How to Evaluate Another Student’s Scientific Claim in PSLE Science

Wait, what? A science question can give you somebody else’s conclusion and still be testing your scientific thinking.

You may see a statement such as “Student A says…”, “Ali concludes…”, “Mei Ling thinks…”, or simply a claim written beneath a table, graph or experiment. The task is not to decide whether you like the statement. It is to evaluate whether the claim is supported by the scientific evidence and conditions given.

Quick Answer

When evaluating a supplied scientific claim in PSLE Science, use this route:

Extract the claim. What exactly is the student saying?
Extract the evidence. What observations, measurements or comparisons are actually given?
Check the conditions. What was changed, measured and kept comparable?
Select the relevant concept. Which scientific relationship should connect the evidence to the claim?
Test the mechanism. Does the proposed explanation make scientific sense in this set-up?
Classify the claim. Supported, contradicted, or not yet established by the evidence?
Repair if needed. State the strongest conclusion that the evidence can actually support.

The goal is not to attack the student. The goal is to test the reasoning.

Owned PSLE Science Learning Job

This guide owns one narrow Primary 5/6 learner job: how to evaluate a scientific claim supplied inside a PSLE Science question by testing it against the given evidence, relevant concept and exact conditions. It does not replace the broader scientific ideas of evidence, uncertainty, competing explanations or claim–evidence reasoning. It applies those ideas to the PSLE Science learner’s specific evaluation task.

A Claim Is Not Evidence

A claim is what someone says the evidence means. Evidence is the observation, measurement, comparison or other information used to support that claim. These must stay separate.

Part	Example
Evidence	Set-up A produced 18 units; set-up B produced 10 units under the stated conditions.
Claim	“Condition X caused the higher result in A.”
Evaluation question	Does the comparison isolate X, and does the relevant concept provide a mechanism that fits the result?

A learner who mixes the claim with the evidence may end up proving the statement by repeating the statement. That is circular reasoning, not evaluation.

Use the PSLE Science Reasoning Law

OBSERVE / READ GIVEN INFORMATION → IDENTIFY THE SCIENTIFIC OBJECT OR RELATIONSHIP → DISTINGUISH OBSERVATION FROM INFERENCE → SELECT THE RELEVANT CONCEPT → EXPLAIN THE CAUSAL MECHANISM → CONNECT TO THE QUESTION’S CONDITION → STATE THE OUTCOME → CHECK AGAINST THE EVIDENCE.

When another student’s claim is supplied, do not start at the conclusion. Rebuild the route from the evidence upward. If the claim survives that reconstruction, it is supported. If it conflicts with the evidence or science concept, it is contradicted. If the evidence does not discriminate strongly enough, it is not yet established.

Three Possible Evaluation Outcomes

Evaluation	Meaning	What the learner should do
Supported	The evidence and relevant scientific relationship are consistent with the claim under the tested conditions.	State why, using specific evidence and mechanism.
Contradicted	The claim conflicts with the evidence, concept, direction of effect or conditions.	Identify the conflict and state the better explanation or conclusion.
Not yet established	The evidence is insufficient or the comparison cannot decide the claim.	State what is known and what further evidence or fairer comparison would be needed.

This third category is important. Science is not always “right” or “wrong” from the information available. Sometimes the scientifically correct answer is that the claim goes beyond what the evidence can decide.

Worked Example 1: A Claim That Is Supported

Original scenario: Two identical containers each hold the same volume of water at the same starting temperature. One container is wrapped with insulating material and the other is not. Both are left in the same surroundings for the same duration. The wrapped container shows a smaller temperature decrease.

A student claims: “The insulating material reduced the rate at which thermal energy was transferred from the water to the surroundings.”

Evaluation: The comparison is designed around the presence or absence of insulating material while other relevant conditions are comparable. The temperature-change evidence is consistent with the relevant heat-transfer concept. The claim is therefore supported within the tested conditions.

Notice the boundary: the evidence supports the claim about this comparison. It does not prove that every possible insulating material performs equally well under every condition.

Worked Example 2: A Claim Contradicted by Direction

Original scenario: A graph shows that as the distance from a light source increases across the tested positions, the measured light intensity decreases.

A student claims: “The farther the sensor is from the light source, the greater the measured light intensity.”

The first job is not to discuss light theory. It is to read the evidence correctly. The direction of the claim is opposite to the plotted pattern. Therefore the claim is contradicted by the data. A repaired statement should match the observed direction across the tested positions.

This shows why evidence reading must happen before explanation.

Worked Example 3: A Claim That the Evidence Cannot Establish

Original scenario: Plant P is kept in a warmer, brighter location and receives more water. Plant Q is kept in a cooler, dimmer location and receives less water. P grows taller over the same period.

A student claims: “The brighter light caused P to grow taller.”

The evidence shows that P and Q differed in growth. But several relevant conditions differed simultaneously. The comparison does not isolate brightness. Therefore the claim may be scientifically plausible, but this investigation alone does not establish brightness as the cause of the observed difference.

A stronger evaluation is: the claim is not established by this evidence because temperature and water amount also differ; a fairer comparison would keep those relevant factors comparable while changing the light condition.

Worked Example 4: True Fact, Unsupported Claim

One of the hardest traps is a statement that contains true science but is not supported by the particular evidence.

Original scenario: A closed container shows a change in an indicator after an organism is placed inside. A student writes a long explanation about a process the organism can perform. The fact is scientifically true, but the question’s conditions do not show that this process is the only possible explanation for the indicator change.

A good evaluator asks two separate questions:

Is the science fact itself correct?
Does this evidence justify using that fact as the explanation here?

Both must pass. Scientific truth and evidential relevance are different tests.

Worked Example 5: Claim About an Experimental Method

Original scenario: A student says, “This experiment is fair because both set-ups use identical beakers.” But one set-up begins with 50 mL of liquid and the other with 200 mL, even though amount is relevant to the investigated outcome.

The statement notices one controlled feature but ignores another relevant condition. The method cannot be judged from one matching feature alone. Evaluation requires the learner to consider whether the design protects the comparison that the conclusion needs.

The repaired reasoning is not “identical beakers do not matter”. They may matter. The repair is: one controlled condition does not make the whole comparison valid if another relevant condition differs.

A Six-Question Claim Test

What exactly is being claimed? Rewrite it mentally in simple words.
What evidence is actually given? Numbers, observations, graph pattern, method or diagram?
What is observation and what is inference? Keep them separate.
What scientific relationship is relevant? Do not choose a fact just because it sounds familiar.
Do the conditions allow the claim? Check variables, fair comparison, measurement and evidence quality where relevant.
How strong should the conclusion be? Supported, contradicted, or not established?

How to Evaluate a Claim From a Table

When the evidence is in a table, first identify the row and column meanings. Then compare only the entries relevant to the claim. Check whether the claim refers to an amount, a rate, a change, a final value or a relationship between variables.

Do not scan for a number that appears to support the statement. The whole relevant comparison must be considered. A claim about a trend cannot be established from one isolated pair if the rest of the tested values do not show that pattern.

How to Evaluate a Claim From a Graph

Read both axes and units.
Identify the tested range.
Look at direction, not just individual points.
Notice plateaus, reversals or regions where the relationship changes.
Do not extend the trend beyond the data unless the question and concept justify prediction.
Separate the observed graph pattern from the mechanism used to explain it.

How to Evaluate a Claim From an Experiment

For inquiry questions, preserve the scientific roles:

Changed variable: what the investigation deliberately changes.
Measured outcome: what is observed or measured in response.
Controlled relevant conditions: what must remain comparable for the intended comparison.
Measurement quality: whether the observations represent the intended quantity sufficiently well.
Evidence boundary: what the data can and cannot support.

Then ask whether the supplied claim respects those roles.

Observable Failure Signatures

The learner agrees because the statement contains a familiar science word.
The learner disagrees because one word looks unfamiliar, without checking the evidence.
The answer repeats the claim as its own justification.
The learner uses one data point while ignoring the rest of the pattern.
The learner decides “correct/incorrect” without explaining why.
The learner treats a plausible claim as proven even though several variables changed.
The learner states a correct fact but never connects it to the given condition.
The learner assumes that if the claim is not proven, it must be false.

Find the Earliest Weak Link

Claim reading: Can the learner state what is being claimed?
Evidence extraction: Can the learner identify what the question actually gives?
Representation reading: Can the learner read the table, graph or diagram accurately?
Concept selection: Does the learner know which science relationship is relevant?
Inquiry structure: Can the learner judge variables and fair comparison?
Mechanism: Can the learner connect condition to outcome?
Evidence limit: Can the learner distinguish “possible” from “supported”?

The repair should begin at the first failed step. If the learner misread the graph, more practice in writing evaluation sentences will not solve the real problem.

Misconception Repairs

“If a claim sounds scientific, it is probably correct.”

Scientific vocabulary is not a substitute for evidence. Test the relationship.

“If the evidence does not prove the claim, the claim is false.”

Not necessarily. The evidence may simply be insufficient. Distinguish contradicted from not established.

“One matching result proves the whole explanation.”

A result may be consistent with an explanation without uniquely proving it. Check whether other plausible explanations remain and whether the design distinguishes them.

“A student’s name in the question means I should judge the student.”

The person is just the carrier of the claim. Evaluate the scientific reasoning, not the person.

A Practice Sequence for Claim Evaluation

Evidence-only round: Hide the claim and ask the learner to describe the evidence neutrally.
Claim-only round: Show the claim and ask what evidence would be needed to test it.
Supported claim: Use a clean example where evidence and mechanism fit.
Contradicted claim: Use data whose direction conflicts with the statement.
Insufficient claim: Use a confounded comparison or incomplete data.
Surface change: Repeat across Diversity, Cycles, Systems, Energy and Interactions contexts without changing the evaluation job.
Delayed return: Retest days later with a fresh representation and no labels.

Unfamiliar Transfer Challenge

Suppose a question shows a new device you have never seen. Four readings are provided. A student claims that a hidden process inside the device becomes faster as a certain condition increases.

Do not panic because the apparatus is unfamiliar. Begin with what you can read: which condition changes, which quantity is measured, what pattern the numbers show, and whether the claim’s direction matches that pattern. Then use the scientific information supplied in the question to test the mechanism. Unfamiliar surfaces do not remove the evidence structure.

Delayed Independent Return Test

Several days later, give the learner three short original cases: one supported claim, one contradicted claim and one not established by the evidence. Do not tell the learner which is which. Require a brief evidence-based classification and one sentence explaining the scientific reason.

The return succeeds when the learner can identify the category from the evidence rather than from surface wording or topic familiarity.

Answer-Checking Receipt

Claim: Did I evaluate the exact statement given?
Evidence: Did I cite the relevant observation, measurement or comparison?
Concept: Did I use the correct scientific relationship?
Mechanism: Did I explain why the evidence supports or conflicts with the claim?
Conditions: Did I check variables and fair comparison where relevant?
Strength: Did I distinguish supported, contradicted and not established?
Boundary: Did I avoid claiming more than the data allows?

Common Traps

Keyword matching instead of evidence testing.
Judging the claim before reading the graph or table.
Using “agree” or “disagree” without scientific justification.
Calling a claim false when the evidence is merely insufficient.
Ignoring another changed variable.
Using a true science fact that does not fit the tested condition.
Extending a trend beyond the observed range without justification.
Inventing a school-specific marking phrase as if it were an official universal rule.

Parent and Tutor Teaching Guide

A useful way to teach evaluation is to remove the emotional language of “right” and “wrong” at first. Ask three neutral questions: What is the claim? What is the evidence? What does the evidence allow us to say?

If the child immediately reaches for a topic fact, redirect to the evidence. If the child reads the evidence correctly but cannot judge the claim, check the relevant concept or inquiry structure. If the concept is known but the learner still overclaims, teach the difference between possibility and support.

For strong learners, use paired claims that are both scientifically plausible but differ in evidential support. Ask which one the current data can defend. This develops restraint as well as explanation.

Useful Internal Routes

Authoritative References

Quiet Return

Scientific thinking is not about agreeing with the confident sentence. It is about asking what the evidence can carry.

Read the claim. Rebuild the evidence. Test the concept. Check the conditions. Then let the strength of the conclusion match the strength of the proof.

Small Group Tutorials