How to Use Scientific Constraints in PSLE Science

Wait, What? A Science answer can contain a true fact and still be impossible for the question in front of you.

That is why familiar words are dangerous. A learner sees “heat”, “roots”, “force” or “energy”, remembers a true statement, and chooses it because the statement sounds scientific. But PSLE Science questions do not ask whether a sentence is generally familiar. They ask what follows from a particular object, condition, observation, relationship or investigation.

One powerful way to reason through an unfamiliar question is to turn the information into scientific constraints. A constraint is something the answer must respect. If a candidate answer contradicts even one decisive condition or observation, it cannot be the answer to that question, even if part of it is scientifically true somewhere else.

Quick Answer

Read the question and translate the important information into statements such as:

This must remain the same.
This quantity increased.
This effect happened only after this condition changed.
This object, not another object, is being measured.
The evidence supports a difference but not necessarily a cause.
The result belongs to this time point or set-up.

Then test each possible answer against those constraints. Reject a candidate only when you can identify the contradiction. If two candidates still fit, do not invent a reason to eliminate one. Return to the evidence and ask what additional condition, mechanism or comparison distinguishes them.

The core reasoning chain is:

READ GIVEN INFORMATION → IDENTIFY THE SCIENTIFIC OBJECT OR RELATIONSHIP → DISTINGUISH OBSERVATION FROM INFERENCE → TURN RELEVANT INFORMATION INTO CONSTRAINTS → TEST EACH CANDIDATE → SELECT THE RELEVANT CONCEPT → EXPLAIN THE MECHANISM → CONNECT TO THE QUESTION’S CONDITION → STATE THE OUTCOME → CHECK AGAINST THE EVIDENCE.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one learner job: how a Primary 5 or Primary 6 student can use the information supplied in a PSLE Science question as boundaries that possible answers must satisfy, so impossible answers can be ruled out scientifically rather than by guessing from familiar vocabulary.

It does not own the underlying scientific concepts. It does not create a universal MCQ trick. It does not claim that “constraint” is an official SEAB command word or marking formula. Existing Science concept pages remain canonical; this page teaches how to use concept knowledge and question evidence together.

Why This Matters in the Current PSLE Science Frame

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

Those jobs require more than recalling a fact. A learner must decide whether a fact is relevant under the conditions actually given. Constraint reasoning helps with that decision because it keeps every possible answer tied to the evidence.

What Counts as a Scientific Constraint?

A constraint is not a magic keyword. It is a boundary created by the question, the evidence or a relevant scientific principle. Different questions create different kinds of constraints.

Constraint source	Example of what it fixes	What an answer must respect
Object identity	Which part, sample or set-up is being discussed	Do not explain the result for a different object
Comparison	A is compared with B under stated conditions	Do not switch the reference halfway through
Direction of change	A measured quantity rises or falls	Do not choose an answer predicting the opposite direction
Time	Before, after, immediately, later	Do not use a result from the wrong stage
Controlled condition	A factor is kept comparable	Do not explain the result by claiming that factor differed
Measurement	What was actually observed or measured	Do not replace the measured outcome with another quantity
Evidence scope	One trial, one set-up, repeated trials or whole investigation	Keep the claim at the same evidence level
Scientific principle	A concept imposes a relationship or mechanism	Reject candidates that violate that relationship

Constraint Is Not the Same as Keyword

A keyword says, “This word appeared.” A constraint says, “Because this information is true in this question, the answer cannot contradict it.”

For example, the word “same” does not automatically tell you what is identical. The question may say that two set-ups contain the same amount of water while their containers differ. The useful constraint is not “same”. It is the amount of water is held comparable between the two set-ups. An explanation that depends on one set-up having more water would violate that constraint.

Worked Example 1: Eliminate a True-Sounding MCQ Option

Imagine an original practice question with two identical containers. Both begin with the same amount of warm water at the same temperature. Container A is wrapped with material; Container B is not. After the same time, the water in A has a higher temperature than the water in B.

Before looking at possible explanations, extract constraints:

The starting amount of water is the same.
The starting temperature is the same.
The comparison is after the same duration.
The wrapping condition differs.
A ends at a higher temperature than B.

Now suppose an option says, “A remained warmer because it started with more water.” The statement might describe a situation that could matter in another investigation, but it is impossible here because it contradicts the stated starting condition.

Notice the method: we did not reject the option because its wording looked unfamiliar. We rejected it because it violated a question-specific constraint.

Worked Example 2: Preserve Uncertainty When Two Explanations Still Fit

Suppose a learner observes that a measured response is lower in Set-up X than Set-up Y. The question tells you that two conditions differ between X and Y.

The evidence constraint is clear: the response differs. But the design does not isolate which changed condition produced the difference.

Two explanations may therefore remain possible. Constraint reasoning should not force one answer merely because the learner wants a single cause. The correct scientific move is to preserve what is undecided and ask what comparison would distinguish the explanations.

This is an important limit: elimination is only as strong as the evidence. Scientific reasoning includes knowing when you cannot yet eliminate enough possibilities.

Worked Example 3: A Method Can Violate the Question Before Any Data Is Collected

Suppose an investigation asks how the amount of light affects a measured plant response. A proposed method changes both the light condition and the amount of water supplied.

The scientific question creates a design constraint: if the aim is to examine the relationship between light and the outcome, other explanation-relevant conditions should be kept appropriately comparable where possible. Changing water at the same time makes the result harder to attribute to light.

Constraint reasoning therefore helps before answering a result question. It can expose a method flaw at the planning stage.

Worked Example 4: An Answer Must Match the Exact Scientific Object

Imagine a diagram containing a whole system and one labelled part. The table reports the temperature of the whole system, while the question asks what happens to the labelled part.

A common error is to transfer the whole-system measurement directly to the part without evidence. Object identity is a constraint. Before using a number, ask: What exactly was measured? The answer cannot silently change the measured object.

The Constraint Board: A Scratch Method for Difficult Questions

During practice, divide a small area of scratch paper into four lines:

OBJECT: What thing, part, sample or set-up is the question about?
GIVEN: What observations, measurements or relationships are directly supplied?
CONDITIONS: What must be held in the reasoning—time, comparison, changed factor, controlled factor, stage?
CANNOT: What would contradict those facts?

This is practice scaffolding, not an examination requirement. As the learner becomes fluent, the board can shrink into a quick mental check.

How to Test a Candidate Answer

State the candidate precisely. Do not test a vague impression.
Match the object. Is the candidate talking about the same object or system?
Match the condition. Does it preserve the exact condition in the question?
Match the direction. Does it predict the observed increase, decrease, presence, absence or equality?
Match the timing. Is it about the correct stage?
Match the evidence level. Does one reading justify the candidate’s broad claim?
Check the concept. Is the relevant scientific principle applied correctly?
Find the contradiction. If rejecting the candidate, be able to name the evidence or principle it violates.

Observable Failure Signatures

The learner chooses the option containing the most familiar scientific words.
The learner rejects an answer because “it sounds wrong” but cannot name a contradiction.
The learner adds an unstated condition and then uses it to eliminate an option.
The learner treats “possible” as “proven”.
The learner keeps only one explanation even though the evidence allows two.
The learner uses a result from the wrong time point.
The learner explains a difference using a factor that the question says was kept the same.
The learner switches from one measured object to another without noticing.
The learner believes every clue must be used, so irrelevant context becomes a false constraint.

Earliest Weak-Link Diagnosis

Failure	Check first	Repair
Uses familiar words to choose	Has the learner extracted evidence first?	Hide options and predict from constraints
Invents a condition	Given vs inferred information	Label every condition as GIVEN or INFERRED
Rejects without reason	Can the contradiction be named?	Require “This conflicts with ___” during practice
Over-eliminates	Does evidence truly contradict the candidate?	Separate impossible from merely unproven
Cannot choose between two	Is there a decisive condition in the stem?	Compare only the scientific difference between the survivors
Chooses a broad claim	Evidence scope	Match claim to reading, trial, set-up or investigation

Misconception Repair: “Not Proven” Is Not the Same as “Impossible”

This is one of the most important boundaries in scientific reasoning.

Impossible under the given constraints means the candidate contradicts the evidence, conditions or relevant scientific principle.

Not proven means the available evidence does not establish it. The statement might still be possible.

If the evidence shows only that A and B differ, it does not automatically prove why they differ. Do not eliminate every alternative cause unless the investigation has actually ruled it out.

Constraint vs Counterexample

These two tools are related but different.

Tool	Main job	Example
Constraint	Tests whether an answer fits this question’s evidence and conditions	Both set-ups start with the same amount, so “A began with more” is incompatible
Counterexample	Tests whether a broad claim is universally true	Find one scientifically valid case where an “always” statement fails

A learner may use both, but should know which job is being performed.

Using Constraints in PSLE Science MCQ

A useful practice sequence is:

Cover the options.
Read the stem and extract two or three decisive constraints.
Predict the direction or type of answer you expect.
Reveal the options.
Test each option against the same constraints.
When two remain, identify their exact scientific difference.
Return to the stem and locate the evidence that distinguishes them.
Check the selected option against every decisive condition once more.

This prevents the options from becoming the learner’s source of Science knowledge.

Using Constraints in Open-Ended Answers

Constraints are not only for eliminating MCQ options. They help construct explanations.

Before writing, ask:

What result must my explanation account for?
What condition must the mechanism include?
What factor cannot be used because it was held comparable?
Which object must remain the subject of the explanation?
How far can the conclusion travel beyond the tested cases?

Then build the answer from evidence through mechanism to outcome rather than writing every fact remembered from the chapter.

Using Constraints in Inquiry Questions

For investigations, the question itself creates design constraints. If the learner wants to test how one factor affects an outcome, the method should allow an informative comparison. Relevant conditions that could otherwise explain the difference may need to be controlled or accounted for. Measurements must actually represent the stated outcome. Repeated evidence should not be confused with a fair comparison.

Always preserve the inquiry roles:

QUESTION → CHANGED OR COMPARED CONDITION → MEASURED OUTCOME → CONTROLLED CONDITIONS → METHOD → OBSERVATIONS → ANALYSIS → CONCLUSION → LIMITS.

A Retrieval Practice Sequence

During revision, do not only memorise which option was correct. Reconstruct why the other candidates failed.

Choose one previously answered question.
Hide the answer.
Write three decisive constraints from memory.
Reconstruct the answer.
For each rejected candidate, name the contradiction.
Change one surface detail while preserving the mechanism.
Ask whether the original constraints still apply.
Return after a delay and repeat without the old annotations.

Unfamiliar Transfer Test

Use a new context that the learner has not seen before. Do not announce the topic. Ask the learner to find the object, evidence, conditions and relationship before choosing a concept.

The transfer test is passed when the learner can say not only “I think this is the answer” but also:

“This candidate cannot fit because it contradicts ___; this candidate is still possible because ___; this one fits both the evidence and the scientific mechanism.”

That sentence is a practice explanation, not a compulsory examination template.

Delayed Independent Return Test

Several days later, present a question with different objects but the same reasoning demand. Remove the old notes. The learner should independently:

extract constraints;
avoid inventing unstated conditions;
distinguish impossible from merely unproven;
select the relevant concept;
build or choose an answer;
check it against the evidence.

Answer-Checking Receipts

Did I keep the same scientific object?
Did I use only conditions that are given or scientifically justified?
Does my answer match the observed direction of change?
Does it belong to the correct time or stage?
Does it require a factor that was kept the same?
Does my claim go beyond the evidence?
If I rejected another answer, can I name the contradiction?
If two answers still fit, have I preserved the uncertainty rather than guessed?

Common Traps

Treating a familiar word as a reason.
Inventing a hidden condition to make a preferred answer work.
Rejecting a candidate because it was not observed, even when it remains possible.
Using every detail in the question whether relevant or not.
Assuming a relationship proves a cause.
Forgetting that one measurement may belong to one part rather than the whole system.
Reversing the comparison reference.
Turning one trial into a universal conclusion.
Believing that elimination removes the need to understand the scientific mechanism.

Parent and Tutor Teaching Guide

When a learner chooses a wrong answer, avoid immediately supplying the right one. Ask, “Which given fact does your answer have to obey?”

If the learner cannot name a constraint, help them sort the stem into object, evidence and conditions. If they can name the constraints but still choose incorrectly, inspect concept knowledge. If they eliminate too aggressively, ask whether the rejected answer is actually impossible or merely not proven.

The teaching goal is a learner who can defend elimination with evidence and who is equally willing to say, “The information does not yet let me eliminate this possibility.” That is scientific discipline, not hesitation.

Useful eduKate Routes

Authoritative External References

Evidence note: The official documents establish the curriculum and assessment objectives. The constraint-board method in this guide is an eduKate reasoning scaffold. It is not an official SEAB answer template or marking rule.

The Quiet Return

Do not ask first, “Which answer sounds most scientific?”

Ask, “What must any possible answer respect?”

Then let the evidence close the impossible doors. If more than one door remains open, keep them open until the Science gives you a reason to close one.

Small Group Tutorials

How to Use Scientific Constraints to Rule Out Impossible PSLE Science Answers