PSLE Stimulus-Based Conversation Performance

SETC-PSLE-ORAL-030

Stimulus-Based Conversation is difficult for a simple reason: the learner cannot know in advance exactly what the examiner will ask, which part of the photograph will matter most, what angle the follow-up will introduce or which personal experience will be most useful. The candidate must therefore do something memorised scripts cannot do well: listen, decide, build and adapt in real time.

This guide treats PSLE Stimulus-Based Conversation as an integrated examination-performance system. The dedicated eduKateSG micro-skill owner, Stimulus-Based Conversation — Turn the Photo into a Conversation, Not a Memorised Speech, remains the canonical owner for the narrow mechanism of using the stimulus and responding conversationally. SETC-PSLE-ORAL-030 owns the larger performance layer: preparation, question-job recognition, evidence discipline, answer development, retrieval, interaction, follow-up adaptation, recovery, unfamiliar-topic transfer and stability under examination pressure.

Official examination format facts should always be checked against current SEAB PSLE English Language materials. The teaching principles here use the PSLE oral environment as an anchor while remaining useful for interviews, viva-style tasks, presentations with questions, seminar discussion and other live communication situations.

The central rule: the question controls the answer

The photograph is important, but it does not decide the answer by itself. The examiner’s question defines the job. A learner can describe every visible object and still fail to answer an opinion question. A learner can tell an excellent personal story and still miss a question asking what can reasonably be inferred from the image.

SETC therefore trains a simple operating sequence: Hear → Classify → Answer → Develop → Connect → Release.

Hear the full question before beginning. Classify what the question asks you to do. Answer the core job early. Develop with a reason, example, consequence, contrast or qualification. Connect to the stimulus, personal experience or wider situation when useful. Release the turn when the answer is complete instead of continuing until it drifts.

Maren notices structure in this sequence. Iona notices evidence and scope. Leonie notices timing, interaction and recovery. All three lenses matter because a strong answer can still fail through a weak opening, unsupported inference or poor follow-up handling.

Why memorised scripts feel safe and become dangerous

A script reduces uncertainty. The learner knows how to begin, which vocabulary to use and what example comes next. On a familiar topic, this can sound polished. The weakness appears when the question is narrower or wider than the script. The learner spends attention trying to force the new task into stored language.

Scripts also create interaction problems. A follow-up that changes the condition feels like interruption because the learner is protecting a prepared route. The examiner becomes someone to get past rather than someone to converse with.

SETC uses prepared language differently. Learners practise portable functions: stating a position, describing evidence, giving a reason, introducing an example, comparing, qualifying, acknowledging uncertainty, changing a view and concluding. Functions can be recombined. Scripts are brittle; functions are modular.

The question-job map

Topic and question job are different dimensions. The same photograph can support several jobs. A photograph of a community garden might lead to observation, personal experience, opinion, recommendation, comparison, fairness, responsibility, environmental effect or future consequence. Learners who prepare only by topic are therefore under-preparing the thinking demand.

Common jobs include:

Observation: what is visibly happening?
Inference: what might be happening, based on evidence?
Experience: what relevant event has happened to you?
Preference: would you choose or enjoy this?
Opinion: what do you think and why?
Explanation: why might this happen or matter?
Comparison: how are two choices different?
Recommendation: what should someone do?
Evaluation: what works well or poorly?
Consequence: what might happen next?
Perspective: how might another person see it?
Condition: would your answer change if a new factor appeared?

Classification should become fast enough that the learner does not need to name the category aloud. Training uses labels to build awareness; examination performance uses the awareness without the label.

The direct answer

Strong responses usually make their relevance visible early. This does not mean every answer must begin with “Yes” or “No”. It means the examiner should not have to wait forty seconds to discover the learner’s position or experience.

A direct answer could be: “I would join the activity because I enjoy practical projects with other people.” It could be: “The photograph suggests that the event is organised for families, because adults and children appear to be participating together.” It could be: “The main problem is accessibility; some people may not be able to use those stairs easily.”

The exact wording changes. The structural principle does not: establish the answer job before adding background.

Development is relationship, not length

A long answer is not automatically developed. Development occurs when each added sentence changes the idea. Useful moves include cause, example, consequence, contrast, qualification, condition, stakeholder perspective and connection to experience.

Maren uses a “new information test”. Before adding a sentence, she asks whether it contributes a relationship not already present. If not, the sentence may be repetition.

Iona uses a “scope test”. She asks whether the new detail still answers the question. An interesting fact can still be irrelevant.

Leonie uses an “airtime test”. She asks whether the extra development can be delivered clearly without making the answer harder to manage than the value it adds.

The answer-development ladder

A thin answer can be expanded in stages.

Position: state the core answer.
Reason: explain why.
Example: make the reason concrete.
Consequence: show why the example matters.
Qualification: add a condition, limitation or exception if useful.
Release: stop before repetition begins.

This is not a compulsory template. Some questions need only two or three moves. The ladder exists to show that depth comes from relationships rather than word count.

The photograph: evidence before story

A photograph freezes one moment. Learners often fill the missing story automatically. That is natural human cognition, but examination answers should distinguish observation from inference.

“Two people are standing beside a bicycle” is observation. “They may be checking whether something is wrong with it” is an inference supported by where they are looking. “They are siblings arguing because one broke the bicycle” is a much stronger story that the image may not justify.

Iona trains three categories: I can see, I can reasonably infer, I cannot know from this image alone. The categories disappear during fluent performance, but the evidence discipline remains.

Evidence language

Useful language for uncertain interpretation includes “appears to”, “seems to”, “may be”, “could suggest”, “from the photograph, I can see…”, and “I would infer this because…”. These are not phrases to force into every answer. They help learners represent uncertainty accurately.

Epistemic control—knowing how certain a claim is—makes conversation stronger because it prevents confident invention. It also transfers to comprehension, science and critical reasoning.

Personal experience without invented drama

When asked for personal experience, learners sometimes search for an impressive story. This creates retrieval pressure. Ordinary experiences are easier to remember truthfully and often easier to explain well.

A useful experience bank includes helping someone, changing a plan, learning a new skill, waiting for something, making a mistake, joining a group, disagreeing politely, taking responsibility, losing an item, choosing between options, receiving advice, trying again after failure and noticing a problem in a public place.

Store each experience as cues, not sentences: situation, decision, result, reflection. A class project may support questions about teamwork, planning, responsibility, communication or learning. The memory is reusable because the lens changes, not because the wording is memorised.

What to do when there is no personal example

Do not invent a false event. The learner can say that the situation has not happened directly, then use a related experience, something observed, what they would probably do or a general example. Honesty does not make an answer weak; inability to develop the answer does.

A useful transition is: “I have not experienced that exact situation, but I have had something similar happen when…” or “I have not done that myself, but I think I would…” The wording should remain natural rather than memorised.

Opinion questions: strength through qualification

Children are sometimes trained to take extreme positions because they seem easier to explain. Real reasoning often benefits from conditions. “I think it is useful, especially when…, although…” shows the learner can hold a position and recognise limits.

A compact opinion path is Position → Reason → Example → Consequence → Condition. Again, not every answer needs every step. The condition becomes especially useful when the examiner adds a follow-up.

Recommendation questions

A recommendation is stronger when the learner identifies a problem, proposes an action and explains who benefits or what risk it reduces. “They should add more signs” is thin. “They should add clearer signs near the entrance because visitors may not know where the activity begins, especially if the venue has several rooms” is more useful.

Recommendations can also include trade-offs. A solution may cost money, take space or inconvenience another group. Recognising those constraints can make the answer more realistic.

Comparison questions

Comparison is not two descriptions placed next to each other. It requires a criterion. Compare cost, convenience, safety, learning value, accessibility, environmental impact, enjoyment or another dimension relevant to the question.

Maren asks, “What exactly am I comparing?” Iona asks, “What evidence supports the difference?” Leonie asks, “Can I explain the difference in one clean sentence before adding detail?”

Perspective questions

Perspective asks the learner to consider how another person may experience the same situation. Avoid mind-reading. Use plausible factors: age, responsibility, access, cost, experience, goals and risk.

“A younger child might need more guidance because the instructions contain several steps” is reasoned. “The child is scared because children always dislike complicated instructions” overclaims.

Follow-up questions are new tasks

The examiner may introduce a new condition, challenge an assumption or ask for an example. The learner must reset. A follow-up is not evidence that the first answer failed.

The general SETC owner Oral Exam Follow-Up Questions | How to Listen, Reset and Answer the New Angle develops this mechanism broadly. In the PSLE lane, follow-up training is integrated into stimulus-based conversation performance.

The follow-up reset

A useful internal sequence is: What changed? Does my position change? What new reason is now relevant?

If the examiner adds cost, discuss cost. If the examiner asks how an older person might feel, shift perspective. If new evidence contradicts the first assumption, update the answer. Do not simply repeat the original point more loudly.

Answer release: know when to stop

Learners who fear silence often keep speaking after the answer is complete. The result can be repetition, contradiction or drift. Strong conversation includes turn release.

Train a completion question: have I answered the job and developed it enough for this turn? If yes, stop naturally. The examiner can continue the interaction.

Release is especially important for learners like Maren, who can keep organising more detail indefinitely. Conversation is not an essay spoken without paragraph breaks.

Fluency without rushing

Fluency is controlled continuity, not maximum speed. A brief pause for thought can make an answer clearer than a stream of fillers. The general SETC fluency owner English Speaking Fluency | How to Keep Meaning Moving Without Rushing provides the broader mechanism.

For stimulus-based conversation, train short intentional pauses before difficult questions, examples or qualifications. Silence becomes a tool rather than evidence of failure.

Vocabulary retrieval

Spoken vocabulary must be available at speaking speed. If a sophisticated word fails to arrive, the learner should preserve meaning through paraphrase. “Accessible” can become “easy for different people to reach or use”. “Sustainable” can become “able to continue without using too many resources”.

Train both routes: precise term and plain explanation. This makes the lexicon resilient.

Grammar under live conditions

Grammar matters because it carries relationships, but conversation does not allow written editing. Focus on high-value spoken structures: complete clauses, stable tense, pronoun clarity, cause, comparison, condition and qualification.

Do not interrupt every error during the first attempt. Record patterns, choose one or two high-cost recurring issues, practise them briefly and retest on a fresh question. Constant interruption can make the learner monitor grammar so intensely that thought stops moving.

Recovery when the sentence goes wrong

A false start is normal speech. Train a clean restart: “Let me put that more clearly…” followed by a simpler sentence. Do not apologise repeatedly or explain the error.

If the idea itself is weak, return to the core answer and add one concrete reason or example. If the question was misunderstood, acknowledge the corrected angle and answer it. Recovery should be brief and forward-moving.

Unfamiliar topics

Unfamiliarity should not force invention. Use a foothold sequence: identify what is visible, identify who may be affected, name one likely benefit or problem, consider one alternative or condition, then form a cautious judgement.

This is where broad world knowledge helps. Reading about community systems, technology, environment, science, transport, museums, education and everyday design gives learners more concepts. But SETC does not turn the oral lane into duplicate subject articles. The wider eduKate ecosystem supplies knowledge; this page owns how that knowledge is retrieved and used in live conversation.

The first-weak-link diagnostic

Observed problem	Possible earliest cause	First repair
Long photo description before answer	Question-job control	Direct-answer opening drills
Relevant but one-sentence response	Development not automatic	One relationship after every answer
Long but repetitive response	Weak answer release	Sentence labelling and compression
Confident invented photo claims	Evidence boundary weakness	Observation/inference/unknown split
Good first answer, weak follow-up	Interaction reset weakness	Changed-condition drills
Pauses while searching for advanced word	Retrieval fragility	Paraphrase training
No personal example	Retrieval strategy too narrow	Ordinary experience bank
Strong familiar-topic performance only	Overfitting	Unfamiliar-topic reasoning practice

The conversation performance floor

A stable floor does not require brilliant vocabulary or unusually long answers. On fresh tasks, the learner usually hears the question, identifies the job, gives a relevant answer, develops at least one useful relationship, distinguishes evidence from inference, responds to follow-up and recovers from a minor difficulty. Once that floor is stable, finer improvements become worth pursuing.

Lane connections

Return to SETC-PSLE-ORAL-000 — PSLE English Oral Examination Performance and SETC-PSLE-ORAL-010 — PSLE English Oral Training. Use SETC-PSLE-ORAL-020 — PSLE Reading Aloud Performance for the other Paper 4 component. Continue to SETC-PSLE-ORAL-040 — PSLE English Oral Practice for diagnostics, feedback and mock design.

Sources and boundary

Official format information should be verified from current SEAB PSLE English Language materials. All answer models, diagnostic systems, case examples, drills and prompt families on this page are original SETC teaching material. They are not official questions, examiner scripts, official marking rubrics or score guarantees. This page owns integrated Stimulus-Based Conversation examination performance and transfer; narrow mechanism pages remain with their canonical eduKate owners.

The Stimulus-Based Conversation casebook: thirty recurring performance problems

Conversation weaknesses often look alike from a distance. A hesitant learner may lack an idea, may be searching for a memorised opening, may have misunderstood the question or may simply fear a short pause. The cases below separate those causes so practice can target the system rather than the surface symptom.

Case 1: the photograph narrator

The learner begins every answer by describing the entire image. If the examiner asks for an opinion, the response remains indirect for too long. The repair is a direct-answer constraint: for ten questions, the first sentence must answer the job before any photo detail is added. The stimulus returns only when it supports the answer.

Case 2: the answer-without-photo learner

The opposite pattern ignores the stimulus even when the question clearly invites evidence from it. The learner gives a generic speech about the topic. Training requires one observable detail or cautious inference to anchor the response before moving wider.

Case 3: the mind-reader

The learner assigns emotions, motives and relationships to people in the image without evidence. The repair is the three-column method: visible, inferable, unknowable. During transfer, the labels disappear but the boundary remains.

Case 4: the one-sentence thinker

The learner gives a correct answer and stops. Knowledge is present; development is not automatic. Use one relationship cue—reason, example, consequence or qualification—after every direct answer until the habit becomes independent.

Case 5: the endless speaker

The learner believes duration equals quality. Answers continue after the idea is complete and begin to repeat. Record one response, label each sentence by function and remove repetition. Practise release after the final useful relationship.

Case 6: the memorised introduction

A polished general opening appears before every topic: “In today’s modern society…” or another fixed phrase. The first ten seconds sound prepared but often fail to answer the task. Ban generic openings temporarily. Rebuild with direct relevance.

Case 7: the framework prisoner

One answer framework is applied to observation, experience, opinion and recommendation questions. Structure looks tidy but task accuracy is weak. Sort questions by job and practise different openings. Frameworks become tools rather than cages.

Case 8: the topic-bank specialist

The learner performs well on rehearsed themes and poorly elsewhere. Reduce exact-topic prediction. Train stakeholders, causes, benefits, risks, trade-offs, alternatives and evidence as general reasoning moves.

Case 9: the advanced-word hunter

Speech stops while the learner searches for a sophisticated term. Train a dual route: precise word if available, plain-language paraphrase if not. Meaning must keep moving.

Case 10: the safe-language minimalist

The learner avoids all complexity to reduce grammar risk. Answers are clear but conceptually thin. Build reliable spoken structures for cause, contrast and condition so the learner can express richer ideas without grammatical overload.

Case 11: the grammar monitor

The learner restarts every sentence after noticing a small error. The repair is selective correction. Continue unless meaning is compromised; review recurring patterns after the turn. Oral communication requires deciding which errors deserve immediate attention.

Case 12: the no-example freeze

An opinion is available but no example comes to mind. Train three example sources: personal experience, observed experience and plausible hypothetical case. The learner does not need a dramatic story.

Case 13: the favourite-story learner

One polished personal story is forced into many questions. Ban the story for a period and build three new ordinary experiences. Practise one experience through different lenses rather than one story for every theme.

Case 14: the weak-return storyteller

The personal anecdote is fluent but never reconnects to the question. Add a return sentence: “That experience taught me…” or an equivalent natural conclusion that makes relevance explicit.

Case 15: the yes-no minimalist

The learner answers directly but never moves beyond the position. Use a two-move minimum during training: answer plus reason. Once stable, add examples and qualifications selectively.

Case 16: the five-reason lister

The learner produces many shallow reasons. Limit the response to one main reason and require depth. Later, additional reasons can be added only if they contribute something genuinely new.

Case 17: the unsupported recommendation

“They should add more signs.” “They should organise more activities.” The proposal lacks a problem and beneficiary. Train Problem → Action → Benefit, then add a constraint when useful.

Case 18: the comparison without criterion

The learner describes Option A and Option B separately without stating the dimension of comparison. Require the criterion first: cheaper, safer, more convenient, more inclusive, more educational or another relevant axis.

Case 19: the first-follow-up collapse

The opening answer is strong, but the examiner changes one condition and the learner repeats the original response. Train follow-up resets with cost, time, fairness, safety, access and perspective changes.

Case 20: the challenge panic

When a follow-up appears to disagree, the learner assumes the first answer was wrong. Practise acknowledging the challenge: “That would change my view slightly because…” or “I would still keep my view, although…”. A conversation can evolve.

Case 21: the premature answer

The learner begins before the examiner finishes and misses a condition near the end. Train a one-beat listening rule. The pause is processing, not hesitation.

Case 22: the question repeater

The learner repeats most of the question to buy time. Replace repetition with a short thinking pause and a direct answer. Natural reformulation is fine; mechanical echo wastes airtime.

Case 23: the filler stream

Every pause is filled with “um”, “like” or another sound. Practise intentional silence before examples and follow-ups. A one-second pause can sound more controlled than uninterrupted filler.

Case 24: the uncertainty inventor

When the topic is unfamiliar, the learner invents facts rather than admit uncertainty. Train epistemic language and evidence-based reasoning. “I am not sure of the exact reason, but from the photograph…” is stronger than fabricated certainty.

Case 25: the uncertainty retreat

The opposite learner gives up as soon as specialist knowledge is missing. Use the foothold sequence: visible evidence, affected people, likely benefit or problem, one alternative, cautious judgement.

Case 26: the examiner-cue dependent learner

The learner waits for nods, smiles or “why?” prompts before developing. Use neutral listeners and delayed prompting. Track whether development appears independently.

Case 27: the correction cascade

A grammar or vocabulary mistake causes apology, restart and loss of the original point. Train local recovery: simplify the sentence, restate the core answer and continue.

Case 28: the first-answer shadow

A weak first answer affects the next two questions because the learner is still evaluating it internally. Train a reset between turns: hear the new question, identify the new job, begin again.

Case 29: the ordinary-day gap

Conversation is strong only when the learner is fresh and heavily prepared. Later training should include some ordinary-day sessions, not to create fatigue but to test whether the core system survives normal conditions.

Case 30: the strategy overload

The learner knows dozens of frameworks and acronyms. Under pressure, choosing a framework becomes harder than answering. Near the examination, compress the active system: listen, answer, develop, adapt, recover.

One hundred and twenty original Stimulus-Based Conversation training prompts

These prompts are original SETC teaching material. They are not official PSLE questions, past-year questions or predictions. They are grouped by thinking job so teachers can vary the surface while keeping the cognitive demand visible.

Observation and evidence

Look at a photograph of a shared workspace. Name five details you can see without interpreting intention.
Identify the detail that seems most important to understanding the situation.
Describe two relationships between objects or people that are directly visible.
State one thing that appears organised and explain what visual detail supports that view.
State one thing that may create difficulty, using only visible evidence.
Separate what is in the foreground from what is in the background and explain why that matters.
Identify an action that appears to be happening and the evidence for it.
Describe one sign, tool or object and explain its likely role without overclaiming.
Identify one detail another viewer might overlook.
State two things the photograph cannot tell you.

Inference and uncertainty

Make one cautious inference from posture or positioning and cite the evidence.
Give two possible explanations for the same visible detail.
State which explanation is better supported and why.
Identify a claim you should not make because the image lacks evidence.
Use “appears to” naturally in one inference.
Use “may be” naturally in one inference.
Explain what additional information would make an inference more certain.
Revise an inference after a new visible detail is pointed out.
Distinguish a plausible inference from a creative story.
Explain why uncertainty can still support a useful answer.

Personal experience

Describe a time you helped someone complete a task.
Describe a time you changed a plan because new information appeared.
Describe a time you joined an activity you were unsure about.
Describe a time you decided not to join something and why.
Describe a time you made a mistake and corrected it.
Describe a time you waited longer than expected and how you handled it.
Describe a time you worked with someone whose ideas differed from yours.
Describe a time a rule seemed inconvenient but useful.
Describe a time you learned a practical skill.
Describe a time you noticed a problem in a public place.
Describe a time you gave or received advice.
Describe a time you had to choose between convenience and responsibility.
Describe a time you used technology to solve a problem.
Describe a time technology made a task more difficult.
Describe a time you helped organise something.
Describe a time you had to communicate clearly because another person misunderstood.
Describe a time you changed your opinion after hearing another person.
Describe a time you had no direct experience but learned by observing others.
Describe a small failure that taught you something useful.
Describe an ordinary event that became memorable because of what you learned.

Opinion and qualification

Should public spaces provide more quiet areas?
Should schools encourage pupils to take part in community projects?
Is convenience always the most important factor when choosing a service?
Should more activities be free if possible?
Is it better to repair an item or replace it?
Should children have more opportunities to make decisions for themselves?
Are rules more effective when people understand the reasons behind them?
Should technology be used for every task it can make faster?
Is group work always better than working alone?
Should events try to include people of different ages?
Is a popular activity necessarily a useful one?
Should people give up convenience to reduce waste?
Is it better to plan carefully or adapt as you go?
Should mistakes be discussed openly when others can learn from them?
Is competition helpful in learning?
Should communities share more resources instead of everyone owning the same items?
Is it important for children to learn practical skills outside school subjects?
Should public information be simplified even if some detail is lost?
Is volunteering valuable even when the impact is small?
Should a good idea still be changed if some people cannot access it easily?

Recommendation and design

How would you improve a crowded noticeboard?
How would you make a community event easier for first-time visitors?
How would you improve a shared study area?
How would you make a library activity more welcoming to younger children?
How would you reduce confusion at a busy registration desk?
How would you encourage people to return borrowed items?
How would you make a recycling point easier to use correctly?
How would you improve a volunteer sign-up process?
How would you make a public space easier for people with different mobility needs?
How would you make an event safer without making it unpleasant?
How would you reduce waiting time fairly?
How would you help visitors understand a new rule?
How would you encourage people to try an unfamiliar activity?
How would you improve communication when plans change suddenly?
How would you design a shared resource so people take responsibility for it?
How would you make a school exhibition easier to navigate?
How would you reduce waste at a large event?
How would you make a digital system easier for people who are less confident with technology?
How would you balance quiet and social use in a common area?
How would you improve a rule that is useful but difficult to follow?

Comparison and trade-offs

Compare a quiet reading room with an open collaborative space.
Compare borrowing an item with buying it.
Compare online registration with in-person registration.
Compare a large event with a small event.
Compare volunteering once with volunteering regularly.
Compare walking a short distance with taking transport.
Compare repairing something yourself with asking an expert.
Compare printed instructions with video instructions.
Compare strict rules with flexible guidelines.
Compare learning alone with learning in a group.
Compare a free service with a paid service that offers more support.
Compare a centralised shared facility with several smaller local facilities.
Compare speed and accuracy when a task is urgent.
Compare convenience and environmental responsibility.
Compare tradition with a new method.
Compare a familiar activity with an unfamiliar one.
Compare giving people more choices with keeping a system simple.
Compare solving a problem immediately with studying it first.
Compare individual responsibility with organisation-level responsibility.
Compare an attractive design with an accessible design when the two conflict.

Follow-up and changed conditions

Would your answer change if the activity cost money?
Would your answer change if participants were much younger?
Would your answer change if older adults were the main users?
Would your answer change if the event were outdoors in bad weather?
Would your answer change if the waiting time doubled?
Would your answer change if the system excluded people without smartphones?
Would your answer change if the cheaper choice created more waste?
Would your answer change if the safer choice took more time?
Would your answer change if volunteers were unavailable?
Would your answer change if the venue had very limited space?
Would your answer change if new evidence showed the problem was rare?
Would your answer change if the rule affected one group more than another?
Would your answer change if the technology sometimes failed?
Would your answer change if people strongly preferred the less efficient option?
Would your answer change if you had to explain it to a younger child?
Would your answer change if the project could help many people but only after a long delay?
Would your answer change if the immediate benefit created a future cost?
Would your answer change if a close friend disagreed with you?
Would your answer change if you learned that your original assumption about the photo was wrong?
Would your answer change if the examiner asked for the strongest argument against your view?

Twenty answer-development laboratories

Lab 1: reason

Give ten positions and add one “because” relationship without using the word because every time.

Lab 2: example

Take five general reasons and make each concrete with an ordinary example.

Lab 3: consequence

Take five examples and explain why they matter.

Lab 4: condition

Add one circumstance that would change each opinion.

Lab 5: qualification

Replace absolute claims such as always or never with more defensible language when appropriate.

Lab 6: stakeholder

Identify one person who benefits and one person who may face difficulty.

Lab 7: alternative

Give a second solution and compare it with the first.

Lab 8: evidence

Connect a photograph detail to one inference explicitly.

Lab 9: uncertainty

State what is not known and continue reasoning responsibly.

Lab 10: perspective

Explain how the same situation may look different to another user.

Lab 11: compression

Cut an overlong answer in half without losing its strongest relationship.

Lab 12: expansion

Turn a one-sentence answer into four sentences without repetition.

Lab 13: release

Stop after the answer is complete even if the allotted time has not been filled.

Lab 14: paraphrase

Explain one key concept without the preferred vocabulary word.

Lab 15: false-start recovery

Restart a tangled sentence in simpler language.

Lab 16: changed-view recovery

Revise a position after new evidence without apologising for changing your mind.

Lab 17: no-example recovery

Use a hypothetical or observed example when no personal event is available.

Lab 18: direct opening

Answer ten different question jobs in the first sentence.

Lab 19: follow-up reset

Practise ten second-turn questions where the condition changes.

Lab 20: integrated turn

Hear, classify, answer, develop, connect and release on fresh material with no coaching.

The central training question

Do not ask whether the learner can memorise enough good answers to cover likely topics. Ask whether the learner can build a relevant answer from the evidence, question and knowledge available at that moment. The first goal produces polished islands. The second produces portable performance.

The ninety-day Stimulus-Based Conversation progression

A conversation programme becomes transferable when it moves from controlled answer construction to genuine interaction. The progression below deliberately alternates skill building, variation and pressure so that learners do not confuse familiarity with readiness.

Days 1–10: establish the cold baseline

Use several unseen photographs and question jobs without coaching during the first attempt. Record where the response first weakens: question interpretation, evidence, idea retrieval, development, language, interaction or recovery. Do not diagnose from one topic alone. A recurring pattern across different surfaces is more informative.

Teach the learner how to review a response without rewriting it completely. One direct question after playback—“Where did the answer first stop serving the question?”—often produces more useful information than a long list of corrections.

Days 11–20: direct relevance

Train the first sentence. Use mixed question jobs and require the core answer early. The learner should experience that a direct opening does not prevent sophistication; it creates a stable platform for development.

Remove generic introductions and question repetition during this phase. Once relevance is automatic, natural variation can return.

Days 21–30: development

Use one relationship at a time. Answer plus reason. Answer plus example. Answer plus consequence. Answer plus qualification. Then combine them selectively. The goal is not to create a fixed six-part answer but to give the learner multiple ways to make thought visible.

Days 31–40: evidence and the stimulus

Use varied images. Separate observation from inference. Require evidence for interpretations. Add questions where the photograph is central and questions where it is only a starting point. This prevents both over-description and complete stimulus neglect.

Days 41–50: personal experience and retrieval

Build an ordinary experience bank. Store cues, not scripts. Practise retrieving one experience under several themes and several experiences under one theme. Introduce no-experience questions so learners can use hypothetical or observed examples honestly.

Days 51–60: follow-up adaptation

Make the second turn the target. Use condition changes, counterexamples, requests for evidence, alternative perspectives and new constraints. The learner must hear what changed before deciding whether the original position still stands.

Days 61–70: unfamiliar topics and world knowledge

Use accessible but less rehearsed material. Encourage cautious reasoning from evidence, stakeholders, cause, consequence and alternatives. Continue ordinary reading and discussion outside oral drills so the language system is fed by a wider world.

Days 71–80: timing, listeners and recovery

Use realistic interaction conditions more often. Add a different listener. Practise intentional pauses, paraphrase, false-start recovery and resetting after a weak turn. Measure cascade length: how many later sentences or questions are affected by one difficulty?

Days 81–90: stability and taper

Use full conversation sequences selectively. Compare first attempts across several days. Reduce active strategies to a small operating routine. Continue fresh tasks, but stop searching for secret predicted topics.

A sixteen-week conversation curriculum

Weeks 1–2: question jobs

Listen fully, classify silently, answer directly.

Weeks 3–4: development

Reason, example, consequence, contrast and qualification.

Weeks 5–6: stimulus evidence

Observation, inference, uncertainty and scope.

Weeks 7–8: experience retrieval

Ordinary memory bank, flexible lenses, honest alternatives when experience is absent.

Weeks 9–10: follow-up interaction

New conditions, challenges, changed evidence and perspective shifts.

Weeks 11–12: unfamiliarity

Less rehearsed topics, wider conceptual knowledge, paraphrase and cautious reasoning.

Weeks 13–14: pressure integration

Realistic preparation, live listeners, no mid-turn coaching, deliberate recovery.

Week 15: stability audit

Several unseen sequences; compare floor and variance.

Week 16: taper

Short fresh tasks, stable routines, minimal new language burden.

The follow-up laboratory: thirty ways the examiner can change the task

These are original training transformations. The first answer can be about any suitable topic; the second turn introduces a new relationship.

1. Add cost

“Would you still choose it if it were expensive?” Train the learner to discuss affordability or value rather than repeat the first benefit.

2. Add time

“Would you still choose it if it took twice as long?” The answer should weigh time against the original advantage.

3. Add safety

“What if the safer option were less convenient?” The learner now has a trade-off.

4. Add fairness

“What if one group benefited more than another?” Shift from personal preference to distribution.

5. Add access

“What if some users could not use a smartphone?” The learner should consider inclusion and alternatives.

6. Add age

“Would your answer be different for younger children?” Shift perspective and support needs.

7. Add older users

“Would your answer be different for older adults?” Consider accessibility without stereotyping.

8. Add scale

“Would this still work for five hundred people?” Move from individual case to system design.

9. Add weather

“What if the activity were outdoors during bad weather?” Consider contingency and safety.

10. Add limited space

“What if the venue were very small?” The learner adapts design rather than repeating the ideal plan.

11. Remove volunteers

“What if fewer helpers were available?” Consider simplification, scheduling or automation.

12. Remove technology

“What if the digital system failed?” Consider backup processes.

13. Add environmental cost

“What if the convenient option created more waste?” Balance immediate and long-term effects.

14. Add financial limit

“What if the organisation had very little budget?” Prioritise essentials.

15. Add conflicting preference

“What if most people preferred the less efficient choice?” Weigh popularity against outcome.

16. Add new evidence

“What if new information showed the problem was rare?” Reconsider proportionality.

17. Reverse the evidence

“What if the photograph detail you noticed actually had another explanation?” Practise revising inference.

18. Ask for strongest counterargument

“What would someone who disagrees say?” Practise perspective and intellectual flexibility.

19. Ask for compromise

“Can you suggest a middle option?” Move beyond binary thinking.

20. Ask for priority

“Which problem should be solved first?” Require selection and justification.

21. Ask for evidence

“What makes you think that?” Return to visual or experiential support.

22. Ask for example

“Can you give a specific example?” Move from abstract claim to concrete case.

23. Ask for consequence

“What could happen if nothing changed?” Extend causal reasoning.

24. Ask for personal action

“What would you do yourself?” Shift from system judgement to individual choice.

25. Ask for organisational action

“What should the organisers do?” Shift responsibility level.

26. Ask for future change

“Would your answer be different in five years?” Consider technology, habit or scale cautiously.

27. Ask for another stakeholder

“How might a parent, worker or visitor see it?” Change perspective.

28. Ask for limitation

“What is one weakness in your own idea?” Practise self-qualification.

29. Ask for fallback

“What would you do if your first plan failed?” Train recovery thinking.

30. Ask whether the learner has changed their mind

“After discussing these conditions, do you still agree with your first answer?” Practise reasoned revision or reaffirmation.

The question-job atlas: forty cognitive jobs

These categories overlap. Their purpose is to expose the learner to the range of thinking a live conversation can demand.

Locate: identify the relevant part of the image.
Describe: state what is visibly present.
Infer: propose a supported interpretation.
Bound: state what cannot be known.
Recall: retrieve a related personal event.
Relate: connect experience to the question.
Prefer: choose between options.
Justify: explain a choice.
Exemplify: make a claim concrete.
Explain cause: describe why something happens.
Explain effect: describe what follows.
Compare: use a shared criterion.
Contrast: identify a meaningful difference.
Evaluate: judge strengths and weaknesses.
Recommend: propose an action.
Prioritise: choose what matters first.
Qualify: limit a claim.
Condition: explain when the claim changes.
Predict: give a cautious future outcome.
Design: improve a system or space.
Allocate: decide how limited resources should be used.
Balance: manage competing goals.
Defend: support a position under challenge.
Revise: change a position with new evidence.
Reframe: look at the issue through another lens.
Empathise: consider another user’s experience without mind-reading.
Distinguish: separate evidence from assumption.
Generalise: draw a broader principle from a case.
Specify: replace a vague statement with detail.
Simplify: express the same idea in easier language.
Paraphrase: recover from unavailable vocabulary.
Clarify: restate a tangled sentence.
Challenge: identify a weakness in an idea.
Compromise: combine useful parts of different options.
Sequence: explain steps or events in order.
Classify: group examples by a useful criterion.
Reflect: explain what was learned.
Transfer: apply a lesson to a different context.
Recover: rebuild after a weak response.
Release: recognise when the turn is complete.

The ordinary experience bank

Personal experiences are easiest to retrieve when organised by human actions rather than predicted topics. Build a bank around verbs.

Helped: assisted a classmate, family member or stranger.
Changed: revised a plan or opinion.
Waited: handled delay or uncertainty.
Lost: misplaced something and responded.
Learned: acquired a skill or corrected a misconception.
Joined: participated in something unfamiliar.
Refused: chose not to participate and explained why.
Shared: used a common resource.
Repaired: fixed an object, process or relationship.
Planned: organised an event or task.
Failed: experienced an unsuccessful attempt.
Retried: changed approach after failure.
Disagreed: managed a difference in view.
Apologised: took responsibility.
Noticed: observed a problem others missed.
Suggested: proposed an improvement.
Adapted: responded when conditions changed.
Protected: acted for safety or care.
Chose: balanced two options.
Explained: helped another person understand.

For each event, remember four cues: situation, choice, result and reflection. Avoid full scripts. The same memory can support different questions because the learner extracts a different relationship.

Vocabulary as conceptual equipment

Topic vocabulary is useful when it makes thought more precise. Organise it by concepts rather than predicted titles.

Access

accessible, barrier, inclusive, convenient, reachable, support, alternative.

Responsibility

accountable, maintain, contribute, obligation, shared, individual, organise.

Fairness

equal, equitable, opportunity, disadvantage, priority, transparent, reasonable.

Environment

waste, reuse, resource, consumption, durable, efficient, impact.

Technology

digital, reliable, privacy, convenient, automated, backup, user-friendly.

Community

participation, volunteer, local, shared space, cooperation, belonging, public.

Learning

practice, feedback, independence, guidance, experience, reflection, improvement.

Each word should have a plain-language route. Vocabulary exists to increase expressive choice, not to create another point of failure.

Answer compression: remove without weakening

Take a ninety-second response and remove every sentence that repeats, drifts or decorates without adding a relationship. The result often becomes stronger at sixty seconds. Then ask whether anything essential was lost.

This drill teaches the learner to recognise informational density. It is especially useful for Maren-style overdevelopment.

Answer expansion: add without repeating

Take a one-sentence response and add a reason, example and consequence. Every sentence must contribute a new relationship. If the answer simply restates the position in different words, the development move has failed.

This drill is especially useful for Iona-style precise but thin answers.

Recovery ladder

Missing word: paraphrase.
False start: restart the sentence more simply.
Weak example: replace it or explain the connection.
Misunderstood question: acknowledge the corrected job.
Unsupported inference: reduce certainty or return to visible evidence.
Changed condition: reset the position and reason.
Blank mind: return to the direct answer and one ordinary example.
Weak first turn: treat the next turn as new.

The ladder turns recovery into a trainable system rather than a personality trait.

Conversation recording review

Listen once for the overall answer. Was the question answered?
Listen for development. Which sentence added the most new information?
Listen for evidence. Which claims came from the stimulus, and which were assumptions?
Listen for language bottlenecks. Where did retrieval or grammar interrupt thought?
Listen for interaction. Did follow-ups change the answer?
Listen for release. Where did the answer become complete, and did speech continue unnecessarily?
Choose one repair.
Retest on a different question.

Conversation without a recording

Recording is useful but not compulsory. A listener can note the first sentence, strongest developed point, first drift point and response to follow-up. The learner then self-reports what felt difficult. Comparing listener and learner perspectives often reveals blind spots.

What strong conversation is not

It is not nonstop speech. It is not memorised sophistication. It is not an encyclopaedic display of topic knowledge. It is not an accent performance. It is not a fixed framework applied to every prompt. It is not agreement with the examiner at all costs.

Strong conversation is relevant, developed, evidence-aware, responsive and recoverable.

The conversation practice ladder

Stimulus-Based Conversation should not move directly from model answers to full mock examinations. A practice ladder allows the teacher to control which demand is being added and to see exactly where a stable skill becomes fragile.

Single-job drill: answer only observation, opinion, experience or another defined job.
Development drill: add one reason, example, consequence or qualification.
Stimulus integration: use visible evidence where relevant.
Two-turn exchange: add one follow-up.
Changed-condition exchange: make the follow-up alter cost, time, fairness, safety or perspective.
Unfamiliar-topic exchange: keep language accessible while reducing topic familiarity.
New-listener exchange: test portability across people.
Mini-mock: use the visual stimulus and a short live sequence with no coaching.
Integrated oral mock: place conversation after Reading Aloud and realistic preparation.
Transfer confirmation: change stimulus, topic and question job while retaining the same performance expectations.

Twenty original mini-mock conversation sets

These are training situations created by SETC. They are not official PSLE materials. A teacher can pair each set with an original or licensed photograph matching the broad scene.

Set 1: community repair table

Opening: Would you be interested in joining an event where people repair household items together? Why or why not?

Follow-up: What if repairing the item took longer than buying a new one?

Second follow-up: How might the event be made useful for children?

Set 2: mobile library

Opening: What do you think is useful about bringing books to different neighbourhoods?

Follow-up: Would a digital library make the mobile library unnecessary?

Second follow-up: Who might still benefit from the physical service?

Set 3: shared garden

Opening: Would you enjoy helping to care for a shared garden?

Follow-up: What if only a few people did most of the work?

Second follow-up: How could responsibility be shared more fairly?

Set 4: quiet carriage

Opening: Do you think public transport should provide quiet areas?

Follow-up: What if passengers disagree about what counts as quiet?

Second follow-up: Would a guideline or a strict rule work better?

Set 5: volunteer information desk

Opening: What qualities would make someone useful at an information desk?

Follow-up: What if the volunteer did not know the answer to a visitor’s question?

Second follow-up: How should the volunteer recover?

Set 6: public recycling station

Opening: What could make this recycling station easy or difficult to use?

Follow-up: What if the clearest signs were expensive to replace?

Second follow-up: What lower-cost improvement could be made first?

Set 7: shared tool cabinet

Opening: Do you think neighbours should share tools that are rarely used?

Follow-up: What problem might arise if an item were returned damaged?

Second follow-up: How could the system remain fair?

Set 8: outdoor performance

Opening: Would you prefer to watch a performance outdoors or indoors?

Follow-up: What if the outdoor performance were free but the weather uncertain?

Second follow-up: What should organisers prepare?

Set 9: museum hands-on station

Opening: Do hands-on activities make museums more interesting?

Follow-up: Could they also distract from the objects being displayed?

Second follow-up: How would you balance activity and attention?

Set 10: school swap table

Opening: What might be useful about a table where pupils exchange books or stationery?

Follow-up: What if some items were much more valuable than others?

Second follow-up: How could the exchange remain fair?

Set 11: community noticeboard

Opening: What makes a noticeboard useful?

Follow-up: What if too many notices are displayed?

Second follow-up: Who should decide what stays?

Set 12: digital check-in point

Opening: Do digital check-in systems make events more convenient?

Follow-up: What if a visitor does not own a smartphone?

Second follow-up: What backup should organisers provide?

Set 13: pedestrian crossing volunteer

Opening: Why might volunteers help near a busy crossing?

Follow-up: Should volunteers be responsible for safety if traffic systems already exist?

Second follow-up: What should remain the responsibility of drivers and pedestrians?

Set 14: community cooking class

Opening: What could people learn from cooking together?

Follow-up: What if participants have very different skill levels?

Second follow-up: How could the activity include everyone?

Set 15: bicycle repair point

Opening: Should public places provide simple repair stations?

Follow-up: What if equipment is damaged or stolen?

Second follow-up: Would that change your view?

Set 16: neighbourhood reading bench

Opening: Would placing books in public spaces encourage reading?

Follow-up: What if books were lost or damaged?

Second follow-up: Is the risk worth the possible benefit?

Set 17: school exhibition route

Opening: What makes an exhibition easy for visitors to understand?

Follow-up: What if the venue is too small for clear routes?

Second follow-up: Which problem should be solved first?

Set 18: shared study table

Opening: Do shared tables help people learn together?

Follow-up: What if some people need quiet?

Second follow-up: How could the space serve both needs?

Set 19: public water refill point

Opening: What could encourage people to use refill stations?

Follow-up: What if users worry about cleanliness?

Second follow-up: How should the organisation respond?

Set 20: intergenerational activity

Opening: What might children and older adults learn from doing an activity together?

Follow-up: What difficulties could arise because of different abilities or preferences?

Second follow-up: How could the activity be redesigned?

How to score practice without inventing an official rubric

Use descriptive evidence rather than pretending to reproduce examiner marking. A training review can track six domains:

Relevance: was the question answered early?
Development: did the learner add a useful relationship?
Evidence: were stimulus claims supported and uncertainty represented accurately?
Language: was the meaning understandable at speaking speed?
Interaction: did the learner respond to the new angle?
Recovery: did difficulty remain local?

Tag each behaviour as stable, variable or not yet available. This supports training decisions without claiming to duplicate official marking.

The self-review card

Did I hear the whole question?
What job did the question require?
Did my first sentence make the answer clear?
What new relationship did I add?
Did I use the photo only when it supported the answer?
Did I distinguish observation from inference?
Did I use a real or honest example?
Did I paraphrase instead of freezing when a word disappeared?
Did the follow-up change my answer?
Did I stop when the turn was complete?
What was the first weak link?
What fresh question will test the repair?

The parent’s role

Parents can help most by being real listeners. Ask the child one question and wait. Do not supply the first idea immediately. If the answer is thin, ask “Why?” or “Can you give an example?” once. If the answer is off-topic, ask what the question required before supplying content.

Avoid turning home practice into a performance court. Descriptive feedback is enough: direct opening, unclear example, unsupported photo claim, strong recovery, repeated idea. The purpose is independence.

The teacher’s role

Teachers control variability. They decide when to isolate a skill, when to change topic, when to remove prompts, when to introduce a new listener and when full simulation is justified. Good task selection is part of teaching.

Keep a ledger by cause rather than by topic. “Weak on recycling” is less useful than “gives recommendations without explaining the problem.” The second description transfers to many topics.

Small-group conversation design

Groups can produce many useful turns when roles rotate quickly. One learner answers; one peer listens for relevance; another listens for development or evidence. After thirty to sixty seconds, roles change. Do not ask peers to invent marks. Ask them to report listener experience.

Use the resident-character lenses if helpful: a Maren lens for structure, an Iona lens for evidence and a Leonie lens for execution. Learners can rotate the lens without being labelled as a type.

World knowledge without encyclopaedic pressure

A broad reading life gives learners more concepts and examples. Use short materials from science, history, museums, technology, environment, communities and everyday systems. Discuss problem, stakeholders, choices, evidence and consequences. Do not require the learner to memorise facts for predicted oral topics.

The broader eduKate ecosystem can feed this knowledge while canonical subject pages retain ownership. The oral lane uses the knowledge to practise reasoning and expression.

Conversation and listening

Stimulus-Based Conversation is partly a listening task. Learners who prepare excellent answers but fail to hear changed conditions are not yet interaction-ready. Practise listening to full questions at natural speed, including questions where the decisive phrase appears at the end.

Use occasional paraphrase checks after practice: “What did the examiner actually ask?” If the learner cannot state the job, answer construction is premature.

Conversation and reading

Reading supplies vocabulary, world knowledge and models of reasoning. But written language can become too formal for spontaneous speech. When learners borrow useful phrasing from reading, convert it into natural spoken language rather than importing essay sentences unchanged.

Conversation and composition

Both involve idea development, but conversation has turn-taking, time pressure and live adaptation. A well-planned composition framework cannot simply be spoken. Oral answers need earlier relevance, lighter structure and space for the examiner to change direction.

Misconceptions

“Longer answers are always stronger.”

Length helps only when it adds new relationships.

“The photo must be described first.”

The question decides whether description is needed.

“A model answer should be memorised.”

Analyse the model, then transfer the behaviour to a new task.

“Every answer needs the same framework.”

Different question jobs need different structures.

“Advanced vocabulary proves ability.”

Precision used naturally is valuable; vocabulary that stalls the answer is not.

“Changing your mind looks weak.”

Changing a view because evidence or conditions changed can demonstrate reasoning.

“A pause means poor fluency.”

Short intentional pauses can improve clarity and control.

“No personal experience means no answer.”

Use related experience, observation or a hypothetical response honestly.

“Confidence means speaking fast.”

Confidence can look calm, measured and responsive.

“One weak answer ruins the conversation.”

The next question is a new task. Recovery matters.

Frequently asked questions

How many topics should a learner prepare?

Use themes to build knowledge, but do not treat a topic list as coverage insurance. Train question jobs and reasoning moves across varied material.

Should learners memorise useful phrases?

Memorise language functions lightly if helpful, not complete answers. A phrase should be available as one option rather than compulsory wording.

How often should conversations be recorded?

Often enough to reveal patterns, not so often that speaking becomes a recording performance. Live listener notes are also useful.

Should every grammar error be corrected?

No. Prioritise recurring errors that interfere with meaning or limit answer development.

What if the learner is shy?

Shyness and weak communication are not identical. Train complete, clear turns without demanding an extroverted style.

What if the learner speaks too much?

Teach answer release and sentence-function review. The goal is substance, not airtime.

What if the learner gives very short answers?

Train one development relationship at a time and retest on fresh questions.

What if the learner cannot think of reasons quickly?

Practise general reasoning categories—benefit, risk, cost, fairness, access, consequence and alternative—across ordinary topics.

How should the final week change?

Reduce new techniques, keep fresh but manageable tasks, practise recovery and use a few high-fidelity sequences rather than endless full mocks.

What should the learner remember on examination day?

Hear the whole question. Answer it. Develop one useful idea. Use evidence honestly. Listen to what changes. Recover and continue.

Twenty-point readiness audit

The learner hears the whole question before answering.
The learner identifies the question job quickly.
The first sentence is usually relevant.
The learner can develop one point without prompting.
The learner can use a reason.
The learner can use a concrete example.
The learner can explain a consequence.
The learner can qualify an opinion.
The learner uses the photograph as evidence rather than as a script.
The learner separates observation from inference.
The learner can acknowledge uncertainty.
The learner can retrieve ordinary personal experiences.
The learner can answer honestly without direct experience.
The learner can paraphrase missing vocabulary.
The learner can restart a tangled sentence.
The learner responds to follow-ups as new tasks.
The learner can change or refine a position when conditions change.
The learner can stop when the turn is complete.
The core behaviours survive unfamiliar topics and listeners.
A weak moment no longer controls the next question.

This is a training checklist, not an official rubric. It converts vague readiness into observable behaviour.

Final principle

Stimulus-Based Conversation is strongest when the learner is prepared without being pre-written. The photograph provides evidence, the question provides direction, world knowledge provides material and the learner provides judgement. Training succeeds when those parts can be assembled live, adjusted when the examiner changes angle and rebuilt when a word or idea briefly fails.

The examination rewards communication in the moment. The transferable skill is organised adaptability.

Deep flexibility laboratory: twenty-four transfer tests

Flexible conversation cannot be verified by one polished response. It has to survive changes in topic, question job, listener, time, evidence and interaction. The transfer tests below are designed to expose hidden dependence without turning practice into a trick. Each test changes one condition and asks whether the learner’s core behaviours remain available.

Transfer test 1: familiar topic, unfamiliar job

Use a topic the learner knows well, but ask a question that requires a less familiar job such as qualification, comparison or recommendation. If the response becomes weak despite topic familiarity, the problem is probably not knowledge. Train question-job recognition and development.

Transfer test 2: unfamiliar topic, familiar job

Keep the job simple—preference or observation—but use a less rehearsed setting. If the learner can still answer directly and develop one useful point, the conversation system is becoming portable.

Transfer test 3: photo central, then photo peripheral

Ask one question that requires close use of visual evidence and another where the photograph is only a starting point. A stable learner knows when to anchor tightly to the stimulus and when to move into experience or wider reasoning.

Transfer test 4: one-turn versus two-turn performance

Compare a strong standalone answer with a conversation that includes a follow-up. If quality drops only on the second turn, train interaction reset rather than first-answer construction.

Transfer test 5: same topic, changed evidence

Give new information that weakens the learner’s first assumption. The learner should revise or qualify the position rather than defend it automatically. This tests evidence responsiveness.

Transfer test 6: ordinary vocabulary versus preferred vocabulary

Ask the learner to answer without several favourite advanced words. If meaning remains clear through paraphrase, vocabulary knowledge is resilient. If speech freezes, train multiple routes to the same concept.

Transfer test 7: personal experience available versus unavailable

Use one question with an obvious personal example and one where no direct experience exists. A stable learner can shift to observation, related experience or a hypothetical response without inventing a false story.

Transfer test 8: familiar listener versus new listener

Keep the task difficulty similar and change the adult asking the questions. Compare directness, volume, pause length, development and recovery. Large differences indicate listener dependence.

Transfer test 9: comfortable time versus realistic time

Allow generous preparation on one task and realistic constraints on another. Identify which useful behaviour disappears first. If evidence checking vanishes, train a faster evidence scan. If answer development collapses, train retrieval of portable development moves.

Transfer test 10: first question versus final question

Use a sequence of several turns. If later answers become longer, less relevant or more repetitive, train cognitive reset and turn release.

Transfer test 11: agreement versus disagreement

Ask the learner first to defend a genuine opinion, then to explain the strongest reasonable opposing view. This tests whether language and reasoning depend on personal certainty.

Transfer test 12: yes versus qualified yes

Take an easy binary question. First answer directly. Then add a condition that makes the answer less absolute. The learner should keep a clear position while acknowledging the boundary.

Transfer test 13: direct question versus indirect wording

Ask the same underlying job in two different ways. If the learner recognises only familiar wording, question classification remains surface-bound.

Transfer test 14: one stakeholder versus several

Begin from personal preference and then introduce another user whose needs differ. The learner should widen the frame without abandoning relevance.

Transfer test 15: benefit versus trade-off

After the learner states a benefit, introduce a cost, inconvenience or limitation. Strong performance can hold two relationships at once rather than merely switch sides.

Transfer test 16: evidence certainty versus uncertainty

Use one photograph with obvious action and one ambiguous photograph. The learner should change the certainty of language rather than speak with the same confidence in both.

Transfer test 17: first attempt versus corrected attempt

Allow one correction on a response, then change the question. The new task reveals whether the learner improved the mechanism or merely polished the first answer.

Transfer test 18: immediate versus delayed retest

Repair a target behaviour and retest it after one or two days. Delayed transfer is stronger evidence than immediate imitation.

Transfer test 19: ordinary day versus high-energy day

Compare performance without deliberately exhausting the learner. The examination target is a dependable floor, not a best-day peak.

Transfer test 20: teacher prompt versus no prompt

Track prompts such as “why?”, “example?” and “what does the question ask?”. Remove them gradually. Improvement should include reduced dependence on external cues.

Transfer test 21: model available versus model closed

Analyse a model answer for structure, then close it and answer a fresh question. If useful behaviour transfers without copying wording, the model has taught a mechanism rather than a script.

Transfer test 22: no-script restriction

Ban the learner’s usual opening, favourite personal example and favourite connector for one short session. If communication remains strong, the system has multiple routes.

Transfer test 23: deliberate false start

Begin one answer awkwardly on purpose, then restart in simpler language. The learner experiences recovery as an ordinary skill rather than an emergency response.

Transfer test 24: integrated no-rescue sequence

Use an unseen stimulus, realistic interaction and no coaching until the sequence ends. Afterwards identify only the first significant cause and retest it on a new task. This is the strongest independence check in the progression.

Thirty no-script constraints

Constraints are temporary training devices. They remove familiar crutches so the learner has to generate meaning from the task. Use one or two at a time, never all at once.

Do not begin with a memorised general introduction.
Do not repeat the examiner’s whole question.
Do not use your favourite personal story.
Do not use the phrase you normally use to start an opinion.
Answer the question in the first sentence.
Use only one main reason.
Develop that reason with one concrete example.
Add one consequence without adding a second reason.
Add one qualification to an otherwise strong claim.
Use one visible photo detail as evidence.
State one thing you cannot know from the photograph.
Use a cautious inference rather than a confident story.
Avoid one preferred advanced vocabulary word and paraphrase it.
Use ordinary accurate language for the entire answer.
Then repeat with two more precise words and no extra length.
Stop after forty seconds even if you could say more.
Compress a long answer to half its length.
Expand a thin answer without repeating the position.
Give the strongest reasonable counterargument to your own view.
Change your view when new evidence justifies it.
Keep your view but change the reason when the condition changes.
Use a hypothetical example instead of a personal one.
Use an observed real-world example instead of a hypothetical one.
Wait one second after every new question before speaking.
Use one silent pause instead of a filler.
Restart one tangled sentence more simply.
Answer a follow-up without referring back to memorised wording.
Give one answer from another stakeholder’s perspective.
Finish the turn as soon as it is complete.
Retest the same behaviour on a new topic immediately afterwards.

The final twenty-one-day stability plan

The last three weeks should convert improvement into availability. They are not the time to multiply techniques. Each week has a different function: expose variance, repair the remaining high-cost causes, then taper.

Days 21–15 before the examination: map the floor

Use three unseen conversation sequences across different topics and question jobs. Include at least one unfamiliar but accessible topic and one different listener if possible. Do not coach during the first attempt. Record the first significant breakdown in each sequence.

If the same cause appears twice, it becomes the main repair target for the week. If causes differ, choose the one with the largest communicative cost. Avoid trying to correct every small weakness simultaneously.

Use short daily transfer work: direct openings, one-point development, evidence discipline, follow-up reset or paraphrase according to the diagnosis. Finish each repair session with a fresh question rather than a polished repetition of the original.

Days 14–8: integrate and pressure-test

Use realistic preparation more often. Combine the conversation with Reading Aloud in selected sessions so the learner practises sequence transitions. Introduce one or two changed-condition follow-ups. Include ordinary-day practice once, but do not deliberately create exhaustion.

Compare variance. The strongest answer may remain similar while the weakest answer improves. That is valuable progress because the floor is rising.

Reduce teacher prompts. If the learner still waits for “why?” before developing, make independence the target. If recovery still requires reassurance, practise one small disruption followed by a fresh turn.

Days 7–4: simplify

Stop adding new frameworks. Use short fresh questions, ordinary experience retrieval and one or two mini-mocks. Keep world-knowledge exposure normal rather than turning every article into a potential oral topic.

The active checklist should now be compact: hear the question, answer it, develop one useful idea, use evidence honestly, listen to the new angle, recover and continue.

Days 3–2: maintain

Use one manageable conversation sequence each day. Practise one paraphrase and one follow-up reset. Review the learner’s strongest recovery evidence: a time when an imperfect answer was followed by a strong next turn. The aim is confidence grounded in demonstrated control.

Final day

One brief warm-up is enough. Do not conduct a marathon of predicted topics or advanced vocabulary. The system is already built. Protect access to it.

What the learner should carry into the room

Not a catalogue of model answers. Not a memorised paragraph for every topic. Not a fear that every pause is visible. The learner should carry a small operating system: listen fully, recognise the job, answer early, develop meaning, use the stimulus as evidence when relevant, adapt to follow-up and recover without turning one imperfect moment into the rest of the conversation.

Final ownership note

SETC-PSLE-ORAL-030 owns the integrated PSLE Stimulus-Based Conversation performance layer: flexible answer generation, evidence-aware stimulus use, live question handling, follow-up adaptation, recovery, transfer and performance stability. Narrow skill mechanisms already owned by eduKateSG remain linked rather than duplicated. General secondary oral mechanisms on SETC remain broader owners outside the PSLE-specific layer. All drills, scenarios, transfer tests and schedules here are original teaching material, not official questions or official examiner guidance.

PSLE Stimulus-Based Conversation Performance | Train Flexible Answers Without Memorised Scripts