12MyCELPIP

Spoke 15 · Listening Parts 3 & 4

CELPIP Listening Parts 3 and 4: Tips and Strategies

Last updated: June 2026

Parts 3 and 4 are the sections most test-takers lose unexpected points on. Part 3 looks straightforward — one speaker, factual information — but information density and the “NOT mentioned” question type trip many candidates. Part 4's multi-speaker format is the hardest attribution challenge in the module.

Part of the CELPIP Listening Module: Complete Guide.

Share:RedditFacebook

Why these parts are harder than they look

Part 3
  • ·No conversational back-and-forth to cue what's important — all information arrives in a single stream
  • ·Information density is high: dates, numbers, names, places appear in quick succession
  • ·The 'NOT mentioned' question type requires real-time tracking of all options simultaneously
  • ·No emotional or tonal variation to signal importance — everything sounds equally factual
Part 4
  • ·Three or more voices mean three sets of potentially contradictory information to track
  • ·Speaker positions evolve during the discussion — you must update your mental model continuously
  • ·Questions often ask about what ALL speakers agree on, which requires sustained tracking across the full clip
  • ·Questions about speaker attribution penalise heavily — one misattribution can affect 2–3 questions

Part 3: Question types and how to handle each

Main purpose

In every Part 3 clip

Example question

What is the main purpose of this announcement?

Answered by the opening sentence. The purpose is stated in the first 10–15 seconds: 'This is a public service announcement about…', 'Today's news segment covers…', 'Residents are reminded that…'. Write the main topic noun from the opening line.

Specific fact extraction

2–3 questions per clip

Example question

When will the new service begin? / What is the fee for late registration?

Numbers, dates, times, and fees are the most tested specific facts. Write every number you hear, immediately. You cannot infer numbers — they must be captured in the moment. If you miss a number, your best strategy is to leave the question and use process of elimination from the answer options.

NOT mentioned

1 question per clip

Example question

Which of the following is NOT mentioned in the announcement?

The four options will all seem plausible — they are all related to the clip's topic. As you listen, physically check off (with a tick mark) options that appear in the audio. The one you never tick is the answer. Do not rely on memory after the clip; you must track in real time.

What happens next / recommended action

1 question per clip

Example question

What does the speaker recommend listeners do? / What will the city do next?

The recommendation or next action almost always appears in the final 20 seconds of the clip. Listen for closing signals: 'Residents are encouraged to…', 'For more information…', 'The city will…', 'Call…'. This is why you must not relax your attention at the end of a clip.

📅

Free 30-Day CELPIP Study Schedule

A day-by-day plan covering all 4 sections — Reading, Listening, Writing, and Speaking. Download the PDF and follow it straight to test day.

Or see what's inside first →

Part 4: Question types and how to handle each

Speaker attribution

2–3 questions per clip

Example question

Which speaker believes that the new policy will save money?

Use the speaker-mapping table. Before the audio starts, draw: A | B | C with rows for 'position' and 'concern'. Fill in each speaker's main point as they speak. Speaker attribution questions are answered directly from your table — you do not need to remember the audio.

Agreement / consensus

1–2 questions per clip

Example question

What do all three speakers agree on? / What point does Speaker B eventually accept?

Consensus is almost always stated near the end of the discussion. Listen for synthesis phrases: 'So we all agree that…', 'I think we can say…', 'That seems fair to everyone'. Add a 'AGREE:' row to your speaker table to capture this moment.

Proposal or suggestion

1 question per clip

Example question

What does Speaker C propose the team do?

Track proposal language: 'How about we…', 'I suggest we…', 'What if we…', 'My recommendation would be…'. In Part 4, proposals often come from the speaker who contributes least — a quiet third speaker often delivers the key proposal that resolves the conflict.

Attitude / emotion

1 question per clip

Example question

How does Speaker A feel about the proposed change?

Attitude is signalled by vocabulary intensity and modals: 'I'm very concerned about…' (worried), 'I'd strongly support…' (enthusiastic), 'I'm not sure that's the best approach' (skeptical). Listen for the first thing each speaker says — attitudes are often established immediately and then elaborated.

Sample scenarios with questions

Read each scenario like a transcript. Then answer the questions before looking at the answers.

Part 3 example

Audio: A recorded city council announcement about changes to recycling collection. 75 seconds.

Transcript

Good evening. The City of Hamilton would like to inform residents that beginning Monday, March 10, blue box recycling collection will move from Tuesday to Wednesday in Wards 4 through 7. Residents in these areas should place their bins at the curb by 7:00 AM on Wednesday mornings. Glass must now be separated from other recyclables and placed in a clear bag beside the blue box. Residents with questions may call 905-546-2489 or visit hamilton.ca/recycling. Thank you.

1. What is the main purpose of this announcement?

Answer

To inform residents of a change to recycling collection schedule.

2. On what day should affected residents now put out their recycling?

Answer

Wednesday.

3. What must residents do with glass items?

Answer

Place them in a clear bag beside the blue box.

4. Which of the following is NOT mentioned in the announcement? A) New collection day B) New pickup time C) Glass separation D) New composting program

Answer

D — composting is not mentioned.

Part 4 example

Audio: Three coworkers discuss whether to switch from in-person to hybrid meetings. 90 seconds.

Transcript

Speaker A (Maya): I think we should move to a fully hybrid model. Attendance has been low for in-person meetings anyway, and people are more productive when they can join from home. Speaker B (David): I disagree — the collaboration suffers. Last quarter's planning session was a disaster because half the team was remote and the other half couldn't hear them. Speaker C (Priya): Both of you make valid points. What if we kept major project kick-offs in person but allowed hybrid for weekly check-ins? That way we get both. Speaker A: I could live with that. Speaker B: That's actually a reasonable compromise. I'd support it.

1. What is Speaker A's initial position?

Answer

Fully hybrid — all meetings allow remote attendance.

2. What concern does Speaker B raise?

Answer

Collaboration suffers when the team is split between remote and in-person.

3. What does Speaker C propose?

Answer

Keep major kick-offs in person; allow hybrid for weekly check-ins.

4. By the end, what do all three speakers agree on?

Answer

Speaker C's compromise — hybrid check-ins, in-person kick-offs.

Share:RedditFacebook

One CELPIP tip per week

Free, from a 12/12 scorer.