How do you use the video clip in the O-Level English oral?

What is the video clip for in the O-Level English oral?

It gives you something to respond to. In the O-Level / SEC G3 English oral (1184 / K300), the Planned Response is a response "to a video clip and accompanying prompt presented on a computer screen". SEAB's 2026 O-Level instructions describe the task as preparing a response "based on a given topic using the video as a stimulus". The prompt is the question. The clip is the material you answer it with.

That sentence is about Part 2, but it tells you what the clip is for in both parts. In our reading, no marks go to remembering what people said. What counts is what you make of the situation the clip shows. The two parts "may be thematically linked", so the clip also sets up the discussion that follows; the follow-up questions candidates recall show how far from the clip that discussion moves.

What have the clips shown?

Everyday situations, four of the six below recognisably in Singapore, each with a prompt that asks for a judgement. These are clips and prompts candidates recalled from the current paper. SEAB publishes no past clips, and the wording is approximate:

• 2023: elderly women making wigs together. "Are the elderly women making good use of their time? Why, or why not?" (corroborated)
• 2023: haze over the city, with the Merlion and Esplanade visible. "How has the haze affected people living in the city?" (corroborated)
• 2024: a street hawker preparing fruit juice with a manual press. "Do you think there should be such street hawkers in Singapore? Why or why not?" (corroborated)
• 2024: the children's corner of a library. "Do you think this particular section of the library appeals to visitors?" (corroborated)
• 2025: cleaning robots at Changi Airport. "Do you think machines like the one shown in the video are useful?" (corroborated)
• 2025: two teenagers doing pottery together. "Who will be attracted to this activity? Why do you say so?" (corroborated)

"Corroborated" means at least two independent compilers recorded it. Even then it is recall, not a document. The full record, including single-source rows and 2026, is on the recalled O-Level English oral topics, year by year. Look at what the prompts ask: is it useful, should there be more, who would it attract, how are people affected. Each needs evidence, and the clip is where it comes from.

What should you notice while the clip plays?

Notice details you can use to answer the prompt, not everything on screen. SEAB's 2026 instructions give you paper to write notes on while you watch, and the English documents do not say whether you can watch the clip twice. So notice with a purpose the first time. Our guidance is to look for four kinds of detail:

1. Who is there. Their ages, whether they are alone or together, whether they are working, playing or being served. "Who would be attracted…?" and "good use of their time" prompts depend on this.
2. What they are doing, and how. By hand or by machine, with skill or for fun. A manual juice press and a cleaning robot are both answers waiting to be used.
3. Where it is. A public space, a home, a school, somewhere recognisably Singapore. The setting can hold the reason something matters.
4. What is not shown. Who is missing, what happens before or after, who pays. In our view this is the detail most likely to lead to a "well-considered" answer, because it takes you past what everyone in the room saw.

Write each one as two or three words. You may refer to your notes during the response, so they only need to remind you. What happens in the preparation time covers the notes paper and the rest of the room.

How do you turn a detail into evidence?

Say what it shows, then link it to your answer. A detail on its own is description. It becomes evidence when the next sentence says what it means. Our guidance is a three-step chain:

1. The detail. "The women are talking to each other as they work."
2. What it shows. "So the afternoon is about company as much as it is about wigs."
3. The link to your answer. "And for older people who spend most days alone, that may be the best use of their time there is."

The Response descriptors reward development: the top band asks for a response that is "well-developed and organised", and Band 3 has "some development". In our reading, a detail with no second step is description rather than development, which is the difference between those bands. How to structure the two minutes shows where these chains go in a whole answer.

How much of your response should be about the clip?

Our guidance: a small part of it. The clip starts the answer, and your thinking finishes it. In both of our top-band worked answers, the clip's details come early and briefly. Most of the answer is spent on what they mean and on people the clip did not show: in the robots answer, the cleaner whose job changes; in the wigs answer, the question of whether older people should have to be "productive" at all. One or two details, used well, do more work than a full retelling.

Two compilers independently recorded this prompt, so it is corroborated recall rather than a SEAB document. We wrote both answers; the stronger one covers only the first part of a response.

What should you avoid when using the clip?

Avoid anything that puts the clip in place of your answer. Our list of four:

• Retelling it in order. The examiners raise the same prompt you saw, and the prompt asks for a judgement, not a summary of the clip.
• Quoting what people said. In Part 2 the syllabus rules out questions about it, and in our reading the same applies in spirit to Part 1: the prompt, not the dialogue, is what you are answering.
• Inventing details. Use what you saw. If you are guessing ("he probably…"), say so, as the stronger answer above does.
• Forgetting the prompt. If the clip confuses you, go back to the prompt. A clear answer to the question, using whatever you did notice, is better than an accurate account of a clip you did not understand.

For how Response and Delivery are marked across every band, see how the O-Level English oral is marked.