Fourteen minutes, four parts, two candidates, under exam conditions. Nothing on this page is marked, because speaking is not marked by a page. Do this one with a partner and a clock: record each part in a single take, at the real timings, without stopping to think or reading anything below the prompts. Only afterwards open the models and listen back against them. The recording is the paper.
Under exam conditions Speaking is timed as the real test is: record each part in one take, without stopping to think, then listen back against the model. There is nothing to hand in — the recording is the paper.
The examiner asks you and the other candidate questions in turn. Answer, then extend — about twenty to thirty seconds each time. Under exam conditions, the partner playing the examiner reads the questions and moves on at the timing, whether or not you have finished.
Debrief — what the examiner was listening for in Part 1: that you answered the question that was asked, that you went past one sentence without being asked twice, and that you stopped. Three of these five questions were fishing for a specific tense: have learnt wants a present perfect handed back and then a past simple for the story; if you could wants I'd choose … because I'd be able to …; grew up wants used to. Listen to your recording and check whether you gave them.
A strong twenty-five-second answer to “Tell us about the area where you grew up”:
“A suburb, about half an hour from the city — the kind of place where every street looks the same and everybody's dad washed the car on a Sunday. It used to feel very small, and I couldn't wait to leave. I've been back twice this year, though, and it's oddly comforting: nothing has moved, including the car.”
Why it works: a place named in the first phrase; one concrete image rather than a list of facts; used to for the past habit and a present perfect for the recent visits; a small turn (though) that shows the candidate can hold two feelings about one place; and a last sentence that closes with a joke rather than a fade.
If your Part 1 went badly, it was almost certainly length: either one sentence and silence, or ninety seconds and the examiner cutting in. The fix is a shape — answer, reason, detail, stop — and you can practise the shape on any question in the world, in the shower, in under a minute.
Each candidate speaks alone for one minute about two photographs, and then the other candidate answers a short question about the same pictures for about thirty seconds. The partner playing the examiner reads the instruction, times the minute, and says “Thank you” at sixty seconds whether or not you have finished.
Debrief — the shape of a one-minute long turn. Fifteen seconds of comparison, thirty-five on the question, ten to finish. Listen to your recording with a clock: if the comparison ran past twenty-five seconds, the question got less than half the minute, and the question is where the marks are.
A strong long turn for the two meals:
“Both photographs show people eating, but one is a meal and the other is fuel. The first is a big family lunch — there must be a dozen people, three generations, and nobody is in a hurry. The second is a woman at her desk with a sandwich in one hand and the mouse in the other, and I doubt she could tell you what was in the sandwich.
As for what they might enjoy: for the family, I'd say the food is almost beside the point. What they're enjoying is the noise, the arguing, the fact that everybody turned up. The grandfather is probably enjoying being the one with the knife. The woman at the desk is harder — she might not be enjoying it at all — but if there is a pleasure there, it could be the twenty minutes with headphones in when nobody can ask her anything. Which is a kind of meal too, I suppose.”
Why it works: a comparison in one sharp sentence (a meal and fuel) and then detail; the question answered for both photographs, including the harder one; and speculation all the way through — must be, I doubt, I'd say, probably, might not, could be. Nothing is stated as fact, because nothing in a photograph is.
If your long turn went badly, listen for one of three things: you described for the whole minute and never reached the question; you answered for one photograph and forgot the other; or you said they are happy instead of they must be. Each is one habit, and each is fixed by one more minute in front of any two pictures on your phone.
You are given a question and five written prompts. Talk to each other for about two minutes; then the examiner asks you to make a decision together and you have about one more. The partner playing the examiner reads the instruction once, says nothing for two minutes, then reads the decision instruction.
Debrief — what was actually being marked here is Interactive Communication: whether you invited the other person in, reacted to what they said rather than to your own next idea, disagreed without stopping the conversation, and built something together. Listen to the recording and count the questions you asked your partner. Fewer than three in two minutes is the commonest reason a Part 3 stays in the middle band.
Thirty seconds of a strong Part 3:
A: Can I start with the one that looks like a joke? The tin can.
B: Go on — I was going to skip it.
A: That's why it's interesting. Before it, food went off. After it, you could feed a city in winter, or an army, or a family with no fridge. It's the least glamorous thing on the list and it might have changed more lives than the television.
B: I take the point, although — hasn't the fridge replaced it? Whereas the washing machine hasn't been replaced by anything. My grandmother says it gave her a day a week back.
A: A day a week is a strong argument. So we're both saying the boring ones matter more than the exciting ones?
B: I think we are. Which is a bit awkward for the mobile phone.
Look at what they are doing: A opens by asking permission, not by stating a view. B asks a question instead of taking a turn. B disagrees with a softener (I take the point, although) and adds new evidence. A summarises what they have both said (so we're both saying…), which is the move that lifts a Part 3 out of the middle band, and B confirms it with a joke. Nobody has made a speech, and both have said something the other could not have said alone.
The decision minute: “So shall we say the washing machine, and put the phone second?” — a proposal, with room to push back. If you agreed instantly, you left a mark on the table; a brief, polite disagreement before the decision is worth more than a fast one.
If your Part 3 went badly, it was one of two things: you went round the five prompts like a list, or one of you spoke and the other waited. Both are fixed by the same rule — never give an opinion without asking one back.
The examiner asks broader questions arising from Part 3, sometimes to one of you, sometimes to both. Answers here are thirty to forty-five seconds, and they want an opinion with reasons. The partner playing the examiner chooses four or five of these, in any order, and directs some to one candidate and some to both.
Debrief — what lifts Part 4: a position, a reason, an example, and one honest sentence against yourself. Four moves, forty seconds. Listen to your recording and find the fourth move; if it is not there, every answer was an opinion, and opinions alone do not reach the top band.
A strong Part 4 answer to “Do you think new technology makes people's lives easier, or just busier?”:
“Busier, mostly — though it's not the technology's fault. Every invention that saves time gets filled straight away with something else: the washing machine gave people a day, and the day got spent. My phone answers any question in two seconds, and so I ask it four hundred questions a day that I'd never have bothered with before.
I'd add that ‘easier’ and ‘busier’ aren't opposites, though. My grandmother's life was harder and quieter; mine is easier and louder. If you asked her which she'd choose, she'd take mine — and then complain about the noise. So easier, yes. But we've used the ease to become busy, and that was our decision, not the machine's.”
Why it works: a position with a limit in the first three words; a mechanism, not just a claim (every invention that saves time gets filled); a concrete example that costs nothing to invent; and a second paragraph that questions the question — which is what the top band sounds like. Notice how much of it is ordinary vocabulary used precisely. Nothing here is a rare word.
If your Part 4 went badly, it was usually one of three habits: answering an abstract question abstractly (bring it home to something you have seen); agreeing with your partner and adding nothing (respond, then add); or answering a twenty-year question in the present simple (will probably, I can imagine, by then most people will have). One habit, one week.
Finally — the recording. This was the exam-conditions mock, so the first take is the one that counts, and you have it. Listen to it once for content and once with your eyes shut, for the gaps. Then re-record only the part that went worst, once, and stop. The exam is the day after tomorrow's version of you, not tonight's fifth attempt.
Once, in real time, with a partner reading the examiner's lines and a clock running — Part 1 two minutes, Part 2 four, Part 3 four, Part 4 four. Then the debrief: listen to each recording once for content and once with your eyes shut, for the gaps. The gaps are the report.
The four criteria are Grammar and Vocabulary · Discourse Management · Pronunciation · Interactive Communication. This is the last mock before the real thing: if one criterion is weak in the recording, that is the one to practise this week, and the models below say how. Do not re-record all four parts tonight. Re-record the one that went worst, once.