Prediction & Pre-Listening Strategy

What you'll learn: how to use the brief preview window before each part's audio to predict content, question focus, and likely distractor types.

The preview window is not just reading time

Before each part's audio plays, CELPIP gives you a short window to look at the upcoming questions (and often the answer choices). Many test-takers treat this purely as "extra reading time" and skim it passively. Used correctly, it's closer to a prediction exercise: every question stem tells you, in advance, what kind of information to listen for, which lets you allocate attention before a single word of audio has played.

Predicting from question stems

Question stems almost always reveal a category of answer before you hear anything:

• "What time does..." → listen for a specific time, and any correction to an original time.
• "Why did [speaker] decide to..." → listen for a "because" clause tied to that speaker.
• "Which of the following was NOT mentioned..." → you'll need to track a full list, not just one fact — flag this as a higher-attention question.
• "What does [speaker] suggest instead?" → listen for a proposal that follows an initial rejection.

Reading even three or four question stems before the audio starts means you walk in with a checklist of specific things to catch, rather than trying to absorb everything equally and sort it out afterward.

Predicting from answer choices

When answer choices are visible during preview, scan them for what they have in common — that overlap tells you exactly what distinction the audio is about to draw. If three answer choices are all times ("2:00," "2:30," "3:00"), you know precisely to listen for one specific time, possibly with a correction ("originally 2:00, but moved to..."). If answer choices are near-synonyms ("frustrated," "disappointed," "annoyed"), you know the audio will require picking up a precise emotional word, not just a general negative tone.

Worked example

Preview text for an upcoming question: "According to the speaker, what is the main reason the trip was postponed?" with answer choices: (A) bad weather (B) a scheduling conflict (C) budget concerns (D) low interest.

Before the audio plays, you already know: there's a trip, it was postponed, and the audio will state a reason that matches exactly one of four plausible categories. You're not listening generally — you're listening specifically for the word or phrase that maps to weather, scheduling, budget, or interest, and prepared to note it the instant it's said (postponed b/c ___).

Prediction checklist

• I read every visible question stem before the audio starts, even if I can't read every word.
• I note what category of answer each question wants (time, reason, person, location, quantity).
• I scan answer choices for a shared pattern that reveals the exact distinction being tested.
• I flag "NOT mentioned" or "all of the following except" questions as needing full-list tracking, not single-fact tracking.

Common mistakes

• Skipping the preview window entirely to "save energy" for the audio, losing a free prediction advantage.
• Reading question stems but not translating them into a specific listening target before the audio starts.
• Ignoring answer choices during preview when they're visible, missing the pattern they reveal.
• Treating every question as equally demanding, instead of flagging list-based ("NOT mentioned") questions as needing extra attention.

Recap

The preview window is a prediction tool, not just reading time — question stems reveal the category of answer to listen for, and visible answer choices reveal the exact distinction being tested. Building the habit of translating a stem into a specific listening target before the audio starts turns passive comprehension into targeted, efficient note-taking.