Primary 3 Science Learning Guide | Evidence Limits, Uncertainty & How Do You Know?

One of the most important Science questions a child can learn to answer is: “How do you know?”

The question forces the learner to separate a claim from the evidence supporting it. Primary 3 pupils do not need advanced ideas about probability or formal uncertainty analysis, but they can already learn that some evidence is stronger than other evidence, one observation may fit more than one explanation, and a conclusion should not claim more than the test can show.

This guide develops evidence limits and scientific restraint across the P3 topics of living things, materials, life cycles and magnets.

Wait, What? A Correct Answer Still Needs a Reason

A pupil writes, “Object A is living.” Perhaps that is correct. But if the question asks “How do you know?”, the label alone is incomplete. The pupil should point to evidence: “Object A grows and reproduces. These are characteristics of living things.”

The scientific structure is:

Claim → Evidence → Scientific idea.

Evidence Supports a Claim

Evidence is information that makes a claim more or less reasonable. At P3, evidence may be an observation, a comparison result, a diagram feature, a table entry or a learned rule that clearly applies.

• Claim: Material A is more flexible.
• Evidence: It bent farther than Material B before breaking under the same test.
• Science idea: Flexibility is the ability to bend without breaking.

The answer becomes stronger because each part has a job.

Support Is Not the Same as Absolute Proof

Primary 3 pupils can begin to use careful language such as “the results show”, “the evidence supports” or “this suggests”. These phrases are useful when the conclusion is tied to a particular test.

If a material did not absorb water during a short classroom test, we can say the test supports classifying it as waterproof in that context. We should not automatically claim that the material can never absorb water under any condition anywhere.

One Observation May Fit More Than One Explanation

An unknown bar is attracted to a known magnet. That observation fits at least two possibilities in the P3 model: the bar may be a magnetic material, or it may itself be a magnet with an unlike pole facing.

The evidence is real, but it does not distinguish the possibilities. A better test is needed. Trying to produce repulsion with a known pole provides more discriminating evidence.

This is an important scientific habit: when several explanations fit, ask what new observation could separate them.

A Stronger Test Answers a More Specific Question

“Is the object attracted to a magnet?” is a useful question about magnetic behaviour. It is not enough to answer “Is this definitely a magnet?” because attraction is shared by magnets and magnetic materials.

Good evidence depends on matching the test to the claim.

Evidence Can Be Relevant but Incomplete

A plant has dry soil and drooping leaves. These observations are relevant to the possibility that the plant lacks water. But they do not automatically prove water shortage is the only possible explanation.

Pupils can learn to say “may” when the evidence supports a possibility but does not uniquely determine the cause.

Uncertainty Does Not Mean “We Know Nothing”

Scientific uncertainty means the evidence has limits. We may know something confidently while still recognising what remains unresolved.

For example, repeated observation can show that a plant grew over a week. The growth itself may be clear. The exact contribution of sunlight, water, temperature and other conditions may still be uncertain if those factors were not separately tested.

The “How Do You Know?” Routine

1. Claim: What am I saying?
2. Evidence: What observation, result or label supports it?
3. Connection: Which scientific idea makes that evidence relevant?
4. Limit: What does the evidence not show?
5. Next test: What extra evidence would help if needed?

The first three steps may be enough for many P3 answers. The last two develop stronger scientific judgement.

Worked Example 1: Living Thing

Claim: Object A is likely to be living.

Evidence: It grows and reproduces.

Connection: Growth and reproduction are characteristics of living things.

The answer is stronger than “It looks alive” because the evidence can be checked.

Worked Example 2: Waterproof Material

Observation: Sample B did not absorb water during the test.

Conclusion: The result supports that Sample B is waterproof under the test conditions.

Limit: The test does not show every possible condition the material could face in real use.

Worked Example 3: Flexible Material

Sample A bends farther than Sample B before breaking, using the same basic test.

The evidence supports the conclusion that Sample A is more flexible in that comparison. It does not automatically tell us which sample is stronger because strength is a different property.

Evidence has a target. One test should not be forced to answer a different question.

Worked Example 4: Life-Cycle Diagram

A diagram shows egg → larva → pupa → adult.

The evidence supports the sequence and the presence of four major stages in the model. It does not show the exact length of every stage unless time information is provided.

Reading what the model does not show is part of understanding the evidence.

Worked Example 5: Magnet Attraction

Observation: Unknown Bar X is attracted to a known magnet.

Safe conclusion: X shows magnetic attraction.

Unsafe overclaim: X is definitely a magnet.

Better next test: Check whether an end of X can repel a known pole.

Worked Example 6: Similar Results Do Not Prove Identical Materials

Two samples both do not absorb water. That evidence supports that both are waterproof in the test. It does not prove the two samples are made from the same material or have identical properties in every other way.

A shared result is evidence for the property tested, not for every possible similarity.

Worked Example 7: Different Results Can Have More Than One Cause

Plant A receives more water than Plant B, but Plant A is also placed in a brighter location. Plant A grows taller.

We cannot conclude that water alone caused the difference because more than one relevant condition changed. The evidence does not isolate a single cause.

Strong Evidence Is Relevant Evidence

A correct fact can still be irrelevant. If the question asks why a material is suitable for a rain cover, transparency may be true but irrelevant if the required properties are waterproofness and flexibility.

The best evidence is the evidence that answers the actual scientific job.

Repeated Evidence Can Increase Confidence

If a result is unclear, repeating the same reasonable test can help. A barely moving object near a magnet can be retested. Repeated plant observations can show a more convincing pattern over time.

But repetition cannot rescue a fundamentally unfair comparison. Repeating a bad method produces more bad evidence.

Evidence Quality Has Several Parts

• Relevant: answers the question.
• Clear: observation or measurement is unambiguous enough to interpret.
• Comparable: samples were tested in a similar way where comparison is intended.
• Recorded: result stays attached to the right sample or time point.
• Repeatable enough: another check could be made where useful.
• Limited honestly: conclusion does not go beyond the test.

Use “Always” Carefully

Words such as always, never, definitely and proves are strong. They should not be used when the evidence supports only a narrower statement.

At P3, pupils can begin replacing unnecessary absolutes with more accurate language such as “in this test”, “the results show”, “this suggests” or “the evidence supports”.

Do Not Confuse Uncertainty With Weakness

“We need another test before deciding” can be a stronger scientific answer than a confident unsupported claim. Knowing what is not yet known is part of scientific thinking.

Ask What Evidence Would Change the Conclusion

This question helps pupils think beyond static answers.

• If the unknown bar repels a known N pole, what can we infer?
• If the material starts absorbing water in a repeat test, what changes?
• If the life-cycle diagram includes no pupa, does the four-stage pattern still apply?
• If an object only moves but shows no other life characteristics, is the living classification still supported?

Good Science is responsive to evidence.

The “Claim Too Big?” Check

1. What exactly was tested?
2. What exactly was observed?
3. Does my conclusion stay inside that evidence?
4. Did I add a cause, material identity or universal rule that was not shown?

This is a quick way to reduce overclaiming in open-ended answers.

Common Evidence Errors

• Using one characteristic as complete proof of life.
• Calling every attracted object a magnet.
• Using transparency evidence to answer a flexibility question.
• Claiming a cause when two conditions changed.
• Assuming a diagram shows exact size or duration without labels.
• Using one trial to make an unlimited claim.
• Treating repeated evidence as a substitute for a fair method.
• Ignoring evidence that disagrees with the preferred answer.

How to Practise “How Do You Know?”

After any answer, ask for one piece of evidence. Then ask what the evidence does not show. For stronger pupils, ask what additional observation would make the conclusion more certain.

This can be done with every P3 topic and requires no extra syllabus content.

How Parents Can Use Evidence Language

Useful prompts include: “Where did that information come from?” “Which result supports that?” “Could there be another explanation?” “What would you test next?”

These questions teach healthy scepticism without teaching the child to distrust everything.

How Teachers Can Model Uncertainty

Use language such as “From this result we can conclude…” and “This result alone does not tell us…” Pupils learn that scientific confidence is tied to evidence rather than personality or certainty of tone.

A Mini Diagnostic

1. Give evidence that an unknown organism is living.
2. Explain why attraction alone does not prove an object is a magnet.
3. State one limit of a waterproofness classroom test.
4. Explain why a flexibility test does not automatically tell us strength.
5. Give an example of one observation fitting two explanations.
6. Explain why repeating an unfair test does not fix it.
7. Rewrite an overconfident statement using more accurate evidence language.

Primary 3 Science Checkpoint

• I can answer “How do you know?” with evidence.
• I connect evidence to the scientific idea.
• I understand that evidence can support a claim without proving every possible version of it.
• I can recognise when one observation fits several explanations.
• I can suggest a better test when evidence is ambiguous.
• I know that repeated evidence helps only if the method is meaningful.
• I can state what a diagram or test does not show.
• I use strong words such as “always” and “definitely” carefully.
• I can keep a conclusion within the limits of the test.
• I know that saying “we need more evidence” can be scientifically strong.

Continue the Primary 3 Science Learning Guide

• Safe Home Investigations & Everyday Observation
• Better Questions & Practice Design
• Year-End Synthesis & Independent Learning Plan

Return to the Primary 3 Science Learning Hub.