A Listening Task Where Students Hear the Wrong Number
You play a two-minute recorded announcement for a class: a community-center schedule that includes a pool closure, a changed class time, and a registration phone number. The recording has mild background noise and a speaker with a fast rate. After one listening, students answer five questions. Results: 90% answer the pool-closure question correctly; 55% get the changed class time; 25% get the phone number. Several students wrote "fifteen" where the speaker said "fifty." Two students insist the speaker said the class moved to "Tuesday" when the recording says "Thursday." The transcript confirms Thursday. The recording cannot be paused with the current equipment, but you can replay it.
1. Diagnose why the three question types produced such different results.
Read the worked answer
The pattern is consistent with bottom-up decoding failure, not comprehension failure.
- Pool closure (90%): supported by redundant context - words like closed, maintenance, and until recur, and the meaning survives even if a word is misheard. Top-down knowledge fills gaps.
- Class time (55%): depends on a single number, with no redundancy. Numbers are phonologically fragile: -teen and -ty differ mainly in stress and a final consonant, which background noise can mask.
- Phone number (25%): a string of digits, each an independent decoding event. One misheard digit corrupts the answer, and there is no semantic context to repair it. This is the hardest listening condition on the recording.
- Tuesday/Thursday: Tuesday and Thursday share stress pattern and length; fast speech and noise blur the distinction. Students who "hear" Tuesday are reconstructing from expectation, a classic top-down override.
The task conflated comprehension with accurate sound discrimination.
2. Redesign the task so it measures comprehension fairly without eliminating the listening challenge.
Read the worked answer
Keep the authentic recording but change the task and the conditions.
- Pre-teach the decoding contrasts before listening: minimal pairs for -teen/-ty (thirteen/thirty, fifteen/fifty) and day names (Tuesday/Thursday). Two minutes of contrastive practice removes an unfair barrier without removing the challenge.
- Set a purpose for listening and allow two plays, with a 30-second pause between them for note-taking. Real-world listening rarely requires one-shot recall of a phone number.
- Change the response format for hard items: instead of open recall, offer the phone number as four written options that differ in one digit, and the class time as a multiple-choice pair (2:15 / 2:50). This isolates discrimination from memory load.
- Add a gist question ("What is the main purpose of the announcement?") so students who miss a digit can still demonstrate comprehension.
This preserves the listening demand while separating decoding accuracy from comprehension and memory.
3. Two students remain convinced the speaker said Tuesday. How do you handle this in class without embarrassing them?
Read the worked answer
Treat it as a perception lesson, not a correction contest.
- Replay the segment and ask the whole class to listen for the vowel quality and the number of syllables. Tuesday (/ˈtjuːzdeɪ/) and Thursday (/ˈθɜːrzdeɪ/) differ in the initial consonant and the vowel; the difference is real but subtle under noise.
- Normalize the error: explain that all listeners use expectation to fill in unclear sounds, and that knowing this makes you a better listener. Cite the class data - a quarter of the room missed the phone number - to show it is a decoding issue, not a failure.
- Give a strategy: when a day or number is critical, repeat it back mentally in a full phrase ("the class moves to Thursday") and check it against the rest of the announcement.
- Confirm privately with the two students afterward so they can save face, and note the segment for future contrastive practice.