“I don’t know” and “I’m not sure” are different learner states. A child who has no route needs teaching. A child who has two plausible answers needs discrimination. A child who knows the answer but lacks confidence needs evidence and calibration—not another full explanation.
This page uses Punggol Primary School only as the school-family context named in the preserved 2019 URL. Its reader job is the uncertainty log: separating “don’t know” from “not sure” before checking.
The old page contained historic Yishun/Punggol centre, tutor, telephone and generic syllabus claims. Those are not carried forward. eduKate Singapore is an independent tuition provider and is not affiliated with Punggol Primary School, MOE or SEAB. This page is not a replacement for the current canonical Punggol English service pages.
Three Uncertainty States
| State | Student experience | Next move |
|---|---|---|
| Don’t know | No workable route | Teach or restore missing prerequisite |
| Not sure | Two or more plausible routes | Compare evidence / discriminating feature |
| Know but low confidence | Route is sound, reassurance sought | Require evidence, then commit |
Why the Distinction Matters
If every hesitation receives a full explanation, the student may become more dependent. If a genuine knowledge gap receives only “try again”, frustration grows.
The tutor needs to know what type of uncertainty is present.
The Uncertainty Log
For selected questions, record:
- question type;
- first answer;
- confidence: high / medium / low;
- reason for uncertainty;
- evidence checked;
- final answer;
- whether the change was justified.
Low Confidence Can Still Be Correct Reasoning
Some students repeatedly change correct answers because another option feels possible. Teach them to ask:
- What evidence supports my first answer?
- What evidence supports the alternative?
- Which option requires fewer assumptions?
- Has new evidence appeared?
Change the answer only when there is a reason, not because anxiety increased.
“Don’t Know” Needs a Blocker Diagnosis
- unknown word;
- question demand unclear;
- cannot localise passage region;
- grammar rule unknown;
- idea not generated;
- cannot encode idea in writing.
Once the blocker is named, give the smallest useful repair.
“Not Sure” Needs Discrimination
When two answers are plausible, identify the feature that separates them.
- evidence relevance;
- word meaning;
- grammar relationship;
- scope;
- intensity;
- question condition.
This teaches judgement rather than dependence on answer keys.
Confidence Should Be Calibrated, Not Maximised
The goal is not to make the child feel certain about everything. The goal is for confidence to match evidence.
- strong evidence → stronger confidence;
- ambiguous evidence → moderate confidence;
- no route → ask for targeted help.
Use the Log to Find Patterns
After several tasks, ask:
- Which question types create low confidence?
- Does the student change correct answers often?
- Does uncertainty come from vocabulary, evidence or task demand?
- Does confidence fall under time?
The pattern becomes a targeted teaching job.
A Confidence-Calibration Diagnostic
| Pattern | Likely issue |
|---|---|
| High confidence, weak evidence | Overconfidence |
| Low confidence, correct reasoning | Underconfidence |
| Changes answer without new evidence | Reassurance/anxiety loop |
| No answer route | Knowledge or strategy gap |
How a 3-Pax Class Uses Uncertainty
In a maximum three-student group, learners can privately rate confidence before discussion. The tutor can compare confidence with evidence and prevent peer confidence from replacing individual reasoning.
Current MOE English Context
The MOE Primary English Language Syllabus 2020 develops strategic and increasingly independent language use. Distinguishing uncertainty states helps students choose whether they need teaching, evidence comparison or simply commitment to a sound answer.
School Context Boundary
This article uses Punggol Primary School only as the school-family context named in the preserved URL. It does not imply endorsement, partnership or affiliation.
Signs Uncertainty Is Better Calibrated
- Students say why they are unsure.
- Correct answers are changed less randomly.
- Specific help requests improve.
- Confidence tracks evidence more closely.
- Repeated uncertainty patterns become teachable targets.
Uncertainty Log: Almost-Code Summary
QUESTION:
answer()
rate_confidence()
IF dont_know:
name_blocker()
repair()
ELSE IF not_sure:
compare_evidence()
find_discriminating_feature()
ELSE:
defend_and_commit()
LOG:
answer_change_reason()
OUTPUT:
calibrated_confidence
better_help_seeking
fewer_random_answer_changes