Primary 6 Science becomes more precise when students learn to ask one question before accepting a conclusion: what does the evidence actually prove? A graph can show a pattern without proving every possible cause. An experiment can support a relationship under tested conditions without proving that the same relationship holds everywhere. A structured answer can sound scientific while quietly claiming more than the evidence allows.
This Sengkang Primary 6 Science small-group tutorial page owns one specific job: evidence boundaries. The wider Sengkang P6 Science estate already covers one-paper/three-repair-plan work and broad PSLE preparation. Here we focus on evaluating graphs, tables, investigations and competing claims under the revised 2026 PSLE Science demands.
The Revised 2026 PSLE Science Paper Rewards More Than Recall
For 2026, Standard Science is examined in one 1 hour 45 minute paper. Booklet A contains 30 multiple-choice questions worth 60 marks; Booklet B contains 10–11 structured questions worth 40 marks. The assessment aligns with the 2023 Primary Science syllabus and includes knowledge with understanding as well as application and scientific inquiry.
Scientific inquiry includes interpreting information, analysing, evaluating and communicating explanations and reasoning. That makes evidence boundaries important. Parents can verify the current examination structure at SEAB and the syllabus at MOE.
Evidence Can Support a Claim Without Supporting Every Version of the Claim
| Evidence | Reasonable claim | Overclaim |
|---|---|---|
| Three tested temperatures show increasing rate | Within the tested conditions, the rate increased as temperature increased | The rate always increases at every possible temperature |
| One material absorbs less water than another | Material A absorbed less water than B in this test | Material A is always the best material for every waterproof use |
| Plant under one changed light condition shows different result | The changed light condition affected the measured outcome in the investigation | Light is the only factor that can affect the plant |
| One circuit arrangement produces brighter bulb | This arrangement produced greater observed brightness under the stated setup | This arrangement is always superior for all circuits |
The strongest answer is often the one that is confident about what the evidence supports and restrained about what it does not.
Step 1: Identify the Claim
Before evaluating evidence, students should state the claim clearly. Vague claims are difficult to test.
- What exactly is being asserted?
- Is it about cause, comparison, prediction or classification?
- Does the claim apply to one tested case or a wider category?
- Does it use absolute language such as “always”, “only” or “all”?
Words such as “always” raise the evidence threshold because they make broader claims.
Step 2: Identify the Evidence
Students then separate what was actually measured, observed or provided from their interpretation.
- Which values are in the table?
- What does the graph show?
- What was changed in the experiment?
- What was kept the same?
- What observation was made?
- How many conditions were tested?
This prevents memory of the topic from replacing the evidence in the question.
Step 3: Ask Whether the Evidence Matches the Claim
A useful evaluation routine is:
- Does the evidence directly address the claim?
- Was the relevant comparison fair enough to interpret?
- Could another changed condition explain the result?
- Does the conclusion stay within the tested range?
- Is there an exception that changes the conclusion?
Graphs: Trend Is Evidence; Cause Needs a Model
A graph can show that two quantities change together. It does not automatically establish why. The student must combine the observed trend with the relevant scientific concept and the investigation design.
| Graph statement | Job |
|---|---|
| “As X increased, Y decreased in the tested range.” | Pattern description |
| “This occurred because…” | Scientific explanation requiring concept |
| “Therefore X is the only factor affecting Y.” | Likely overclaim unless method supports exclusivity |
Tables: One Number Rarely Proves the Whole Relationship
Students should compare the relevant rows or columns rather than select the number that looks most dramatic.
- Read headings and units.
- Identify the changed condition.
- Compare the measured outcome.
- Check whether the pattern is consistent.
- Notice any exception.
- State the conclusion at the right level of generality.
Experiments: A Conclusion Is Only as Strong as the Comparison
If two important variables change at once, the result may not tell us which one caused the difference. P6 students should be able to explain that limitation rather than merely repeat “not a fair test”.
A stronger response identifies the confounding condition and explains how it prevents a clean causal interpretation.
Booklet A: Distractors Often Exploit Evidence Overreach
In MCQ, a distractor may contain a true Science fact that is not supported by the particular evidence. The student should ask:
- Is this statement scientifically true in general?
- More importantly, does it answer this question?
- Does the diagram/data support it?
- Is it broader than the evidence?
- Does another option fit the evidence more precisely?
This is why strong MCQ work is not keyword recognition.
Booklet B: Evaluation Must Be Written Clearly
Structured responses may ask students to justify, explain, suggest or evaluate. A useful evaluation answer often contains three elements:
- identify the evidence or method feature;
- state the limitation or strength;
- explain how that feature affects the conclusion.
For example, “The experiment is unfair” is incomplete. “The two setups used different starting amounts, so the difference in result may be caused by starting amount rather than the tested factor” makes the reasoning visible.
Competing Explanations: Which One Needs Fewer Unsupported Assumptions?
A small-group tutorial can present three plausible explanations for the same evidence.
| Explanation | Evaluation question |
|---|---|
| A fits all observed results | Does it use an appropriate scientific concept? |
| B fits some results but ignores one condition | What evidence contradicts it? |
| C sounds scientific but adds an untested factor | Where did that factor come from? |
Students learn that explanation quality depends on both scientific correctness and evidential fit.
Three Students Makes Evidence Evaluation Social but Verifiable
eduKate’s three-student format can make claim evaluation visible without turning it into majority voting.
- Each student chooses or writes a conclusion independently.
- Each student marks the evidence used.
- The group compares how broad each claim is.
- The tutor asks what evidence would be needed for a stronger claim.
- Each student then evaluates a fresh case alone.
The final authority is not the most confident student. It is the relationship between evidence, method and scientific model.
Ask “What Evidence Would Change Your Mind?”
This question develops scientific flexibility. A student should not defend an answer simply because it was the first answer.
- What new result would weaken your conclusion?
- What additional comparison would strengthen it?
- What measurement would distinguish two explanations?
- What variable should be controlled next time?
The learner begins treating conclusions as evidence-responsive rather than identity statements.
Separate Confidence From Certainty
Science often works with degrees of support. At primary level, students do not need advanced statistical language, but they can learn to avoid false certainty.
“The results support…” is often more scientifically disciplined than “This proves forever that…”. The exact wording should remain natural and appropriate to the question.
The Five Syllabus Themes Still Need the Same Evidence Discipline
- Diversity: does the classification evidence support the grouping?
- Cycles: does the observation support the proposed stage or change?
- Systems: does changing one part support the predicted system effect?
- Energy: does the evidence support the proposed transfer or conversion?
- Interactions: does the observed effect support the claimed interaction?
The content changes, but the evidence question remains.
Common P6 Evidence Errors
| Error | Repair |
|---|---|
| Overgeneralises from one case | Limit conclusion to tested conditions |
| Uses true fact unrelated to data | Reconnect answer to question evidence |
| Ignores uncontrolled variable | Explain alternative cause |
| Describes graph but claims causation | Add method and concept reasoning |
| Changes answer because peer sounds confident | Require evidence before revision |
What Parents Can Notice
- Can your child say exactly what a graph or table shows?
- Can they distinguish pattern from explanation?
- Can they identify when an investigation does not isolate the intended factor?
- Can they recognise an overclaim?
- Can they explain why one competing claim is better supported?
- Can they state what additional evidence would strengthen a conclusion?
For Sengkang Families
Families searching from Sengkang should confirm current class location, lesson timing and availability directly. eduKate’s location-targeted pages organise tuition information but do not imply a physical branch in every named neighbourhood.
The P6 Goal: Strong Claims, Proper Boundaries
PSLE Science rewards students who can use scientific ideas with evidence. The strongest learner is not the one who makes the largest claim. It is the one who can explain exactly what the evidence supports, what it does not yet support, and why.
About eduKate
eduKate uses very small groups to compare claims, inspect evidence boundaries and retest scientific evaluation independently. Our core values are Integrity, Empathy, Critical Thinking and Responsibility.
