Five Questions Per Lesson: Drills That Transfer to OET Tasks
Problem: candidates grind hundreds of identical gap-fills — then freeze on real letters, because drills never resembled tasks. Answer in brief: drill in five formats per lesson (explain, rewrite-to-form, diagnose, judge, reflect), accept valid variants instead of matching models, stay inside supplied facts, and scope edits and speaking turns honestly. Method, models and 12 tasks below.
Learning outcomes
By the end you will be able to: work five drill formats per lesson (explain, rewrite, diagnose, judge, reflect) with purpose; distinguish form-specified items from open items and treat variants correctly; stay inside supplied facts without inventing clinical advice; scope paragraph edits as editing exercises and speaking turns as focused practice; and code every attempt (G/E/S) for answer-guide verification. All guidance follows the source book, OET Grammar: Practice workbook (OET Grammar Practice Workbook) (Jobin Thomas, September 2026 edition) — 120 focused questions (five per lesson), six paragraph edits and six speaking tasks with question codes, verified by full-text decode and lesson sampling. Independent preparation; not affiliated with or endorsed by OET — fictional patients, language examples only.
The five drill formats and what each trains
Each lesson’s five questions vary by design. Explain (“Is X supported? Explain”) trains judgement — the reason habit from the answer-key volume. Rewrite-to-form (“rewrite using might”) trains controlled production under constraint. Diagnose (“what is wrong with…?”) trains error detection, the review-pass skill. Judge (“does this establish…?”) trains claim-vs-evidence discipline. Reflect (“why can a correct sentence still mislead? Give your own example”) trains generalisation — the transfer moment. Rotate formats deliberately: candidates who only rewrite never learn to judge, and judging is what review passes require. The fifth question of each lesson is usually the transfer test — never skip it.
Form-specified vs open items
Some tasks specify a form (“rewrite using might”); others allow any accurate version. The distinction matters: form-specified items must be practised in that construction even when another sentence could convey the meaning — the constraint is the training. Open items accept equally valid wording: do not treat every difference from the model as an error — compare reasoning (facts preserved? qualifier kept? reference clear?) instead. Mark attempts with their codes (G01-1, E03, S05) so answer-guide comparison stays aligned; code discipline across separate books prevents the common mess of verified-but-unfiled practice.
Supplied facts only: the clinical-advice boundary
Workbook drills supply facts; candidates must not invent clinical advice. “Walking possible with assistance” never becomes “walks independently” (limit dropped) or “should walk hourly” (advice invented). The boundary has three rails: use only supplied facts; preserve every qualifier; convert form without adding content. This is exam realism, not just honesty — OET letters are assessed on accurate use of case notes, and invented content fails content criteria regardless of grammar. When notes are thin, write thin accurate sentences; padding with plausible medicine is the costliest habit in OET preparation.
Scoping edits and speaking turns honestly
Paragraph tasks (E01–E06) are short language-editing exercises, not complete OET Writing tests — they train repair (status, actor, reference, articles) inside one paragraph, without purpose/content/genre demands of full letters. Spoken tasks (S01–S06) are focused communication practice, not complete roleplays — they train turns (modals, questions, checks) without full consultation arcs. Honest scoping prevents two failures: mistaking a clean edit for letter-readiness, and mistaking fluent turns for roleplay-readiness. Graduate deliberately: edits → full letters with official tasks; turns → full roleplays with partners.
Worked models: drills in action
Labelling: all items below are newly written in the workbook’s five formats with codes. The book’s 120 items, edits and tasks are not reproduced. Fictional patients; language examples only.
Model 1 — Five formats on one lesson (newly written, G-style)
Explain (G-X1): Notes say “discharge possible tomorrow.” Is “She will be discharged tomorrow” supported? → No — possible ≠ certain; “She may be discharged tomorrow” preserves modality. Rewrite (G-X2): rewrite “He needs rest” using “should” → “He should rest” / “He should get rest.” Diagnose (G-X3): what’s wrong with “The patient walks independent”? → adjective for adverb: “independently.” Judge (G-X4): does “afebrile today” establish recovery? → No — one sign, not a verdict. Reflect (G-X5): why can fluent turns still fail? → fluency without checks leaves understanding unproven; give your own example.
Model 2 — Form-specified discipline (newly written)
Task: rewrite using the passive (form specified). Notes: “The nurse changed the dressing.” Required: “The dressing was changed (by the nurse)” — passive practised even though the active is equally accurate. Note: the constraint trains the construction; choosing the easier active dodges the drill. Form-specified means form-first, always.
Model 3 — Facts boundary held (newly written)
Notes: “Appetite reduced; weight stable; review requested.” Thin-accurate paragraph: “His appetite is reduced though his weight remains stable. A review has been requested.” Rejected padding: “…suggesting early malnutrition; commence supplements…” — diagnosis and prescription invented. Note: three supplied facts, three sentences, zero additions — thin accuracy beats padded invention every time.
Model 4 — Edit scoped honestly (newly written, E-pattern)
Flawed paragraph: “Confirmed follow-up Friday. Fully recovered. Family will collect medicines.” Repair: “A Friday follow-up has been proposed (status). She continues to improve and remains under review (progress calibrated). Her daughter will collect her medicines (actors/reference fixed).” Scope note: repair complete as an edit; letter-readiness still requires purpose, content selection and genre work on official tasks — say so in the log.
Five-format drill card (tape to the desk)
| Format | Prompt stem | Trains |
|---|---|---|
| Explain | Is X supported? Explain. | reason habit |
| Rewrite | Rewrite using [form]. | controlled production |
| Diagnose | What is wrong with…? | review-pass detection |
| Judge | Does this establish…? | claim-evidence discipline |
| Reflect | Why can…? Give your example. | transfer |
Practice bank: 12 tasks with answers
Attempt with codes before opening each answer. All items are newly written; fictional patients; language examples only.
1 (Formats): Name the five formats and their training targets. — Show answer
Answer: explain (judgement) / rewrite (controlled production) / diagnose (error detection) / judge (claim-evidence) / reflect (generalisation). Note: never skip the fifth — transfer lives there.
2 (Explain): “Discharge possible” — is “will be discharged” supported? — Show answer
Answer: No — “may be discharged” preserves modality. Note: explain-format trains the reason habit.
3 (Rewrite): Rewrite with “might”: “Transport is needed.” — Show rewrite
Rewrite: “Transport might be needed.” Form specified → form first. Note: dodging the constraint dodges training.
4 (Diagnose): What’s wrong with “walks independent”? — Show diagnosis
Diagnosis: adjective for adverb — “independently”. Note: diagnose-format is review-pass training.
5 (Judge): Does “afebrile today” prove recovery? — Show judgement
Judgement: No — one sign is not a verdict. Note: claims need evidence proportionate to their size.
6 (Reflect): Why can accurate sentences still mislead? — Show reflection
Reflection: accuracy without qualifiers/limits overstates — “can walk” without “with frame” misleads. Note: reflect with your own example.
7 (Variants): Model says “may require”; you wrote “might need”. Verdict? — Show verdict
Verdict: accept — facts and modality preserved; difference is not error. Note: compare reasoning, not words.
8 (Codes): Code these: lesson 7 question 3; third edit; fifth speaking task. — Show codes
Codes: G07-3; E03; S05. Note: code discipline aligns attempts with the guide.
9 (Facts): Hold the boundary: notes “pain 6/10, paracetamol given”. — Show sentences
Sentences: “Pain is rated 6/10. Paracetamol has been given.” No advice, no diagnosis added. Note: thin accuracy beats padded invention.
10 (Scope): State both scoping limits. — Show limits
Limits: edits ≠ full letters; turns ≠ full roleplays. Note: graduate deliberately to official tasks.
11 (Graduate): Plan graduation from E-edits to letters. — Show plan
Plan: six edits clean → one official letter with purpose/content checks → compare; turns → partnered full roleplays. Note: scoping honesty schedules graduation.
12 (Mixed): One lesson’s five formats now. — Show setup
Setup: pick weakest lesson → attempt all five formats coded → verify against guide → reflect in own example. Note: five formats, one lesson, full transfer.
Common errors and corrections
- Gap-fill grinding: identical formats endlessly. Correction: five formats per lesson, weakest format first.
- Constraint dodging: easier actives for passive drills. Correction: form-specified means form-first.
- Difference-as-error: valid variants “corrected”. Correction: reasoning comparison, not word matching.
- Clinical invention: padded diagnoses and prescriptions. Correction: supplied-facts-only boundary.
- Scope confusion: clean edits mistaken for readiness. Correction: honest scoping with graduation plans.
- Codeless practice: unverifiable attempts. Correction: G/E/S codes on everything.
Independent task and self-check
Task (newly written): drill weakest lessons five-formats weekly; verify coded against the guide; graduate edits→letters and turns→roleplays on official tasks.
Self-check: are all five formats attempted? Are variants accepted correctly? Are facts uninvented? Are scopes honest with graduation logged? Four yeses = drills transferring.
Study sequence with the book
- Read the how-to-use and codes; fix formats, variants, facts boundary.
- Drill weakest-lesson five-format sets with codes.
- Verify against the guide; bank reasons as rule cards.
- Repair E01–E06 as edits; log letter-readiness as pending.
- Perform S01–S06 as turns; graduate to full roleplays.
- Re-drill cleared lessons with new facts monthly; confirm transfer on official tasks.
FAQs
What if my answer differs from the model? Compare reasoning: facts, form, qualifiers, reference. Equal on all four means correct.
Why five formats instead of more items? Formats train different muscles — judging and reflecting transfer where tenth gap-fills don’t.
Can I invent likely details to practise? No — practise thin accuracy. Invention trains the costliest OET habit.
When am I letter-ready? When edits are clean and official-task letters hold purpose, content and genre — edits alone never certify.
How does this volume fit the set? Coursebook teaches, this workbook drills, diagnostic maps, key explains. Drill here; verify there.
The matching book
OET Grammar: Practice workbook (OET Grammar Practice Workbook) contains 120 focused questions across 24 lessons, six paragraph edits and six speaking tasks with codes — the drilling ground for the OET Grammar set. If five-format drilling helped here, the book is where you repeat it 120 times. Product page link to be added when the authenticated staging catalogue mapping is available — no URL is invented here.
Continue your practice with OET Grammar: Practice workbook (OET Grammar Practice Workbook)
Want more practice? This article introduces the key principles, but OET Grammar: Practice workbook (OET Grammar Practice Workbook) provides additional practice, worked examples and structured preparation. View the book →
Sources and editorial note
Teaching follows the verified source PDF (37 pages; 120 items + 6 edits + 6 tasks with codes; confirmed by stream-decode and lesson sampling). All practice items in this article are newly written; no book item is reproduced. Fictional patients; language examples only, never clinical guidance. Not affiliated with or endorsed by OET. AI-assistance disclosure from the source set applies equally here.
Put this into practice
Compare these course outlines and choose the practice that matches your next goal.
