Listening Comprehension
Listening comprehension
This domain checks whether you can follow spoken English in announcements, short dialogues, and teacher-style monologues and choose the response that matches the situation. Gunma-style listening items often test who is speaking, where the exchange happens, what will happen next, and which detail corrects an earlier statement.
Core rules
- Read the question and choices before the audio starts when the format allows; predict the kind of answer (time, place, reason, action).
- Track speaker roles and turn changes. A question from one speaker is often answered indirectly by the other.
- Listen for discourse markers: but, actually, instead, not until, only, except, so, because. These often carry the correcting detail.
- Keep listening after a number or name appears. The first number is often a plan; the final number is often the confirmed fact.
- For picture/response items, identify the function first: request, apology, invitation, permission, suggestion, or information question.
Worked example
Audio: “The library was supposed to open at nine, but it won’t open until ten thirty because of a staff meeting.” Question: When will the library open? Choices: 9:00, 9:30, 10:00, 10:30. Answer: 10:30. “Was supposed to” presents the old plan, and “but … not until” corrects it. The confirmed opening time is 10:30; 9:00 is a trap for students who catch only the first time.
Traps that cost marks
Matching a word in the audio to a choice without checking meaning; missing negation (“not,” “never,” “hardly”); confusing similar sounds (thirteen/thirty, walk/work); choosing an option that answers a different wh-word; and stopping at the first detail instead of the final decision.