Checked: 2 September 2026. This reader guide is aligned to MOE’s English Language Syllabus 2020 Primary and current Secondary English expectations. It does not claim a current eduKateSG Jurong West branch.
Reading comprehension becomes much easier to diagnose when three levels are kept separate. A literal question asks what the text directly says. An inferential question asks what can be concluded from clues. An evaluative question asks the reader to make a judgement about an idea, action, source or effect using evidence and criteria. Students often lose marks because they use the wrong reasoning level. For Jurong West families, this legacy 2017 URL now owns one precise reading mechanism: the evidence ladder from literal to inferential to evaluative understanding.
Quick read
- Literal reading stays close to explicitly stated information.
- Inference combines textual evidence with justified reasoning.
- Evaluation adds criteria and judgement while remaining evidence-based.
- Do not answer a literal question with speculation.
- Do not answer an inferential question by merely copying a line.
- The goal is a reader who knows how far the evidence allows them to travel.
Why the ladder matters
The three levels are related but not interchangeable.
Imagine a passage states:
Marcus checked the clock three times, packed his notes hurriedly and left before the meeting ended.
Literal:
What did Marcus do?
Inferential:
What can we infer about his state of mind?
Evaluative:
Was leaving early a reasonable decision? Use evidence.
The same sentence supports three different intellectual jobs.
Level 1: literal reading
Literal reading asks:
- who;
- what;
- where;
- when;
- what happened;
- what was said;
- which detail appears;
- what sequence is stated.
The answer should be anchored in explicit information.
Literal does not mean careless copying
A literal answer may still need:
- selection;
- paraphrase;
- precise reference;
- complete sentence formulation;
- removal of irrelevant details.
The cognitive demand is lower than inference, but language control still matters.
Literal failure pattern: copying too much
The question asks for one reason and the learner copies three lines.
Repair:
Identify the minimum sufficient information that directly answers the task.
This connects to answer economy.
Literal failure pattern: adding assumptions
The passage says a character left quickly. The student writes that she was angry even though no evidence supports it.
Repair:
Ask:
Is this stated, or am I adding it?
Literal answers should not cross into inference unless required.
Level 2: inferential reading
Inference asks the reader to connect clues with background understanding.
A useful formula is:
evidence + relevant knowledge → justified conclusion.
The conclusion is not directly stated, but it must be supported.
Inference needs a bridge
Weak inference:
Marcus was nervous.
Stronger reasoning:
Marcus repeatedly checked the clock and packed hurriedly, suggesting that he was anxious about time or an upcoming commitment.
The evidence makes the inference defensible.
Inference should match claim strength
Clues may support:
- possible;
- likely;
- strongly suggested;
- almost certain
conclusions at different strengths.
If the evidence is ambiguous, avoid absolute claims.
Language such as suggests, likely or appears may be more accurate.
Inference failure pattern: plausible but unsupported
The student gives an idea that could be true in real life but has no textual support.
Repair:
Require one clue and an explanation of why the clue supports the conclusion.
If the bridge cannot be built, the inference is weak.
Inference failure pattern: evidence copied with no conclusion
The student writes:
He checked the clock three times.
but never explains what that suggests.
Repair:
Add the interpretive relationship.
Evidence alone is not inference.
Level 3: evaluative reading
Evaluation asks the reader to make a judgement.
Common tasks include:
- Was the decision reasonable?
- How effective is the writer’s argument?
- Which source is more convincing?
- Was the character’s response appropriate?
- How successful was the strategy?
Evaluation requires more than personal preference.
Evaluation needs criteria
A strong evaluative answer includes:
- a criterion;
- evidence;
- analysis;
- judgement.
Example:
Leaving early was understandable because Marcus appeared to be under time pressure, but it was not fully responsible because the passage gives no indication that he informed the organiser before leaving.
The answer weighs evidence against a standard.
Possible evaluation criteria
Depending on the task:
- effectiveness;
- fairness;
- relevance;
- practicality;
- clarity;
- credibility;
- safety;
- impact;
- appropriateness.
The criterion should fit the question.
Evaluation failure pattern: opinion only
Weak:
I think he should not have left.
Repair:
Add:
- why;
- according to what criterion;
- what evidence supports that judgement.
Evaluation is reasoned judgement.
Evaluation failure pattern: evidence with no weighing
The student lists two facts but never decides what they mean for the judgement.
Repair:
Use language such as:
- therefore;
- this makes the decision more/less reasonable because…;
- however;
- overall.
The reasoning should culminate in a position.
The evidence ladder
A practical classroom model:
- Literal: What does the text say?
- Inferential: What does the evidence suggest?
- Evaluative: What judgement can I defend?
Each step depends on the one below it.
Evaluation without accurate literal reading is fragile. Inference without evidence is speculation.
Use one paragraph to train all three
Tutors do not need separate passages for every level.
One rich paragraph can generate:
- literal detail question;
- inference question;
- evaluation question.
This helps students feel how the intellectual demand changes while the text stays constant.
The evidence-highlighter method
Use three annotations:
- L: direct textual fact;
- I: inference built from one or more clues;
- E: evaluation using criteria.
Students should mark which layer their answer belongs to.
This makes overreach visible.
Primary progression
Primary students can begin with:
- direct detail retrieval;
- simple inference about feelings, motives and causes;
- basic evaluative responses using evidence.
The teacher can use simple labels:
said → suggested → judged.
The formal terminology can grow later.
Secondary progression
Secondary learners face more complex:
- tone;
- writer viewpoint;
- source reliability;
- argument quality;
- implicit assumptions;
- counterarguments;
- evaluative comparison.
The same evidence ladder still applies.
Literal and inferential vocabulary questions
Vocabulary-in-context can also move up the ladder.
Literal:
What word is used?
Inferential:
What does the word mean here based on context?
Evaluative:
How effective is the word choice in creating a particular effect?
One lexical item can support different reading depths.
Author’s purpose questions
Students should distinguish:
- what the writer says;
- what the language suggests about purpose;
- how effectively that purpose is achieved.
This is the same ladder applied to writer craft.
Character questions
Literal:
What did the character do?
Inferential:
What does this suggest about the character?
Evaluative:
Was the character’s decision justified?
Again, the reader climbs by evidence.
Comparative reading
With two texts:
- literal: identify each text’s claim;
- inferential: interpret stance or implication;
- evaluative: compare which is more convincing and why.
Students should not jump directly to preference.
Three students and three evidence levels
In a 3-pax class, students can answer three questions on the same passage, each at a different level.
Then rotate.
The group compares:
- how evidence use changes;
- how much interpretation is allowed;
- where judgement enters;
- which claims are too strong.
This teaches reading depth without requiring three separate lessons.
Catch Up, Keep Up, Move Ahead
Catch Up: stabilise literal understanding and evidence location.
Keep Up: build reliable inference and basic evaluation.
Move Ahead: evaluate source quality, competing interpretations and argument effectiveness with explicit criteria.
Every mode stays anchored to evidence.
A 90-minute evidence-ladder lesson
A useful session can:
- read one short text;
- answer literal questions;
- identify clues for inference;
- build one evaluative judgement;
- compare claim strength;
- rewrite one overreaching answer;
- apply the ladder to an unseen passage.
The learner should know how far each question allows them to move from the exact wording.
How parents can help
- Ask “Is that stated or inferred?”
- Ask “What clue supports that?”
- Ask “What criterion are you using to judge?”
- Encourage children to say “suggests” when evidence is not absolute.
- Avoid accepting a plausible opinion as an evaluative answer without evidence.
Current official source
MOE’s English Language Syllabus 2020 Primary explicitly develops comprehension at literal, inferential and evaluative levels. Secondary English extends those habits into richer texts, argument and source work.
What we removed from the old 2017 page
The historical Jurong West article was a generic locality clone with stale service claims, “first language” marketing and unrelated galleries.
This rebuild removes the false locality implication and gives the URL one useful reading job: the literal→inferential→evaluative evidence ladder.
Boundary and locality
This guide is written for Jurong West families and other Singapore English learners. It does not claim a current eduKateSG Jurong West branch. Current physical locations and class availability should be checked on the contact page.
Frequently asked questions
Is every comprehension question one of these three types?
Not neatly. Some tasks combine levels, but the ladder is a useful diagnostic model for how far the answer moves beyond explicit text.
What is the biggest inference mistake?
Giving a plausible idea with no textual clue or making a claim stronger than the evidence supports.
What makes evaluation different from inference?
Evaluation adds a judgement based on criteria and evidence; inference mainly constructs an unstated conclusion from clues.
Should literal answers always copy the passage?
No. Select and paraphrase where appropriate while preserving the explicit meaning.
What is the long-term goal?
A reader who knows what the text says, what it reasonably suggests and what judgement can be defended—without confusing those three levels.
The larger point
Reading depth is controlled distance from the text. Strong readers know when to stay close, when to infer, and when they have earned the right to judge.