How to Track What Stays the Same in PSLE Science When Something Changes

Wait, What? Change Is Easier to See Than Continuity

A piece of paper is crumpled.

Its shape changes immediately. The learner notices that.

The same paper is then unfolded. It looks different again.

Now ask a quieter question:

What have we been following through all those changes?

The same piece of paper did not become a new scientific object merely because its form changed. Its identity continued while one of its properties changed.

This distinction appears everywhere in PSLE Science. Water can change state. An object can change position. A plant can grow. A circuit can be redrawn. A quantity can rise while the system being measured stays the same. A multi-part question can move from “before” to “after” while still referring to one setup.

Learners often concentrate so hard on what changed that they accidentally lose what stayed the same. Then the explanation breaks. They compare different objects as if they were one, treat the same object as if it were new, or assume a property remained unchanged when the evidence never said so.

Good scientific reasoning tracks change and continuity at the same time.

Quick Answer

When a PSLE Science question describes change, first identify the object, system or relationship you are following. Then separate its identity from its changing state, property, position, condition or outcome. Mark what the question states or validly implies remains the same, and mark what is not established.

Use this route:

IDENTIFY WHAT IS BEING FOLLOWED → MARK THE STARTING STATE → MARK THE CHANGED CONDITION OR PROCESS → LIST WHAT DEFINITELY CHANGES → LIST WHAT DEFINITELY STAYS THE SAME → LEAVE UNSUPPORTED “SAME” CLAIMS OPEN → EXPLAIN THE MECHANISM → STATE THE OUTCOME → CHECK THAT THE SAME OBJECT, QUANTITY AND SYSTEM ARE STILL BEING TRACKED.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one learner job: how a Primary 5 or Primary 6 learner tracks scientific identity and other relevant invariants through a change without assuming that everything stayed the same or that everything became different.

It does not own the scientific mechanisms of states of water, forces, circuits, plant growth, ecosystems or any other concept. Existing concept pages remain canonical.

It also does not replace the fair-test idea of deliberately keeping controlled conditions the same. That is a different job. Here, the learner is asking:

Across this question, what is still the same thing, what has changed about it, and what has not been established either way?

Why This Matters for the 2026 PSLE Science Frame

For examination from 2026, the revised PSLE Science paper assesses attainment in the 2023 Primary Science syllabus. The official assessment objectives include knowledge with understanding, application of scientific facts, concepts and principles, and scientific inquiry such as prediction, interpretation and analysis, evaluation of observations, information and methods, and communicating explanations and reasoning.

All of those jobs become harder if a learner loses track of what the question is referring to. Scientific reasoning depends on continuity: the learner must know whether two measurements belong to the same object, whether a changed diagram still represents the same system, whether a later observation follows from an earlier state, and whether a comparison is between like quantities.

The five Primary Science themes—Diversity, Cycles, Systems, Energy and Interactions—also invite learners to follow relationships across change. A cycle changes state while preserving a connected sequence. A system changes when one part or condition changes. An interaction produces effects while the learner must still track the objects involved. Energy questions require careful tracking of sources, transfers and effects. Diversity asks learners to distinguish identity and grouping criteria rather than treating every visible difference as a new scientific category.

Identity, State, Property and Relationship Are Not the Same Thing

Layer	Question	Example
Identity	What object, organism, material or system are we following?	The same metal spoon.
State	What condition is it in now?	Warmer now than before.
Property	What characteristic is being described or measured?	Temperature, shape, mass, length, brightness.
Relationship	How is it connected to another object or condition?	Connected in the same circuit path.

A change in one layer does not automatically mean every layer changes.

A spoon can remain the same spoon while its temperature changes. A plant can remain the same individual while its height changes. A diagram can be redrawn while representing the same circuit connections. Water can remain water while changing between familiar physical states.

But do not turn this into a universal conservation rule. Whether mass, amount, number, energy or another quantity remains the same depends on the system, evidence and process.

The Three-Column “Same / Changed / Not Established” Test

For difficult before-and-after questions, make three mental columns:

Same	Changed	Not established
Same container	Temperature	Exact amount of heat transferred
Same plant	Height	Mass, unless measured or stated
Same piece of paper	Shape	Surface area exposed in a particular setup unless defined

The third column is essential. Strong Science does not force every feature into “same” or “changed” when the question gives no evidence.

What Counts as “the Same” Must Be Defined

Two objects can be the same in one sense and different in another.

Two cups can contain the same volume of water but have different shapes.
Two bulbs can be the same type but be placed in different circuit arrangements.
Two plants can be the same species but be different individuals.
The same individual plant can have a different height on two days.
The same material can be cut into different shapes.

Whenever you write “same”, attach the noun or relationship:

same object, same material, same mass, same time, same type, same connection, same condition, same measured value.

Do not let “same” float any more than “more” or “faster” should float.

Worked Example 1 — The Same Water, a Different State

Original practice situation: Ice in a container is allowed to melt under suitable conditions.

A learner sees solid ice first and liquid water later and says, “It became a different substance.”

That loses identity.

The scientific target is water. Its physical state changes from solid to liquid. The material identity being tracked remains water.

What else stays the same? Do not automatically claim every quantity remains unchanged. If the question provides measurements—for example, a closed setup whose total mass is measured before and after—use those measurements. If it does not, do not add unsupported conservation statements simply because the word “melting” appears.

The durable reasoning is:

same substance being followed → different physical state → use the correct change-of-state mechanism → state only the quantities the evidence supports.

Worked Example 2 — The Same Paper, a Different Shape

A sheet of paper is crumpled into a ball.

The same sheet is being followed. Its overall shape changes dramatically.

If the question compares how the flat and crumpled forms fall, the learner should not treat them as two different materials. The useful comparison is that the same material has been given a different shape and orientation to the surrounding air.

That helps isolate the relevant changed feature.

However, saying “everything else is exactly the same” may still be too strong. Crumpling changes several geometric features at once, including projected area and surface arrangement. The question and existing concept owner determine which difference matters.

Worked Example 3 — The Same Object, a Different Position

A toy car is at Point A, then later at Point B.

The car’s position changes. The car does not become a new car merely because it moved.

This sounds obvious, yet wordy questions can make learners lose object continuity. They may describe “the car at B” as though it were a separate object from “the car at A”, especially when diagrams are split across panels.

Mark the object identity first:

CAR X: A → B

Then reason about what caused the motion, what changed, and what evidence is supplied.

Worked Example 4 — The Same Plant, a Different Measured Height

A plant is measured on Day 1 and Day 7.

The height increases from 12 cm to 17 cm.

Identity: same individual plant.

Changed quantity: measured height.

Not automatically established: its mass, number of leaves, water content or health, unless those are measured or described.

This prevents a common overreach: one measured change becomes a story about the whole organism.

Worked Example 5 — The Same Circuit Relationship Can Survive a Different Drawing

Two circuit diagrams look different because one is drawn horizontally and the other vertically. The wires bend in different places on the page.

Do not decide that the circuits differ by appearance alone.

Trace which terminals and components are connected. If the connectivity is the same, the diagrams may represent the same electrical arrangement despite looking different.

Here, the invariant is not page shape. It is the connection relationship.

If one connection genuinely changes, then the circuit state can change. Track the relationship, not the drawing style.

Worked Example 6 — Same Final Value, Different Histories

Setup P and Setup Q both finish at 30°C.

Does that mean they experienced the same change?

Not necessarily.

Setup	Start	Finish	Change
P	60°C	30°C	−30°C
Q	40°C	30°C	−10°C

The final measured value is invariant across the two setups. The amount of change is not.

“Same” must always be attached to the correct scientific quantity and time.

Worked Example 7 — Same Quantity, Different Mechanism

Two setups produce the same final measured value.

That does not prove they reached it by the same mechanism. Different starting states, pathways or simultaneous processes can lead to the same observed outcome.

The invariant is the measured final value. Mechanism identity remains unestablished unless the evidence distinguishes it.

This connects directly to the existing guide on two setups giving the same result.

Worked Example 8 — A Multi-Part Question Changes the Setup but Not the Story

Part (a) describes a container before heating. Part (b) changes one condition. Part (c) asks about a later observation. Part (d) compares the same setup with a new one.

A learner who mentally resets at each sub-question loses information that still applies.

Instead keep a small state record:

Stage	Object/system	What changed?	What still applies?
(a)	Container P	Initial state only	All stated setup details
(b)	Same P	One named condition	Other stated conditions unless changed
(c)	Same P later	Measured outcome	History from (a) and (b)
(d)	P versus new Q	Comparison target expands	P’s accumulated state remains relevant

The learner is carrying one scientific story across several prompts.

Controlled Conditions Are Not the Same as Invariants

In a fair investigation, a learner deliberately keeps relevant conditions comparable so the effect of the changed variable can be interpreted.

That is a design decision.

An invariant is a broader reasoning idea: something that remains the same across the states or representations you are comparing.

Sometimes they overlap. For example, the same mass of starting water may be a controlled condition and also an invariant across compared setups at the start.

But do not assume every invariant was deliberately controlled, and do not assume every controlled condition remains unchanged after the process begins.

Identity Can Stay the Same While Several Properties Change

The “same object” does not mean “same everything”.

A metal object heated for several minutes can remain the same object while its temperature changes and while other properties may also change depending on the conditions.

A growing plant remains the same individual while height, number of leaves and mass may change.

A moving object remains the same object while position, speed or direction may change.

Identity continuity is not property constancy.

A Property Can Stay the Same While the Object Changes

The reverse also occurs.

Two different objects can have the same mass. Two different organisms can have the same height. Two different circuits can produce the same observed bulb state. Two different materials can show the same measured temperature at one moment.

Therefore, “same measurement” does not prove “same object”, “same system” or “same mechanism”.

Representation Can Change While the Science Stays the Same

A relationship may be shown as:

a paragraph;
a labelled diagram;
a table;
a line graph;
a sequence of pictures;
a simple model.

The representation changes. The underlying scientific relationship may remain the same.

This is why transfer practice should change representation. If a learner knows a concept only when it appears in one familiar diagram, the learner may have memorised the representation rather than the relationship.

The “Redraw It” Test

Take a diagram and redraw it with:

different orientation;
different object labels;
different spacing;
different but equivalent symbols where allowed;
a changed surface context.

Then ask:

Which scientific relationships survived the redraw?

Those surviving relationships are often closer to what the learner truly needs to understand.

Do Not Claim Conservation Without a Boundary

One dangerous version of “what stays the same?” is to assume that a quantity must be conserved without checking the system.

An open container can lose water to the surrounding air. A living organism exchanges matter with its environment. A system can gain or lose energy across its boundary.

Before saying a quantity remains constant, ask:

What exactly is inside the system?
Can matter or energy cross the boundary?
Was the quantity actually measured?
Does the relevant Primary Science model justify the claim?

If those answers are unclear, place the quantity in “not established”, not automatically in “same”.

Do Not Confuse “Not Mentioned” With “Unchanged”

A question may mention only one changed variable because that is the focus. That does not logically prove every unmentioned property stayed constant unless the setup or fair-test conditions establish it.

For example, a plant-height graph that does not mention leaf number does not tell you that leaf number stayed the same.

Silence is not evidence of constancy.

Do Not Confuse “Same Label” With “Same State”

Setup P can appear in several diagrams while its temperature, water level, connection state or position changes.

The label preserves identity. It does not freeze the state.

Do Not Confuse “Different Label” With “Different Scientific Role”

Two diagrams may call objects X and Y in one question and A and B in another. The labels change, but the relationship may be identical.

This is why strong learners recognise the same concept when the surface example changes.

The Before–Change–After State Table

When a question feels confusing, use three columns:

Before	Change/process	After
Object identity	What acts or changes?	Same object or new object?
Starting property/value	Mechanism	Final property/value
Starting conditions	Which condition changed?	Which conditions still apply?
Evidence available	New evidence produced	Supported conclusion

This table prevents two common errors: resetting the whole world after a change, and carrying forward a condition that no longer applies.

The Earliest-Weak-Link Diagnostic

Failure signature	Earliest weak link	Repair
“The ice became a different substance.”	State change was confused with material identity.	Name the substance first, then the state.
“It is the same object, so everything stays the same.”	Identity continuity was confused with property constancy.	List each relevant property separately.
“The question did not mention mass changing, so mass stayed the same.”	Silence was treated as evidence.	Move unsupported quantities to “not established”.
“These two diagrams are different circuits because they look different.”	Representation appearance replaced relationship tracking.	Trace actual component connections.
“The final values are the same, so the processes were the same.”	Outcome equality was confused with mechanism identity.	Compare starting states, paths and evidence.
“Part (c) starts a new problem.”	Multi-part state continuity was lost.	Carry forward unchanged information from earlier parts.
“Same species means same organism.”	Category identity was confused with individual identity.	Distinguish type/group from individual object.
“The quantity must be conserved.”	System boundary and exchange were ignored.	Check what can enter or leave and what was measured.

Misconception Repair — Change Does Not Mean Replacement

A change of position, state, temperature, shape or measured value often describes a new state of the same tracked object or system.

Ask “what are we still following?” before inventing a new object.

Misconception Repair — Continuity Does Not Mean Nothing Happened

The same system can undergo a large scientific change. Keeping identity does not minimise the change.

A water sample can freeze. A circuit can become open. A plant can grow substantially. A moving object can reverse direction. The story remains continuous even when the state changes sharply.

Misconception Repair — “Same Amount” Needs Measurement or Justification

Do not use conservation vocabulary as decoration. If a question asks whether the amount of something remains unchanged, look for measurements, a closed-system condition or the relevant canonical concept.

If none is available, say only what the evidence supports.

Misconception Repair — Same Outcome Does Not Mean Same Cause

Two routes can reach one destination. Two scientific processes can produce the same measured endpoint. One invariant value cannot identify the entire history.

Question-Reading Protocol — Find the Invariant Before You Explain the Change

Circle the scientific object, organism, material or setup being followed.
Mark the time or stage: before, during, after.
Underline the condition that changes.
List the measured or observed properties.
Mark what the question explicitly keeps the same.
Mark any scientifically justified invariant.
Put unsupported features in “not established”.
Select the relevant canonical concept.
Explain the mechanism linking change to outcome.
Check that your answer still refers to the correct object and quantity.

How This Helps With Diagrams

When two diagrams show “before” and “after”:

match the same objects across panels;
ignore decorative movement of labels;
note which connections change;
note which parts remain connected;
track arrows according to their scientific meaning;
do not infer changes merely because the drawing is rearranged.

Redraw the two panels using consistent positions if necessary. The goal is to make change visible against a stable reference.

How This Helps With Tables

In a table, each row may refer to:

the same object at different times;
different objects under the same condition;
the same setup under different conditions;
different setups altogether.

Do not compare values until you know which of these structures the table uses.

How This Helps With Graphs

A line graph often tracks one measured quantity as another variable changes.

The plotted point changes. The identity of the measured series may remain the same.

If two lines represent two setups, keep the line identities separate even when they cross. A crossing does not swap the setup labels.

This sounds trivial until a learner follows “the upper line” rather than “Setup P” and accidentally switches identity after the crossing.

How This Helps With Cycles

A cycle contains change and recurrence.

The learner should ask:

what changes at each stage;
what returns;
what material, organism or process is being tracked;
whether the next cycle involves the same individual or a new generation;
which features truly repeat and which may vary.

For example, a life cycle can repeat as a pattern without the exact same organism returning to an earlier life stage.

How This Helps With Systems

A system question may change one component, pathway or input while preserving the rest of the system.

Strong reasoning isolates the changed relation against the stable background.

But if the change alters the system boundary, connection or operating condition, update the model. Do not carry old assumptions forward simply because most labels stayed the same.

How This Helps With Interactions

When one object affects another, keep the identities and direction of effect clear.

If the objects swap position in a second diagram, the causal direction does not automatically swap. Track the scientific relationship, not page location.

How This Helps With Energy Questions

Energy questions often require tracking a system across stages: source, transfer, transformation or effect.

Do not confuse “the same system” with “the same energy form everywhere”. The system identity and the energy description are different layers.

Use the existing Energy learning owner for the canonical mechanism. This guide’s job is simply to keep the tracked system and changing energy description distinct.

The Invariant Check Before a Prediction

A prediction often changes one condition while assuming other relevant conditions remain comparable.

Before predicting, ask:

What is being held constant?
What is deliberately changed?
What scientific relationship is assumed to continue?
Is the prediction still within the tested or justified range?

If too many background conditions change, the prediction may no longer follow from the original evidence.

The Invariant Check Before a Causal Claim

A causal comparison needs a stable reference.

If Setup P and Q differ in five ways, there is no single invariant background against which one changed factor can be isolated. The claim becomes confounded.

This is why invariant tracking connects naturally to fair-test reasoning without replacing it.

The Invariant Check Before a Conclusion

Before writing a conclusion, verify:

same measured quantity?
same object or clearly different objects?
same time point or justified interval?
same relevant conditions?
same system boundary?
same meaning of the labels?

If one answer is no, the conclusion may need a narrower comparison.

Practice Sequence — Same, Changed, Unknown

Take ten before-and-after PSLE Science practice situations.
Write the object or system identity.
Create three columns: SAME / CHANGED / NOT ESTABLISHED.
Place every relevant property or relationship into one column.
Explain why each item belongs there.
Change one extra condition and update the columns.
Redraw the representation without changing the underlying relationship.
Use a near-miss where the same final value comes from a different starting value.
Use another where two different objects share one property.
Return several days later with an unfamiliar mixed-theme question.

Unfamiliar Transfer Challenge

A mystery object is observed at three stages.

Stage	Position	Temperature	Shape
1	A	25°C	Long
2	B	40°C	Long
3	B	40°C	Bent

The question states that it is the same object throughout.

What stays the same?

object identity across all three stages;
shape from Stage 1 to Stage 2;
position from Stage 2 to Stage 3;
temperature from Stage 2 to Stage 3.

What changes?

position from Stage 1 to Stage 2;
temperature from Stage 1 to Stage 2;
shape from Stage 2 to Stage 3.

What is not established?

mass;
material identity unless supplied elsewhere;
the cause of each change;
whether one change caused another.

The table contains observations. The mechanism needs additional scientific context.

Delayed Independent Return

Four to seven days later, solve a fresh question without the three-column scaffold.

Ask yourself:

What exact object, system or relationship am I following?
Which properties change?
Which properties are explicitly kept the same?
Which apparent invariants are only assumptions?
Has the representation changed while the relationship stayed the same?
Has the system boundary changed?
Am I still referring to the same quantity at the same time?
Could the same outcome arise from a different mechanism?
Can I explain the change without losing the continuity?

The learning receipt is that the learner can preserve the scientific thread even when the surface story changes.

The Answer-Checking Receipt

Did I identify the object or system being followed?
Did I distinguish identity from state or property?
Did I name exactly what stayed the same?
Did I name exactly what changed?
Did I leave unsupported quantities as “not established”?
Did I avoid treating a redraw as a new scientific system?
Did I avoid treating a new state as a new substance without evidence?
Did I keep the system boundary in view before making conservation claims?
Did I preserve information from earlier parts of a multi-part question?
Did I check the final explanation against the evidence?

Evidence and Model Limits

“What stays the same?” can sound like a simple memory question, but scientifically it depends on definitions and boundaries. Some invariants are direct observations. Some are controlled conditions. Some are consequences of a scientific model. Others are only assumptions.

This guide therefore does not teach a universal list of conserved quantities. That would exceed the learner job and could create false certainty in open systems or poorly specified questions.

The safer rule is:

preserve identity when the question preserves identity; preserve a quantity only when the evidence or scientific model justifies it.

That distinction is enough to prevent many hidden reasoning errors while keeping Primary Science appropriately bounded.

Useful Internal Routes

Parent and Tutor Teaching Guide

Children are naturally good at noticing visible change. Teach continuity deliberately.

When a setup changes, ask three questions before asking for the explanation:

What are we still following?
What definitely changed?
What are you assuming stayed the same?

The third question is especially useful. It separates evidence from habit.

Use simple physical demonstrations. Move one object without changing it. Warm one object. Crumple one sheet. Redraw one circuit. Change one graph scale. Each time ask the learner to name identity, changed property and unsupported assumptions.

Then use near-misses:

two different objects with the same measured value;
the same object with several changed properties;
two diagrams with the same relationship;
two setups with the same final value but different starting values;
an open system where a quantity can leave.

Do not reward “everything else stays the same” unless the learner can say what “everything else” means and why it is justified.

After the learner succeeds with the SAME / CHANGED / NOT ESTABLISHED table, remove it. Return days later with a mixed question. The skill is mature when continuity survives without the scaffold.

Authoritative and Research References

Singapore Examinations and Assessment Board — PSLE Formats Examined in 2026.
Singapore Examinations and Assessment Board — PSLE Science syllabus, for examination from 2026.
Singapore Ministry of Education — Science Teaching and Learning Syllabus, Primary, 2023.
Research on relational reasoning, representational competence and transfer is used here as broader learning evidence rather than PSLE-specific marking policy.

The Quiet Ending

Science is not only the study of what changes.

It is also the discipline of knowing what you are still talking about after the change.

Hold the object. Track the state. Respect the boundary. Leave unknowns unknown.

Then the story remains scientifically continuous, even when the world inside it changes.

Small Group Tutorials