make the topic and conditions comparable: same group, measure, time and setting. Second, isolate the propositions that remain. Third, ask whether both could be true. If they can both be true, explain the distinction. If they cannot, identify the actual conflict without guessing which source is correct.
For example, “All 30 registered students attended the morning session” and “Five registered students missed the morning session” genuinely conflict if both refer to the same 30 and the same morning. It is possible that one report contains an error, but the texts alone might not reveal which. The best answer states the incompatibility and the limits of what can be concluded.
16. Avoid inventing a compromise
Students sometimes write, “The truth must be somewhere in between.” That is not a universal principle. If one text says a station opened on Tuesday and another says it opened on Friday, the true date is not automatically Wednesday or Thursday. Conflicting claims require additional evidence, not arithmetic averaging of statements.
The same applies when a source alleges an effect and another denies it. Your response may need to explain what each source says and how the methods differ. Do not blend incompatible reports into a fictional middle position merely to sound balanced. Balance means fair representation, not automatic compromise.
17. How to phrase agreement without erasing differences
Useful patterns include: “Both texts identify the same obstacle, but they differ on its severity”; “Text A reports a trial outcome, whereas Text B describes a later decision”; and “The sources agree on the attendance figure but offer different explanations for it.” These constructions make the exact common ground visible.
Replace vague lines such as “Both texts talk about the library” with a subject-specific overlap. Mention whether they share a fact, goal, concern, method or conclusion. General topic overlap is rarely enough to show meaningful agreement.
18. How to phrase disagreement without exaggerating it
Write “Text A attributes the delay to late delivery; Text B attributes it to a scheduling error” when the explanations conflict. Avoid “Text A proves that the supplier caused the delay” unless the source offers decisive evidence. When the text is less certain, use language such as “suggests”, “reports” or “argues” according to its actual degree of commitment.
A useful contrast does not need aggressive vocabulary. “The accounts differ on the cause” is direct. “The authors are fighting over a scandal” imports emotion and perhaps an event never described. Calm prose is an examination advantage because it leaves more room for correct distinctions.
19. Respect the question’s requested action
Some questions ask what a particular detail means in one text. Others ask how the visual reinforces a message, why a writer selects a phrase, or what may be inferred. A cross-source comparison is useful only if the question calls for that comparison. A well-informed answer can still be off-task if it answers a broader, unasked question.
Identify the command word, the required source and the requested level of response. If the task asks for one inference from a specified paragraph, support that inference with the relevant details. Do not write an elaborate discussion of a second source unless it clarifies the requested point.
20. Short-answer construction: claim, difference, evidence
When comparison is requested, a compact answer can work in three sentences. Sentence one names the substantive common ground or contrast. Sentence two gives a precise detail from the first source. Sentence three adds the relevant second-source detail and explains the distinction. Not every answer needs all three sentences, especially where marks and space are limited.
A weak answer says, “Text A is optimistic while Text B is pessimistic.” A stronger answer says, “Text A focuses on favourable ratings among trial participants, whereas Text B concentrates on the number of later users and the school’s response to noise complaints.” The stronger version grounds the difference in what the texts contain.
21. Summary writing is not the same as source comparison
K300 Paper 2 Section C contains a summary task of about 80 words, excluding introductory words supplied. It is based on the relevant non-narrative text, not automatically on all the other texts in the paper. Do not combine details from Texts 1 and 2 into a summary of Text 4 unless the instructions explicitly demand that.
The skill overlap is fidelity: identify distinct relevant points and preserve qualifiers. Volume 0148 on summary writing owns the 80-word response mechanics. This chapter focuses instead on detecting when two apparent claims are not actually about the same thing.
22. A worked examination-style visual interpretation
Fictional poster: “Eight out of ten participants would return!” Its footnote says “20 respondents from one Saturday workshop.” A separately supplied report says “Only 25 out of 100 invited pupils attended the workshop.” At first glance, the poster suggests overwhelming popularity while the report suggests low uptake. The statements can both be true: 16 of the 20 respondents intended to return, while 25 of 100 invitees attended.
Do not infer that 80 of the original 100 invitees would return. Even among the 25 attendees, only 20 answered the question. A strong response notes the smaller respondent base and the difference between intention and observed behaviour. “Would return” reports stated intention; it does not prove a future attendance count. The footnote is as important as the headline.
23. A four-step correction routine
First, mark precise nouns and quantities. Second, circle scope and time conditions. Third, classify each claim as observation, interpretation, expectation or instruction. Fourth, write a response that states any common ground, genuine conflict and uncertainty in separate clauses. This sequence is intentionally short enough to use under timed conditions.
During revision, explain why each qualifier matters. If removing “surveyed” changes a claim from respondents to the whole school, keep it. If removing “after repairs” turns a historical statement into a current one, keep it. The goal is not cautious wording for its own sake; it is a conclusion whose limits match the available material.
24. Twenty-six original practice questions and answers
Each item below includes a small invented claim pair or a question about the worked sources. Name the distinction before reading the explanation. They are training exercises, not extracted SEC examination questions.
1. Scope
Question. A says “Some pupils participated”; B says “Most pupils participated.” Contradiction? Worked response. No. Most implies some in the same relevant group. Reasoning. The quantifiers differ in strength but can be simultaneously true.
2. Negation
Question. A says “Not all rooms are occupied”; B says “No rooms are occupied.” Identical? Worked response. No. Reasoning. Not all leaves at least one unoccupied; none asserts every room is unoccupied.
3. Time
Question. A says the hall was closed in May; B says it reopened in July. Conflict? Worked response. No. Reasoning. The status changed between dates.
4. Location
Question. A reports heavy rain in Sengkang; B reports dry weather in Jurong. Conflict? Worked response. No. Reasoning. Different places are being described.
5. Measure
Question. A says total users rose; B says users per session fell. Conflict? Worked response. Not necessarily. Reasoning. Totals and per-session averages have different denominators.
6. Audience
Question. A says “Nine in ten club members approved”; B says “Half the school did not join.” Conflict? Worked response. No. Reasoning. Approval among members differs from uptake among all pupils.
7. Prediction
Question. A says a pilot may reduce queues; B says queues did not fall during an earlier week. Conflict? Worked response. Not automatically. Reasoning. Prediction and measured history may concern different stages.
8. Cause
Question. A attributes delay to supplies; B attributes delay to staffing. What is the difference? Worked response. Competing causal explanations. Reasoning. They could describe independent contributing factors unless each claims sole cause.
9. Number
Question. A reports 24 students, B reports 42 on the same day and same list. What needs checking? Worked response. The count and its basis. Reasoning. If both use identical definitions and timing, the numbers conflict.
10. Visual scale
Question. A chart begins its vertical axis at 90 rather than zero. What risk? Worked response. The apparent difference may be exaggerated. Reasoning. Read printed values rather than judging only the height contrast.
11. Footnote
Question. Poster says 90% support; footnote says 9 of 10 respondents. What can be claimed? Worked response. Nine respondents supported. Reasoning. A sample of ten does not establish support across the whole population.
12. Opinion
Question. A describes a scheme as fair; B calls it unfair. Are their raw facts necessarily different? Worked response. No. Reasoning. A value judgement may use different criteria while sharing the same facts.
13. Quantifier
Question. Do “all” and “at least one did not” conflict for the same group? Worked response. Yes. Reasoning. They cannot both hold under the same definition and time.
14. Uncertainty
Question. A says “might improve”; B says “will certainly improve.” What differs? Worked response. Strength of prediction. Reasoning. The second is a much stronger future claim.
15. Source independence
Question. Three posts quote the same survey. Are there three independent surveys? Worked response. No. Reasoning. Repetition of one survey is not independent measurement.
16. Selection
Question. 16 of 20 respondents liked a workshop; 25 of 100 invitees attended. Is 80% school-wide approval supported? Worked response. No. Reasoning. The rating denominator is the 20 respondents, not 100 invitees.
17. Intention
Question. Does “would return” mean participants returned? Worked response. No. Reasoning. It reports intention, not observed future action.
18. Shared fact
Question. Two texts cite 20 early arrivals but disagree about whether that is good. What can be stated? Worked response. They agree on the count and differ in evaluation. Reasoning. Separate observation from a benchmark-dependent judgement.
19. Chronology
Question. A says a park is being built; B says it was completed. Can both be accurate? Worked response. Yes, at different times. Reasoning. Check publication dates and milestones.
20. Comprehension task
Question. A question asks about Text 3’s character motive. Should you cite Text 1’s chart? Worked response. Usually no. Reasoning. Use the requested narrative evidence unless directed otherwise.
21. Summary scope
Question. Paper 2 Section C summary is about Text 4. Should you import Text 2 ideas? Worked response. No, unless required. Reasoning. The specified source and task control the selection.
22. Comparing claims
Question. A says “more people came”; B says “attendance rate fell.” What extra information matters? Worked response. Population size and denominators. Reasoning. Counts can rise while percentages decline.
23. Experimental scope
Question. A result applies to one device under one load; an advert says it applies to all devices. What changes? Worked response. Generalisation beyond tested conditions. Reasoning. The evidence’s scope has expanded without support.
24. Correction
Question. Rewrite “The report proves every student benefited” if it surveyed 18 participants who reported improvement. Worked response. “Eighteen surveyed participants reported improvement.” Reasoning. This preserves the actual group and type of evidence.
25. Meeting record
Question. One source says “the plan was discussed”; another says “the plan was approved.” Are these equivalent? Worked response. No. Reasoning. Discussing and approving are distinct events or stages.
26. Best answer
Question. Two sources disagree on a date but neither supplies a document. What should you conclude? Worked response. The dates conflict; the correct date is not established. Reasoning. Do not invent a compromise or claim certainty beyond the sources.
25. Turn each mistake into a specific revision exercise
If you keep treating samples as populations, practise twenty sentences with different denominators and underline the group after every percentage. If you confuse intention and action, transform claims about plans, hopes and predictions into sentences that clearly distinguish future expectations from completed outcomes. If you overlook dates, build short chronological timelines from multi-stage announcements.
A good revision notebook does not need pages of copied definitions. Record the wrong inference, the missing qualifier and a corrected sentence. Repeat the problem with new details a day later. This targets the decision you actually made incorrectly rather than rewarding yourself for rereading the explanation.
26. A seven-day reading routine
On day one, compare two notices with different dates. On day two, compare a headline and its footnote. On day three, practise denominators and scope words. On day four, separate observed facts from interpretations in a short report. On day five, write three-sentence supported comparisons. On day six, attempt a timed K300 comprehension section with the official task requirements in view. On day seven, review every overstatement and missing qualifier.
This is an illustrative routine, not an official SEAB schedule. A pupil who already handles time and scope may devote more days to multimodal reading or inference. Parents and tutors can listen for explanations: “They are not contradictory because they discuss different groups” is stronger evidence of learning than a guessed label of “agreement”.
27. Questions families commonly ask
Must every pair of texts be compared? No. The syllabus describes specific texts and tasks; follow what the question asks. Comparison is a transferable thinking skill, not an added compulsory examination section. Should a student call a writer biased? Only when identifiable evidence supports a meaningful explanation of selection or perspective. Can two different percentages both be correct? Yes, when their groups, measures or times differ.
Is a visual always more reliable than prose? No. Both need careful reading. A chart can use misleading scales or unspecified sampling; prose can accurately define limitations that a graphic omits. How much should an answer say? Enough to make the specific distinction and support it, not an essay that travels beyond the asked evidence.
28. Links for the next stage
Continue through the G3 SEC Learner’s Guide, the PSLE Learning Guide for foundational source reading and the Secondary English Learning Hub for subject practice. Review Volume 0185 on evidence verbs and Volume 0188 on science-gallery claims.
Above all, use the official 2027 K300 syllabus for assessment details. Practise reading the full sentence, not just the assertion that sounds strongest. When two sources appear to disagree, compare the actual claims before deciding what the reader may responsibly conclude.
