How to Test a PSLE Science Explanation by Asking What Evidence Would Count Against It

Wait, what? A good Science explanation should be able to lose.

That sounds strange at first. If you have worked hard to build an explanation, why would you immediately ask how it could fail? Because an explanation that can survive only when you look for supporting clues is fragile. Scientific reasoning becomes stronger when you can say not only, “This evidence fits my explanation,” but also, “If I observed this instead, I would have to weaken, change or reject it.”

For a Primary 5 or Primary 6 learner, this does not mean becoming a philosopher of science or trying to disprove every answer. It means learning one practical habit: before you trust an explanation, ask what result would make it less believable.

Quick Answer

To test a PSLE Science explanation, first identify the exact mechanism you are claiming. Then ask what that mechanism predicts under the question’s conditions. Finally ask: what observation, measurement or comparison would not fit that prediction?

If such counter-evidence is actually present in the question, you cannot simply ignore it. You must decide whether the explanation is wrong, incomplete, too broad, applied to the wrong condition, or still possible but not well supported.

The Exact PSLE Science Learning Job This Guide Owns

This guide teaches one distinct learner job: how to stress-test a proposed PSLE Science explanation by identifying evidence that would count against its mechanism.

It is not a general guide to MCQ elimination, not a generic lesson on counterexamples, and not a replacement for learning scientific concepts. It comes after you already have a plausible explanation. The question is now: how do you test whether your own explanation deserves to survive?

The Core Reasoning Chain

Keep the usual PSLE Science reasoning chain intact:

Read the given information.
Identify the scientific object or relationship.
Separate observation from inference.
Select the relevant concept.
Explain the causal mechanism.
Connect the mechanism to the exact condition in the question.
State the expected outcome.
Ask what evidence would not fit that expected outcome.
Check the actual evidence again.

That eighth step is the focus of this article. It turns an explanation from something that merely sounds sensible into something that has been tested against the evidence.

Why “It Sounds Right” Is Not Enough

Many Science mistakes survive because the answer contains true scientific words. A learner remembers a fact, sees a familiar diagram, and writes a plausible mechanism. The problem is that several mechanisms can sometimes fit the same first observation.

Suppose a bulb does not light. An open circuit is one possible explanation. But so is a flat cell. A faulty bulb could also explain it. A loose connection could explain it. The observation “the bulb does not light” is therefore not enough by itself to establish one cause.

A stronger learner asks a second question: What additional observation would make my explanation harder to defend?

If the circuit path is visibly complete and a known-working bulb still does not light, “the circuit is open” is no longer a good explanation. You now need to inspect other possibilities. The point is not to produce a long list of causes. The point is to stop treating the first plausible cause as automatically correct.

Support, Fit and Test Are Different

Three ideas are often mixed together:

Support: the evidence is consistent with the explanation.
Fit: the explanation can account for what was observed.
Test: you can identify what evidence would make the explanation less credible.

An explanation may fit without being uniquely supported. For example, “the cloth dried faster because the air was moving” may fit a setup with a fan. But if that setup was also warmer, the evidence may not isolate air movement as the cause. The explanation sounds reasonable, yet the comparison is not strong enough to establish it.

This is why fair-test thinking matters even outside a question that explicitly asks about variables. You are always asking whether the evidence really discriminates between competing explanations.

The Counter-Evidence Test

Use this compact protocol:

Name the explanation. What exactly are you claiming caused the result?
Name the mechanism. What process links the cause to the outcome?
State the prediction. If the mechanism is operating under the stated conditions, what should you expect to observe?
Reverse the expectation. What result would be surprising if your explanation were correct?
Check the evidence. Is that counter-evidence present, absent, uncertain or not measured?
Update the explanation. Keep it, narrow it, revise it, or reject it.

The useful move is step four. Many learners never perform it. They ask only, “What supports me?” rather than, “What could prove me too confident?”

Worked Example 1: Evaporation and Exposed Surface

Imagine an original practice question with two containers holding equal amounts of water. One container exposes a larger water surface to the air. Both are kept in the same room for the same period. The larger-surface container loses more water.

A learner explains: “The container with the larger exposed surface lost more water because a larger surface allowed more water to evaporate over the same period.”

That is a plausible causal explanation. Now test it.

Prediction: under comparable conditions, the larger exposed surface should generally show greater water loss over the same period.
Counter-evidence: if repeated fair comparisons consistently showed the smaller exposed surface losing more water, the proposed surface-area explanation would need to be questioned.
Important boundary: one unexpected reading does not automatically destroy the explanation. You would first check measurement, setup differences and whether the comparison remained fair.

This is scientific discipline. Counter-evidence is not a magic “wrong” button. It is a reason to inspect the explanation and the quality of the evidence more carefully.

Worked Example 2: A Bulb That Does Not Light

Observation: a bulb in a simple circuit does not light.

Proposed explanation: “The bulb does not light because the circuit is open.”

Now make the explanation earn its place.

What would you expect if the circuit were open? There should be no complete conducting path.
What would count against that explanation? Evidence showing a complete path would weaken the claim that the circuit is open.
What then? You would keep other possibilities alive, such as a cell or bulb problem, rather than forcing the first explanation to survive.

The lesson is larger than circuits: one outcome can sometimes have several possible causes. Testing an explanation means looking for evidence that distinguishes among them.

Worked Example 3: A Plant Comparison

Suppose two similar plants are observed after a week. Plant A is taller than Plant B. A learner immediately says, “Plant A grew taller because it received more light.”

That explanation is possible only if the question actually gives relevant information about light and the comparison does not contain uncontrolled differences that could also matter.

Ask:

Was light exposure actually different?
Were other important conditions comparable?
Was the observed height change measured from the same starting point?
Would a result showing equal growth under the different light condition count against the proposed explanation?

Do not add a cause just because it is scientifically relevant in general. The cause must be connected to the question’s evidence.

A Simple Table for Stress-Testing an Explanation

Question	Your note
What is my explanation?	State one causal claim.
What mechanism am I using?	Name the scientific relationship, not just a keyword.
What should I observe if it is correct?	State the expected outcome under the given condition.
What would not fit?	Name a result that would weaken the mechanism.
Is that evidence present?	Yes / No / Not measured / Uncertain.
What should I do now?	Keep / Narrow / Revise / Reject.

The Earliest Weak Link to Diagnose

If you struggle with this skill, do not immediately practise more difficult questions. Find the first broken step.

Weak link 1: You cannot state your explanation in one sentence. Your idea may still be too vague.
Weak link 2: You can state the explanation but not the mechanism. You may know a topic label without the causal relationship.
Weak link 3: You know the mechanism but cannot say what it predicts. Your concept knowledge may be disconnected from observable outcomes.
Weak link 4: You can make a prediction but cannot imagine counter-evidence. You may be treating your first idea as something to defend rather than something to test.
Weak link 5: You see counter-evidence but ignore it. You need evidence discipline, not more memorised facts.

Common Misconceptions

“If I can find one piece of counter-evidence, the explanation is definitely wrong.”

Not always. A surprising result might come from measurement error, an uncontrolled condition, an incomplete model or genuine evidence against the explanation. You must inspect the quality and context of the evidence before deciding.

“I should list every possible reason.”

No. PSLE Science reasoning still needs relevance and control. Keep alternatives alive only when the evidence has not yet separated them.

“A true fact can never count against my explanation.”

A fact may be true but irrelevant. The key question is whether it bears on the mechanism you are claiming under the conditions in the question.

“Science answers should sound certain.”

They should sound as strong as the evidence allows. Sometimes the correct scientific move is to say that the evidence supports one explanation better than another, or that more evidence is needed.

How This Helps With Open-Ended Questions

Before finalising an open-ended explanation, perform a ten-second stress test:

What relationship am I claiming?
What evidence in the question supports it?
What condition makes the mechanism relevant?
What evidence would make this explanation fail?
Is any of that counter-evidence already present?

This can catch answers that are scientifically plausible but inconsistent with the actual setup.

How This Helps With MCQ

For an MCQ option that looks attractive, do not ask only, “Can this be true?” Ask, “What would have to be true for this option to explain the given result?” Then check whether the stem supplies or contradicts those conditions.

This is different from eliminating an option because one word looks unfamiliar. You are testing the scientific relationship.

Practice Sequence: From Easy to Independent

Stage 1: Teacher gives the explanation

You identify one observation that would support it and one that would count against it.

Stage 2: Teacher gives the evidence

You generate the explanation, prediction and counter-evidence yourself.

Stage 3: Changed context

Use a different science topic and repeat the same reasoning pattern. The surface story changes; the evidence-testing job stays the same.

Stage 4: Delayed return

Three to seven days later, solve a fresh question without notes. If you can still produce both a mechanism and a possible disconfirming observation, the reasoning is beginning to transfer.

Your Answer-Checking Receipt

Before you move on, your explanation should leave these receipts:

I can point to the evidence I used.
I can name the scientific relationship.
I can explain the mechanism in ordinary language.
I can connect the mechanism to the exact condition.
I can state what outcome the mechanism predicts.
I can name evidence that would weaken the explanation.
I checked that this counter-evidence is not already present.

If you cannot do the last two, your explanation may still be right. But it has not yet been properly tested.

Useful Internal Routes

How Do We Know This Is a Useful Way to Learn Science?

The official 2026 PSLE Science syllabus states that the paper assesses not only knowledge with understanding, but also application of scientific knowledge and inquiry, including interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. The 2023 Primary Science syllabus likewise places inquiry and the use of evidence inside the development of scientific understanding.

Research in science education also treats explanation, evidence evaluation, argumentation and model revision as important parts of scientific reasoning. Studies of refutation and conceptual change suggest that learners can benefit from explicitly confronting evidence that conflicts with an initial idea, but also warn that contradiction alone is not enough: learners need support to compare explanations, inspect evidence and rebuild a more useful model. This guide therefore does not teach “find one contradiction and throw the answer away.” It teaches a slower and more scientific sequence: predict, test, inspect, update.

Evidence Boundaries

This is a learning and reasoning guide, not an official marking rubric. There is no claim here that PSLE candidates must write a particular phrase such as “evidence against” in their answer. The educational purpose is to strengthen the thinking that happens before and during answer construction.

Also remember that a single surprising observation may not be decisive. Scientific evidence has quality, scope and conditions. Measurements can vary. Methods can be weak. A model can be useful without being complete. Good learners do not become less confident in everything; they become better at matching confidence to evidence.

Teaching Guide for Parents and Tutors

The easiest way to teach this skill is not to lecture about falsification. Use a learner’s own explanation and ask one quiet question: “What result would make you change your mind?”

If the learner cannot answer, do not supply the counter-evidence immediately. First ask them to state the mechanism and its prediction. Once the prediction is clear, the counter-evidence often becomes visible.

A useful teaching progression is:

Ask the learner to state the explanation.
Ask what should happen if it is correct.
Ask what should not happen if it is correct.
Show or invent one changed result.
Ask whether the explanation survives unchanged, needs narrowing, or should be replaced.
Return days later with a different topic.

Watch for the earliest weak link. A child who cannot name counter-evidence may actually have a more basic problem: the mechanism was never clear enough to generate a prediction. Repair that first.

Authoritative References

Quiet Return

The goal is not to become suspicious of every answer. It is to stop confusing familiarity with strength.

A mature PSLE Science explanation should be able to tell you what it explains, what evidence supports it, what conditions it depends on, and what evidence would make you reconsider it.

That is a powerful change in how a learner studies Science. You are no longer collecting answers that sound correct. You are building explanations that can survive being tested.

Small Group Tutorials