Punggol Primary 4 Science Tuition | 3-Pax Cause, Evidence & Competing-Explanation Control

---

Phase 4 Deepening — From a Possible Cause to a Defensible Scientific Explanation

Deepened: 4 September 2026. This page remains the specialist owner for cause, evidence and competing-explanation control in Primary 4 Science.

The canonical year-level parent is Primary 4 Tuition Punggol | English, Mathematics & Science Hub. The protected Science heroes remain untouched at Primary 4 Science Tutors in Punggol and Punggol Primary 4 Science Tuition. This specialist does not attempt to own the whole subject. It owns the moment when several explanations appear possible and the student must decide which one the evidence allows to survive.

Featured Snippet — How Should Primary 4 Students Choose Between Competing Scientific Explanations?

Primary 4 students should first separate observation from explanation, identify what changed and what remained comparable, generate a small set of plausible causes, state what each cause would predict, compare those predictions with the evidence, reject explanations that contradict the setup, build the mechanism connecting the surviving cause to the outcome, and limit the conclusion to what the evidence actually supports. The strongest explanation is not the most familiar one; it is the one that accounts for the relevant evidence with the fewest unsupported assumptions.

The Explanation-Selection Architecture

OBSERVATION → CANDIDATE CAUSES → PREDICTIONS → DISCRIMINATING EVIDENCE → SELECTION → MECHANISM → BOUNDARY → RETEST

This is a teaching architecture, not an official examination formula.

It protects students from two opposite errors.

• Premature certainty: the first familiar cause is accepted without testing.
• Unlimited possibility: every imaginable cause is treated as equally valid.

Science operates between those extremes. It begins open enough to consider alternatives and becomes narrow through evidence.

Possible Is Not Yet Supported

A plant can grow poorly for many real-world reasons.

• too little water;
• too little light;
• unsuitable temperature;
• damaged roots;
• poor soil conditions;
• disease;
• insufficient time.

But a particular question may state that water, temperature, soil and time were kept comparable while only light differed.

Too little water remains possible in the world.

It is not admissible as the explanation for this comparison unless the question provides evidence that water differed.

REAL-WORLD POSSIBILITY ≠ EVIDENTIAL SUPPORT INSIDE THE QUESTION.

Observation Before Explanation

Students should state what the setup, diagram, table or graph directly shows before naming a cause.

Sentence	Role	Why the distinction matters
The water level was lower after two hours.	Observation	Reports a visible or measured change.
Some water evaporated.	Inference or explanation	Names a process that may account for the observation.
The water vanished.	Vague description	Hides the scientific process and may imply matter ceased to exist.
The warmer condition increased the rate of evaporation, so more liquid water changed into water vapour.	Causal explanation	Connects condition, process and outcome.

If observation and explanation are mixed too early, a memorised chapter answer can replace the actual evidence.

The Evidence Inventory

Before generating causes, list what is given.

• what was observed or measured;
• the direction of change;
• which condition differed;
• which conditions remained comparable;
• the time sequence;
• labels and arrows;
• units and scale;
• what happened before and after;
• what the question states as known.

The inventory keeps later explanations attached to the problem rather than to whichever topic word feels familiar.

Generate a Small Candidate Set

Students do not need ten possible causes.

Two or three plausible candidates are often enough to teach discrimination.

For example, a bulb does not light.

• Candidate A: the circuit path is open.
• Candidate B: a material in the path does not conduct electricity.
• Candidate C: the bulb or cell is not functioning.

The question is not merely which cause is possible.

It is what each cause would predict and what evidence would distinguish them.

Every Explanation Makes a Prediction

A useful explanation should tell us what else we would expect to observe.

Candidate explanation	Prediction	Discriminating check
Open circuit	Completing the path should allow the bulb to light if components work.	Close the switch or repair the break.
Non-conducting material	Replacing it with a conductor should change the outcome.	Use a known conducting material in the same position.
Faulty bulb	The bulb should remain unlit in another working circuit.	Test the bulb in a known functioning setup.

Prediction turns an explanation into something that can be tested.

The Discriminating Clue

A discriminating clue is evidence that different explanations would not all predict.

• If both plants received equal water, water difference cannot explain the comparison.
• If the bulb lights in another circuit, a faulty bulb becomes less likely.
• If two materials have the same mass but different volumes, size alone does not determine mass.
• If a shadow changes while the object’s position remains fixed and the light source moves, the light-source relationship becomes central.

The student should learn to ask:

Which detail would be different if the other explanation were true?

Supporting Evidence and Excluding Evidence

Evidence can strengthen one explanation.

It can also remove another.

A student who only searches for support may overlook a decisive contradiction.

Use two columns:

Evidence for	Evidence against
What observation does this explanation predict correctly?	Which observation is difficult for it to explain?
Which changed condition fits the mechanism?	Which controlled condition rules it out?
What additional pattern becomes understandable?	What result would have appeared if it were true but did not?

Scientific judgement becomes stronger when students can both admit and reject possibilities for stated reasons.

Negative Evidence Must Be Used Carefully

Not observing an effect can weaken an explanation, but only when the method was capable of detecting the effect.

For Primary 4, the language can remain simple:

• Was the observation long enough?
• Was the measuring tool suitable?
• Was the expected change large enough to notice?
• Did the setup actually test the prediction?

Absence of observed evidence is not always evidence of absence.

Controlled Conditions Remove Alternative Causes

A controlled condition is not merely something pupils memorise under the instruction “keep everything the same”.

It performs an evidential job.

If this condition also changed, could it produce the measured outcome?

If yes, it becomes a competing explanation.

Keeping it comparable protects causal clarity.

Confounding Conditions

A confounding condition changes alongside the intended factor and may also affect the outcome.

Suppose one plant receives more light but is also given more water.

If it grows more, the evidence cannot cleanly tell us which difference produced the result.

The correct Primary 4 conclusion may be:

The investigation does not provide a fair comparison of light alone because the amount of water also differed.

Recognising when a cause cannot yet be isolated is part of scientific reasoning.

Correlation Is Not Automatically Cause

Two changes may occur together without one necessarily producing the other.

At Primary 4 resolution, students can ask:

• Was the proposed cause deliberately changed?
• Did it happen before the outcome?
• Could another condition explain both?
• Does a mechanism connect them?
• Would changing the cause predictably change the outcome?

The vocabulary need not become advanced. The distinction can be taught through careful comparison.

Cause Needs a Mechanism

“It happened because the temperature was higher” may identify a relevant condition without explaining the Science.

The complete route is:

CHANGED CONDITION → AFFECTED PROCESS OR RELATIONSHIP → DIRECTION OF CHANGE → OBSERVED OUTCOME.

The causal middle may involve:

• faster or slower evaporation;
• energy moving from a warmer to a cooler region;
• a complete or broken path;
• a structure allowing a function;
• water or another substance moving through a system;
• a force changing movement;
• an environmental condition affecting a biological process.

The mechanism is what makes the explanation scientific rather than merely associative.

The Causal Chain Test

1. Name the condition that differed.
2. Name the component, process or relationship affected.
3. State whether it increased, decreased, opened, closed, moved or changed.
4. Connect that change to the observed outcome.
5. Check that every link belongs to the evidence and concept.

If a link is missing, the answer may jump from beginning to end.

Counterexamples Test Over-Broad Rules

A student says:

Bigger objects are heavier.

Introduce a large, light sponge and a smaller, dense object.

The counterexample does not merely show the answer is wrong.

It forces a better rule:

• Which property was confused?
• What must be measured rather than judged visually?
• Under what limited conditions might the original observation have seemed true?

Counterexamples reveal model boundaries.

Counterfactual Questions

Change one condition mentally.

• What if the switch were closed?
• What if both plants received the same light?
• What if the material were replaced?
• What if the object were moved but the light source remained fixed?
• What if the temperature difference were smaller?

A model that predicts the changed outcome is more likely to be understood rather than memorised.

Explanation Strength Has Levels

Level	What the student does	Remaining weakness
Names a cause	states one possible factor	may not be supported
Uses evidence	links factor to a changed condition	mechanism may be missing
Builds mechanism	connects cause to outcome	alternative explanations may remain
Discriminates	shows why one explanation fits better	claim may still be too broad
Bounds	limits conclusion to tested conditions	needs transfer and delayed return
Transfers	reconstructs the reasoning in a changed setup	move into maintenance when stable

Claim Size Should Match Evidence Size

One comparison may support a conclusion about the tested conditions.

It may not prove an absolute rule for every object, organism or environment.

Useful boundary language includes:

• in this setup;
• under these conditions;
• compared with;
• the results support;
• may have caused;
• does not by itself show;
• more evidence would be needed.

Students do not need to weaken every answer with uncertainty. They need to match confidence to evidence.

Confidence Calibration

Before discussion, students can mark:

• High confidence: one explanation predicts the evidence clearly and alternatives are contradicted.
• Medium confidence: one explanation fits best but another remains possible.
• Low confidence: evidence is insufficient or the setup is not understood.

After analysis, compare confidence with the quality of the evidence chain.

The goal is not to make every child sound certain.

It is to make certainty earned.

The Competing-Explanation Matrix

Candidate	Mechanism	Prediction	Supporting evidence	Contradicting evidence	Decision
A	How would A produce the outcome?	What else should occur?	Which observations fit?	Which observations do not?	keep, narrow or reject
B	How would B produce the outcome?	What else should occur?	Which observations fit?	Which observations do not?	keep, narrow or reject

The matrix need not appear in every answer. It is a temporary external support for learning how evidence selects.

The Three-Student Explanation Lab

Three students can propose three causes for one observation.

The room becomes useful when every learner first records an independent observation and candidate.

• Student A identifies a familiar cause but ignores a controlled condition.
• Student B selects the correct cause but cannot state the mechanism.
• Student C builds a complete mechanism but claims it always happens.

One model answer cannot repair all three.

Peer comparison should ask:

• What would each explanation predict?
• Which condition rules one out?
• Which link is missing?
• Which claim is too large?
• What additional evidence would decide?

The First-Attempt Protocol

OBSERVE → RECORD CANDIDATE → MARK CONFIDENCE → STATE PREDICTION → COMPARE → REVISE → RETEST ALONE.

The fastest speaker should not decide which explanations everyone else appears to know.

The 90-Minute Cause-and-Evidence Runtime

1. 10 minutes — Retrieval: revisit observation, inference and cause.
2. 10 minutes — Evidence intake: describe a new setup without explaining it.
3. 10 minutes — Candidate generation: propose two plausible causes.
4. 10 minutes — Prediction: state what each cause would imply.
5. 10 minutes — Discrimination: identify the clue that separates them.
6. 15 minutes — Mechanism: build condition, process and outcome.
7. 10 minutes — Boundary: limit the conclusion to the evidence.
8. 5 minutes — Peer calibration: compare reasoning after first attempts.
9. 5 minutes — Counterfactual: change one condition and predict.
10. 5 minutes — Delayed-return plan: record the active error owner.

Catch Up, Keep Up, Move Ahead

Catch Up

Use one clear observation and two visibly different explanations. Ask the learner to eliminate one using a stated condition.

Keep Up

Use diagrams, tables and everyday contexts where evidence is complete enough to select, then introduce cases where the conclusion must remain limited.

Move Ahead

Ask what additional observation would distinguish two models, whether the method isolates the cause and how the conclusion changes when one assumption is removed.

Strong Students: Test the Model, Not Just the Answer

• generate a credible alternative;
• state what evidence would falsify the preferred explanation;
• identify a confounding condition;
• qualify the conclusion;
• design one discriminating observation;
• find a counterexample to an over-broad rule;
• compare which explanation accounts for more evidence with fewer assumptions.

Students Who Are Struggling: Build One Defensible Chain

Use:

I OBSERVE ___.

ONLY ___ CHANGED.

THIS AFFECTED ___, CAUSING ___.

Begin with one cause and one contradicted alternative.

Complexity should increase only after the evidence-to-mechanism route becomes stable.

The Error Ledger

Error owner	Visible pattern	Repair
Observation–explanation mix	cause stated as though directly seen	describe first, explain second
Possibility error	any real-world cause accepted	test against stated conditions
Single-cause reflex	first familiar explanation chosen	generate one alternative and prediction
Control blindness	kept-same conditions ignored	state which alternative each control removes
Confound blindness	two changed factors treated as one	identify why cause cannot be isolated
Mechanism gap	condition jumps to outcome	insert affected process and direction
Contradiction blindness	only supporting evidence noticed	find the hardest observation to explain
Overclaim	one setup becomes an absolute rule	limit claim to tested conditions
Transfer block	reasoning works only in original context	change surface, diagram and condition

The Stop Rule

A cause-and-evidence target can move into maintenance when the student can:

• separate observation from explanation;
• generate a small candidate set;
• state predictions;
• use a discriminating clue;
• reject contradiction;
• build the causal middle;
• bound the claim;
• repeat the process on a changed setup after delay and with reduced prompting.

The Primary 4 → Primary 5 Science Handoff

Primary 5 should receive a learner who can increasingly:

• begin from observation rather than a memorised topic answer;
• distinguish possible from supported;
• use controlled conditions as evidence;
• recognise confounding changes;
• compare two explanations through predictions;
• identify evidence that supports and evidence that excludes;
• build a complete mechanism;
• keep claim size within evidence size;
• calibrate confidence;
• ask what further evidence would decide.

The next level owner is Primary 5 Tuition Punggol | English, Mathematics & Science Hub. The Primary 4 readiness specialist is Punggol Primary 4 Science Tuition | Primary 5 Readiness Bridge.

The Competing-Explanation Principle in One Sentence

Primary 4 Science becomes disciplined when the learner treats a cause as a candidate rather than an answer, asks what it predicts, compares those predictions with the changed and controlled conditions, rejects contradictions, builds the mechanism and states only the conclusion the evidence has earned.

Continue Through the Punggol Primary 4 Science Spine

Primary 4 parent hub: Primary 4 Tuition Punggol

Protected Science tutor hero: Primary 4 Science Tutors in Punggol

Protected broad Science hero: Punggol Primary 4 Science Tuition

Readiness specialist: Primary 5 Readiness Bridge

Structure–function specialist: Structure–Function Mapping & Diagram Annotation

Experimental specialist: Variables, Fair Tests & Experimental-Setup Reading

Next level: Primary 5 Tuition Punggol