How CELPIP Listening Works: 6 Parts, 38 Questions
Six parts, 38 scored questions, and audio that plays exactly once. Before any technique is worth learning, you need a map of what is coming and when.
The section in one page
CELPIP Listening runs six parts and 38 scored questions, and takes somewhere between 47 and 55 minutes depending on how long the clips run. Before Part 1 there is an unscored practice task. It does not count toward anything, and it is the only point in the section where nothing is at stake.
The six parts are not variations on one task. They change speaker count, they change register, and they change how the questions are delivered. Parts 1 to 3 give you the questions one at a time, spoken aloud, with about thirty seconds each. Parts 4 to 6 put the questions on screen after the recording, with one time limit covering the whole screen.
That split is the single most useful thing to know about the section, because it changes what your notes are for. In Parts 1 to 3 you are answering from memory a few seconds after hearing the answer. In Parts 4 to 6 you are answering from notes, minutes later, with several questions competing for the same recall.
What each part actually asks
Part 1, Listening to Problem Solving. Two people, a problem, and a decision that moves across the conversation. Eight questions. The answer to an early question is often revised later in the same clip, which is the trap.
Part 2, Listening to a Daily Life Conversation. Shorter, friendlier, five questions. Candidates relax here and lose marks to detail rather than difficulty.
Part 3, Listening for Information. One speaker, dense with facts. Six questions. Numbers, dates and proper nouns arrive faster than you can write them.
Part 4, Listening to a News Item. A register shift into broadcast English, five questions. News clips signal importance with structure, and learning that structure is most of the work.
Part 5, Listening to a Discussion. Three speakers, eight questions, and the hardest part of the section by a clear margin. People change their minds mid-discussion and the answer is rarely in one place.
Part 6, Listening to Viewpoints. One speaker giving an extended opinion, six questions, all of them on screen at once. Separating what is asserted from what is merely mentioned is the skill being tested.
Where the time actually goes
The clock is not evenly distributed, and the parts that feel hardest are not the ones that cost the most time. Parts 1 to 3 are paced for you: the audio runs, a question is read, you get roughly thirty seconds, and the test moves on whether or not you are ready. You cannot fall behind in these parts because there is nothing to fall behind on.
Parts 4 to 6 are where time is lost. The whole screen of questions shares one budget. Spending ninety seconds on a question you were never going to get leaves the last two questions on that screen unanswered, and those were probably the easy ones.
The rule that follows is unpopular and correct: in Parts 4 to 6, answer everything you are sure of first, then return. The questions are not ordered by difficulty and there is no reward for going in sequence.
The three habits that separate the bands
They write less, not more. Weak candidates transcribe. Strong candidates write six or seven anchors per clip: a name, a number, a decision, a reversal. Transcription steals the attention the listening itself needs.
They abandon a lost question immediately. The audio does not wait. Four seconds spent recovering a missed detail costs the next question too, and that is how one miss becomes three.
They never leave a blank. There is no penalty for a wrong answer in CELPIP Listening. A blank is a mark handed back. Even an uninformed guess on a four-option question returns one mark in four, and an informed one does considerably better.
In Parts 4 to 6 the questions appear after the audio, so write down anything numeric the moment you hear it. You will not get a second chance, and numbers are the first thing memory drops.
How the six parts are scored, and why that matters to you
Every one of the 38 questions is worth the same. Part 5 is the hardest part of the section and it is worth no more per question than Part 2, which most candidates find easy. That single fact should change how you practise: the marks are not where the difficulty is, they are where the questions are.
Part 1 carries eight questions and Part 5 carries eight. Between them that is sixteen of the 38, or roughly forty two per cent of the section, sitting in the two parts with the most speakers and the most changes of position. Parts 2 and 4 carry five each. If you have limited practice time, the arithmetic tells you where to spend it.
Your raw score out of 38 is converted to a CELPIP level from 1 to 12, and that conversion is not linear. The gap between 30 and 32 correct sits across a band boundary for many candidates, while the gap between 20 and 22 often does not move the level at all. Near the top of the range, two questions is the difference between a CLB 8 and a CLB 9, which is the difference that decides most immigration applications.
This is also why chasing a perfect score is the wrong target. Nobody needs 38 out of 38. Work out the level you actually need, look at what raw score that corresponds to, and aim two or three questions above it so that a bad clip on the day does not cost you the band.
What this section is not testing
It is not testing your memory for trivia, and it is not testing whether you can understand a Canadian accent. The speakers are ordinary Canadian voices at ordinary speed. If a clip is hard, it is hard because of what is being argued, not how it is pronounced.
It is also not testing note-taking as a skill in itself. Your notes are never submitted and never scored. They exist only to get information from the clip to the question, and any notation that does that is correct, including one in your first language.
What to do with this before your next test
Knowing the shape of the section is worth a mark or two on its own, but only if it changes what you do. Three things follow from the map above, and each of them is a decision to make before the audio starts rather than during it.
First, decide your Parts 4 to 6 order now. You will answer the questions you are sure of and come back, and you will do that every time, so it does not need deciding under a clock. Candidates who make that decision live lose the last two questions on a screen to a single stubborn one.
Second, decide what your notes are for. In Parts 1 to 3 they are almost useless, because you are answering seconds after hearing the answer and writing steals the attention listening needs. In Parts 4 to 6 they are the whole game. Writing in both places, at the same density, is the commonest habit that caps a score in the sevens.
Third, agree with yourself that a blank is never acceptable. There is no penalty for a wrong answer, so the arithmetic is not close: a guess returns one mark in four at worst, a blank returns nothing at all. Candidates leave blanks anyway, usually because they ran out of time on a screen they had not planned an order for, which is the first decision arriving late.
Set up your three columns for Part 5 before the clip starts. You are told there are three speakers before you hear a word, and building the grid while the audio runs costs you the opening exchange, which is where positions are stated.
Practise it
Five questions on the structure itself. Knowing the shape of the section is worth a mark or two on its own.