Three students discuss whether their university should replace some live lectures with recorded lessons. One says recordings give students flexibility. Another agrees about flexibility but warns that attendance may fall. The third says a small pilot worked well, although it involved only one department.
A weak summary sounds like this:
The first speaker talks about recordings. The second speaker talks about attendance. The third speaker talks about a pilot. They all have different ideas.
The facts are not completely wrong, but the relationships have disappeared. There is a shared point, a warning, and a limit on the evidence. The lexical decision in Summarize Group Discussion is therefore not “Which impressive word can I add?” It is “Which verb shows what each idea is doing?”
For tests taken after 7 August 2025, Pearson places this task in Speaking & Writing. You listen to three people for up to three minutes, have 10 seconds to prepare, and make one recording of up to two minutes. The task assesses Listening and Speaking. Your response needs the main topic, the speakers’ contributions, and the connections among their ideas—not a memorized list of speaker labels.
Build relationships, not a transcript
Good notes separate content from function. A short symbol can preserve a relationship:
=shared point≠disagreement+ evidencesupportbutlimit or qualification
Then build a simple path: topic → common ground → disagreement → evidence → limit. You do not have to name every speaker mechanically. You do have to attribute important views accurately. “The group agreed” is false if only two speakers agreed. “The evidence proved” is too strong if a speaker only described one small example.
This is the difference between a list and a synthesis. A summary selects key points briefly. To synthesize is to combine material from separate sources into a coherent whole. A synthesis may reveal a pattern or support a conclusion, but it does not have to create a new conclusion.
Weak: “Speaker A mentioned access. Speaker B mentioned attendance.”
More precise: “While both speakers valued access, the second warned that recorded lessons could weaken attendance.”
The second version still summarizes. It also synthesizes because it connects the ideas.
Six useful lexical boundaries
Synthesize and summarize
To summarize is to state the main points briefly. To synthesize is to combine points from more than one source into a whole. They overlap: a strong spoken synthesis contains concise summaries. Do not claim that every synthesis produces a new insight. In this task, the practical move is to connect the speakers rather than report them in three isolated blocks.
Useful frame: “Taken together, the speakers suggest that flexibility is valuable, but its effects on participation need testing.”
Contend and claim
To contend that something is true is to put it forward as a position. It does not prove the position, and it does not show that you agree with it. Claim overlaps broadly, so it is not wrong in the same slot. Contend is formal and works best when you are neutrally reporting a developed position.
Use the full clause: “One speaker contends that recordings improve access.” Do not say “contends the benefit” or “contends with the benefit.”
Concede and agree
To agree is simply to share a view. To concede is to accept a disputed point, often after resistance or while keeping a broader position: “The second speaker concedes that recordings help commuters, but still favors live teaching.” The two words can overlap when a person moves toward agreement. Use concede only when the discussion really contains that movement. Ordinary common ground does not need debate-like language.
Understanding is only the first step.
Lyra Practice helps you retrieve and use high-value workplace expressions in realistic situations until they feel natural.
Start a practice session →Corroborate and confirm
To corroborate an account or finding is to strengthen it with additional evidence. It does not necessarily prove it. Confirm overlaps and can suggest that doubt has been removed by an authoritative fact or check. In a discussion, use corroborate only when a later speaker supplies evidence that supports an existing point.
Precise: “The third speaker’s pilot results corroborate the view that recordings can improve access.”
Too strong: “The pilot confirms that recordings work everywhere.”
Qualify and hedge
In discourse, to qualify a statement is to limit or modify it. A hedge lowers commitment with language such as may, perhaps, or appears to. These moves often overlap. Neither is automatically evasive. Because the assigned profile treats statement-modifying qualify as a recognition sense, the main listening skill is to notice when a speaker narrows a claim: “The pilot was successful, although it covered only one department.” In your response, plain language such as “She added that the sample was small” is safe and clear.
Juxtapose and compare
To compare is to examine similarities or differences. To juxtapose is to place two things side by side so a relationship becomes visible. The word itself does not say whether they agree, differ, or conflict. A response can juxtapose two views by putting them next to each other, but it must still state the actual relationship.
Weak: “The response juxtaposes the speakers.”
Precise: “The response juxtaposes the pilot’s positive result with the warning that its sample was narrow.”
Worked discussion: from notes to a connected response
Imagine this original discussion about keeping a campus library open later.
Speaker 1: Evening access would help students who work during the day. The whole building should stay open later so students can use every collection and study area. A student survey found strong demand for later hours.
Speaker 2: Later access would be useful, but the library should not open every floor. Staffing and energy costs would be high, so one study floor is enough at first.
Speaker 3: I agree that later access can help. A nearby university extended its hours and evening use increased. However, that campus used volunteer assistants, so its costs may not match ours.
Notes:
- Main topic: extend library hours
- Shared point: all see value in evening access
- Disagreement: full building versus limited floor
- Evidence: local survey + nearby university’s higher use
- Qualification: other university used volunteers; cost comparison is limited
Here is the weak speaker-by-speaker list:
The first speaker says working students need the library. The second speaker says one floor. The third speaker says another university had more use. The discussion is about library hours.
It drops the shared point, the reason for the disagreement, and the limit on the outside evidence.
A precise 90–120-second model
The discussion focused on whether the campus library should stay open later. All three speakers saw a benefit in giving students more evening access, especially those who work during the day. The first speaker supported a broad extension and cited a student survey showing strong demand. The second speaker agreed that later access would help, but opposed opening the whole building because of staffing and energy costs. Instead, this speaker proposed opening one study floor as a trial. This approach would test demand locally without committing the library to the cost of operating every floor. The third speaker added evidence from a nearby university, where evening use rose after hours were extended. That example corroborated the general case for later access, but the speaker also qualified it by noting that the other university relied on volunteers, so its cost experience may not apply directly. In other words, the outside result supports the access argument more clearly than the cost estimate. Overall, the speakers shared the goal of improving access, while differing over the scale of the change and how confidently the outside example could guide the decision.
The response attributes the survey to the first speaker, the limited trial to the second, and the outside example and its limit to the third. It does not pretend the evidence proves a universal result. It also connects ideas instead of attaching one sentence to each person by formula.
Common mistakes
- Flattening all views into agreement. Shared support for later access does not erase disagreement about opening the full building.
- Turning a position into a fact. “The speaker contends that costs will rise” reports a position; it does not establish the forecast.
- Calling every agreement a concession. Use concede only when someone accepts a contested point or limitation.
- Calling repeated information corroboration. A second source must add support. Mere repetition is not new evidence.
- Treating a qualification as retreat. A limit can make a claim more accurate without canceling it.
- Using juxtapose as a verdict. Placement alone does not mean conflict. Name the agreement or contrast.
- Forcing target words. These words help you hear and describe relationships. A plain, accurate verb is better than a formal word used in the wrong frame.
Quick check
Two speakers say recorded lessons help commuters. One gives survey data from a second campus. Which sentence is safest?
- A. “The survey proves that recorded lessons help every commuter.”
- B. “The second campus data corroborates the speakers’ view that recordings can improve access.”
- C. “The data contends that access improved.”
A speaker says, “The plan is useful, although the trial involved only 20 students.” What happened to the claim?
- A. It was qualified.
- B. It was rejected.
- C. It was confirmed conclusively.
Two speakers both support a pilot. Neither previously opposed it. Which verb is the safer description?
- A. conceded
- B. agreed
Answers and explanations
- B. The second dataset adds support to an existing view. It does not prove a result for every commuter. Data can support a contention, but data itself does not normally contend that something is true.
- A. The speaker keeps the positive view but limits its reach. That is qualification, not rejection or conclusive confirmation.
- B. The facts show shared opinion, not movement from opposition. Conceded would add a dispute that the prompt never supplied.
Practice prompt
Three students discuss replacing printed course packs with digital files. One values lower cost. A second agrees about cost but worries about eye strain. A third reports that downloads rose in a small pilot, while noting that the pilot covered only first-year students. Take five relationship notes, then record a 90-second summary that preserves the shared point, the concern, the evidence, and its limit.
For more task practice, explore the PTE Academic vocabulary hub and the broader Exam English hub.
Sources
Related
- Summarize a Meeting in English Professionally
- Juxtapose vs Compare: What's the Difference?
- Hedge vs Qualify: What's the Difference?
- Resonate vs Corroborate, Confirm, Reflect, or Mirror
Lyra Practice is not affiliated with or endorsed by Pearson.