VIEW THIS AS

Auto mode follows the Route Engine until you choose a viewpoint.

YOU ARE HERE

ROUTE CHECK

CONNECTED TO

WHAT NEXT

Use the canonical route for this room, or HELP if you are unsure.

Toa Payoh Secondary English Tuition | Fact → Inference → Evaluation: Keep the Layers Separate • 3-Pax

eduKate Secondary students reviewing open books for How Super Intelligence Works: Embeddings.

Toa Payoh Secondary English tuition should help students separate three layers that are often collapsed into one: what the text states, what the evidence allows the reader to infer, and what judgement the reader makes about that information.

This rebuilt page now owns one distinct job inside the eduKateSG Secondary English estate: fact → inference → evaluation. It does not duplicate general comprehension or argument pages. Its purpose is to stop students from presenting interpretation as fact or personal judgement as though the passage itself said it.

This URL serves families searching from Toa Payoh. eduKateSG’s current centres are at 83 Punggol Central, Singapore 828761 and 8 Fourth Avenue, Singapore 268674; we do not claim a current Toa Payoh branch. Families can enquire about suitable 3-pax class placement according to level, timetable and travel fit.

The Direct Answer

Use three questions:

  • Fact/textual observation: what does the source explicitly state or show?
  • Inference: what conclusion is reasonably supported by those details?
  • Evaluation: how strong, useful, convincing, fair or important is the claim/evidence?

State first. Infer second. Judge third.

Why the Layers Get Mixed

English questions often move quickly from evidence to interpretation. Students learn to “read between the lines” and sometimes begin treating every inference as something the writer directly said.

Later, they add personal judgement and forget that this is a third layer.

Separating the layers gives the student a cleaner evidence trail.

Layer 1: Textual Fact or Observation

Passage: “Amir checked the clock three times, packed his bag before the lesson ended and waited near the door.”

Textual facts: he checked the clock three times, packed early and waited near the door.

These details are observable within the passage. They do not yet state why he behaved this way.

Layer 2: Inference

From the same details, a reasonable inference is that Amir is eager or anxious to leave.

A stronger claim such as “Amir is escaping because he did something wrong” is possible in imagination but unsupported by the available evidence.

Inference should be calibrated to the evidence.

Layer 3: Evaluation

Evaluation asks a different question: how convincing is the inference? Is the evidence sufficient? Are there alternative explanations?

A student might evaluate: “The details make anxiety plausible, but they do not reveal why Amir wants to leave, so a more specific motive would be uncertain.”

The student is now judging evidence quality rather than simply interpreting behaviour.

Fact Does Not Mean “True in the Real World”

In a fictional passage, “the dragon landed on the roof” can be a textual fact even though dragons are not real-world animals.

For comprehension, fact often means explicitly represented by the text, not independently verified reality.

This distinction helps students stay inside the reading task.

Opinion Is Not Automatically Evaluation

“I dislike the policy” is a personal opinion. “The policy argument is weak because the writer provides only one anecdote for a broad national claim” is an evaluation.

Evaluation needs a criterion: relevance, evidence strength, logic, fairness, completeness, feasibility or another defensible standard.

Worked Example: Advertisement

Advertisement: “9 out of 10 users said the product felt easier to use.”

  • Fact from source: the advertisement reports this user response.
  • Inference: many surveyed users preferred or found the product easier.
  • Evaluation: the strength of the claim depends on sample size, question wording and who the users were.

The student avoids jumping from “9 out of 10” to “the product is objectively the best”.

Worked Example: News-Like Text

Text: “After the new bus route opened, average waiting time at three stops fell from twelve minutes to eight.”

  • Fact: the reported waiting time fell at those three stops.
  • Inference: the new route may have improved waiting times there.
  • Evaluation: the evidence does not by itself establish what happened across the whole transport network or whether other changes contributed.

Each layer becomes more interpretive and requires more caution.

Worked Example: Character Judgement

Character gives up a seat to an older passenger without being asked.

  • Fact: the character gave up the seat voluntarily.
  • Inference: the character appears considerate.
  • Evaluation: the action supports consideration in this situation, but one action alone does not fully define the character in every context.

This helps students write precise character analysis.

Language-Effect Questions Cross the Layers

Fact: the writer uses the verb clawed.

Inference: the branches are presented as animal-like and aggressive.

Evaluation/analysis: this choice intensifies the storm scene by making the environment appear threatening.

Students learn that quoting the word is only Layer 1.

Fact–Inference Separation in Visual Text

Observation: the poster places “REGISTER NOW” in the largest text.

Inference: the design prioritises immediate action.

Evaluation: the call to action is clear, though whether it is persuasive depends on the rest of the information and audience.

Visual analysis becomes more disciplined when observation and interpretation are labelled mentally.

Fact–Inference Separation in Argumentative Writing

Students often introduce examples as though their interpretation were part of the evidence.

Evidence: “The school extended library opening hours by two hours.”

Inference: this may increase access for students who cannot stay earlier.

Evaluation: whether the change improves study outcomes would require further evidence about use and learning.

This creates more careful argumentation.

Source Language Can Signal the Layer

Words such as reported, observed, suggests, may, appears, argues and demonstrates carry different evidential force.

Students should preserve those distinctions when paraphrasing. “The study suggests” should not become “the study proves” without justification.

Alternative Explanations Belong in Evaluation

Suppose attendance improved after a new programme began. The student may infer the programme contributed.

Evaluation asks whether another factor could explain the change: timetable, incentives, seasonal variation or a different policy.

This does not mean refusing all conclusions. It means matching confidence to the evidence.

The Three-Column Method

For difficult texts, we sometimes use three columns:

Text says/showsI inferI evaluate
Two examples of delaysService may be unreliableTwo examples alone may not establish typical performance
Writer says “despite improvements”Writer acknowledges progressConcession suggests later criticism is qualified, not absolute
Large warning imageRisk is foregroundedEffect is strong but may simplify other aspects

The visible scaffold fades once students can keep the layers separate mentally.

Common Layer-Confusion Failure Modes

SymptomUnderlying issueRepair
“The text says he is guilty” when guilt is inferredInference presented as factReturn to exact textual evidence
Personal dislike presented as critiqueOpinion lacks criterionEvaluate using evidence/logic/relevance
“Proves” used for suggestive evidenceClaim strength inflatedPreserve evidential verbs
Quote given without interpretationStops at Layer 1Add supported inference/analysis
Alternative explanations ignoredEvaluation shallowTest at least one plausible alternative
Everything labelled “opinion”Textual evidence/interpretation distinction weakSeparate source statement from reader judgement

Why Three Students Helps

Three students can classify the same statement differently. The tutor can ask which words are explicitly in the text and which part is interpretation.

Students can also evaluate each other’s inference strength. This makes the evidence chain visible without reducing discussion to one model answer.

A Typical 1.5-Hour Fact–Inference–Evaluation Lesson

  • Extract: identify what is explicitly stated/shown.
  • Infer: form the strongest justified conclusion.
  • Mark certainty: choose appropriate wording.
  • Evaluate: test relevance, sufficiency and alternatives.
  • Write: express the layers clearly.
  • Challenge: identify where a peer overreaches.
  • Transfer: repeat in visual text, argument or a new passage.

Current Secondary English Context

MOE’s English Language Syllabus 2020 Secondary emphasises critical response, perspective-taking and evaluation of information across varied texts and contexts. Separating textual observation, inference and evaluation is a foundational routine for that work. Official reference: MOE English Language Syllabus 2020 Secondary.

When Toa Payoh Families May Consider This Support

  • The student presents inference as though the passage stated it.
  • Comprehension answers overreach beyond evidence.
  • Evaluation becomes personal opinion.
  • Words such as “proves” are used too strongly.
  • Visual-text observations are not turned into analysis.
  • Argumentative writing confuses examples with conclusions.
  • Secondary reading needs clearer evidence chains and critical judgement.

What Progress Should Look Like

A stronger student can point to what the text actually says, make a calibrated inference and then evaluate the strength or significance of that inference separately.

Current class enquiries: selected eduKateSG Secondary English 3-pax small groups, typically 1.5 hours. Current centres: 83 Punggol Central, Singapore 828761 and 8 Fourth Avenue, Singapore 268674. Contact +65 8823 1234.

Evidence has levels of force

Students need language that matches how much the evidence can support. Observed, reported, suggests, supports, demonstrates and proves do not carry the same strength.

“Proves” should be rare when the evidence is limited, indirect or open to alternative explanations. Strong reading includes calibrating the verb that connects evidence to conclusion.

A reported fact is still a source claim

If an article says “the company reported a 20% improvement”, the textual fact is that the company reported the figure. Treating the 20% as independently verified reality may require additional evidence outside the passage.

This distinction becomes important in media literacy and source comparison.

Inference confidence should be explicit

  • strongly supported: several details converge and alternatives are weak;
  • plausible: evidence supports the reading but alternatives remain;
  • possible: compatible with the text but weakly supported;
  • unsupported: requires information the text does not provide.

Students do not need to write these labels every time, but the distinction should guide claim strength.

Evaluation needs a criterion

“The evidence is weak” is incomplete unless the student can say why. Useful criteria include:

  • relevance to the claim;
  • sufficiency of quantity;
  • representativeness;
  • source authority for that claim;
  • causal strength;
  • fairness to alternatives;
  • feasibility of the conclusion.

Correlation is not automatically causation

If two changes occur together, the text may support an association. A causal inference needs stronger grounds: mechanism, controlled comparison, temporal order and plausible exclusion of alternatives.

Students do not need advanced statistics to learn the language boundary: happened after is not automatically happened because of.

Alternative explanations belong to evaluation, not paralysis

Considering another explanation does not mean no conclusion is possible. It helps determine how strongly the conclusion should be stated.

If one alternative fits the evidence almost as well, use cautious language. If the text rules it out, confidence can rise.

Worked example: programme and attendance

Text: attendance rose after a mentoring programme started.

  • Fact: attendance rose after the programme began.
  • Inference: the programme may have contributed.
  • Evaluation: the timing supports the possibility but does not alone rule out timetable changes, incentives or seasonal effects.

Worked example: source claim versus source evidence

A poster claims an activity “builds confidence” and shows two participant testimonials. The claim is explicit. The testimonials are evidence selected to support it. Evaluation asks whether two positive accounts are enough to justify a broad claim about all participants.

Two plausible inferences can coexist

A character who repeatedly checks the door may be anxious, impatient or expecting someone. If the evidence does not distinguish among them, a careful answer should preserve that uncertainty rather than pretend only one interpretation exists.

The layer-reversal test

Take an evaluation and ask what inference it depends on. Then ask what textual facts support that inference. If the chain cannot be reconstructed backwards, the judgement may be floating free of evidence.

The confidence-reduction test

Replace “proves” with “suggests”, or “is” with “may be”. Does the answer become more faithful to the evidence? If yes, the original wording may have overreached.

AI can silently turn inference into fact

AI often produces smooth summaries that remove uncertainty markers or choose one plausible explanation as though it were certain. A bounded use is: student separates fact, inference and evaluation first → AI may offer a competing interpretation → student compares evidential force → tool removed → new text analysed independently.

What stronger transfer looks like

  • Students distinguish what a source reports from independently verified reality.
  • Inference wording matches confidence.
  • Evaluation names a criterion rather than giving preference.
  • Correlation is less often described as cause.
  • Alternative explanations refine claim strength.
  • Source claims and source evidence are separated.
  • Students can reconstruct evaluation back to textual facts.
  • New texts are read with controlled uncertainty rather than automatic certainty.

The deeper principle

Critical reading is an evidence chain. Keep observation, inference and evaluation separate long enough to see how one becomes the next. Match confidence to evidential force, evaluate with explicit criteria and preserve uncertainty when alternatives remain. Good judgement is not certainty; it is knowing exactly what the evidence allows you to claim.

Discover more from eduKate Singapore

Subscribe now to keep reading and get access to the full archive.

Continue reading