Sengkang Science Tuition | From Knowing Facts to Using Science

The Science Learning System: What the Tutor Is Actually Trying to See

A useful Science lesson is not a race to identify the chapter name. The tutor is trying to see the student’s internal model of the situation: what the learner thinks the parts are, which relationship they believe is active, what evidence they are using and where the reasoning becomes unstable.

That distinction matters because the same wrong answer can be produced by completely different causes.

Same visible symptom	Possible underlying cause	Different teaching response
Wrong open-ended answer	Concept genuinely missing	Teach the scientific relationship
Wrong open-ended answer	Question condition overlooked	Reconstruct the question before reteaching content
Wrong open-ended answer	Evidence selected incorrectly	Train observation and evidence control
Wrong open-ended answer	Reasoning chain incomplete	Repair the missing causal link
Wrong open-ended answer	Science understood but language imprecise	Repair answer construction, not the concept
Wrong open-ended answer	Method collapses only under time	Train execution after the underlying method is stable

The mark tells us that something failed. The teaching job is to locate where.

A Single Science Question Can Contain Several Hidden Jobs

Consider a question with a diagram of an investigation, a table of results and an open-ended prompt. A child may experience it as one question. Cognitively, it may contain many jobs:

• understand the language of the prompt;
• reconstruct what the apparatus is doing;
• identify what changed;
• read the table accurately;
• compare the relevant values;
• retrieve the scientific relationship;
• decide whether the evidence supports that relationship;
• construct the causal explanation;
• express it in sufficiently precise language;
• check that the answer addresses the exact question.

This is why “careless” is often an inadequate diagnosis. The student may have lost the question at any one of these stages.

Worked Example: The Difference Between Knowing Heat and Reading a Heat Question

Suppose two identical metal spoons are placed in hot water. One has a wooden handle covering part of it. The question asks why the covered part feels less hot after the same amount of time.

A student may know the phrase “wood is a poor conductor of heat” and still produce a weak answer such as “Wood does not conduct heat.” The issue is not the keyword. It is the relationship.

• Situation: thermal energy is transferred from the hotter region toward cooler regions.
• Material relationship: wood transfers thermal energy less readily than metal.
• Consequence: less thermal energy reaches the hand through the wooden covering over the same period.
• Observed result: the covered part feels less hot.

The stronger answer is produced by reconstructing the mechanism, not decorating the response with more keywords.

Primary 3 to Primary 6 Is a Change in the Shape of Reasoning

Level	Dominant developmental job	What often breaks
Primary 3	Observe, classify, compare and explain simple relationships	Vocabulary is memorised without a usable concept
Primary 4	Connect structure, function, condition, effect and evidence	Correct facts are attached to the wrong relationship
Primary 5	Manage systems, variables, data and multi-step interactions	The learner loses intermediate links or experiment control
Primary 6	Reconstruct unfamiliar situations and perform reliably under PSLE conditions	Knowledge exists but recognition, transfer, timing or answer construction fails

The progression therefore should not be “more facts every year”. It is a gradual expansion of what the learner can reconstruct and control.

Three Gates: Depth, Load and Transfer

When a student appears to know a topic, we still need to know how robust that knowledge is.

Depth

Can the learner explain the mechanism, distinguish it from nearby ideas and identify what evidence would support it?

Load

Can the learner still use the concept when a diagram, table, variable and written explanation have to be handled together?

Transfer

Can the learner recognise the same relationship when the objects, wording or context change?

A concept that works only in the example where it was taught is not yet a dependable Science capability.

The Science Error Ledger

Instead of keeping only marks, we can keep a compact record of how marks are being lost. Useful categories include:

• Knowledge: the scientific relationship is missing or incorrect;
• Recognition: the learner knows it but does not notice that it applies here;
• Representation: a diagram, table or graph is misread;
• Evidence: the wrong observation or comparison is selected;
• Reasoning: a causal link is absent or reversed;
• Expression: the idea is understood but the written answer is vague or incomplete;
• Transfer: the method works only on familiar surface forms;
• Execution: timing, fatigue or checking causes a method that normally works to disappear.

Over several pieces of work, the ledger begins to show whether the learner has twenty unrelated errors or one recurring mechanism appearing twenty times.

Practice Volume Should Follow Diagnosis

More practice is valuable when the student is practising an improved process. If a child repeatedly misreads variables, ten more full papers can simply reproduce the same error in ten new contexts.

A stronger sequence is:

sample performance → classify the failure → isolate the weak operation → repair → reattempt → vary the context → mix with other topics → return under time.

This does not reduce rigour. It increases the amount of information we obtain from each piece of work.

Why Three Students Can Function as a Science Reasoning Laboratory

The small-group model is most useful when the three learners are not merely doing the same worksheet quietly.

One student can reconstruct the system. A second can challenge the evidence: “Which part of the diagram proves that?” A third can receive the explanation and identify whether a causal step is missing. Then the roles rotate.

This creates three visible positions:

• Builder: form the scientific explanation;
• Challenger: test whether the evidence and relationship support it;
• Receiver: decide whether the explanation is complete and understandable.

The tutor remains the controller of the instructional process, but the group makes reasoning observable. A student who can only repeat another person’s answer becomes visible very quickly.

What Parents Can Ask Instead of “How Many Papers Did You Do?”

• What kind of Science mistake repeats most often?
• Does my child know the concept but fail to recognise it in unfamiliar questions?
• Can they explain what a graph or experiment is showing before answering?
• Do model answers get copied, or does the child reattempt after correction?
• Can the student explain why the original answer failed?
• Does a repaired method still work one week later?
• Can the same scientific relationship be identified when the surface context changes?
• Is the gap between untimed explanation and timed performance getting smaller?

These questions move the conversation from activity to capability.

The Long-Term Science Handover

Primary Science should eventually hand Secondary Science a learner who does more than remember correct facts. The student should increasingly know how to inspect a representation, isolate relevant evidence, identify a relationship, build a causal explanation, challenge an unsupported conclusion and revise when new evidence appears.

That is why the best Primary Science tuition does not make the tutor permanently necessary. The tutor initially carries more of the questioning: What changed? What is the evidence? What relationship applies? Later, those questions should start appearing inside the learner.

The long-term objective is not a child who remembers the tutor’s Science answer. It is a learner who increasingly knows how to reconstruct one.