Multiple-choice questions change the way a person listens. A learner who reads “Why did the train stop?” before the audio may wait for signal failure, fallen tree, or staff shortage and neglect the rest of the announcement. In a teaching lesson, give learners one uninterrupted encounter with the whole message before revealing detailed comprehension questions. Show the questions after that first listen, then replay the audio for focused checking.
This sequence suits a B1 office worker preparing for meetings or TOEIC while commuting on the Taipei MRT. The learner needs two abilities: building a reasonable account of a message as it unfolds and locating a requested detail under time pressure. A question-first routine practises the second ability every time. The first can disappear from the lesson.
Questions are part of the task
Listening questions look like neutral measurement. They also provide vocabulary, possible answers, and a prediction about what deserves attention. Jane Sherman's experiment compared four versions of listening tests with 78 participants: questions before the recording, questions after it, questions between two hearings, and no questions. Her Language Testing study treated question preview as a construct issue because it could focus attention, supply content, increase cognitive load, or encourage shallower processing.
That study is about test method. It does not establish a universal classroom winner. Its useful lesson is simpler: moving questions changes the listening activity. Teachers should choose the position deliberately instead of treating the worksheet layout as inevitable.
Consider a voicemail that says a supplier can deliver on Thursday, provided the revised order reaches them before noon on Tuesday. A previewed item about the delivery day can make “Thursday” feel like the destination of the whole message, although the Tuesday deadline controls what the listener must do. An uninterrupted first pass gives the condition and the result a chance to remain connected. That matters. The later question can still test the date, with the message structure already present in memory.
Listen first.
When the learner has to hold the caller's purpose, the conditional deadline, the requested action, and the consequence of missing it in one provisional account while speech keeps arriving, the exercise reaches beyond keyword detection and begins to resemble the work demanded by an actual voicemail.
On the first listen, ask for a representation with very little writing. Learners can choose a headline, arrange four event cards, or write seven words that preserve the message's route. “Delay, weather, platform change, refund after 10:30” is enough. Full sentences would pull attention toward spelling and grammar while the recording continues.
Build a two-pass lesson
A 96-second recording can support two distinct passes.
First pass: hide the questions. Give one purpose such as “Decide what changed and who is affected.” Play the complete recording without pauses. Learners record a short headline and mark their confidence from 1 to 4.
Second pass: reveal two to five questions. Allow 38 seconds for reading. Learners mark which answer their first representation already supports and circle the detail they still need. Play the recording again. Pair discussion comes after the replay, when each learner has committed to an interpretation.
This is a preference, and it has an edge. I'd rather see a learner misunderstand a complete message once and then repair it than watch them catch four isolated answers without knowing what happened. For a one-play public examination, however, practice must sometimes mirror the test. TOEIC candidates need experience previewing items quickly because that is part of managing the format.
Avoid turning the first pass into guessing
“Listen for the gist” is too vague. Learners may finish with “It was about transport,” which offers little to inspect. A constrained global prompt works better:
- Choose the best 8-word headline.
- Put the speaker's problem and decision in two boxes.
- Mark the relationship: complaint, warning, invitation, or report.
- Draw one arrow from cause to consequence.
Four choices appear here because different recordings need different representations. Use one for a lesson, not all four. A narrative wants event order. A workplace proposal wants problem, evidence, and recommendation. A station announcement wants change, affected passengers, and action.
Toapanta's 2022 intervention study compared instructional approaches that included a metacognitive cycle, awareness work, and vocabulary plus guiding questions. Its design does not isolate the exact timing rule proposed here, so it cannot prove that hiding questions on pass one causes better listening. It supports a broader point: listening instruction can include planning, monitoring, and reflection rather than a repeated play-answer-check routine.
Use questions to direct the replay
After the first listen, detailed questions become a diagnostic lens. If a learner's headline is accurate yet item 3 is wrong, the gap may be one number or negation. If the headline is already off course, replaying a ten-second fragment around item 3 treats the symptom.
Ask learners to label each uncertain answer before the replay: sound, word, reference, or idea. “Fifteen” versus “fifty” is a sound problem. Hearing it without knowing whether it refers to the contract or the deadline is a reference problem. Understanding every clause while missing the speaker's recommendation is an idea problem. The labels can be rough. Their purpose is to decide where attention goes next.
The confidence mark also matters. Our article on marking listening confidence before pressing replay explains how a prediction turns repetition into a test of an interpretation. Here, the same mark separates “I had no model” from “I had a model that the second sentence overturned.” Those require different teacher responses.
Keep assessment and teaching distinct
Question preview can be entirely appropriate in an assessment designed around selective listening. A traveller scans a departure board before attending to an announcement. A colleague enters a meeting already looking for a budget figure. Real purposes often precede speech.
Other situations arrive without a question sheet. A manager leaves a voice message. A classmate tells a story. A customer explains why an order is wrong. The listener must construct the situation before deciding which detail deserves another question. Lessons dominated by previewed items provide too little practice for that first encounter.
Try the change with one recording tomorrow. Cover the lower half of the worksheet. Give learners a headline box and one cause-to-consequence arrow, play the audio, and collect their first representations. Reveal the printed questions only then. The difference between those two sets of marks will tell you what the original worksheet was measuring.