How to Answer a PSLE Science Question When Several Conditions Change at Once

Wait, What? More Than One Change Can Make a Question Easier to Misread

Plant A receives more light and more water than Plant B. Plant A grows taller.

Can you conclude that the extra light caused the greater growth?

Not from that comparison alone. Water changed too.

Now suppose the question is not asking you to identify one cause. Suppose it asks you to reason about what may happen to a plant when both light and water conditions change. Then noticing both changes is exactly what you need to do.

When several conditions change, your first job is to decide whether the question wants causal isolation or whole-system reasoning.

Quick Answer

When several PSLE Science conditions change at once, make a difference list before explaining anything. Identify the measured outcome. Mark which changed conditions could affect it. Then decide whether the question is an investigation that requires a fair comparison or an application question where several causes may operate together. If the changes are confounded, do not claim that one factor alone caused the result unless the evidence isolates it.

This page owns the learner job of reasoning under multiple simultaneous changes. It does not replace the existing variables and fair-tests guide, method-evaluation guide or scientific concept pages.

The Current PSLE Science Frame

For examination from 2026, SEAB states that PSLE Science assesses attainment in the 2023 Primary Science syllabus. The assessment objectives include applying scientific facts, concepts and principles, interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning.

MOE’s Primary Science syllabus also treats the five themes—Diversity, Cycles, Systems, Energy and Interactions—as connected. Real PSLE reasoning can therefore involve several conditions and more than one scientific relationship.

Owned PSLE Science Learning Job

Notice every relevant condition that differs between set-ups.
Identify the outcome or measurement being compared.
Distinguish a fair-test question from a whole-system application question.
Recognise when two or more changed variables are confounded.
Track several possible causal pathways without mixing them into one vague explanation.
Decide whether changed conditions reinforce, oppose or cannot be separated.
State a conclusion no stronger than the evidence allows.
Design the next comparison that would isolate one factor if needed.

The Multiple-Condition Reasoning Chain

READ ALL GIVEN INFORMATION → LIST WHAT CHANGED → IDENTIFY WHAT WAS MEASURED → MARK WHICH CHANGES COULD AFFECT THE OUTCOME → DECIDE WHETHER CAUSAL ISOLATION IS POSSIBLE → TRACE EACH RELEVANT MECHANISM → COMBINE ONLY WHAT THE EVIDENCE JUSTIFIES → STATE THE BOUNDED OUTCOME → CHECK AGAINST THE DATA.

Step 1 — Make a Difference List Before You Explain

Imagine two evaporation set-ups.

Set-up A: water in a wide shallow tray, placed in moving air.
Set-up B: the same starting amount of water in a narrow container, placed in still air.

Two relevant conditions differ: exposed surface and air movement. If A loses water faster, you cannot use this pair alone to measure how much of the difference was caused by surface area and how much by air movement.

A learner who notices only the first difference may write a scientifically true sentence about exposed surface area and still make an invalid causal claim about this comparison.

Step 2 — Identify the Question Job

Job A: Isolate the Effect of One Variable

If the question asks whether one factor affects an outcome, the comparison must allow that factor’s effect to be distinguished from other changing conditions.

If several relevant variables differ, causal attribution to one variable is weakened or impossible from that comparison alone.

Job B: Reason About a System With Several Changes

Some questions deliberately change several conditions because the system itself is changing. You may need to reason about several mechanisms together.

In that case, the presence of multiple changes is not a flaw. The learner must simply avoid pretending that one factor was isolated experimentally.

Worked Example 1 — Light and Water Both Change

Two similar plants are observed over the same period.

Condition	Plant A	Plant B
Light received	Higher	Lower
Water supplied	Adequate	Much less
Measured height gain	Greater	Smaller

Observation: Plant A gained more height under the tested conditions.

What the comparison does not isolate: the separate effect of light alone or water alone, because both differed.

A stronger conclusion is bounded: the plants experienced different combinations of light and water conditions, and Plant A showed greater height gain. This comparison alone cannot determine which one of the two changed conditions caused the difference.

If the question instead asks for scientifically plausible reasons for the difference, both changed conditions may need to be considered, with each explanation tied to the relevant plant process and the evidence.

Worked Example 2 — Surface Area and Moving Air

A wide tray in moving air loses water faster than a narrow container in still air.

Larger exposed surface can favour faster evaporation.
Moving air can also favour faster evaporation under appropriate conditions.
Both differences point in the same broad direction.

You may say that the combined conditions in A are consistent with faster evaporation. You may not use this one comparison to say how much each factor contributed, or that only one factor caused the difference.

When two causes change together, a result can fit both causes without separating them.

Worked Example 3 — A Circuit With Two Changes

Set-up X has two cells and one bulb. Set-up Y has one cell and two bulbs arranged differently. The brightness observed differs.

It is unsafe to say, “The bulb is brighter because there are more cells,” without checking the whole arrangement. The number of bulbs and the connections changed too.

Systems reasoning comes first: identify the complete circuit, component arrangement and observable outcome. If the job is to isolate the effect of cell number, create a comparison where the rest of the relevant arrangement is kept the same.

Worked Example 4 — An Ecosystem With Several Environmental Changes

A habitat becomes drier, warmer and less shaded. One population falls.

It may be scientifically reasonable that several environmental changes affect survival, food, shelter or reproduction. But unless the evidence isolates them, do not state one condition as the proven sole cause.

A strong response can separate:

what changed;
what outcome was observed;
which relationships are scientifically plausible;
what extra evidence would be needed to distinguish the causes.

Three Multiple-Change Patterns

Pattern 1 — Changes Reinforce Each Other

Two conditions both push the same process in the same direction. The combined outcome may be stronger, but the data may still not isolate each contribution.

Pattern 2 — Changes Oppose Each Other

One condition tends to increase an outcome while another tends to decrease it. The final result can be small even though both effects are real.

Pattern 3 — One Change Alters the Effect of Another

Sometimes a condition matters only when another requirement is met. Primary learners do not need advanced interaction statistics. They do need to recognise that “more of factor A” does not guarantee the same effect under every possible condition.

Confounding: The Child-Friendly Meaning

“Confounding” is not a required PSLE term. The useful idea is simpler:

If two important things changed together, the comparison may not tell you which one produced the difference.

You can use that idea without using the word.

How to Repair the Investigation

If you need to isolate one variable, design a second comparison.

For the plant example, compare two similar plants where light differs but the relevant water condition is kept the same. Then perform another comparison for water while keeping light comparable.

The repair principle is:

ONE TARGET CHANGE → COMPARABLE OTHER RELEVANT CONDITIONS → MEASURED OUTCOME.

Do Not Accidentally “Control Away” the Real Question

Fair-test logic is powerful, but not every PSLE Science question is asking for a controlled experiment.

If a real system changes in several ways, your job may be to reason about the system as it is. Do not answer, “This is not a fair test” and stop if the question actually asks for likely consequences of the combined changes.

When Several Conditions Change, Build Separate Mini-Chains

For each relevant change, build:

CHANGED CONDITION → RELEVANT CONCEPT → MECHANISM → POSSIBLE OUTCOME.

Only after the mini-chains are clear should you ask how they combine.

Do the Mini-Chains Reinforce or Compete?

Suppose one change would tend to increase a process while another would tend to reduce it. Without additional evidence, you may not know the final direction.

A scientifically strong answer can say that the available information is insufficient to determine the net outcome. “Cannot determine from the given information” is not a failure when the evidence really is insufficient.

Observation vs Attribution

Keep these separate:

Observation: Set-up A produced a larger change.
Attribution: Factor X caused the larger change.

The first can be read directly from data. The second requires a comparison that makes the causal claim defensible.

Observable Failure Signatures and Earliest Repairs

“I underline only one difference.” Repair: make a full difference list before choosing a concept.

“I see a trend and immediately name one cause.” Repair: separate observation from attribution.

“I say every multi-change question is unfair.” Repair: decide whether the job is experimental isolation or system application.

“I write one giant explanation mixing all causes.” Repair: build separate mini-chains first.

“I know two causes matter but cannot tell the final result.” Repair: check whether the effects reinforce, oppose or depend on each other. If the evidence is insufficient, say so.

Misconception Repair — More Variables Do Not Automatically Mean More Marks

Learners sometimes mention every changed condition to sound thorough. This can produce a long but unfocused answer.

Use only changes that are scientifically relevant to the measured outcome and supported by the Primary-level concept. Relevance matters more than quantity.

Model Limit — Real Systems Can Be More Complicated Than the Question

In real life, many variables can interact. PSLE Science questions simplify reality so learners can reason about selected relationships. Do not invent hidden variables unless the question asks you to evaluate alternatives or limitations.

At the same time, do not pretend a comparison isolates one cause when the given design clearly changes several relevant conditions.

Original Practice Protocol

Take any two set-ups.
Write every difference in a table.
Circle the measured outcome.
Cross out differences irrelevant to the outcome.
For each remaining difference, write one causal mini-chain.
Decide whether the chains reinforce, oppose or cannot be combined confidently.
Write the strongest justified conclusion.
Design one extra comparison that would isolate the most important uncertain factor.

Unfamiliar Transfer Challenge

Two sealed boxes contain identical small devices. Box A is warmer and contains more moisture. Box B is cooler and drier. The device in A operates for a shorter time.

You are not told which environmental factor affects the device. A good learner can still say what the evidence shows, why one factor cannot be isolated from this comparison, and what follow-up comparisons would distinguish temperature from moisture.

Delayed Independent Return Test

Two days later, solve a new multi-condition comparison. Before writing the explanation, produce three things from memory: the difference list, the measured outcome, and the causal-isolation judgement.

If you can do that before reaching for a keyword, the reasoning routine is becoming durable.

Answer-Checking Receipt

Did I identify every relevant condition that changed?
Did I identify what was measured?
Am I describing the result or claiming a cause?
Can this comparison isolate the cause I named?
Did I trace each changed condition through a scientifically valid mechanism?
Do the effects reinforce, oppose or remain uncertain?
Did I state only what the evidence supports?
If isolation is needed, can I propose a cleaner comparison?

Useful Internal Routes

Parent and Tutor Teaching Guide

Before asking “What is the answer?”, ask the child to draw a two-column difference table. This slows premature causal guessing and makes hidden changes visible.

If the child names one cause immediately, ask: “What else changed?” If the child says “not a fair test” to every complex question, ask: “Is the question asking us to isolate one cause, or to reason about the whole situation?”

The teaching goal is not to make Primary learners use advanced causal-inference vocabulary. It is to make their conclusions match the design and the evidence.

Authoritative and Research References

Singapore Examinations and Assessment Board — PSLE Science syllabus, for examination from 2026.
Singapore Ministry of Education — Science Teaching & Learning Syllabus, Primary, 2023.
Education Endowment Foundation — A systematic review of approaches to primary science teaching, 2023.
Deanna Kuhn, “Reasoning about multiple variables: Control of variables is not the only challenge,” Science Education 91 (2007), used as broader evidence about elementary learners’ multivariable reasoning.
Recent elementary data-reasoning research showing that children can reason critically about evidence involving multiple variables, used as broader educational evidence rather than a Singapore assessment rule.

The Quiet Ending

One changed condition gives you one clean line to follow. Several changed conditions demand a better map.

List the differences. Decide what the question is asking. Trace the causes separately. Combine them only when the evidence lets you. That discipline protects you from one of the most common reasoning errors in Science: seeing one true cause and assuming it must be the only cause.

Small Group Tutorials