What this is

A simple system that captures your mistakes in Listening, Reading, Writing, and Speaking, spots patterns, and converts them into tiny actions you can finish today. It keeps entries short, decisions clear, and progress visible.

What you will achieve

• Fewer repeat errors in all sections
• A small set of high value targets each week
• Faster decisions about what to study next
• A personal error bank that upgrades your language

10 minute setup

Create one page or sheet with these columns:

Date | Section | Task/Item | Symptom | Cause | Fix (rule or cue) | Drill (2-5 min) | Priority | Status | Retest date | Result 

Color code Priority: 1-3 green, 4-6 amber, 7-10 red.

Priority score
Frequency 1-5 × Severity 1-5, then add Fixability 0-2

• Frequency: how often it happens
• Severity: how many marks it risks
• Fixability: 0 hard, 1 medium, 2 easy
  Example: TFNG trap weekly (4) that costs many marks (4) and is easy to re-train (2) → 4×4+2 = 18 red.

Error tags by skill

Listening
Numbers and names, directions on maps, spelling variants, dates and times, distractor phrases, multi choice stems, pace loss.

Reading
TFNG logic, heading match, reference words, vocabulary in context, location scanning, table note completion, time overuse on one item.

Writing
Task focus drift, paragraph unity, topic sentence clarity, overlong sentences, verb tense control, articles, comma splices, over formal word choice.

Speaking
Long pauses, short answers, repair overuse, tense slips, weak linking, mis-stress, confusable consonants or vowels, flat intonation.

The Auto-Log workflow (30 seconds per entry)

1. Capture
   Write the exact item or sentence. Add a screenshot or timestamp if possible.
2. Tag
   Choose 1 skill tag and 1 subtype tag from the lists above.
3. Diagnose
   Name the cause in plain words: missed keyword, rushed scan, grammar gap, unclear plan.
4. Fix rule or cue
   One line only. Examples:
   • Reading TFNG: If no direct proof, choose Not Given.
   • Listening numbers: Say the number aloud before writing.
   • Writing articles: If noun is general and singular, try a or an.
   • Speaking length: Part 3 30 to 40 seconds, close with one line.
5. Drill
   Pick a 2 to 5 minute micro-task that trains the fix.
6. Priority
   Score Frequency, Severity, Fixability. Color it.
7. Schedule retest
   Add the next date. Use 48 hours, then 7 days.

Target selection rule

Work on at most 3 targets at a time: one language target, one reading or listening process target, one writing or speaking structure target. Everything else sits in the log until you finish these.

Micro-drill menu

Listening (2 to 5 min each)

• Numbers string dictation: 10 items, check, repeat missed ones.
• Distractor hunt: replay 1 segment, pause at distractor words, say why wrong.
• Map pre-view cues: north, opposite, across from, first left. Shadow 30 seconds.

Reading

• TFNG proof scan: highlight words that prove or disprove. If none, mark NG.
• Headings fast map: write a 5 word gist for each paragraph, then match.
• Location sprint: practice finding names, dates, numbers before reading details.

Writing

• Topic sentence lab: rewrite 3 topic sentences so each has topic, angle, and result.
• Article and plural pass: scan 150 words, fix a or an, the, plurals.
• 6 line paragraph cap: stop at 6 lines, add a one line close.

Speaking

• 3 step answer: answer, reason, tiny example. Record 3 prompts.
• Pacing with timer: 35 seconds responses for Part 3.
• Minimal pairs: 60 seconds on a pair like ship vs sheep, record, compare.

Sample entries

2025-10-08 | Reading | Passage 2 Q13 | Picked T not NG | Searched outside paragraph | If no proof words in same paragraph choose NG | TFNG proof scan 4 Qs | 18 | Doing | 2025-10-10 | 3/4 correct 2025-10-08 | Writing | Task 2 body | 3 ideas in one para | No topic sentence plan | Start each para with claim + scope | Topic sentence lab 3 lines | 14 | Doing | 2025-10-11 | clear 2025-10-08 | Listening | Sec 1 | Wrong phone number | Wrote before hearing check digit | Say number before writing | Numbers string dictation 10 items | 12 | Queue | 2025-10-10 | 9/10 

Mini diagnostics to find true causes

• Symptom I misread TFNG
  Likely cause searching outside the proof zone
  Check did I highlight proof words
• Symptom paragraph too long
  Likely cause no topic sentence or two ideas mixed
  Check can I label claim and support
• Symptom short speaking answers
  Likely cause no example layer
  Check did I add a tiny example after reason

Daily routine using the log (15 to 30 minutes)

1. Open the log. Read your 3 active targets.
2. Run the linked micro-drills back to back.
3. Do one short timed set in the weak skill.
4. Add new entries for any mistakes.
5. Tick Status when a target holds in two retests.

Weekly review in 10 minutes

• Sort by Priority then by date.
• Close green items that survived two retests.
• Promote one amber to active if it is frequent and fixable.
• Rewrite any Fix that is vague. Fix lines must be testable.
• Add one sentence plan for the coming week.

Stoplight summary

• Red: train daily
• Amber: train alternate days
• Green: spot check only

Writing and Speaking upgrade lines

Writing fix lines

• One idea per paragraph.
• Keep examples 1 to 2 lines.
• Prefer clear verbs over heavy nouns.
• After drafting, do a verbs and articles pass only.

Speaking fix lines

• Use a 4 word hook in Part 2, then points A B C D.
• Close with In short or Overall and one result line.
• Replace very with a precise adjective when possible.

Metrics that matter

Track once a day in the log footer.

• Listening accuracy percent and average revisits per section
• Reading score per passage and words per minute on skim
• Writing error density per 100 words and paragraph count
• Speaking words per minute and pauses longer than 1 second per minute

Formula hints:
Error density = total errors divided by words times 100
Pace = words divided by minutes

If-then rules to protect time

• If an item takes more than 60 seconds in Reading, guess, mark, move.
• If you lose place in Listening, skip one answer and rejoin at the next keyword.
• If a Writing paragraph reaches 7 lines, stop and close it.
• If a Speaking answer passes 45 seconds in Part 3, wrap with a short line.

Common pitfalls

• Logging paragraphs instead of one line fixes
• Collecting errors without scheduling retests
• Chasing low frequency mistakes
• Changing too many targets at once

14 day target cycle

• Day 1 to 2: Fill the log, choose 3 targets.
• Day 3 to 5: Daily micro-drills and one timed set.
• Day 6 to 7: Retest and update Priority.
• Day 8 to 10: Replace any target that is green twice.
• Day 11 to 12: Mixed set LRW and one speaking session.
• Day 13: Compact mock and full log review.
• Day 14: Lock next 3 targets and schedule retests.

Final note

Keep entries short, fixes testable, and targets few. The log is your map. Train red every day, amber on alternate days, and green only to check. Repeat this cycle and your mistakes will fade while your score rises with less effort.