How to perform in the new G2 SEC examinations under pressure depends partly on what the learner says internally between one decision and the next. The useful language is not a motivational speech and not a running commentary on every feeling. It is a small set of short cues that return attention to the task: What is the job? What is given? What is the first valid move? What must I check?
This forty-first Learner’s Guide develops self-coaching under time across G2 English, Mathematics and Science. The aim is to turn good examination habits into compact internal prompts that remain usable when attention is tired. The learner should not need a tutor beside them asking the next question. They should be able to generate the next useful question themselves.
For 2027, SEAB lists G2 English Language as K200, Mathematics as K210 and G2 Science combinations as K223, K224 and K225 for school candidates. Use the SEAB 2027 G2 syllabus directory for current official details. The cues in this article are eduKateSengkang learning tools rather than official examination instructions.
Self-coaching is not positive thinking
Useful self-coaching changes an action. “I can do this” may be encouraging, but it does not tell the learner what to read, represent or verify. “Underline the command word” creates a concrete next move. Under time, operational language is more reliable than broad reassurance.
The learner can still use encouraging language if it helps, but the core system should remain behavioural. The purpose is not to feel perfect. The purpose is to keep decision quality high enough for the next mark.
The one-sentence rule
A self-coaching cue should usually fit into one short sentence. Long internal speeches compete with the same working memory needed for comprehension, algebra and scientific reasoning.
- English: “What exactly must this answer do?”
- Mathematics: “What relationship is hidden here?”
- Science: “What is the evidence, and what operation does the command require?”
- Whole paper: “If I am not progressing, mark it and move.”
Short cues are easier to rehearse, easier to recall and less likely to become another source of distraction.
Cue one: What is the job?
This is the most transferable cue in the system. It asks for the task before the content.
In English, the job may be infer, compare, summarise, persuade or respond to an audience. In Mathematics, it may be model, calculate, show, compare or decide feasibility. In Science, it may be describe, explain, predict, calculate, suggest or evaluate. Naming the job reduces the chance of producing correct content in the wrong answer form.
Cue two: What is given?
This cue prevents the learner from solving the problem they expected rather than the problem on the page.
English gives wording, source position, task details and audience. Mathematics gives quantities, units, constraints and relationships. Science gives observations, values, diagrams, conditions and experimental design. Before recalling everything you know, identify what this question actually supplies.
Cue three: What is missing?
A problem becomes easier when the unknown is named precisely. “I do not understand” is too broad. “I need the original price”, “I need the writer’s reason”, or “I need the dependent variable” reduces the problem to a target.
This cue is especially useful after a block. It converts a vague feeling of difficulty into a gap that can be filled.
Cue four: What is the first valid move?
Do not demand the whole solution immediately. Ask for one valid first move.
- locate the relevant lines;
- define the unknown;
- write the formula;
- draw the diagram;
- calculate the change;
- name the scientific concept;
- state the position before developing it.
A valid first move restores structure and often makes the second move obvious.
Cue five: What could make this answer wrong?
This is a targeted checking cue. It is better than rereading everything equally.
For a percentage problem, the risk may be the base. For a Science conclusion, it may be overclaiming beyond the data. For an English response, it may be missing a required point. The best check is chosen from the likely failure mode.
Cue six: Is this still the same question?
Long questions can drift. A learner begins by solving for one quantity and later continues calculating even after the final target has changed.
Return to the wording after an intermediate result. This cue is particularly useful in multi-part Mathematics and Science, and in English writing where an example can pull a paragraph away from its original job.
Cue seven: Am I making progress?
Activity is not the same as progress. Re-reading the same sentence, recalculating the same expression and writing increasingly vague material can consume time without changing the state of the problem.
Useful progress creates structure: a narrowed inference, a valid equation, an identified variable, a paragraph job, a correctly read graph or one eliminated option.
Cue eight: Change representation
When the same route has failed twice, change the form. Turn prose into a table, a Mathematics situation into a sketch, a Science context into evidence plus mechanism, or a long English sentence into shorter clauses.
The cue is not “try something random”. It is “show the same relationship in another form”.
Cue nine: Mark and return
Leaving a question is easier when the learner has a script. “Mark and return” treats the decision as procedure rather than defeat.
The return mark externalises memory. The learner does not need to keep rehearsing the skipped problem while answering later questions.
Cue ten: Next question clean
This cue belongs immediately after a difficult question. It means the next item gets a complete fresh read rather than an anxious half-read performed while mentally arguing with the previous problem.
The cue connects directly to Vol 0021: Recovery After Mistakes. Recovery succeeds when the difficulty remains local.
English cue set
- “What is the command?”
- “Which words in the text carry the answer?”
- “What does this evidence mean?”
- “How many distinct points are required?”
- “Who is the audience?”
- “What job does this paragraph do?”
- “Am I repeating or developing?”
- “Can I stop without losing meaning?”
These cues should not all run on every question. The learner chooses the one matching the task.
Mathematics cue set
- “What is the unknown?”
- “What relationship connects the quantities?”
- “Which representation makes it clearer?”
- “What constraint must the answer satisfy?”
- “What unit should I end with?”
- “Can I estimate the size?”
- “Can I substitute or reverse-check?”
- “Is this result feasible?”
The Mathematics cues turn silent habits into inspectable decisions. They are especially useful during mixed-topic practice where the method is not announced.
Science cue set
- “What is observed?”
- “What is the command word?”
- “Which evidence matters?”
- “What scientific model applies?”
- “What mechanism connects condition to outcome?”
- “What alternative explanation remains?”
- “How strong can the conclusion be?”
- “What improvement fixes this exact weakness?”
The Science cue set should stay tied to the learner’s actual G2 combination. Use the relevant Physics, Chemistry or Biology model rather than blending all three.
The cue hierarchy
Use the smallest cue that resolves the current problem. If task interpretation is wrong, there is little value in jumping ahead to grammar or calculation checks.
- Task cue.
- Evidence or quantity cue.
- Method or mechanism cue.
- Execution cue.
- Checking cue.
- Recovery cue.
This hierarchy protects attention by solving the earliest failure first.
Do not stack five cues at once
A learner can turn self-coaching into a new burden by asking every question on every item. That defeats the purpose.
The mature system is selective. One cue should unlock one next decision. If the answer becomes clear, continue working rather than running the entire checklist.
Self-coaching during reading
Reading cues should slow only the parts that deserve deliberate attention. A useful sequence is “command, condition, target”.
The learner identifies what must be done, what limits apply and what final information is needed. Then the rest of the wording can be read around that structure.
Self-coaching during a block
A block needs a different cue sequence: “What do I know? What is missing? Change representation. Move if needed.”
This prevents the learner from confusing a temporary failure to see the route with total absence of knowledge.
Self-coaching during checking
Checking begins with a risk, not a vague request for certainty. “Check the percentage base”, “check the pronoun”, “check the graph scale”, “check the unit”.
Specific checking is faster because it knows what kind of evidence would justify a change.
Bad cue: I always get this wrong
This sentence feels descriptive, but it gives the learner no action and turns one question into a statement about identity. Replace it with a local cue: “Which step usually breaks?” or “Check the sign after substitution.”
The replacement keeps the problem bounded. It also makes review possible because the learner can later inspect whether the nominated step really was the failure.
Bad cue: This question is impossible
A judgement such as “impossible” closes the search too early. Replace it with “What part is unfamiliar?” The context may be new while the underlying operation is familiar.
Science can hide a known mechanism inside an unfamiliar device. Mathematics can hide a standard relationship inside a new story. English can place a familiar inference demand inside a different topic. The cue should separate surface novelty from deep structure.
Bad cue: I need to write more
More writing is not necessarily more complete. Replace it with “What required meaning is still missing?”
This protects English from repetition, Science from keyword dumping and Mathematics from unnecessary explanation. The learner adds content only when the task remains incomplete.
Bad cue: Check everything
“Check everything” is too large for the final minutes. Replace it with a risk queue: unfinished items, marked uncertainty, high-risk transitions, units and final statements.
This turns checking from anxious rereading into prioritised evidence gathering.
Bad cue: Go faster
Speed without a defined target usually reduces reading quality. Replace it with “Protect the next decision” or “Use the shortest valid route.”
Fluency comes from recognition and stable methods. Panic-speed usually creates additional work through corrections.
Cue design rule one: name an observable action
A strong cue points to something the learner can do now: underline, define, compare, draw, calculate, state, check, mark, move or return.
Avoid cues built only from abstract qualities such as “be careful”, “focus” or “think harder”. Those qualities matter, but they become useful only when translated into behaviour.
Cue design rule two: attach the cue to a trigger
A cue becomes more reliable when the learner knows when to use it. “Check the unit” can be triggered after a conversion or final numerical answer. “Who says this?” can be triggered whenever two English sources are being compared.
This if–then structure reduces the need to remember a long checklist. The question itself activates the cue.
Cue design rule three: one failure, one cue
Do not use a general cue for a specific recurring error. If the learner repeatedly reverses percentage bases, the cue should mention the base. If pronouns become ambiguous, the cue should mention reference.
Specific cues can later be retired once the behaviour stabilises.
Cue design rule four: retire cues that are no longer needed
Self-coaching should become lighter as performance becomes more automatic. A learner who still recites eight steps for a routine calculation may be carrying a scaffold longer than necessary.
Keep cues for fragile transitions and unfamiliar problems. Let stable routine work run with less verbal interference.
Cue design rule five: preserve subject language
A cue should lead into the correct disciplinary model. Mathematics needs quantities, units and relationships. Science needs evidence, variables and mechanisms. English needs task, source, audience and meaning.
Generic phrases such as “use logic” are less useful than a subject-specific first move.
Self-coaching before the paper starts
Before starting, the learner does not need a long ritual. A compact plan can be enough: “Read instructions. Bank secure marks. Mark uncertainty. Protect checking time.”
The purpose is orientation, not emotional performance. Once the paper begins, attention should move to the actual task rather than continue monitoring how prepared the learner feels.
Self-coaching in the first five minutes
The early risk is rushing because the paper is fresh and adrenaline is high. Use a reading cue: “Complete question before calculation” or “Command before content.”
A controlled start establishes the quality standard for the rest of the paper.
Self-coaching in the middle of the paper
The middle risk is unnoticed pace drift. A learner can spend too long on medium questions because there is still plenty of paper left.
Use the cue “Is this producing marks?” This does not mean abandon any difficult item immediately. It asks whether another minute is likely to produce a valid step or whether the question should enter the return queue.
Self-coaching in the final fifteen minutes
The final risk is global panic. Replace it with a hierarchy: “Blank first. Incomplete second. Marked risk third. Polish last.”
This sequence protects opportunity. A missing answer is usually a more urgent target than stylistic improvement to a response already complete.
Self-coaching after discovering an error
Use “Local repair.” Find the first invalid point and correct only what depends on it.
This cue connects with Vol 0027: Mathematics Error Containment and generalises to English and Science. A wrong inference may require one sentence change, not a whole paragraph rewrite. A Science mechanism may be wrong while the observation remains valid.
Self-coaching after a blank
Use “Topic, evidence, first fact.” Identify the topic family, then write one fact or relationship that is definitely correct.
This can reactivate memory without forcing recall through panic. If the route remains absent, mark and move rather than filling the page with unsupported material.
Self-coaching after a difficult MCQ
Use “Why not the others?” When two options remain, identify the exact distinction between them.
This is more useful than changing because one option “looks familiar”. Elimination becomes evidence-based.
Self-coaching for English comprehension
A compact sequence is “command → evidence → meaning → stop”.
The learner first identifies the operation, then locates the relevant textual evidence, explains what it means and stops once the answer is complete. This protects both relevance and economy.
Self-coaching for English writing
Use “purpose → paragraph job → reader need”.
Before each paragraph, the learner knows what the paragraph must accomplish. If the paragraph begins to collect interesting but irrelevant detail, “reader need” returns attention to the communication task.
Self-coaching for English oral
Use “answer first → reason → example → extend”.
This keeps the spoken response direct while still allowing development. If the learner begins to circle the topic, “answer first” resets the interaction.
Self-coaching for Mathematics word problems
Use “target → quantities → relationship → constraint”.
This prevents calculator-first behaviour. The learner knows what is being found, what information matters, how the quantities connect and what conditions the final answer must satisfy.
Self-coaching for Mathematics algebra
Use “one valid change per line”.
This cue protects sign, factor and equation structure. It also creates checkpoints for later checking.
Self-coaching for Mathematics graphs
Use “axes → scale → feature → meaning”.
Graph errors often begin before calculation. This cue makes the learner read the representation before extracting an answer.
Self-coaching for Science structured questions
Use “discipline → command → evidence → operation → answer”.
This is the compact version of Vol 0032. It is especially useful when the context is unfamiliar but the syllabus principle is not.
Self-coaching for Science experiments
Use “weakness → consequence → fix”.
This prevents generic improvements. The learner names what is wrong, explains how it affects evidence and then proposes a change that addresses that exact problem.
Self-coaching for Science conclusions
Use “What does the data support?”
This cue protects the strength of the claim. It discourages universal conclusions from narrow evidence and discourages vague hesitation when the data clearly supports a bounded relationship.
The cue card
During revision, a learner may keep a small cue card containing only the highest-value prompts. A useful first version might contain six:
- What is the job?
- What is given?
- First valid move.
- Is this progress?
- What could make it wrong?
- Mark and return.
Subject-specific cues can be added below only if they target repeated errors.
The cue card should shrink
The goal is not to carry a longer and longer card. Each month, remove cues that have become automatic and keep the ones still preventing real errors.
A smaller card is evidence that the learner is internalising the process.
The 10-minute cue drill
- Choose five mixed questions.
- Before each, say or write one cue only.
- Attempt the question.
- Afterwards, decide whether the cue helped.
- Replace any cue that was too vague.
This trains cue selection rather than rote recitation.
The wrong-cue drill
Give the learner a difficult question and offer three possible cues. Choose which cue addresses the earliest real failure.
For example, if the learner misread “at least”, “check arithmetic” is too late. The better cue is “mark the constraint”.
The cue-to-action drill
During practice, the tutor says a cue and the learner must perform the action immediately without further explanation.
If “check the unit” produces a general reread rather than an actual dimensional check, the cue has not yet been operationalised.
The fade-the-prompt drill
At first, a tutor may supply the cue. Later, the tutor asks, “What cue do you need?” Finally, the learner generates the cue without prompting.
This gradual fading is crucial. A cue that works only when someone else says it has not become self-coaching.
The post-paper cue audit
After a timed paper, identify the moments where a cue could have prevented an error. Do not create a cue for every individual mistake.
Look for recurring triggers: command-word drift, sign changes, overstrong Science conclusions, ambiguous pronouns, missed constraints or time traps. One high-value cue can protect many future marks.
Parents and tutors: ask for the cue, not the answer
When a learner is stuck during practice, ask “What is the next useful question you can ask yourself?” rather than immediately giving the method.
This keeps support focused on independence. The learner practises retrieving the process that will be needed when no adult is present.
When self-coaching becomes self-interruption
Some learners can over-monitor every step and slow themselves down. Signs include repeating the cue after the action is already complete, checking low-risk work repeatedly and feeling unable to proceed until every checklist item has been consciously named.
The repair is simplification. Use cues only at transition points or known risks. Automatic skills should be allowed to remain automatic.
When self-coaching becomes self-criticism
If cues become judgemental—“Don’t be stupid”, “Don’t mess up again”—they are no longer serving the task. Replace them with neutral operational language.
“Check the copied number” identifies a behaviour. It can be tested and improved.
The four-week self-coaching build
Week 1 — identify triggers
Review recent papers and find the recurring moments where decisions fail. Write one short cue for each major trigger.
Week 2 — practise cue to action
Use short mixed sets. The learner chooses and applies one cue before each high-risk item.
Week 3 — fade external prompting
The tutor or parent stops supplying the cue. The learner identifies which prompt is needed.
Week 4 — timed transfer
Use longer sections. Audit whether the cues remain short, selective and useful under fatigue.
Use the existing G2 performance system
Use Vol 0009 for attempt, skip, return and check decisions, Vol 0025 for mark security and Vol 0040 for the most recent Science practice clinic.
Use the subject hubs when the cue reveals a concept gap
A cue can reveal that the learner does not know the underlying content. At that point, stop treating the problem as an examination-control issue.
Use the Secondary English Learning Hub, Complete Mathematics Index or Complete Science Index to repair the underlying skill.
The PSLE bridge
The earlier habit Read Before You Solve can itself become a self-coaching cue. G2 adds more independent decisions, so the cue system becomes more specialised.
The principle remains the same: pause only long enough to make the next decision better.
Use Examination Craft for whole-paper control
For pacing, return queues and final checks, continue through the Examination Craft hub. Self-coaching works inside that larger paper-management system.
Final rule
The best self-coaching sentence is not the most inspiring one. It is the one that changes the next action.
Ask what the job is. Identify what is given. Choose the first valid move. Check the likely failure. Mark and return when progress disappears. Keep the cue short enough that the Mathematics, English or Science still has room to think.
Calibrate cues by error family
A cue should be judged by whether it changes the repeated error, not by whether it sounds sensible. Review several papers and group errors into families. Then test whether one cue reduces that family across different questions.
- reading error → “mark the condition”;
- task mismatch → “name the job”;
- algebra sign error → “bracket negatives”;
- unit error → “unit before box”;
- overstrong Science conclusion → “what does the evidence support?”;
- English repetition → “new meaning or stop?”;
- time trap → “is this producing progress?”
If the error persists, change the cue or repair the underlying concept. Do not keep repeating an ineffective phrase simply because it has become familiar.
Calibrate cues by paper phase
The same learner may need different cues at different stages of a paper. Early, the main risk may be rushing. In the middle, it may be time drift. Late, it may be global panic or excessive checking.
Build one cue for each phase rather than carrying a long list all the way through.
Opening cue
“Read the whole task.” This establishes the quality standard before speed takes over.
Middle cue
“Progress or return.” This keeps medium questions from expanding into time traps.
Closing cue
“Blank, incomplete, high-risk.” This gives the final minutes a priority order.
Calibrate cues by confidence pattern
Some learners under-trust good work. Others over-trust the first answer. The cue should compensate for the actual pattern.
For a chronic second-guesser: “Change only with evidence.” For a learner who refuses to revisit mistakes: “What check could prove this wrong?” The same cue should not be given to both students.
Calibrate cues by subject transition
When switching from English to Mathematics or Mathematics to Science during revision, use a discipline reset. “New subject, new model.”
This cue tells the learner not to carry the previous subject’s dominant habits into the next task. The reset is especially useful in examination weeks where several subjects are sat close together.
Calibrate cues by question length
Short questions need light cues. Long structured questions may justify a slightly richer sequence. A one-mark item should not trigger a seven-step routine.
For short work, “job → answer → check” may be enough. For long work, “target → evidence → representation → method → interpretation” can protect the structure.
When a cue reveals a memory problem
If the learner can name the correct cue but cannot retrieve the required fact or method, the issue is not self-coaching anymore. Stop adding process prompts.
Return to retrieval practice, spaced review and concept teaching. A process cue cannot manufacture content that was never learned or has been forgotten.
When a cue reveals a recognition problem
If the learner knows the method once prompted but cannot recognise when it applies, use mixed practice. The cue may be “Which relationship is this?” rather than “Use percentage change.”
The first version asks the learner to classify. The second gives away the classification.
When a cue reveals an execution problem
If recognition is correct but the working fails, the cue should target the fragile transition: sign, bracket, unit, scale, copied value or sentence boundary.
Execution cues work best when placed immediately before the usual failure point.
When a cue reveals an interpretation problem
A learner may calculate correctly and still answer the wrong practical question. Use “return to context” or “what does this number mean?”
This is especially useful in Mathematics constraints and Science conclusions.
When a cue reveals a communication problem
The thinking may be correct but invisible. Use “show the bridge” in Mathematics, “state the evidence” in Science or “answer before expanding” in English.
The goal is not longer answers. It is visible reasoning where the task requires it.
The cue fading ladder
- Stage 1: tutor supplies cue and learner acts.
- Stage 2: tutor asks learner which cue fits.
- Stage 3: learner names cue before acting.
- Stage 4: learner acts without naming cue aloud.
- Stage 5: cue appears only when the high-risk trigger occurs.
A successful cue eventually becomes part of the performance system rather than a separate sentence that must always be spoken.
The over-cue warning
Too many cues can fragment attention. A learner who pauses after every line to ask a checklist question may lose fluency and working memory.
If accuracy is stable, reduce verbal monitoring. Preserve cues only for decisions that remain fragile or high-risk.
The cue replacement test
Every two or three weeks of serious practice, review the cue list. Remove any cue that no longer prevents a real error. Rewrite any cue that produces vague action.
A good system becomes shorter as the learner becomes stronger.
The cue-transfer test
A cue has transferred when it works on a different topic. “What is the base?” should help in discount, increase, reverse percentage and comparison contexts. “What does the evidence support?” should help in graphs, experiments and unfamiliar contexts.
If the cue works only on the exact worksheet where it was learned, it remains too tied to surface features.
The cue-stress test
Use the cue during a timed mixed set after a difficult question. Can the learner still retrieve and apply it without tutor prompting?
A cue that disappears under mild pressure needs more rehearsal or further simplification.
The cue-success metric
Do not measure cue success by how often it is used. Measure whether the target error decreases and whether the learner needs less external prompting.
The strongest outcome is often that the cue becomes almost invisible while the protected behaviour remains.
A sample personal cue set
A learner whose recurring errors are missed conditions, algebra signs, English repetition and overstrong Science conclusions might keep only four:
- “Mark the condition.”
- “Bracket negatives.”
- “New meaning or stop?”
- “What does the evidence support?”
This set is more useful than a generic page of twenty examination tips because every cue corresponds to observed performance.
Final calibration
Self-coaching should eventually feel like part of solving rather than a separate activity. The learner sees “at least” and automatically marks a boundary. They see a graph and automatically check axes and scale. They reach a difficult question and automatically ask for the first valid move.
That is the target: not more internal language, but better internal control. The cue earns its place only when it protects the next decision and then gets out of the way.
One final self-coaching test
At the end of a timed practice, identify one moment where your internal cue changed the next action. If “mark the condition” prevented a boundary error, keep it. If “be careful” produced no specific behaviour, replace it. The best cue should be visible in what you did differently, not merely in what you remember saying to yourself.
Over time, the cue should shorten while the protected behaviour remains. That is the sign that self-coaching has become genuine examination control rather than another checklist to carry.
