How to Tell What Must Be True From What Could Be True in PSLE Science

Wait, What? A Possible Answer Can Be Scientifically Sensible and Still Not Be Something You Know

A PSLE Science question gives you a diagram, a few observations and a set of conditions. One answer choice says something that could happen. It sounds scientific. It fits what you remember from your notes.

But the question is not asking whether the statement is possible. It is asking what must be true from the evidence.

That changes the job completely.

“Could be true” means the evidence allows the claim. “Must be true” means the evidence and conditions leave no supported alternative.

A child can know the Science very well and still lose the reasoning because the strength of the claim changes. The mistake is not always a missing fact. Sometimes the learner has quietly promoted a possibility into a certainty.

Quick Answer

When a PSLE Science question asks what must, could, cannot, most likely or is supported, keep the evidence strength visible.

Must be true: every scientifically valid interpretation that fits the given information requires the claim.
Could be true: at least one scientifically valid interpretation fits the claim, but other possibilities may also fit.
Cannot be true under the stated conditions: the claim conflicts with the evidence, condition or relevant scientific relationship.
Best supported / most likely: the evidence favours one explanation more strongly, but that wording does not automatically mean absolute certainty.

Use this route:

READ THE EXACT CLAIM WORD → EXTRACT ONLY THE GIVEN EVIDENCE → IDENTIFY THE SCIENTIFIC OBJECT/RELATIONSHIP → LIST THE PLAUSIBLE INTERPRETATIONS → TEST WHETHER THE CLAIM SURVIVES ALL OF THEM OR ONLY SOME → KEEP CONDITIONS ATTACHED → STATE THE CLAIM AT THE SAME STRENGTH AS THE EVIDENCE → CHECK THAT YOU DID NOT TURN POSSIBILITY INTO PROOF.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one learner job: how a Primary 5 or Primary 6 learner distinguishes a claim that the evidence forces from a claim that the evidence merely allows, so PSLE Science reasoning stays calibrated to “must”, “could”, “cannot”, “supported” and similar evidence-strength language.

It does not replace the general guides on necessary conditions, “most likely”, negative-stem MCQs, absolute words such as always and never, or insufficient information. Those pages own their own learner jobs. This page owns the modal comparison:

Does the evidence force this claim, merely permit it, or rule it out?

Why This Matters in the 2026 PSLE Science Frame

For examination from 2026, PSLE Science assesses attainment in the 2023 Primary Science syllabus. SEAB’s assessment objectives include knowledge with understanding, application of scientific facts, concepts and principles, and scientific inquiry involving prediction and hypothesis, interpretation and analysis of information, evaluation of observations, information and methods, and communication of explanations and reasoning.

That means a learner is not only expected to remember Science. The learner must also decide what the given information actually supports.

The Core Difference: One World Versus Every World That Still Fits

Here is a useful way to think about it.

Imagine the question evidence still allows three scientifically possible explanations: A, B and C.

If a statement is true in A, B and C, it may be a must claim.
If it is true in A but not B or C, it is only a could claim.
If it conflicts with A, B and C, it is a cannot claim under those conditions.

You do not need formal logic symbols. You need the habit of asking whether an alternative scientific explanation still survives.

Worked Example 1 — Same Observation, More Than One Possible Cause

Original practice situation: A sealed transparent container becomes cloudy after some time.

Suppose the question gives no information about what is inside the container, its temperature history or whether particles were already suspended in it.

Can you say the cloudiness must have been caused by one particular process?

No. Several scientifically different explanations may still fit. One process may be plausible, but plausibility is weaker than necessity.

What must be true? Only what the observation itself and the given conditions force—for example, the appearance changed from less cloudy to more cloudy if that is explicitly shown.

The reasoning lesson is:

Observation → possible mechanism is not automatically observation → unique mechanism.

Worked Example 2 — A Fair Comparison Can Strengthen “Must”

Two otherwise comparable set-ups differ only in the amount of exposed wet surface. After the same time, Set-up P has lost more water than Set-up Q.

What is forced by the data?

P lost more water than Q over the tested time — must, because it is directly given by the measurements.
P evaporated faster during every single instant of the entire interval — not necessarily must unless the evidence shows the time path.
The difference is consistent with the tested surface-area relationship under the stated comparable conditions — supported.

A final total can force a conclusion about total change without forcing every detail of how the process unfolded between measurements.

Worked Example 3 — “Could” Does Not Mean “Random Guess”

A learner sees an unexpected low reading in one trial.

Could a timing difference have contributed? Perhaps—if timing affects the measured outcome and the question does not rule the difference out.

Could a dragon have changed the apparatus? No. Scientific possibility is not unlimited imagination. A “could” claim must still be:

consistent with the known scientific relationship;
compatible with the stated conditions;
not contradicted by the evidence;
relevant to the measured outcome.

Possibility still has scientific boundaries.

Worked Example 4 — When “Must” Comes From a Definition or Relationship

Suppose two objects travel the same distance, but Object A takes less time than Object B.

For that same distance, A’s average speed is greater. The evidence forces the comparison because the relationship between distance, time and speed fixes the conclusion.

You do not need to invent a new mechanism. You need to identify the quantity relationship the evidence constrains.

Worked Example 5 — “Must” Can Fail Because One Condition Was Not Given

Two lamps look equally bright in a drawing. A learner says they must produce the same amount of light.

But the diagram may be schematic, not a measured brightness display. If brightness was not measured or defined, the visual drawing does not force the claim.

The missing condition is the link between the representation and the quantity.

Worked Example 6 — A “Cannot” Claim Needs a Contradiction

A learner sometimes treats “not proven” as “impossible”. That is another reasoning error.

If the evidence does not prove that a plant received less water, it does not follow that the plant could not have received less water. The fact may simply be unknown.

To say “cannot”, you need a conflict with the stated conditions, evidence or scientific relationship.

Unsupported is not the same as false. Unknown is not the same as impossible.

The Four Evidence States

State	What it means	What the learner should say
Forced	The evidence and conditions require the claim.	Must / is established under the stated conditions.
Compatible	The claim fits, but another explanation also fits.	Could / is possible / may.
Favoured	The evidence supports one explanation more strongly than alternatives.	Most likely / best supported, if that is the task.
Contradicted	The claim conflicts with evidence or conditions.	Cannot / is not supported under those conditions.

The Alternative-Survival Test

Before calling a claim “must”, ask:

Can I construct another scientifically valid explanation that still fits every piece of given evidence but makes this claim false?

If yes, the claim is not yet forced.

This test is especially powerful when:

only one observation is given;
several conditions changed;
an indicator is indirect;
a diagram omits information;
measurements are sparse;
one plausible mechanism is familiar but not uniquely established.

Do Not Confuse “Must” With “Usually”

A pattern that usually occurs in textbook examples does not become logically necessary in every new question.

For example, a familiar environmental change may often affect a process in one direction. But if another limiting condition changes at the same time, the observed outcome may differ.

Read the current system, not the memory of the last worksheet.

Do Not Confuse “Could” With “Most Likely”

Five explanations may be possible while one is much better supported.

If the question asks what could explain the result, one scientifically compatible explanation may be enough. If it asks what is most likely, you must compare the strength of support.

The command changes the evidence job.

Do Not Confuse “Must” With a Necessary Condition

A process may require Condition X. That does not mean observing X proves the process definitely occurred.

Example structure:

If Process P requires X, then P occurring implies X was available. But X being available does not by itself prove P occurred.

This is why “necessary” and “sufficient” are different from “must” and “could”, even though the reasoning can interact.

Do Not Confuse “Must” With a Strong-Sounding Keyword

Words such as therefore, because, always and definitely do not make a claim stronger scientifically. Evidence does.

Whenever the language becomes more certain, ask what new evidence earns that certainty.

The Evidence Ladder

A useful learner ladder is:

OBSERVED → COMPATIBLE → SUPPORTED → STRONGLY SUPPORTED → FORCED BY THE GIVEN CONDITIONS.

Not every PSLE Science question uses these exact words. The ladder is a thinking tool. It helps you keep the conclusion from climbing higher than the evidence.

The PSLE Science Reasoning Chain

Read the given information.
Identify the scientific object or relationship.
Separate direct observation from inference.
Select the relevant scientific concept.
Explain the causal mechanism if the question asks why.
Reconnect the mechanism to the exact condition.
State the outcome.
Calibrate the claim: must, could, favoured, unknown or contradicted?
Check against every piece of evidence.

MCQ Protocol — Must, Could and Cannot

Circle or mentally hold the modal word: must / could / cannot / most likely.
Read each option as a full scientific claim.
Attach the conditions from the stem.
For a must question, try to break the option with one valid alternative.
For a could question, check whether at least one valid scientific path fits.
For a cannot question, look for a contradiction, not merely lack of proof.
Reject generally true Science statements that do not answer the exact evidence job.
Re-read the modal word before submitting.

Structured-Answer Protocol — Calibrate the Sentence

Useful language depends on what the evidence supports.

Directly established: “The data show that…”
Supported possibility: “One possible explanation is…”
Favoured but not certain: “The evidence supports…” or “The most likely explanation is…” when appropriate.
Insufficient: “The information is insufficient to determine whether…”
Contradicted: “This cannot be concluded because…”

These are not mandatory marking phrases. They are examples of language that keeps claim strength visible.

The Earliest-Weak-Link Diagnostic

Failure signature	Earliest weak link	Repair
“It sounds possible, so it must be the answer.”	Possibility promoted to necessity.	Run the alternative-survival test.
“We cannot prove it, so it is false.”	Unsupported confused with contradicted.	Separate unknown from impossible.
“This usually happens, so it must happen here.”	Familiar pattern substituted for current conditions.	Re-read the system and every changed condition.
“All the keywords are correct.”	Vocabulary used instead of evidence logic.	Test the whole claim, not individual words.
“There is one observation, therefore one cause.”	Unique-cause assumption.	Generate at least one alternative mechanism that could produce the observation.
“Could” means any imaginable story.	Scientific compatibility lost.	Require concept, condition and evidence consistency.
“Most likely” means definitely true.	Evidence ranking converted to certainty.	Keep ranked support separate from proof.

Misconception Repair 1 — “Possible Means Proven”

Give the learner one observation and ask for three scientifically possible causes. Then ask what extra evidence would distinguish them. The moment several explanations survive, the learner can see why possibility is weaker than proof.

Misconception Repair 2 — “Not Proven Means Impossible”

Use a hidden box. The box could contain a red object or a blue object. If you have no evidence about colour, neither colour is proven—but neither is impossible. Bring the same distinction back to scientific claims.

Misconception Repair 3 — “One Cause Must Explain Everything”

Use an outcome such as a lower final temperature. Ask for several scientifically relevant causes: different starting temperature, different heat transfer, different duration, different material. Then reveal which conditions were controlled. Evidence removes alternatives one by one.

Misconception Repair 4 — “If It Is in My Notes, It Must Apply”

Notes describe general scientific relationships. A question supplies a particular system. Require the learner to connect the remembered concept to the current condition before accepting the claim.

Practice Sequence — From Obvious to Subtle

Start with direct observations where the must-claim is obvious.
Add one plausible but unproven explanation.
Add two explanations that both fit.
Add a fair-test condition that eliminates one explanation.
Add an indirect indicator so observation and inference must be separated.
Add sparse data so intermediate events remain unknown.
Add a “cannot” option that is merely unsupported and another that is actually contradicted.
Add “most likely” so the learner ranks evidence without claiming certainty.
Mix must/could/cannot stems without announcing which distinction is being tested.

Unfamiliar Transfer Challenge

A mystery device is tested twice. Trial 1 gives 12 units. Trial 2 gives 8 units. The question states that only one named condition was deliberately changed, but it does not say whether every other relevant condition was controlled successfully.

What must be true?

The recorded outcome in Trial 1 is greater than in Trial 2.

What could be true?

The deliberately changed condition contributed to the difference—if the relevant scientific mechanism fits.
An uncontrolled difference may also have contributed.

What cannot you yet claim?

That the named condition alone definitely caused the entire difference.

The content is unfamiliar. The evidence-strength reasoning still transfers.

Delayed Independent Return

Three to five days later, take a fresh PSLE Science item and answer without notes:

What is directly observed?
What is inferred?
What scientific relationship matters?
Which claims are compatible with the evidence?
Which claim is forced?
Which claim is contradicted?
What alternative explanation still survives?
What extra evidence would turn a “could” into a stronger claim?
Did the question ask must, could, cannot or most likely?

The Answer-Checking Receipt

Did I preserve the exact modal word in the question?
Did I use only information actually given?
Did I distinguish observation from inference?
Did I test alternative explanations?
Did I keep the question conditions attached?
Did I separate unsupported from contradicted?
Did I avoid turning “could” into “must”?
Did I avoid turning “most likely” into certainty?
Did I reject generally true statements that are irrelevant to this evidence?
Is my final sentence exactly as strong as the evidence allows?

Evidence and Model Limits

Real science often works with degrees of confidence, statistical uncertainty and models that are more sophisticated than the simple must/could distinction used here. Primary Science does not require formal probability or symbolic logic for this learner job.

The durable habit is simpler: do not claim more than the evidence earns.

Also remember that the exam wording itself controls the task. This guide does not invent a universal marking rule or claim that every PSLE Science question will use the words “must” or “could”. It teaches the reasoning when evidence strength matters.

Useful Internal Routes

Parent and Tutor Teaching Guide

Use three cards: MUST, COULD and CANNOT.

Give one original Science situation at a time. Ask the learner to place each claim under a card and defend the placement with evidence. Do not tell them the topic name first.

When the child places a claim under MUST, ask: “Can you invent one scientifically valid alternative that keeps all the given evidence but makes this statement false?” If the child can, move it down to COULD.

When the child places something under CANNOT, ask: “What exactly contradicts it?” If there is no contradiction, the claim may simply be unknown.

Then remove the cards. The learner is ready when the distinction survives inside unfamiliar diagrams, tables and mixed-topic questions.

Authoritative References

The Quiet Ending

Good Science is not only knowing what might happen.

It is knowing how strongly you are allowed to say it.

Possible. Supported. Forced. Contradicted.

Keep the claim at the level the evidence earns.

Small Group Tutorials