Part 4 plays one news report read by a single speaker, about 1.5 minutes, then shows all 5 questions at once as sentence completions with drop-down menus and about 3 minutes on the clock. The statements stay hidden until the clip ends, so your notes are the only bridge between the audio and the answers.
In Part 4 you hear a short news report read by one speaker, about 1.5 minutes, and it plays one time only. The questions work differently from Parts 1 to 3: nothing is spoken after the clip. Instead, all 5 sentence completions appear together on one screen with drop-down menus and about 3 minutes on the clock, answered in any order. The demo above is the exact interface you get on test day.
Listening is scored by computer, one point per correct answer and no penalty for wrong ones. Part 4 contributes 5 of the 38 scored Listening questions and opens the harder half of the test, where questions appear on screen instead of being spoken.
Each drop-down is simply right or wrong, and a blank scores zero. There is no penalty for guessing, so make sure every menu shows an answer before the shared countdown runs out.
Part 4 has no separate score, its points join the Listening total out of 38, which converts to a CELPIP Level from 1 to 12 with difficulty balanced across test versions.
The mapping is one to one, CELPIP 9 equals CLB 9. Parts grow harder from 1 to 6 and Part 4 sits past the midpoint, so holding your accuracy here is what separates Level 8 from Level 9.
One experimental part can be mixed into the Listening test without counting, and it cannot be identified. Give the news item full effort like everything else, you will never know which part was the extra.
Your only preview is one line on screen, such as You will hear a news item about a community project. The questions stay hidden until the clip ends, so use those seconds to set up a note frame: who, what, when, where, why, plus a lane for names and one for numbers.
News items open with the outcome in the first sentence or two, and question 1 usually targets exactly that. Then track the arc: background, the event timeline with its time markers, one quoted voice, and how the story ends. Write last names, reports introduce people by full name and then switch to the surname.
After the clip, the 5 sentence completions appear together with one shared countdown of about 3 minutes. Skim every stem, they follow the order of the audio, then answer the ones you are sure of first. Any order is allowed, and answers can be changed until the timer ends.
Correct completions reword the audio, so treat options that echo exact phrases with suspicion, and reject details that were never said no matter how sensible they sound. When the countdown hits zero the test moves on for good, so leave no menu empty.
These are the exact on screen instructions plus the signal patterns that structure official news items. The attribution lines matter most, they mark the quoted second voice that questions love to target.
News items put the main outcome in the opening sentence, before many test takers have started listening seriously. Question 1 usually hangs on that lead, so be locked in from the very first word.
Correct completions paraphrase what was said, they never repeat it word for word. An option built from the audio's exact phrases is usually a distractor wearing the news as a disguise.
Most news items quote one person, an expert or a witness, and questions test who said what. Note the surname next to their point, a stem that starts with a name is unanswerable without it.
Distractors reuse real names, numbers, and places from the clip in combinations that were never said. A detail is only correct if the pairing matches your notes, not just the vocabulary.
All 5 statements share about 3 minutes. Sinking a minute into one stubborn item starves the rest, so bank the confident answers first, come back, and never let the timer catch an empty menu.
Yes. The official instruction says you will hear the news item only once, it starts automatically and cannot be paused, so your notes are the only replay. In real practice here, exam mode enforces the single play exactly like test day.
No. The statements appear only after the clip ends, and no question is spoken aloud. Your only preview is the one line introduction on screen, so squeeze it for topic clues and set up your note frame before playback starts automatically.
A 5W frame works best: who, what, when, where, why, plus a lane for names and one for numbers. Write surnames, dates, and amounts as you hear them, in point form with abbreviations. The questions follow the order of the audio, so tidy vertical notes double as an answer map.
