A pupil reads, “Our new reminder system reduced missed equipment on three Fridays.” The next sentence says, “This proves the system is the best possible method for every class.”
The first sentence may report evidence. The second turns that evidence into a much larger claim.
Critical non-fiction reading asks whether the conclusion is proportionate to the evidence, whether the evidence is relevant and whether important limits have been hidden.
This guide extends Inference Across Paragraphs, Viewpoint, Character Voice and Perspective and Text Structure, Main Idea and Paragraph Function.
Quick answer: separate claim, support and scope
For any important statement, ask:
- Claim: What is being said?
- Evidence: What supports it?
- Scope: How far does the claim reach?
- Source: Who provides the information, and what do they know?
- Limit: What remains uncertain?
A compact routine is CLAIM → EVIDENCE → SOURCE → SCOPE → LIMIT.
The 2026 PSLE boundary
SEAB’s 2026 English Language document assesses literal, inferential and evaluative understanding of written and multimodal texts. This guide uses original non-fiction passages to strengthen those reading habits. It is not a separate official PSLE component.
Official reference routes: SEAB PSLE Formats Examined in 2026 and the 2026 PSLE English Language document.
Fact and opinion are useful starting categories
Fact-like claim: “The workshop begins at 3.00 p.m.” This can be checked against a schedule.
Opinion: “The workshop is exciting.” This expresses judgement.
But advanced reading needs more than the fact/opinion label. A factual-sounding statement can still be unsupported, outdated or too broad.
Verifiable does not mean verified
“Twenty pupils preferred the new timetable” is a claim that could be checked. Until we know where the number came from, it is not automatically reliable.
Ask for the evidence source: survey? estimate? attendance list? memory?
Evidence should match the claim
Claim: “The new sign reduced confusion.”
Relevant evidence: fewer visitors asked where to enter after the sign was moved.
Weak evidence: the sign was printed in blue.
Evidence can be true and still irrelevant.
One example does not prove a universal rule
“One pupil improved after using flashcards” does not prove flashcards work best for every pupil.
The evidence supports a smaller conclusion: the method may have helped that pupil under those conditions.
Scope words matter
Watch:
- all;
- every;
- always;
- never;
- most;
- many;
- some;
- may;
- can;
- likely.
Changing some to all changes the claim.
Original Source Pack 1: The Earlier Reminder
Source A — class record: Over four school events, 12 pupils forgot reusable bottles at Event 1, 9 at Event 2, 6 at Event 3 and 5 at Event 4.
Source B — organiser note: Before Event 3, the reminder was sent two days earlier instead of being shown only at the hall entrance.
Source C — pupil comment: “The earlier message definitely solved the problem.”
Question 1: what can Source A support?
It supports the claim that the number of pupils forgetting bottles decreased across the four recorded events.
It does not by itself prove why.
Question 2: what does Source B add?
It identifies a change in reminder timing before Event 3.
That makes earlier reminders a plausible explanation, but other factors could also have changed.
Question 3: evaluate Source C
“Definitely solved” is stronger than the evidence warrants. Forgetting decreased but did not disappear, and the source pack does not rule out other causes.
Better conclusion: The earlier reminder may have contributed to the reduction, but the available data do not prove it was the only cause.
Correlation and cause at Primary 6 level
If two things change together, one may cause the other—but not always.
Reminder timing changed and forgetting decreased. This is suggestive.
Maybe teachers also reminded pupils in class. Maybe pupils became familiar with the event. Maybe weather changed what they carried.
Strong reading keeps alternative causes visible.
Source reliability depends on the question
A pupil who attended an event may be a good source for what they personally saw.
The same pupil may be a weak source for the total number of attendees unless they counted them.
Reliability is task-specific.
Primary versus secondary evidence
Direct record: attendance sheet, measured time, original survey response.
Reported summary: “The teacher said attendance was high.”
Direct evidence is not automatically perfect, but it often gives the reader more precise access to what was recorded.
Evidence quality questions
Ask:
- Was it measured or guessed?
- How many cases?
- Was the same method used each time?
- Does the evidence directly address the claim?
- Could another explanation fit?
- Is the source in a position to know?
Original Source Pack 2: Quiet Study Zone
Source A: A library survey of 40 pupils: 26 preferred having a quiet zone, 8 preferred no special zone and 6 had no preference.
Source B: Two pupils wrote that the quiet zone made project discussion harder.
Source C: The library also has two bookable discussion rooms.
Claim: “The quiet zone is good for every library user.”
Evaluate the claim
The survey shows majority support among the 40 respondents, not universal support. Two pupils report a drawback. The presence of discussion rooms may reduce the drawback, but we do not know whether rooms are always available.
A stronger judgement is: The quiet zone appears useful to many surveyed pupils, but it does not meet every user’s needs equally.
Numbers can strengthen and distort
“65% preferred the quiet zone” is another way to express 26 out of 40.
The percentage is not wrong, but the denominator matters. Sixty-five per cent of 40 surveyed pupils is not 65% of all pupils in the school.
Denominator awareness
Always ask: 65% of what?
“Nine out of ten pupils improved” sounds strong. Was the group ten pupils or one thousand? How was improvement measured?
Missing comparison
Claim: “The new route is faster.”
Evidence: “The new route takes four minutes.”
We still need the old route time.
Comparative claims require comparative evidence.
Missing baseline
“Errors fell to five.”
From fifty? From six? Without the earlier value, the size of improvement is unclear.
Cherry-picking
A report may show only the best day.
Monday: 2 mistakes. Tuesday: 8. Wednesday: 7. Thursday: 9. Friday: 6.
Claim: “Our system produces only two mistakes a day.”
The writer has selected the exceptional result.
Anecdotes
“My friend loved the workshop” is useful evidence about one friend’s experience.
It is weak evidence for “Everyone loved the workshop.”
Expertise and relevance
A librarian may be a strong source about library procedures.
The same librarian is not automatically an expert on medical treatment.
Authority should match the topic.
Opinion can still be informed
“I think the new system is fairer because both groups now receive the same amount of time.”
This is an opinion, but it is supported by a stated criterion and evidence.
Do not treat all opinions as equally unsupported.
Original passage: The Faster Queue
Paragraph 1. The school fair used two registration tables last year and three this year.
Paragraph 2. Last year, the average wait recorded during the first thirty minutes was eight minutes. This year, it was five minutes.
Paragraph 3. An organiser wrote, “The extra table reduced waiting time by three minutes.”
Paragraph 4. However, this year’s event also had fewer visitors during the first thirty minutes.
Paragraph 5. The organisers decided to keep three tables but noted that another event would be needed before they could judge how much difference the extra table made.
Questions
1. Which statement is directly supported by the recorded data?
2. Why is the organiser’s sentence in paragraph 3 too confident?
3. What alternative explanation appears in paragraph 4?
4. Why is paragraph 5 a responsible conclusion?
Worked answers
1. The recorded average wait during the first thirty minutes was lower this year: five minutes instead of eight.
2. The sentence assumes the extra table caused the entire reduction even though other conditions also changed.
3. There were fewer visitors this year during the measured period, which could also reduce waiting time.
4. The organisers keep the promising change while acknowledging that more evidence is needed to estimate its effect.
Responsible uncertainty
Words such as may, suggests, appears, likely, cannot yet determine are useful when evidence is limited.
Qualification is not weakness. It is accuracy.
Headlines can overstate
Evidence: “7 of 10 pupils preferred the new seating arrangement.”
Headline A: “Most surveyed pupils prefer new seating.”
Headline B: “New seating is loved by everyone.”
A matches the evidence more closely.
Visual texts and evidence
A poster may use a large statistic. Ask:
- what was measured?
- who was measured?
- how large was the group?
- what comparison is implied?
- does the image make the claim feel stronger than the data?
Persuasive framing
“Only 3 pupils disagreed” and “30% disagreed” can describe the same group of ten.
Different framing changes emphasis.
Fact-checking within the text
Sometimes a passage supplies enough internal evidence to test a claim.
If a character says “Nobody uses the side entrance”, but the next paragraph reports six families entering there, the claim is contradicted.
Claims in oral communication
When a pupil says, “Everyone should do this”, ask whether a narrower claim is more defensible.
“This could help pupils who often forget deadlines” may be stronger reasoning than an unsupported universal statement.
Claims in writing
Situational Writing should not exaggerate source material.
Source: 18 of 30 pupils interested.
Weak: “The whole class wants this event.”
Accurate: “A majority of surveyed classmates expressed interest.”
Source Pack 3: Reading Corner
Source A: The class introduced a reading corner for six weeks.
Source B: Borrowing from the class shelf rose from 20 books in the previous six weeks to 32 books.
Source C: The teacher also began recommending one book each Friday.
Claim: “The reading corner caused a 60% increase in borrowing.”
Evaluate
The borrowing count increased from 20 to 32, which is indeed a 60% increase. But causation is uncertain because teacher recommendations began during the same period.
Accurate statement: Borrowing rose by 60% during the six weeks when the reading corner and weekly recommendations were introduced.
One sentence can separate observation from explanation
Observation: borrowing rose.
Explanation: why it rose.
Keep them separate until evidence supports the causal link.
Independent transfer: Water Refill Station
Source A: A refill station was installed near the hall.
Source B: During one school event, 48 bottles were refilled.
Source C: At the previous event, no refill count was recorded.
Source D: One organiser says, “The station prevented 48 plastic bottles from being used.”
Questions:
1. Which source gives a measured fact?
2. Why does Source D go beyond the evidence?
3. What additional evidence would help test the organiser’s claim?
4. Write a cautious conclusion.
Transfer answers
1. Source B records 48 refills.
2. A refill does not prove a new plastic bottle would otherwise have been used.
3. Information about what participants would have done without the station, or comparable bottle-use data from similar events.
4. The station was used 48 times during the event, suggesting it provided a useful refill option, but the data do not show exactly how many disposable bottles it prevented.
Three pupil profiles
The fact/opinion sorter: can label sentences but cannot evaluate evidence. Repair claim-support matching.
The number believer: assumes statistics prove the conclusion. Repair denominator, baseline and alternative cause checks.
The universaliser: turns some into all. Repair scope-word monitoring.
A four-week critical-reading cycle
Week 1: claims, facts and opinions. Week 2: evidence relevance and scope. Week 3: source reliability and alternative explanations. Week 4: mixed source packs and cautious conclusions.
What parents can observe
- Can the pupil identify the claim?
- Can the pupil name the evidence?
- Can the pupil spot a scope word?
- Can the pupil ask “of what?” after a percentage?
- Can the pupil separate correlation from cause?
- Can the pupil explain why a source is or is not in a position to know?
- Can the pupil write a qualified conclusion?
Secondary 1 bridge
Secondary English increasingly asks students to read claims, evidence, bias and source reliability. The Primary 6 habit to carry forward is proportionate judgement: conclusions should not exceed the evidence.
Continue with Secondary 1 Non-Fiction Reading: Claims, Evidence and Bias.
Continue Batch 9
Use Grammar for Meaning to control certainty and condition, Punctuation and Sentence Boundaries for written clarity, and Selective Listening Note-Taking for real-time evidence capture.
Return to the Secondary English Learning Hub.
The quiet return
Strong non-fiction reading is not suspicious of everything. It is precise about what the evidence can and cannot support.
Name the claim. Find the evidence. Check the source. Watch the scope. Then stop the conclusion where the evidence stops.
