Four or Fourteen: A Listening Item Lost in Station Noise
An open-school listening test plays a recorded station announcement twice to 25 candidates. Item 4 asks: 'Which platform will the 5:40 express to Ambala depart from?' The recording says: 'The 5:40 express to Ambala will now depart from platform four instead of platform fourteen.' The key is 4. In the scripts, 14 candidates wrote 14, 9 wrote 4, and 2 left the answer blank. The recording has a 3-second burst of station noise over the words 'four instead of'. The invigilator reports that several candidates asked for a third playing, which the instructions do not permit. The item is one of six, each worth 1 mark; the listening section contributes 10% of the final grade. The audio file cannot be edited before the next sitting, but a transcript is available. The examination committee must decide whether the item is sound and what to do with the current scripts.
1. What does the split between '14' and '4' suggest about what Item 4 is actually measuring? Use the recording details.
Read the worked answer
The split suggests the item is measuring bottom-up sound discrimination under noise and short-term memory, not comprehension of the announcement. The key contrast is 'four instead of fourteen', and the noise burst falls exactly over 'four instead of'. That is the contrastive phrase carrying the change of platform. Candidates who heard the more salient number 'fourteen' or who missed the word 'instead' would reasonably write 14. The 14 who chose 14 were not necessarily careless; they may have lost the one cue that reverses the meaning.
Construct problem: a listening item should test whether candidates understand the message, including the change. Here the acoustic confound makes the item depend on hearing a suffix through noise. The 9 correct responses may reflect either good comprehension or lucky attention to the second number. With 14 choosing the distractor, the item has poor discrimination and is contaminated by an irrelevant factor. The transcript confirms the intended content, but the audio condition, not the transcript, determined performance. That is a validity problem, not a candidate problem.
2. Should the item be retained, discarded, or re-administered for the current 25 candidates? Give a defensible decision and the reasoning.
Read the worked answer
Discard Item 4 for this cohort and rescore the listening section out of 5. Retaining it would penalise 14 of 25 candidates for an acoustic confound that the test instructions did not allow them to resolve with a third playing. Re-administering only Item 4 would be unfair because those who already heard the recording twice would gain an advantage, and a third playing was not permitted in the original conditions. Re-administering the whole section is disproportionate for a 10% component and would require a new audio file, which is not available before the next sitting.
Rescoring: multiply each candidate's raw score out of 5 by 6/5 so the section remains comparable with previous cohorts. Record the original Item 4 responses, the noise timing, the transcript, and the decision. Flag the audio file for re-recording before reuse: move the noise away from the contrastive phrase or remove it. If the board insists on retaining the item, the only defensible alternative is to accept both 4 and 14, but that would make the item unable to distinguish comprehension from mishearing, so discarding is cleaner.
3. Rewrite Item 4 and/or the audio condition so that the item tests listening comprehension rather than a number trap.
Read the worked answer
Rewrite the item: 'The 5:40 express to Ambala has been moved. Which platform will it now depart from?' Options: (A) 4 (B) 14 (C) 5:40 (D) Ambala. Add a second item: 'Why has the platform changed?' with options about engineering work, a delay, a different train, or a cancelled service, so the section tests both detail and reason.
Change the audio condition: record the announcement with the noise placed after the key phrase, not over 'four instead of'. Alternatively, keep the noise but add a 1-second pause before the contrastive phrase and ensure the words 'instead of' are clearly audible. Do not rely on a transcript during the test; the transcript is for review and teacher use.
Reasoning: the original item turned on a suffix contrast masked by noise, so it measured hearing under adverse conditions. The rewrite still requires candidates to follow the change, but the answer options make the contrast meaningful: 14 is the old platform, 4 is the new one. The added reason item checks whether candidates understood the announcement as a message, not just a number.