PSLE Science Reality Lab Vol No.006 | “It Happened on Camera” — Does a Demonstration Test Why It Happened?

Series ID: PSLE-SCI-REALITY-0006

Wait, What? A Video Can Be Completely Real and Still Not Prove the Explanation

A cup suddenly fogs. A paper strip changes colour. A bottle collapses after being cooled. A balloon expands when a container is warmed. A magnet appears to pull something from a distance. The camera records the event clearly. Nobody has edited the clip. The effect genuinely happened.

And yet the explanation written above the video can still be wrong.

This is one of the most useful scientific habits a Primary 5 or Primary 6 learner can build: separate the phenomenon from the proposed cause. A demonstration can show that something happened under certain conditions. It does not automatically show which condition caused it, whether the suggested mechanism is correct, whether another explanation fits the same observation, or whether the result would survive a fair comparison.

That distinction is already part of good PSLE Science reasoning. The 2026 PSLE Science assessment objectives require students to apply scientific inquiry by interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. The 2023 Primary Science syllabus also asks learners to exercise healthy scepticism, consider more than one explanation and evaluate scientific communication in different forms and media. The same habits that help with an unfamiliar structured question can therefore help with a dramatic video on a phone.

Quick Answer

When a video or live demonstration shows a surprising effect, ask two separate questions:

What did I actually observe?
What evidence would be needed to decide why it happened?

The first question protects the observation. The second protects the explanation.

The Owned Learner Job

This Reality Lab owns one narrow transfer job: how to evaluate a real-world or viral science demonstration when the visible event is stronger evidence than the claimed explanation.

It does not replace the existing eduKateSengkang guides on observation versus inference, fair testing, variables, alternative explanations, investigation design or causal reasoning. Those pages remain the canonical skill owners. Here, we combine those skills around a communication object that learners increasingly meet outside worksheets: a demonstration presented as proof.

Reality Lab Case: The Collapsing Bottle

Imagine an original video made for this lesson. A student shows a thin plastic bottle. The bottle is open and looks normal. Warm water is poured into it, swirled around and emptied. The cap is tightened. The bottle is then placed in a basin of cool water. After a short time, the sides bend inward.

The caption says:

“Cold water pulls the bottle inward.”

The bottle really did collapse. But did the video test the explanation in the caption?

What Was Actually Observed?

The bottle was initially not collapsed.
Warm water contacted the inside of the bottle and was then removed.
The bottle was capped.
The capped bottle was placed in cooler surroundings.
Its sides later bent inward.

Those are observations about objects, actions and visible changes. They are not yet a complete explanation.

What Was Claimed?

The caption claims that the cool water itself pulled the bottle inward. That statement contains a proposed causal mechanism. It goes beyond what the camera directly records.

A learner should therefore resist two opposite mistakes. Do not say, “The video proves the caption.” But also do not say, “The caption must be false because a video cannot prove anything.” A demonstration can provide evidence. The important question is which claim the evidence supports.

The Demonstration Test: Phenomenon or Cause?

A useful first move is to classify the job of the set-up.

A phenomenon demonstration shows that an effect can happen under a set of conditions.
A causal investigation deliberately compares conditions so that competing explanations can be separated.

The collapsing-bottle video is strong evidence that this sequence of actions can be followed by a collapse. By itself, however, it is weaker evidence for the exact cause because several things changed between the starting state and the final state: the bottle was warmed internally, water was removed, the bottle was sealed, the surroundings became cooler, time passed and the gas or vapour inside could change as temperature changed.

When several conditions change together, the final effect does not identify one cause automatically.

A Four-Layer Reading of Any Demonstration

Layer 1: Observation

Describe only what could be recorded without knowing the explanation. What moved? What changed colour? What reading changed? What appeared or disappeared? What happened first and what happened later?

Layer 2: Comparison

Ask what the demonstration is being compared with. Was there a second set-up? Was only one variable changed? Were the starting conditions comparable? If there is no comparison, the demonstration may still be informative, but causal confidence should be lower.

Layer 3: Explanation

Name more than one plausible mechanism before choosing. The purpose is not to invent silly alternatives. It is to avoid locking onto the first story that fits the picture.

Layer 4: Discriminating Test

Ask what new comparison would produce different predictions under the competing explanations. A good follow-up test is not merely “do it again”. It changes the evidence so the alternatives stop making the same prediction.

Why Repeating the Same Demonstration Is Useful but Not Always Decisive

Suppose the bottle collapses five times out of five. That strengthens the claim that the phenomenon is repeatable under similar conditions. It does not necessarily distinguish between two explanations that both predict collapse under those same conditions.

This is a subtle but powerful scientific distinction:

Repeating an observation asks, “Does the effect return?”
Changing a discriminating condition asks, “Which explanation survives?”

Both jobs matter. They are not the same job.

Worked Reasoning Case 2: The Floating Pepper Video

An original demonstration shows pepper flakes floating on water in a shallow dish. A cotton bud touches the centre of the water and the pepper rapidly moves outward. The caption says, “The cotton bud pushes the pepper away.”

A good learner does not need advanced chemistry to reason scientifically here. Start with the evidence job.

Observation: after the cotton bud touches the centre, pepper pieces move away from that region.
Not directly observed: an invisible pushing force coming from the cotton bud.
Question: what changed when the cotton bud touched the water?
Alternative possibilities: the bud may physically touch the pepper; a substance on the bud may alter the water surface; the motion could depend on where the bud touches; the effect could be caused by a hidden change not stated in the caption.
Discriminating tests: compare a clean dry bud with a bud carrying the active substance; touch the water at different locations; check whether contact with pepper is required; repeat with controlled amounts.

The point is not that every learner must know the final microscopic mechanism immediately. The point is that a surprising effect can be investigated without pretending the first verbal explanation is already established.

The Earliest Weak Link: When the Caption Becomes Part of the Observation

A common failure happens before any scientific concept is recalled. The learner reads the caption first and then watches the demonstration through that caption.

If the caption says “air pressure crushes the bottle”, the learner may later report, “I saw air pressure crush it.” But air pressure was not seen. The bottle changing shape was seen. Air pressure is part of the explanation.

This is why a useful Reality Lab habit is:

Watch or read once without accepting the explanation.
Write the observable change.
Write the proposed explanation separately.
List at least one alternative that could also fit the observation.
Ask what comparison would separate them.

Evidence Can Be Strong for One Claim and Weak for Another

The same video can support different claims to different degrees. Consider a clip showing a seedling bending toward a bright window over several days.

Strongly supported: the recorded seedling changed orientation over the recorded period.
Possibly supported: light direction was associated with the direction of bending in this set-up.
Needs a stronger investigation: light direction alone caused the bending.
Not justified from one seedling: every plant species always behaves this way under all light conditions.

This is what evidence boundaries look like in practice. Scientific thinking is not simply “believe” or “disbelieve”. It asks how far the evidence can carry the conclusion.

The Missing Control Is Not Always the Only Problem

Students often learn to ask, “Was there a control?” That is useful, but a checklist can become shallow. A comparison only helps if it actually discriminates the claim.

Imagine two bottles. Bottle A follows the warm-water-and-cooling procedure. Bottle B is left untouched on the table. Bottle A collapses; Bottle B does not.

This tells us more than the one-bottle demonstration, but several conditions still differ between A and B. If the claim is specifically about cooling after sealing, a stronger comparison would keep the preparation similar while changing the cooling condition. The right control depends on the causal question.

What Would Strengthen the Explanation?

A comparison where only the proposed causal condition differs.
Repeated trials that show the result is not a one-off event.
Measurements rather than only visual impressions where measurement is possible.
A method detailed enough for another person to reproduce.
Results under changed conditions that match predictions from the proposed mechanism.
Evidence that plausible alternatives fail where the chosen explanation succeeds.

What Would Weaken the Explanation?

The effect disappears when the claimed cause is present but another condition changes.
The same effect occurs when the claimed cause is absent.
The demonstration depends on an unstated preparation step.
Different starting conditions produce the result even without the proposed mechanism.
The caption predicts one pattern but repeated measurements show another.
The explanation changes after each failed prediction merely to preserve the original claim.

Model and Measurement Limits

A camera is a measurement device of a kind, but it has limits. It records selected angles, times, colours and frames. It may not show temperature, mass, pressure, concentration, invisible gases, forces or events outside the frame. Slow change may be compressed by time-lapse; fast change may be missed between frames. A close-up can hide the rest of the apparatus. Editing can remove unsuccessful attempts even when the successful clip itself is genuine.

Therefore, “I can see it” is not the same as “I can see every relevant variable”.

A PSLE-Style Transfer Case

A student places identical ice cubes on two surfaces. One is a metal tray and the other is a wooden board. The cube on the metal tray melts faster. A video caption says, “Metal produces heat, so it melts ice faster.”

Try to answer before reading on:

What was observed?
What explanation is being claimed?
What scientific knowledge is relevant?
What alternative explanation better fits the mechanism?
What additional measurement could strengthen your explanation?

Worked answer: The observed result is that the ice cube on the metal surface melted faster. The claim that metal “produces heat” is not established by the observation. A more appropriate explanation is that the metal transfers thermal energy from the warmer surroundings and tray to the colder ice more readily than the wood does, under the conditions of the set-up. Measuring the starting temperatures and keeping cube size, surface area, room conditions and starting time comparable would strengthen the interpretation.

Delayed Independent Return

Come back to this two days later. Find any harmless science demonstration in a book, school resource or public video. Do not ask whether you “believe” it. Instead write five lines:

I observed…
The source claims…
Another possible explanation is…
The comparison I would want is…
The conclusion I can justify now is…

If you can do that without a template beside you, the reasoning is beginning to transfer.

Useful eduKateSengkang Routes

Parent and Tutor Teaching Guide

When a child watches a striking demonstration, avoid immediately asking, “What is the correct explanation?” Begin one step earlier: “Tell me exactly what you saw.” Then ask, “What did the presenter say caused it?” Keeping those two statements separate reveals whether the learner is preserving evidence or absorbing interpretation as though it were an observation.

If the learner proposes an alternative explanation, do not reward novelty for its own sake. Ask whether that alternative predicts the observation and what new test would distinguish it from the original explanation. This keeps scepticism disciplined rather than reflexive.

A useful teaching sequence is: observation first, two plausible explanations second, discriminating test third, conclusion last. Over time, reduce the prompts. The goal is not to make the child suspicious of every video. The goal is to make the child capable of asking what the evidence actually did.

Authoritative Sources

The Quiet Rule to Keep

A dramatic event deserves attention. It does not deserve an automatic explanation.

Keep the observation intact. Keep more than one possible cause alive long enough to compare them. Then ask for the evidence that can make one explanation stronger than the others. That is not distrust. It is Science doing its job.

Small Group Tutorials