Yishun Primary 6 Science tuition should help students recognise not only what a table, graph or investigation supports, but also what it does not justify. Many upper-primary errors come from turning a limited result into an absolute statement.
This rebuilt legacy URL now owns one distinct P6 Science job: observation → supported inference → unsupported inference → boundary → calibrated conclusion. Old result promises, mixed-level copy and implied Yishun-centre claims have been removed.
eduKateSG does not claim a current Yishun branch. Current centres are at 83 Punggol Central, Singapore 828761 and 8 Fourth Avenue, Singapore 268674. Selected Primary Science classes run in focused 3-pax small groups, typically 1.5 hours.
The Direct Answer
Every piece of evidence has a boundary. Before accepting a conclusion, ask:
- What was actually observed?
- What conditions were tested?
- How many samples or trials were used?
- Was only one factor changed?
- Does the conclusion go beyond the tested group, time or condition?
A strong Science answer knows where to stop.
Worked Example: One Plant
Observation: one plant grew taller in brighter light than another plant in dimmer light over seven days.
Supported conclusion: under these tested conditions, the plant in brighter light grew more in height.
Unsupported conclusion: “All plants always grow best in the brightest possible light.”
The test involved limited plants, a specific time period and specific light conditions.
Worked Example: One Material
Observation: Material A absorbed more water than Material B in one fair test.
Supported: Material A was more absorbent than Material B under the test.
Unsupported: Material A is the most absorbent material in the world.
Worked Example: Correlation
Graph shows that as temperature rises, evaporation amount also rises.
If the investigation changed only temperature while controlling relevant conditions, the evidence can support a relationship in the tested range.
If the graph merely compares different days with many changing conditions, saying temperature alone caused the difference may be unsupported.
Worked Example: Food Web
A predator population decreases after one prey population falls.
Supported: reduced prey availability may contribute to the predator decline if that prey is an important food source.
Unsupported: the prey decrease is definitely the only cause. Disease, migration or other food changes may also matter.
Sample Size Boundary
One seed, one animal or one trial provides limited evidence.
Students should be careful with words such as:
- all;
- always;
- never;
- must;
- proves.
These require stronger evidence than may, suggests, under these conditions.
Range Boundary
If an investigation tests 20°C, 30°C and 40°C, it does not automatically tell us what happens at 80°C or 0°C.
Do not extend a trend beyond the tested range without evidence.
Population Boundary
If the subjects are one type of plant, do not generalise immediately to all living things.
If data comes from one classroom setup, keep the conclusion tied to that context unless the question supplies broader knowledge.
Time Boundary
A seven-day result does not necessarily predict a seven-month result.
Short-term patterns and long-term outcomes can differ.
Variable Boundary
If two factors changed together, the evidence cannot isolate either factor cleanly.
Example: one plant received more light and more water. Greater growth cannot be attributed confidently to light alone.
Measurement Boundary
If the outcome measured was height, do not claim overall “health” improved unless the task defines health through that measure or provides other evidence.
Worked Example: “Better”
Data shows Container A kept water warmer for longer.
Supported: Container A was a better insulator under the test.
Unsupported: Container A is a better product overall. Cost, durability, safety and other functions were not measured.
The “What Was Not Measured?” Test
After reading the conclusion, identify which important words were never actually observed or measured.
Claim: “Students were healthier.”
Data: only pulse rate was measured.
The conclusion may be too broad.
The “Only Cause?” Test
If the answer says “because of X”, ask whether another factor could also explain the observation.
If yes, the wording may need to become “may be due to” or “supports the idea that”, unless the experimental design isolates X.
The Counterexample Test
Claim: “All metals are magnetic.”
One non-magnetic metal counterexample breaks the universal claim.
Counterexamples are powerful checks against overgeneralisation.
The Extrapolation Test
Graph trend rises steadily from 1 to 5 units.
Can we assume it will rise forever?
No. Outside the tested range, another factor may become limiting or the relationship may change.
The Model Boundary Test
Food webs, circuit diagrams and system models simplify reality. They show selected relationships, not everything that exists.
Students should answer from the model while recognising what it omits.
MCQ Distractor Pattern
Wrong options often contain true words with an unjustified absolute:
- always;
- only;
- all;
- proves;
- definitely.
Do not reject these words automatically; reject them when the evidence does not support that strength.
Open-Ended Answer Calibration
Weak: “The experiment proves that heat always makes evaporation faster.”
Stronger: “Under the tested conditions, the water at the higher temperature showed greater evaporation over the same period.”
The second answer states exactly what the evidence established.
From Observation to Claim Ladder
Possible levels:
- “In this trial…”
- “Across the repeated trials…”
- “Under these tested conditions…”
- “The results support the relationship that…”
- broader generalisation only when syllabus knowledge and evidence justify it.
PSLE Context
The 2026 PSLE Science examination assesses the 2023 Primary Science syllabus and requires students to interpret evidence, reason from investigations and communicate conclusions. Official references: MOE Primary Science Syllabus and SEAB 2026 PSLE Science syllabus.
Common Prohibited-Inference Failure Modes
| Symptom | Overreach | Repair |
|---|---|---|
| One trial → always | Sample scope exceeded | Limit to tested result |
| One plant type → all plants | Population widened | Keep group explicit |
| 20–40°C trend → all temperatures | Range extrapolated | Stay inside tested range |
| Two variables changed → one cause | Causality unsupported | Identify confounding factor |
| Height measured → health improved | Outcome broadened | Claim only measured property |
Why Three Students Helps
One student states the observation, one proposes the strongest defensible inference and the third tries to break it with a boundary or alternative explanation. The tutor calibrates the final claim.
When Yishun Primary 6 Families May Consider This Support
- Open-ended answers use “always” or “proves” too often.
- Graphs are extended beyond the tested range.
- One observation becomes a universal conclusion.
- Causal claims ignore alternative factors.
- P6 Science needs stronger evidence discipline before PSLE.
What Progress Should Look Like
A stronger P6 learner can state what the evidence supports, identify what was not tested and write a conclusion whose scope and certainty stop at the correct boundary.
Independent location notice: eduKateSG does not claim a current Yishun branch. Current centres: 83 Punggol Central, Singapore 828761 and 8 Fourth Avenue, Singapore 268674. Contact +65 8823 1234.