Many PSLE mistakes begin before the learner writes a single word or performs a single calculation. The question is read, but the learner has not yet decided what decision the question actually requires. English becomes a search for any related sentence. Mathematics becomes a hunt for an operation. Science becomes a search for a remembered keyword. Work begins before the job is defined.
This Learner’s Guide develops one cross-subject habit: name the decision before you start the work. The learner should be able to say, in plain language, what must be decided by the end of the question. Is the job to identify a motive, compare two reactions, calculate the original amount, choose the relevant model, decide which variable caused a result, or explain why one set-up behaved differently?
This volume builds on the PSLE Learning Guide, Vol 0001: Read Before You Solve, and Vol 0026: Name the First Useful Move. Vol 0035 sits one step earlier: before choosing a move, define the decision.
QUESTION → DECISION → REQUIRED EVIDENCE OR RELATIONSHIP → METHOD → ANSWER.
The quick answer: what is a decision?
A decision is the exact outcome the learner must settle. It is narrower than the topic. “This is a ratio question” names a topic. “I need to find how many red counters remain after the ratio changes” names the decision. “This is about heat” names a topic. “I need to explain why the wrapped cup cooled more slowly” names the decision.
A good decision statement contains a target and, when needed, the basis for deciding it. In English: decide which interpretation the passage supports. In Mathematics: decide the value of the requested quantity. In Science: decide what the evidence shows or which mechanism explains it.
Why topic labels are not enough
Topic labels are useful for retrieval, but they can also trigger premature answers. A learner sees percentage and immediately multiplies. A learner sees evaporation and writes about surface area. A learner sees an English question about feelings and writes a familiar emotion. The label has activated content before the actual decision is clear.
The repair is simple: keep the topic label, but add the decision. “Percentage — find the original price.” “Evaporation — compare the two tested conditions.” “Inference — decide which emotion is best supported by the character’s actions.”
The decision sentence
Before solving, complete one short sentence: I need to decide ______. During training, say or write the sentence. In the examination, it becomes a mental cue.
- English: I need to decide why the character changed her mind.
- Mathematics: I need to decide the original amount before the 20% increase.
- Science: I need to decide whether the comparison isolates the effect of light.
Three parts of a strong decision statement
- Target: what must be found, explained, compared or selected?
- Scope: which part of the passage, problem or investigation controls the answer?
- Standard: what evidence, relationship or condition makes one answer better than another?
Not every question requires all three to be written. But when the learner is confused, expanding the decision statement often reveals what is missing.
English: decision before evidence
In English comprehension, the learner often searches the passage too early. If the decision is not defined, almost any sentence can feel relevant. First decide the job: identify a stated detail, explain a motive, infer a trait, compare reactions, interpret a phrase, or support a claim.
Worked English case: motive
Question: Why did Aisha return to the classroom? Weak launch: search for the word classroom. Better launch: “I need to decide the reason for Aisha’s return.” Now the learner looks for a trigger, need, intention or earlier event that explains the return.
Worked English case: comparison
Question: How were Ben’s and Ryan’s reactions different? Weak launch: summarise everything each character did. Better launch: “I need to decide the difference between their reactions on the same basis.” That decision points toward a comparable dimension such as confidence, willingness, speed or emotion.
English: decision before inference
Inference questions are especially sensitive to decision framing. “What does this show about her?” may ask for a temporary feeling, a motive, a trait or an attitude depending on context. Before naming a trait, decide what the evidence is actually being used to establish.
Use Vol 0006 after the decision is clear: decision first, evidence zone second, inference third.
Mathematics: decision before operation
In Mathematics, the decision is usually the target quantity. Learners who skip this step often perform valid arithmetic that solves the wrong quantity. Before choosing an operation, write or say the target in noun form: original price, number remaining, area shaded, total distance, value of one ratio unit, average speed, number of pupils in Group B.
Worked Mathematics case: percentage
A price increases by 20% to $360. The question asks for the original price. Weak launch: calculate 20% of 360. Better launch: “I need to decide the original amount that became $360 after a 20% increase.” The decision reveals that $360 is the final 120%, not the base 100%.
Worked Mathematics case: ratio
Two groups are in the ratio 3:5 and the question asks for their difference. Weak launch: add 3 and 5 and stop. Better launch: “I need the numerical difference between the groups, so I need the value of one ratio unit first.” The target determines the intermediate quantities.
Mathematics: decision before representation
Vol 0003 teaches representation before calculation. Vol 0035 adds a prior checkpoint: represent the decision, not just the numbers. A bar model should show the quantity being asked for. A table should organise the quantities needed to decide the target. An equation should define the unknown as the target or a useful intermediate.
Science: decision before recall
In Science, the learner may recall a correct concept too early. The question might ask to describe a trend, explain a mechanism, identify a variable, predict a result, evaluate a method or justify a conclusion. These are different decisions.
Worked Science case: describe versus explain
A graph shows temperature decreasing over time. If the question asks Describe the trend, the decision is to state what the graph shows. If the question asks Explain why, the decision is to identify the relevant mechanism. Writing a mechanism for a describe question can waste time and miss the requested job.
Worked Science case: fair comparison
Two plants grow differently, but both light and water changed. The decision is not simply Which plant grew more? if the question asks whether light caused the difference. The learner must decide whether the method isolates light strongly enough to support that causal claim.
A four-question launch routine
- What exactly must I decide?
- What information could settle that decision?
- What information is related but not necessary?
- What first move will expose the needed relationship or evidence?
The routine is deliberately short. It prevents the learner from doing expensive work before the direction is stable.
Decision errors versus knowledge errors
A learner may know the content yet answer the wrong job. That is a decision error, not a knowledge gap. Vol 0009’s error classification is useful here. Before reteaching a topic, check whether the learner could have succeeded by framing the target correctly.
The decision card drill
Create cards with questions but hide the answer space. The learner must not solve. They may only write the decision sentence, identify the likely evidence or relationship, and name the first useful move. This isolates question framing from execution.
Use English, Mathematics and Science cards in one mixed set. The learner must switch decision types without a topic heading telling them what to do.
Twenty-four decision cases
English stated detail
Question type: What did Mira place under the desk?
Decision sentence: Decide which object the passage states Mira placed there.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
English inference
Question type: What does Ben’s repeated hesitation suggest?
Decision sentence: Decide which interpretation is best supported by the repeated behaviour.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
English motive
Question type: Why did Aisha apologise?
Decision sentence: Decide the reason or intention that explains the apology.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
English comparison
Question type: How did the two speakers respond differently?
Decision sentence: Decide the difference on one comparable dimension.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
English language-in-context
Question type: What does sharp mean here?
Decision sentence: Decide which contextual meaning fits the surrounding language.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
English evidence
Question type: Which detail shows that Ryan was reluctant?
Decision sentence: Decide which passage detail directly supports the claim of reluctance.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Mathematics percentage
Question type: Find the original price before discount.
Decision sentence: Decide the 100% base from the discounted amount.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Mathematics ratio
Question type: Find the number of red counters.
Decision sentence: Decide the value of the red share, usually through one ratio unit.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Mathematics fractions
Question type: Find the remainder after two removals.
Decision sentence: Decide the changing whole at each stage and the final remainder.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Mathematics speed
Question type: Find average speed for the whole journey.
Decision sentence: Decide total distance and total time before dividing.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Mathematics geometry
Question type: Find the shaded area.
Decision sentence: Decide which component areas combine to form the shaded region.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Mathematics data
Question type: Find the new average after one value changes.
Decision sentence: Decide the new total and unchanged count or the appropriate changed count.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Science describe
Question type: Describe what happened as light intensity increased.
Decision sentence: Decide the observed relationship shown by the data.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Science explain
Question type: Explain why the wrapped cup cooled more slowly.
Decision sentence: Decide the mechanism linking insulation to the observed temperature change.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Science variable
Question type: What was changed in the investigation?
Decision sentence: Decide which condition the experimenter deliberately varied.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Science controlled variable
Question type: State one factor kept the same.
Decision sentence: Decide which relevant condition was held constant across set-ups.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Science predict
Question type: Predict what happens if the surface area increases.
Decision sentence: Decide the expected outcome under the stated new condition.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Science infer
Question type: What can be inferred from the observation?
Decision sentence: Decide the most supported interpretation that goes one step beyond observation.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Science evaluate
Question type: Why is the test unfair?
Decision sentence: Decide which extra changing factor prevents a clean comparison.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Cross-subject MCQ
Question type: Which option is correct?
Decision sentence: Decide what property makes an option satisfy the actual task, not which option looks familiar.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Cross-subject two-part
Question type: State and explain.
Decision sentence: Decide both the required statement and the explanation; neither can replace the other.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Cross-subject checking
Question type: Which answer should you review first?
Decision sentence: Decide where a known risk or flagged uncertainty makes checking valuable.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Cross-subject recovery
Question type: What should you do when stuck?
Decision sentence: Decide whether the blockage is reading, knowledge, method, execution or checking.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Cross-subject final review
Question type: What must be true before you change an answer?
Decision sentence: Decide what new evidence, contradiction or method justifies the change.
The learner should then name one first move that could produce the evidence or relationship needed to settle that decision.
Decision drift
Decision drift occurs when the learner begins with one target and gradually solves something else. In English, an inference becomes a summary. In Mathematics, the value of one part replaces the requested total. In Science, a describe question becomes an explanation.
Prevent drift by checking the decision sentence at major transitions: after locating evidence, after finding an intermediate quantity, or before writing the final Science sentence.
The final-answer identity check
Before moving on, ask: Does this final answer actually settle the decision I named at the start? If yes, stop. If the answer is related but does not settle the target, one more step is needed.
Decision versus deliverable
The decision is what must be settled; the deliverable is the form the answer must take. A learner can understand the decision yet still lose marks by producing the wrong deliverable. In English, the decision may be which reaction is stronger, while the deliverable requires a comparison sentence using evidence from both characters. In Mathematics, the decision may be the original price, while the deliverable requires a numerical value with dollars. In Science, the decision may be which factor caused the observed change, while the deliverable requires an explanation linking condition, mechanism and outcome.
Before writing the final response, ask two separate questions: Have I settled the decision? and Have I delivered it in the form requested? This prevents a learner from solving mentally but writing an incomplete answer.
Ten fast decision cues
- Why → decide the reason or mechanism.
- How → decide the process, method or relationship.
- Compare → decide the difference or similarity on the same basis.
- Explain → decide the mechanism that accounts for the evidence.
- Find → decide the requested quantity, not an intermediate one.
- State → decide the direct result or fact required.
- Infer → decide the best-supported meaning beyond the observation.
- Predict → decide the expected outcome under the new stated condition.
- Evaluate → decide whether the method or claim is justified.
- Justify → decide which evidence or reasoning makes the answer defensible.
One more control: distinguish the decision from the route
Two learners can make the same decision using different routes. An English inference can be supported by two clues, a Mathematics target can be reached by a bar model or equation, and a Science explanation can be built from the same mechanism using different wording. The decision stays fixed even when the route changes. This matters during recovery: change the route if needed, but do not accidentally change the target.
When a method fails, repeat the decision sentence before trying another method. The sentence acts as an anchor. A new route is useful only if it still aims at the same job.
Decision framing under time pressure
The habit must become fast. Most questions need only a two- to five-second mental statement. Difficult questions may need the decision written in a few words. The aim is not to add ceremony; it is to prevent minutes of work on the wrong job.
A seven-day decision-framing cycle
- Day 1: English question jobs without answering.
- Day 2: Mathematics target quantities without calculation.
- Day 3: Science command words and decisions.
- Day 4: mixed EMS decision cards.
- Day 5: first-move pairing after each decision.
- Day 6: timed mixed set with decision statements only for hard questions.
- Day 7: review wrong answers and ask whether the first error was the decision itself.
What parents and tutors should ask
Ask: What are you actually deciding? What would a complete answer settle? Which information could decide it? Which information is merely related? Why is this your first move? These questions reveal whether the learner has a target before beginning work.
Common mistakes
- Topic instead of decision: saying ratio or heat without naming the actual target.
- Operation instead of decision: saying divide by 5 before explaining what quantity is being found.
- Evidence instead of decision: copying a passage line before knowing what the line must prove.
- Keyword instead of decision: recalling a Science term before identifying the command word.
- Over-framing: spending too long writing elaborate decision statements for easy questions.
Foundation recap: read before you solve
PSLE performance begins before the first answer is written. A learner can know the content and still lose marks because the question was read too quickly, the wrong task was selected, a useful piece of information was ignored, or the answer was never checked against what the question actually asked. This first learner’s guide builds the foundation beneath English, Mathematics and Science: read before you solve.
The goal is not to read slowly for the sake of being slow. The goal is to convert a question into a clear job. Once the job is clear, the learner can choose the right knowledge, method and answer form. This guide connects to the PSLE Learning Guide for English, Mathematics and Science, where each subject has its own deeper route.
READ → NAME THE JOB → FIND THE GIVEN INFORMATION → CHOOSE THE METHOD → ANSWER → CHECK AGAINST THE JOB.
The quick answer: what does “perform” mean in PSLE?
Performing well in PSLE is not the same as knowing many facts or completing many worksheets. In an examination, knowledge has to survive a chain of decisions. The learner must recognise what is being asked, retrieve the relevant knowledge, select a suitable method, carry it out accurately, communicate the answer in an assessable form and recover quickly when a question is difficult.
- English: understand meaning, identify the required response, use textual or task evidence, and communicate clearly.
- Mathematics: identify quantities and relationships, choose a valid representation or strategy, calculate accurately and verify.
- Science: identify the scientific job, use the evidence given, retrieve the relevant concept, connect condition to mechanism to outcome, and check that the explanation fits the case.
- Across all three: manage attention, time, error checking and recovery without letting one difficult item damage the next one.
The shared skill is therefore task control. A learner should know what the question is asking before spending effort on an answer.
Why capable learners sometimes answer the wrong question
A familiar topic creates a strong pull. The learner sees a word such as “percentage”, “character”, “heat”, “compare” or “explain”, remembers a method, and starts. Sometimes that works. Sometimes the remembered method belongs to a different task. Familiarity can create speed without control.
The safest response is not hesitation. It is a short launch routine. Before solving, ask three questions: What must I produce? What information controls the answer? What would count as a complete response?
1. What must I produce?
Look at the command and the answer space. Are you being asked to identify, explain, compare, calculate, infer, support, describe, write, transform or evaluate? The same topic can require different jobs.
2. What information controls the answer?
Mark the values, conditions, words, evidence, diagram labels, passage details or constraints that the answer must respect. Do not treat every detail as equally important. Find the information that changes the decision.
3. What would count as complete?
A correct idea can still be incomplete. A comparison may need both sides. A mathematical answer may need a unit. A Science explanation may need the mechanism linking the condition to the result. An English response may need support from the text rather than a plausible personal opinion.
A 20-second launch routine
- Read the final instruction first so you know the destination.
- Read the whole question from the beginning.
- Underline or mentally mark the command word and decisive conditions.
- Say the job in simple language: “I need to find…”, “I need to explain…”, or “I need to show…”
- Choose the smallest valid method that can complete that job.
- Only then begin the answer.
During practice, this routine may take twenty seconds. With repetition, much of it becomes automatic. The aim is not to add a ritual to every item. The aim is to train the mind to stop launching blindly.
Worked example: one habit, three subjects
English
Imagine a comprehension question asking why a character changes a decision after reading a note. A learner who recognises the scene may begin retelling the story. The actual job is narrower: identify the relevant change in thinking and support it with the passage. The useful question is not “What happened in the story?” but “What caused the decision to change, and what evidence shows that?”
Mathematics
Imagine a word problem giving an original amount, a percentage increase and a final comparison. A learner may immediately multiply by the percentage. Before calculating, identify which quantity is the base, whether the percentage describes the change or the final amount, and what the question ultimately asks for. The numbers are not yet the problem; the relationship between them is.
Science
Imagine a table showing two set-ups and asking which has a faster rate of change, followed by an explanation. The first job is to compare the evidence. The second is to explain the difference using the relevant concept. A remembered Science sentence should not replace the data. The data establish what happened; scientific knowledge helps explain why.
The same cross-subject habit appears each time: identify the job before producing the response.
The four most common launch errors
Launch error 1: answering the topic instead of the question
The learner recognises “fractions”, “character”, or “electric circuit” and writes everything remembered. Repair: state the job in one short sentence before solving.
Launch error 2: solving the first visible number
This is common in Mathematics. The learner treats the first available operation as the first required step. Repair: represent the relationship before computing. The dedicated Mathematics guide in this batch develops this in detail.
Launch error 3: using knowledge without evidence
This appears frequently in English comprehension and Science. The learner writes something true or plausible but does not anchor it to the passage, data, diagram or condition. Repair: point to the evidence that licenses the answer.
Launch error 4: finishing without returning to the question
A learner can work accurately and still answer a different quantity, use the wrong unit, explain only one half of a comparison or omit a required second reason. Repair: after writing, reread the final instruction and compare it with the answer.
Build the habit in practice before expecting it under time pressure
Do not wait until a full mock paper to train question launch. Full papers are expensive places to discover basic reading errors because many skills are changing at once. Use small sets first.
- Choose four questions from one subject.
- Before answering each one, write a five-to-ten-word job statement.
- Circle or underline the decisive information.
- Answer normally.
- After marking, classify any error as knowledge, reading, method, execution or checking.
- Repeat the set later without writing the job statement; see whether the launch habit survives mentally.
Once the routine is stable in one subject, mix English, Mathematics and Science tasks. The surface changes, but the control question remains: What is my job here?
How to know the habit is becoming independent
- The learner can explain what a question wants without the tutor paraphrasing it.
- The learner notices when two similar questions require different response types.
- The learner stops beginning calculations before identifying the target quantity.
- The learner can point to the evidence that supports an English or Science answer.
- The learner detects incomplete answers during checking.
- The learner can recover from a difficult question and launch the next one cleanly.
Independence matters because the examination removes the tutor’s prompts. Training should therefore remove prompts gradually too.
A simple checking loop
QUESTION → ANSWER → QUESTION AGAIN.
After answering, reread only the key instruction. Ask: Did I answer this exact thing? Did I use the correct unit or response form? Did I include every requested part? Is my claim supported by the information given? If not, repair the answer before moving on.
Checking is not redoing the whole question. It is comparing the output with the job.
Where to go next
Use Vol 0002: English — Answer the Actual Task when the difficulty is comprehension, writing, oral, listening or language-task control. Use Vol 0003: Mathematics — Represent Before You Calculate when the learner rushes into operations without making the relationship visible. Use Vol 0004: Science — Evidence Before Explanation when the learner knows Science but struggles to connect evidence, concept and mechanism.
For the wider map, return to the PSLE Learning Guide. The subject libraries are the English Tuition Learning Guide, Primary 6 Mathematics Learning Hub and PSLE Science Learning Guide.
Official examination references
Use the current examination-year documents as the authority for assessment structure and requirements. For 2026, consult the Singapore Examinations and Assessment Board documents for PSLE English Language, PSLE Mathematics and PSLE Science. School instructions and the correct examination-year documents should take priority over any generic study advice.
Independence indicators
- The learner can state the target in plain language.
- Topic labels are followed by a specific decision.
- First moves connect directly to the decision.
- Intermediate work does not replace the final target.
- English evidence is searched with a purpose.
- Science content is recalled only after the reasoning job is identified.
Next route
Continue to Vol 0036: English — Track the Writer’s Revision From First Impression to Final Meaning, Vol 0037: Mathematics — Track the Base Through Successive Percentage Changes, and Vol 0038: Science — Separate a Cause From a Necessary Condition.
Official PSLE reference
Use the current official SEAB PSLE page and PSLE Formats Examined in 2026. Official examination documents and school instructions take priority over generic study advice.
Series: How to Perform in PSLE | Learner’s Guide · Vol 0035 · Advanced cross-subject task control
