The Barangay Meeting Recording
You are reviewing a 6-minute audio recording of a barangay assembly for a listening-comprehension task. The recording was made on a phone placed near the front table. Two speakers talk at once for about 40 seconds around the 2:10 mark. The barangay captain speaks Taglish, switching to Tagalog when he discusses the budget, and a vendor from the audience answers in rapid Tagalog with strong Bicol intonation. At 4:30 a rooster crows and a tricycle passes. Your student, who has studied Filipino for two years, says she understood 'the general idea' but cannot answer who proposed the additional P15,000 for drainage. She also misheard 'pahintulot' as 'pahinga'.
1. The student claims general comprehension but fails on the drainage proposal. What does this reveal about her listening profile, and what should you assess next?
Read the worked answer
General-idea comprehension can coexist with poor selective decoding. She is likely using top-down prediction (budget talk = someone proposes money) rather than bottom-up phoneme recognition. The specific failure is a minimal-pair confusion: pahintulot (permission/approval) vs pahinga (rest) differ in stress and final syllables, and the noisy overlap at 2:10 plus the rooster at 4:30 masked cues.
Next steps: (1) Give a gist question first to confirm she has the frame, then a detail question targeting the exact speaker and amount. (2) Isolate the 3:50-4:20 segment, replay it twice, and ask her to transcribe only the noun phrase around the proposal. (3) Test whether she can identify who proposed it, since speaker attribution under overlap is a separate skill from word recognition. (4) Distinguish accent effects (Bicol intonation) from vocabulary gaps by asking her to paraphrase the vendor's sentence in her own words. If she can paraphrase but not quote, the issue is phonetic, not lexical.
2. How would you redesign the listening task to be fair given the recording's flaws, without discarding authentic material?
Read the worked answer
Keep authenticity but scaffold the noise. A fair design separates comprehension into layers:
- Pre-listening (30 sec): preview three key names and the topic 'drainage budget' so the student has a schema, reducing top-down load.
- First listen: one gist question only ('What was the meeting mainly about?').
- Second listen: pause before the overlapping section and ask a prediction question ('Who do you think will object?').
- Third listen: replay the 3:50-4:20 segment twice for the detail question about the P15,000 and the proposer.
- Transcript check: give the transcript only after answers, and ask her to mark where she misheard.
This preserves real-world listening (overlap, accent, background noise) while ensuring each question tests one skill. Marking the misheard point turns the task into metacognitive training, which is more valuable than a single right answer.