Listening is about prediction and pattern recognition. Build habits that scale across levels and measure first‑pass accuracy instead of replaying endlessly.

Core model

Preview the choices or question stems first; predict plausible answers.
Listen once for structure (who/where/what changes), then verify details.
Track discourse signals (でも、しかし、つまり、要するに、一方で) to anticipate turns.
Avoid pausing during timed practice; train attention and prediction.

N5–N4

Read choices before audio; predict likely answers.
Focus on key words: times, locations, numbers, prices, preferences.
Short shadowing daily (5–8 minutes) with slow, clear dialogs.
Dictation for 1–2 lines helps connect sound to kana.

N3

Track topic shifts and contrast markers (でも、しかし、一方で).
Alternate intensive (loop + shadow one line at a time) with extensive (play through without pausing) days.
Summarize in one sentence after each clip; check against transcript if available.

N2–N1

Anticipate endings from setups; verify with audio.
Note paraphrases and distractors that echo the passage but miss the question.
Practice light note‑taking for multi‑step items (names, numbers, decisions) without trying to write full sentences.
Shadow fast segments to get used to connected speech and reductions.

Weekly plan

4 days: short shadowing sessions (5–10 minutes) focused on rhythm and timing.
2 days: timed mini‑sets; no pausing, immediate review post‑set.
1 day: extensive listening (news briefs, graded material) for flow.

Metrics

First‑pass accuracy (no replay) on short sets.
Words correctly predicted before the final line is spoken.
Recall of numbers/names without transcript support.

Pitfalls

Over‑reliance on replaying; train attention, not memory of the clip.
Writing too many notes; capture only anchors (names, times, decisions).
Ignoring connectors; they signal contrast and conclusions.

How Kanji Koi helps with listening

Build cards with short audio‑anchored phrases you actually hear often.
Adaptive SRS keeps high‑frequency collocations fresh.
Offline access enables daily micro‑sessions and streaks.

Use Kanji Koi to attach a one‑sentence summary to any phrase card you add from listening practice. When it resurfaces, you recall both the phrase and the situation—improving answer prediction on similar JLPT items.

Short daily shadowing (5–10 minutes) plus one weekly timed set is enough to build speed.