The single biggest change in PSLE Oral preparation over the last three years — across both Chinese and English — is that memorised model answers no longer score well. SEAB has reshaped both exams, formally in English and informally in Chinese, to specifically defeat the kind of pre-scripted, "list three reasons" answers that dominated the previous decade. A child who walks into the exam with 30 memorised answers and slots them into questions is now likely to score lower than a child who has practised a flexible four-sentence opinion structure on a wide range of topics.
This article explains the shared shift — what changed in each language, why memorisation stops working, the four red flags examiners are trained to spot, and the single answer structure that works in both Mandarin and English. It is written for parents who already have a child in PSLE oral preparation and are wondering whether their current approach is still the right one.
A shared shift in two languages
The two PSLE Oral exams sit in different syllabi (English Language 0001, Chinese Language 0005) and are run by different examiner pools. They have evolved on different timelines. But anyone watching closely will see that both are moving in the same direction: away from rote, toward thinking.
On the English side, the change is formal and dated. In 2025, SEAB published a major overhaul of the English Language oral component: total marks rose from 30 to 40, the new PACT preamble was added to Reading Aloud, the Stimulus-Based Conversation moved from posters with text to real photographs with no text, and sub-prompts were removed entirely so that all three conversation questions are now opinion-based. The full set of changes is documented in The 2025 PSLE English Oral overhaul.
On the Chinese side, the change has been informal and gradual. The format on paper has not changed — the Chinese oral remains a 50-mark exam split 20/30 between reading aloud and a video-based conversation. But since around 2023, examiners have increasingly framed conversation questions as 你同意吗？ (Do you agree?) opinion challenges that cannot be answered with a memorised template. We unpack this trend in 你同意吗: the opinion question shift.
Two languages, two timelines, one direction. SEAB wants students to think on their feet, take a position, defend it with personal examples, and do all of this without scaffolding from the examiner.
Why memorisation stopped working
Memorisation worked, for years, because the oral exam was structurally predictable. There was a small set of recurring topics — helping others, friendship, environment, healthy living — and a small set of standard question patterns. A student who memorised 30 model answers covering the recurring topics had a good chance of landing close enough to the actual question that the rehearsed material could be slotted in. Sub-prompts gave the student a second chance if the first answer wandered.
Three things broke that approach:
- Sub-prompts were removed (English, 2025). A scripted student now gets one shot at each question. If the rehearsed answer does not actually fit, there is no recovery.
- Opinion questions broke topic-matching (both languages). A question that asks "do you agree that helping others teaches you new things?" cannot be answered by reciting a "helping others is good" speech. The student has to engage with the specific claim being made.
- Examiners are now trained to detect scripts. Tutors and teachers across Singapore report that examiners actively follow up rehearsed-sounding answers with specific probing questions: "You said you have done this. When? With whom? What did you learn?" A student answering from a script cannot answer those follow-ups in real time.
The combined effect is that the average memorised-answer student now scores meaningfully lower than they did five years ago — in both languages, on both components.
Four red flags examiners are trained to spot
Based on consistent reports from tutors and teachers across Singapore, here are four signals that mark an answer as "scripted" in the eyes of an examiner. If your child does any of these, they are probably losing marks.
1. Topic pivots that ignore the actual question. The student hears a familiar word and launches into a rehearsed speech about that word, instead of answering the specific question being asked.
2. Generic examples with no specific details."One time I helped someone and they were happy." No name, no place, no detail. Examiners are trained to ask for specifics — when, where, who, what happened next — and scripted students cannot improvise these.
3. "Three reasons" lists that all sound rehearsed."Firstly... secondly... thirdly..." in fluent but characterless prose. The structure is fine; the content reads like it was lifted from a model answer book.
4. Inability to handle a follow-up. The student gives a confident initial answer and then freezes when the examiner asks a probing question that takes the conversation in an unexpected direction. This is the clearest possible signal of a scripted approach.
The single structure that works in both languages
The good news is that the same answer structure works for opinion questions in both Mandarin and English. It is short, disciplined, and built from four sentences in this order:
- Position. "I agree." or "我同意。" One sentence. Commit.
- Reason. "Because..." or "因为..." A single clear reason. Not three.
- Example. "For example, last week..." or "例如，上个星期..." A specific personal experience with at least one detail an examiner could ask about.
- Extension. "So I think..." or "所以我觉得..." A small reflective conclusion that ties the example back to the original question.
The structure is the same in both languages because the underlying logic — take a position, give a reason, ground it in a real example, reflect on what it means — is universal. Drilling this structure for ten minutes a day on a wide range of topics, in both English and Mandarin, builds the reflex that both exams now reward.
This is essentially the same advice as the Chinese-specific P.E.E. framework (Point, Example, Elaboration), with the addition of an explicit position-taking sentence at the start. The position sentence matters because, under the new format, fence-sitting ("both sides have a point") is the worst possible answer.
One tool, both languages
PSLEPrep covers both PSLE Chinese Oral and PSLE English Oral in a single subscription. The conversation modes are built around the new opinion-question format in both languages, with AI follow-ups that probe scripted-sounding answers — exactly the way real examiners do.
Try 10 free practice sessions →What to actually do at home
A practical translation of all of the above into a daily routine looks like this:
- Stop drilling model answers. Put the model answer book away. It is not helping. Use it only as a vocabulary reference, never as a script.
- Drill the four-sentence structure on new topics every day. Pick any opinion statement — it does not have to be about a PSLE-style topic. Any statement works. The goal is to make the structure reflexive.
- Reward specific examples. When your child gives a vague example, ask for details: when, where, who, what happened next. Train them to include those details automatically.
- Practise both languages on the same day. The structure is identical, so practising in both languages reinforces the reflex twice over. A 20-minute daily session split 10 minutes English / 10 minutes Mandarin is more effective than 20 minutes in one language.
- Welcome disagreement. Coach your child that 不同意 / "I disagree" is a perfectly good starting position. Examiners reward defended views, not "correct" views.
Frequently asked questions
Are memorised model answers really failing in PSLE Oral?
Yes, in both languages. SEAB's 2025 English Oral overhaul removed sub-prompts entirely so all three conversation questions are now opinion-based. PSLE Chinese Oral examiners have been increasingly asking 你同意吗 (Do you agree?) opinion questions since around 2023. Both shifts reward students who can take a position and defend it with personal examples in real time, and penalise students who recite rehearsed material that does not engage with the specific question.
What is the single best answer structure for PSLE Oral opinion questions?
Four sentences in order: Position (commit to a view), Reason (one clear reason, not three), Example (a specific personal experience with details an examiner could probe), Extension (a small reflection that ties the example back to the question). The same structure works in both Mandarin and English.
How do PSLE examiners detect scripted answers?
Four common red flags: (1) topic pivots that ignore the actual question; (2) generic examples with no specific details; (3) "firstly, secondly, thirdly" lists in characterless prose; (4) freezing when asked a follow-up question that takes the conversation in an unexpected direction. Examiners are now trained to actively probe for these signals and ask follow-up questions designed to expose scripts.
Should my child still use a model answer book?
Only as a vocabulary reference, never as a script. Model answer books are useful for seeing how connectors like 因为 (because) and 例如 (for example) are used in context. They are actively harmful when used to memorise content. The goal is for the child to internalise the structure, not the words.
Is the trend toward opinion questions formal or informal?
On the English side, formal. The 2025 SEAB English Language syllabus (0001) explicitly removed sub-prompts and made all three conversation questions opinion-based. On the Chinese side, informal. The Chinese Language syllabus (0005) format has not changed on paper, but examiners have shifted in practice toward 你同意吗-style questions. The practical implication for parents is the same in both languages: prepare for opinion questions, not memorised topic essays.
Can my child practise both PSLE Chinese Oral and PSLE English Oral with the same routine?
Yes — and they should. The four-sentence opinion structure is the same in both languages, so practising in both reinforces the reflex twice over. A 20-minute daily session split 10 minutes English / 10 minutes Mandarin is more effective than 20 minutes in one language. The languages require different vocabulary and pronunciation work, but the underlying conversational thinking is shared.