Originally published 20 February 2015 as an eduKate Yishun tuition page. Rebuilt in 2026 as a narrow PSLE Science companion focused on open-ended responses and examination execution.
Quick answer: This page has one job: help Primary 6 students convert Science understanding into precise open-ended answers. The broader PSLE Science concept/inquiry guide is Primary 6 PSLE Science — Scientific Inquiry, Evidence, Explanation and Exam Transfer.
Archive boundary: this URL previously duplicated an older Yishun PSLE Science tuition page and included outdated centre/result claims. It is not a current centre listing, and this companion is intentionally not indexed to prevent search cannibalisation with the canonical PSLE Science guide.
What the 2026 PSLE Science paper actually asks students to do
SEAB’s 2026 PSLE Science syllabus assesses both knowledge with understanding and application of knowledge with scientific inquiry. That includes interpreting and analysing information, evaluating observations and methods, and communicating explanations and reasoning. Open-ended questions therefore test more than recall.
The answer chain
For many explanation questions, students should make the relationship explicit:
Given condition → relevant scientific process → intermediate effect → answer to the question.
The exact chain changes with the problem. A memorised sentence should never override the evidence shown.
Answer the command word
- State: give the required fact, feature or observation.
- Explain: supply the scientific relationship that makes the result happen.
- Compare: make the relevant difference explicit.
- Predict: state an expected outcome grounded in a scientific relationship or data pattern.
- Suggest: propose a reasonable method or explanation consistent with the situation.
A scientifically correct sentence can still miss the mark if it performs the wrong job.
Use the evidence provided
Before writing, identify what the question gives you:
- diagram;
- table;
- graph;
- experimental setup;
- changed condition;
- observed result.
Then ask which part of that evidence your explanation must account for. This prevents generic model-answer dumping.
Observed versus inferred
Write down what is directly shown before explaining why it happened. A graph that rises is an observation. A claim about the biological or physical mechanism is an interpretation. Keeping these separate improves accuracy.
Open-ended error classes
- Knowledge: the concept itself is missing.
- Evidence selection: the student ignores the relevant clue.
- Causal link: correct terms but the relationship is missing or reversed.
- Question fit: the response does not answer the command.
- Precision: wording is too vague to establish the scientific idea.
- Overclaim: conclusion goes beyond what the evidence supports.
- Execution: a known idea is lost through rushing or incomplete reading.
Model answers: extract the reason, not the sentence
When reviewing a model answer, ask:
- Which relationship earns the mark?
- Which detail is specific to this experiment?
- Which scientific term is needed for precision?
- How would the answer change if the condition changed?
If a student cannot adapt the answer to a changed scenario, the sentence has been learned more strongly than the Science.
Timed execution
Open-ended sections expose time management because students must both reason and compose. A practical routine is:
- read the command and evidence;
- identify the scientific relationship;
- write the shortest complete chain;
- check that every clause contributes;
- move on if the question is not yielding new information;
- return later if time remains.
Longer is not automatically better. Completeness and precision matter more than decorative phrasing.
Paper review
- mark the response;
- locate the first missing scientific link;
- classify the error;
- rewrite one complete answer;
- attempt a changed question using the same underlying concept;
- retest after a delay.
Another full paper without this repair can reproduce the same mistake at higher volume.
What not to conclude
- Do not use this page as a current Yishun centre listing.
- Do not treat it as the canonical PSLE Science article; use the linked 6 February guide for the full route.
- Do not memorise open-ended answers as fixed scripts.
- Do not assume more words earn more marks.
- Do not infer more than the data supports.
- Do not promise a grade outcome.
PSLE Science Open-Ended Questions Are Explanation Problems
Open-ended Science questions are not solved by writing the greatest number of keywords. They require the student to identify the evidence, select the relevant scientific concept, explain the mechanism and connect that mechanism to the result asked.
Use the Current 2026 Science Format
SEAB’s revised 2026 PSLE Science paper consists of one written paper with two booklets. Booklet A has 30 multiple-choice questions worth 60 marks. Booklet B has 10–11 structured questions worth 40 marks. The full paper lasts 1 hour 45 minutes. Candidates are assessed on knowledge with understanding and on application of knowledge with scientific inquiry.
What Booklet B Actually Demands
Structured questions can combine diagrams, tables, graphs, experimental setups and several linked parts. The student must often carry information from one part into the next.
The Evidence–Concept–Mechanism–Result Chain
This four-part structure is a reliable way to diagnose incomplete answers. Not every response needs four sentences, but the reasoning should contain the necessary links.
Evidence
What observation, value, trend, diagram feature or experimental condition matters? Mark it before writing.
Concept
Which scientific fact, principle or relationship explains the evidence? The concept should fit the question rather than merely belong to the same topic.
Mechanism
How does the concept produce the observed change? This is the link students most often omit when they rely on keywords.
Result
What outcome does the question ask the student to explain? Complete the chain all the way to that result.
Question Scope
Before writing, identify whether the task is to state, compare, predict, conclude, suggest an improvement, explain or evaluate. Correct Science can still lose marks if it answers the wrong task.
Comparison Questions
If A and B are compared, keep both conditions visible. Students often write only about the more obvious condition and leave the comparison incomplete.
Prediction Questions
State the prediction and justify it using the relevant mechanism or pattern. A prediction without reasoning is weak when the question asks why.
Conclusion Questions
Conclude only what the investigation supports. Avoid turning one limited comparison into a universal statement.
Variable Questions
Identify what is changed, what is measured and what is kept the same where relevant. The student should understand why controlled conditions make causal interpretation stronger.
Improvement Questions
Name the weakness and propose a matching correction. If measurement is inconsistent, improve measurement. If a variable is uncontrolled, specify what should be kept the same.
Evaluation Questions
Students may need to judge observations, information or methods. The answer should identify a specific strength, limitation or consequence rather than provide a generic phrase.
Graph Questions
Read title, axes, units and scale first. Describe the trend before explaining it. Distinguish observed pattern from inferred cause.
Table Questions
Compare relevant rows or columns and preserve units. Use the actual values if the question requires evidence.
Diagram Questions
Arrows, labels, positions and relative sizes may contain evidence. Students should inspect the visual rather than treat it as decoration.
Open-Ended Language Must Be Precise
Ambiguous pronouns and vague verbs can weaken a correct idea. Name the object, material or process when reference could be unclear.
Keywords Need Verbs
Scientific nouns become explanations when linked through verbs: transfers, absorbs, decreases, prevents, allows, increases. Students should practise relationships, not vocabulary piles.
Do Not Overwrite
Extra statements increase the chance of contradiction. Write enough Science to complete the reasoning and then stop.
Model Answers Should Be Analysed, Not Memorised
Identify what makes the model answer complete. Which evidence was used? Which mechanism was stated? Then close it and reconstruct the answer for a changed context.
Correction Should Be Rewritten
After feedback, the student should produce the improved answer from scratch. Copying does not prove the reasoning can be generated independently.
Build an Open-Ended Error Ledger
- scope;
- evidence;
- concept;
- mechanism;
- comparison;
- variable control;
- graph or table reading;
- language precision;
- overwriting;
- time.
Retest Error Families
If the student missed the comparison, retest comparison in a different topic. Process errors often travel across content.
Use Timed Open-Ended Sets
Short sets of three or four structured parts can reveal whether the student is slow because of reading, concept selection or writing. This is more diagnostic than immediately doing another full paper.
The 2026 Paper Requires Integration
With 30 MCQs and 10–11 structured questions in 1 hour 45 minutes, students need both knowledge fluency and disciplined written reasoning. Practice should train switching between question types without losing precision.
The Final Open-Ended Routine
- read the exact task;
- mark the evidence;
- identify the concept;
- explain the mechanism;
- connect to the result;
- check scope and clarity.
Final Guide
PSLE Science open-ended questions become more manageable when students stop treating them as memory tests and start treating them as evidence-and-explanation problems. The strongest answer is not the longest; it is the one whose reasoning is complete and easy to follow.
A Complete Open-Ended Answering Handbook
Structured Science questions are difficult because the student must do several things at once: read unfamiliar information, identify the relevant concept, decide which evidence matters, construct the mechanism and express the answer precisely. The best preparation therefore separates these demands during training and then recombines them under timed conditions.
Step 1: Identify the Task Verb
State, describe, compare, predict, explain, conclude, suggest and evaluate are not interchangeable. The first reading should establish what kind of response is required.
Step 2: Mark the Evidence
Underline the relevant value, trend, label, condition or observation. This prevents the answer from drifting into generic topic knowledge.
Step 3: Select the Concept
Ask which scientific relationship explains the evidence. The concept should fit the specific situation, not merely belong to the same chapter.
Step 4: Build the Mechanism
Explain how the concept produces the observed result. This middle link is what distinguishes explanation from keyword recall.
Step 5: Return to the Exact Result
The final sentence should close the loop and answer what was asked. Students should not leave the reasoning one step short.
Worked Example Type: Comparison
If two plants grew differently, the student should identify the relevant condition for both plants, compare the evidence and explain why the difference in condition produces the difference in growth. Describing only the better-growing plant is incomplete.
Worked Example Type: Prediction
If a condition changes, the student should state the expected result and connect the change to the mechanism. Predictions should not be guesses detached from Science.
Worked Example Type: Conclusion
A conclusion should match the tested relationship and stay within the range of evidence. Students should avoid universal claims from a limited setup.
Worked Example Type: Experimental Improvement
Identify exactly what weakens the method. If measurements are inconsistent, standardise or improve measurement. If a condition is uncontrolled, keep it the same. The improvement should answer the flaw.
Worked Example Type: Evaluation
When judging a method, observation or claim, state a specific limitation and why it matters. Generic phrases such as “not accurate” are weak without the mechanism.
Worked Example Type: Graph Interpretation
Describe the pattern first. Then explain. This prevents students from inventing causes before establishing what the graph actually shows.
Worked Example Type: Table Evidence
Use the relevant values and units. Students should not say “A is better” when the table supports a precise statement such as “A had a higher temperature at every recorded time”.
Worked Example Type: Diagram Evidence
Arrows, labels, relative positions and structures may provide the evidence needed. Train students to extract information from visuals before writing.
Avoid the “Keyword = Mark” Myth
Keywords help because Science needs precision, but marks depend on whether the scientific relationship is communicated. Students should think in complete causal relationships rather than lists.
Use Mechanism Verbs
Transfers, absorbs, increases, decreases, blocks, allows, prevents, carries and causes are examples of verbs that connect scientific nouns. Strong explanations need relationships.
Use Pronouns Carefully
“It”, “they” and “this” can become ambiguous in dense explanations. Name the object, substance or process when reference could be unclear.
Use Answer Length Proportionally
A two-mark response rarely needs an essay. A higher-mark structured part may need several linked ideas. Students should complete the reasoning without adding unrelated content.
Teach Editing of Science Answers
After writing, remove any sentence that does not help answer the question. Then check whether the remaining answer still contains evidence, concept and mechanism.
Use Model Answers as Rubrics
Compare the student’s response with a model and identify the missing relationship rather than memorising the exact wording. Different wording can still express correct Science.
Build Error Families Across Topics
A student may repeatedly omit comparison in heat, plants and materials. That is one process problem, not three topic problems. Repair the process across several contexts.
Use Open-Ended Error Codes
- S = scope;
- E = evidence;
- C = concept;
- M = mechanism;
- V = variable control;
- G = graph/table reading;
- L = language precision;
- T = time.
A short code helps students see patterns across papers.
Use Delayed Rewrite
After correction, wait a day or two and ask the student to answer a related question without the model. This tests whether the reasoning can be regenerated.
Use Topic Transfer
If the student learns comparison structure in one topic, test it in another. Process skills should become topic-independent.
Use Question Reconstruction
Give a correct answer and ask what kind of question it could answer. This reverse task makes scope and evidence requirements visible.
Use Evidence-Only Drills
Show a graph, table or diagram and ask only what can be directly observed. Then add the explanation. Separation helps students avoid mixing observation and inference.
Use Mechanism-Only Drills
Provide cause and outcome and ask for the missing process. This isolates the layer many open-ended answers lack.
Use Concision Drills
Give an overlong answer and ask the student to reduce it without losing scientific meaning. This teaches efficient communication.
Use Timed Sets Before Full Papers
Time three or four structured questions. Diagnose whether delay comes from reading, concept selection, writing or checking. Then target the bottleneck.
Use Full Papers for Integration
Full papers are useful once enough component skills are stable. They test switching between MCQ and structured reasoning, endurance, pacing and recovery.
The 2026 Time Demand
The revised 2026 PSLE Science paper gives 1 hour 45 minutes for 30 multiple-choice questions and 10–11 structured questions. Students therefore need efficient reading and concise structured responses as well as scientific knowledge.
Build Personal Time Checkpoints
Students can practise approximate checkpoints that fit the official duration and their own working style. The purpose is to notice major pacing drift before the final minutes.
Recovery After a Difficult Structured Question
Mark what is known, answer any accessible subparts, decide whether more time is justified, and return later if needed. One difficult question should not damage the rest of the paper.
Final Open-Ended Checklist
- I know what the question asks.
- I marked the relevant evidence.
- I selected the right concept.
- I explained the mechanism.
- I connected to the result.
- I used precise language.
- I removed irrelevant material.
- I kept the answer within time.
Final Perspective
PSLE Science open-ended success comes from disciplined reasoning. Students should be able to reconstruct explanations from evidence rather than reproduce memorised paragraphs. That is exactly the kind of capability the revised examination is designed to test.
A Final 2026 PSLE Science Open-Ended Operating Manual
The revised 2026 PSLE Science paper requires students to answer 30 multiple-choice questions for 60 marks and 10–11 structured questions for 40 marks within 1 hour 45 minutes. Open-ended preparation must therefore combine scientific reasoning with time control.
Before Writing: The Five-Second Scope Check
- What exactly is being asked?
- What evidence matters?
- Which concept applies?
This brief pause prevents many off-scope answers.
While Writing: Complete the Scientific Relationship
Do not stop at the keyword. Explain the process that links cause and result. Use precise nouns and verbs so the relationship is easy to follow.
After Writing: The Three-Point Check
- Did I answer the task?
- Did I use the evidence?
- Did I complete the mechanism?
Comparison Operating Manual
Identify the relevant difference between A and B, connect that difference to the mechanism, then explain the difference in result. Keep both sides visible.
Prediction Operating Manual
State the expected outcome, identify the changed condition and explain the mechanism. Predictions are scientific when they are reasoned.
Conclusion Operating Manual
Use the observed relationship and stay within the tested conditions. Avoid claims the evidence cannot support.
Improvement Operating Manual
Name the specific method weakness and propose a matching correction. Explain why the change improves the test where required.
Evaluation Operating Manual
Identify a strength or limitation and connect it to the quality of the evidence or conclusion. Avoid generic “accurate” or “unfair” language without explanation.
Graph Operating Manual
Read title, axes, units and scale. Describe the trend. Use the pattern as evidence. Only then explain.
Table Operating Manual
Compare the relevant values and preserve units. If the question asks for evidence, use the data explicitly.
Diagram Operating Manual
Mark arrows, structures, relative positions and labels that affect the answer. Visual information is part of the evidence.
Variable Operating Manual
Identify what changes, what is measured and which important conditions stay the same. Then explain why the comparison becomes more interpretable.
Concision Operating Manual
After writing, remove any sentence that does not contribute to the required scientific relationship. Concision reduces contradiction risk.
A Weekly Open-Ended Practice Plan
- one comparison question;
- one graph or table question;
- one variables or method question;
- one mechanism explanation;
- one timed structured set;
- one delayed rewrite of an earlier error.
A Four-Week PSLE Open-Ended Cycle
- Week 1: scope and evidence.
- Week 2: mechanism and explanation.
- Week 3: inquiry and evaluation.
- Week 4: timed mixed structured questions.
The Final Error Ledger
- off-scope answer;
- evidence not used;
- wrong concept;
- mechanism missing;
- comparison incomplete;
- variable misidentified;
- graph/table misread;
- vague language;
- overwriting;
- time loss.
The Final Transfer Test
Take the same process error into another topic. If the student can now compare correctly in heat, plants and electricity, the repair has transferred. If it works only in one chapter, more varied practice is needed.
The Final Timed Test
Use a short structured section under realistic conditions. The goal is not merely completion; it is to preserve evidence reading and mechanism quality while time pressure increases.
Final PSLE Open-Ended Standard
A strong candidate can reconstruct the answer from unfamiliar evidence, communicate the mechanism precisely and manage the structured section without relying on memorised model paragraphs. That is durable Science performance.
A Full PSLE Science Open-Ended Training System
Structured Science questions should be trained in layers. Students need content knowledge, but they also need task recognition, evidence use, explanation, inquiry reasoning and time control. A full training system separates these layers before recombining them under examination conditions.
Layer 1: Command-Word Recognition
Give the student question stems and ask what type of response is required before showing the content. State, compare, predict, conclude, suggest and explain should trigger different response structures.
Layer 2: Evidence Selection
Show a graph, table or diagram and ask which evidence matters. Do not allow explanation yet. This isolates the reading layer and reveals whether the student is scanning accurately.
Layer 3: Concept Selection
Once evidence is identified, ask which scientific concept explains it. This prevents students from grabbing the first familiar topic word.
Layer 4: Mechanism Construction
Ask what process connects cause and result. Students should be able to explain the middle rather than jump from keyword to outcome.
Layer 5: Response Compression
After a complete explanation is written, remove sentences that do not help answer the task. Concision becomes an editing skill, not an excuse to omit mechanism.
Layer 6: Time Control
Use short structured-question sets under time. Track whether the student loses time reading, deciding, writing or checking. The intervention should match the bottleneck.
Worked Case: Comparison Question
Two plants are shown under different conditions. The weak answer describes only Plant A. The repair is to state the relevant difference between A and B, then explain how that difference affects the result. The comparison must remain visible throughout the answer.
Worked Case: Prediction Question
A graph shows a trend and the student is asked what happens next. The response should state the predicted direction and justify it using the observed pattern or scientific mechanism, depending on the question.
Worked Case: Conclusion Question
The investigation tests one factor over a limited range. The student should conclude only within those conditions. Overgeneralisation creates a claim larger than the evidence.
Worked Case: Experimental Improvement
A measurement method is inconsistent. The answer should name the weakness and propose a specific improvement that reduces inconsistency. Generic “repeat more times” is only appropriate when repetition addresses the identified problem.
Worked Case: Controlled Variable
Two setups differ in light and water, but the question aims to test light. The student should identify water as an uncontrolled variable and explain why it should be kept the same.
Worked Case: Graph Scale
The student reads one grid square as one unit when the axis increases by five. The error is not topic knowledge; it is data-reading execution. Repair with several scale-reading drills across topics.
Worked Case: Keyword Without Mechanism
The student writes “conduction” or “evaporation” but does not explain how the process produces the observed result. The repair is a cause-process-result sentence.
Worked Case: Overwriting
The answer contains correct facts plus one incorrect extra statement. The tutor should show that more Science is not always better Science. Remove everything that does not serve the exact question.
Use Answer Surgery
Take a long answer and cut it down to the smallest complete version. Then compare what remained: evidence, concept, mechanism and result. This teaches concision through structure.
Use Missing-Link Questions
Provide a partially complete answer with one missing causal step. Ask the student to supply only that step. This directly targets explanation quality.
Use Wrong-Answer Comparison
Show two wrong answers: one ignores evidence, the other has evidence but no mechanism. Ask which error family each represents. Students learn to diagnose, not just imitate.
Use Delayed Reconstruction
After reviewing a model answer, close it and reconstruct the reasoning later. If the student cannot rebuild it in a changed context, the learning remained surface-level.
Use Cross-Topic Structured Sets
Mix heat, forces, plants, electricity and data questions. The student should identify the relevant concept from evidence, not from worksheet headings.
Use Section Timing Before Full Papers
Train a small structured section under time. Review whether the student spends too long writing, interpreting diagrams or choosing concepts. Then repair the bottleneck before another timed set.
Use Full Papers for Integration
Full papers should test the interaction of knowledge, MCQ efficiency, structured reasoning, time allocation and recovery. They are most useful after the main process weaknesses have been identified.
The 2026 Time Budget
The official 2026 paper gives 1 hour 45 minutes for both booklets. Students need a personal pacing plan that leaves enough time for the 40-mark structured section without rushing the 60-mark MCQ section. Exact pacing can differ by student, but it should be rehearsed rather than improvised.
A Personal PSLE Science Error Ledger
- scope;
- evidence;
- concept selection;
- mechanism;
- comparison;
- variables;
- graph/table reading;
- language precision;
- overwriting;
- time management.
A Four-Week Open-Ended Cycle
- Week 1: diagnose error families.
- Week 2: practise evidence and mechanism separately.
- Week 3: mix topics and reconstruct answers.
- Week 4: time structured sections and review.
The Final Structured-Question Checklist
- I know what the question asks.
- I marked the relevant evidence.
- I selected the right concept.
- I explained the mechanism.
- I connected to the result.
- I removed irrelevant statements.
- I checked units, labels and scope.
Final Perspective
Open-ended Science becomes less intimidating when students understand that most lost marks come from identifiable reasoning gaps. Evidence, concept, mechanism, result, scope and timing can all be trained separately and then integrated into one disciplined response.
A Final PSLE Science Structured-Response Workshop
Once students understand the evidence–concept–mechanism–result chain, practise it under controlled variation. Use the same scientific relationship across a graph, a table, an experimental setup and a short written scenario. The student should recognise the concept even when the representation changes.
Workshop Step 1: Identify the Task
Before solving, label the question as comparison, prediction, conclusion, improvement, variable control or explanation. This prevents the learner from producing a generic chapter answer.
Workshop Step 2: Mark the Evidence
Highlight the two values, trend, arrow, condition or observation that matter. The student should be able to explain why that evidence was selected.
Workshop Step 3: Build the Minimum Complete Answer
Write only what is required to connect evidence to result. Then ask whether any sentence can be removed without breaking the reasoning. This trains concision without sacrificing scientific completeness.
Workshop Step 4: Transfer the Same Reasoning
Change the context or representation. If the student can rebuild the explanation rather than repeat the original wording, the concept is becoming portable.
Final Examination Standard
For the revised 2026 paper, the strongest structured-response student is not the one who memorises the most model answers. It is the one who can read unfamiliar evidence, identify the scientific relationship, explain the mechanism precisely and manage the process within the overall 1 hour 45 minute paper.
In the final examination phase, students should also practise switching cleanly between MCQ and structured-response modes. MCQ rewards efficient elimination and concept recognition; structured questions require evidence selection, explanation and disciplined writing.
The student should not carry the same pace into every question type. A personal pacing plan should be tested in full papers and adjusted from actual evidence.
After each full paper, classify the lost marks before doing another one. If losses come mainly from structured-response scope or mechanism, targeted open-ended practice is more useful than repeating the entire paper.
The final aim is controlled flexibility: the student can move from quick recognition to deeper explanation without losing accuracy or time awareness.
A final structured-response drill is to remove the chapter title and mix questions from several topics. Before answering, the student labels only the task type and the evidence source. Then the student chooses the concept and writes the minimum complete explanation. This forces recognition and scope control at the same time. When repeated under realistic timing, it prepares the learner for the actual 2026 demand of switching between quick MCQ decisions and deeper structured reasoning within one 1 hour 45 minute paper.
A final PSLE Science routine is to reserve the last moments of each structured section for targeted checking rather than rereading everything. The student should scan personal high-risk areas: whether both sides of a comparison were addressed, whether the mechanism is complete, whether graph or table evidence was used accurately, and whether an extra sentence introduces contradiction. This makes checking efficient and evidence-led.
The strongest final PSLE Science check is whether the student can explain why every sentence in an open-ended response is necessary. If a sentence does not contribute evidence, concept, mechanism, comparison or result, it may not belong. This discipline improves clarity and protects time in the revised 2026 structured section.
The final PSLE Science principle is disciplined reasoning: use the evidence given, explain only the mechanism required and keep the response within the exact scope of the structured question.
Structured Science answers should remain concise, evidence-led and mechanism-complete.
Evidence-led explanation completes the structured response.
Reasoning must remain precise.