Train the Criteria, Not the Test: Drills for Each IELTS Speaking Band Descriptor

Problem: candidates practise “speaking” as one blob — so fluency improves while grammar stagnates, or vocabulary grows while coherence collapses, and the average refuses to move. Answer in brief: isolate the four criteria and drill each with its own exercises: extension ladders for fluency, precision sets for lexis, skeleton sentences for grammar, stress marking for pronunciation — then integrate on detailed three-part sets.

Learning outcomes

By the end you will be able to: extend any answer fluently with the 1→2→4 sentence ladder; build topic-precise lexical sets with substitution-proofing; deploy skeleton complex sentences (although, relative, conditional) accurately in speech; mark and deliver word/sentence stress with natural rhythm; and diagnose recordings criterion by criterion. All guidance follows the source book, 50 IELTS Speaking Practice Sets for Guaranteed High Scores (IELTS SP – 189) (Jobin Thomas), whose modules overview, criteria/skills framing and fifty detailed three-part sets were verified by full-structure scan with set sampling. (The book’s title wording about guarantees is the publisher’s marketing; no score outcome is promised here — only trainable skills.)

Fluency and coherence: extension ladders

Fluency is sustained, organised speech — not speed. The ladder drill: take any one-sentence answer and climb it — 1 sentence (fact), 2 (fact + reason), 4 (fact + reason + example + close). “I enjoy reading (1). It lets me escape and learn simultaneously (2). Books give both relaxation and knowledge — when stressed, reading calms and recharges me (4).” Coherence joins fluency through audible joints: and/but/so for turns, first/then/finally for sequences, although/however for qualifications. Drill rule: no answer stops below its target rung (Part 1: rung 3–4 sentences; Part 2: continuous 110 seconds; Part 3: 4–6 sentences). Hesitation devices (“well…”, “that’s an interesting one…”) buy planning time legitimately — silence buys nothing.

Lexical resource: precision sets, not word lists

Lexis scores precision and flexibility, not rarity. Build per-topic sets of six: three precise nouns/adjectives, two collocations, one idiom used sparingly and correctly. Reading topic: engrossing, dog-eared, plot-driven; “broaden one’s perspective”; “lose track of time”. Substitution-proof each item: it must fit its sentence exactly (picturesque views pass; picturesque stories fail). Paraphrase reflex: restate the question’s key words in your opener (“Do you enjoy reading?” → “Reading has been a constant habit…”) — flexibility demonstrated in the first five seconds. One new item per answer is the density that displays range without forcing.

Grammatical range and accuracy: skeleton sentences

Range means demonstrated variety: simple, compound, complex — plus conditionals and passives where natural. The skeleton method pre-builds three complex frames and spends them deliberately: although-skeleton (“Although fees deter, deferred support preserves access”); relative-skeleton (“a routine that survived travel and illness”); conditional-skeleton (“Had I waited, I would have lost a year”). Accuracy protects the display: tense consistency inside narratives (past stays past), article discipline (society, not the society, for general claims), subject-verb agreement at speed. One skeleton per part, monitored in transcription — range shown on purpose beats range attempted everywhere and achieved nowhere.

Pronunciation: stress marking and rhythm

Intelligibility rests on stress and rhythm, not accent. The marking drill: before speaking any answer, underline stressed words in the script — content words full, function words reduced (“I REALLY enJOY READING beCAUSE it CALMS me”). Deliver with the marks exaggerated in rehearsal; exam delivery lands naturally. Word-stress accuracy (PHOtograph vs phoTOgraphy) outranks sound perfection — mis-stressed words baffle, mispronounced sounds merely colour. Intonation shapes meaning: fall for statements and wh-questions, rise for yes/no questions and uncertainty. Record one minute weekly and audit marks-vs-delivery: the gap is the drill target.

Worked models: one set, four criteria lenses

Labelling: the set below follows the book’s books-and-reading pattern with newly written answers, each annotated per criterion. The book’s questions and samples are not reproduced. Suggested models — other valid answers exist.

Model 1 — Part 1: reading answers, criteria-flagged (newly written)

Q: Do you enjoy reading? “I really enjoy reading because it lets me escape into different worlds while learning (fluency: rung-2 opening; grammar: while-clause). Books give me relaxation as well as knowledge — when stressed, reading calms and recharges me (coherence: dash-expansion; lexis: recharge used precisely).” Q: What types do you read? “Mostly non-fiction on psychology and habits, which give practical ideas (relative-skeleton deployed). Although I also enjoy relationship novels from time to time (although-skeleton) — the variety keeps the habit alive (lexis: variety, alive).” Flags: F (ladder climbed), L (recharge/alive precise), G (two skeletons), P (REALLY, CALMS, RECHARGE stressed).

Model 2 — Part 2: book card, criteria-built (newly written)

Card (newly written pattern): a useful book — what / why read / what learned / why useful.

Talk (suggested model, ~105 seconds): “A psychology of habits that a mentor pressed on me (lexis: pressed, precise). It argues identity precedes results — become a reader, and pages follow (grammar: nominalisation + parallel). What changed my routine: stacking ten words onto breakfast tea — the streak survived travel and illness (fluency: Extra scene; coherence: dash-joint). Why it lasted: motivation fades; systems persist (pronunciation: MOtivation FADES, SYStems perSIST — contrastive stress carries the aphorism).” Flags: all four criteria visibly served — design answers this way, then the band follows.

Model 3 — Part 3: reading culture, criteria-checked (newly written)

Q: Do people read less than in the past? “On paper, yes; in pixels, enormously more (fluency: contrastive open). Although book sales wobble, total text consumed — messages, articles, captions — has exploded (grammar: although + present perfect). The loss isn’t reading but sustained reading: depth traded for skimming (lexis: sustained, traded). If schools taught stamina-reading explicitly, the balance might recover (conditional).” Check: F (4 sentences, joints audible), L (sustained/stamina precise), G (although + perfect + conditional), P (PAper/PIxels contrast stressed).

Model 4 — Criteria diagnosis on a weak transcript (newly written)

Transcript: “I like books. They are good. I read daily. Sometimes novels.” Diagnosis: F — rung 1 throughout, no joints. L — good/daily generic, no set. G — simple only, no skeleton. P — unmarked, flat. Prescription (one per criterion): F — ladder to rung 3 with reason+example. L — bank six reading items, deploy two. G — one relative-skeleton (“books that…”). P — mark and stress three content words. Note: four criteria, four work orders — the diagnostic habit this article installs.

Second set: technology topic, criteria-built (newly written)

Part 1 — “How often do you use your phone?” “Constantly, but deliberately (F: qualified open). Mornings are messaging and news; evenings, one film or a long call (L: deliberately, precisely placed). Although the screen-time report scolds me weekly (G: although-skeleton), the alternative — missed family news — costs more (P: SColDS, MISSED, COSTS stressed).”

Part 2 — object card (gadget), ~100s: “A refurbished e-reader, bought certified-resale two winters ago (L: certified-resale precise). I had been annotating PDFs on a glaring laptop for a year (G: past perfect background). From the first week it changed evenings — calm reading where scrolling ruled (F: Extra scene; coherence: where-joint). Overall, had I waited for cheaper, I would have lost another reading year (G: conditional close inside conclusion-frame).”

Part 3 — “Has technology improved leisure?” “Access, yes; attention, no (F: split verdict). Although options multiplied, sustained enjoyment thinned — scrolling replaces savouring (G: although + L: sustained/savouring). If platforms priced attention honestly, usage would halve (G: second conditional). Design the default better and leisure recovers (P: deFAULT, reCOVers).” Note: second topic, same flag discipline — criteria training is topic-independent.

Weekly circuit timetable

Practice bank: 12 tasks with answers

Attempt aloud and recorded before opening each answer. All items are newly written; the book’s sets are not reproduced.

Common errors and corrections

• Blob practice: “speaking” trained as one skill. Correction: isolate criteria; drill each with its own exercises.
• Speed-as-fluency: rushing mistaken for flowing. Correction: organised extension at natural pace; joints, not velocity.
• Rarity-as-lexis: obscure words misused. Correction: precision sets with substitution-proofing.
• Safety grammar: simple-only sentences. Correction: pre-committed skeletons, one per part.
• Accent-as-pronunciation: imitation over intelligibility. Correction: stress marking and rhythm audits.
• Single-verdict diagnosis: “weak speaking”. Correction: four lenses, four findings, four fixes.

Independent task and self-check

Task (newly written): run the 20-minute circuit daily for a fortnight plus two fully flagged sessions weekly. Keep per-criterion tallies (rungs hit, items deployed, skeletons used, marks delivered).

Self-check: does every answer hit its rung target? Is one precise item deployed per answer? Is one skeleton present per part? Are stresses marked before speaking? Four yeses = criteria training installed.

Study sequence with the book

1. Read the modules overview and criteria/skills framing; fix the four lenses.
2. Drill one criterion weekly with this article’s tasks while reading the book’s sets.
3. Attempt five sets cold with full F/L/G/P flagging; prescribe four fixes each.
4. Run daily 20-minute circuits on the weakest criterion; re-flag fortnightly.
5. Re-speak weakest parts with deployed fixes; chart rung/item/skeleton/mark tallies.
6. Final loop: integrated flagged sessions only; last diagnosis 48 hours before the exam.

FAQs

Which criterion improves fastest? Usually fluency (ladders work in days) — but the band averages, so harvest fluency early then fund the slowest.

Can criteria really be isolated? For drilling, yes; in performance they merge. Isolate to train, integrate to perform — circuits then sessions.

How do examiners weight the four? Equally in the average. A 5 in one criterion costs as much as a 7 earns — triage accordingly.

Is there really a guarantee? No honest material can guarantee scores — outcomes depend on candidacy, preparation and day. What detailed sets guarantee is coverage: every part, practised in detail. Train the skills; let the band follow.

How does this book differ from its 189-series siblings? 189-1 gives the session workout; 189-2 gives cross-module transfer; this 189 volume gives detailed three-part coverage framed by the assessment criteria — the drill ground for criterion-targeted training.

The matching book

50 IELTS Speaking Practice Sets for Guaranteed High Scores (IELTS SP – 189) contains the modules overview, criteria and skills framing, and fifty detailed sets covering Parts 1, 2 and 3 — the drill ground for this article’s criterion training. If flagged practice helped here, the book is where you repeat it fifty times. Product page link to be added when the authenticated staging catalogue mapping is available — no URL is invented here.

Take this further with 50 IELTS Speaking Practice Sets for Guaranteed High Scores (IELTS SP – 189)

If this guide helped, 50 IELTS Speaking Practice Sets for Guaranteed High Scores (IELTS SP – 189) gives you the full training volume behind it: more practice, worked examples and step-by-step preparation. Get the book →

Sources and editorial note

Teaching follows the verified source EPUB (modules overview; criteria/skills framing; 50 detailed sets, confirmed by structural scan and set sampling including the books-and-reading set). All answers in this article are newly written in the book’s patterns; no book question, sample or phrase is reproduced, and no third-party material is copied. New answers are suggested models. The book’s guarantee wording is not repeated as fact. Test-format summaries are editorial paraphrase; current exam rules should be confirmed against the official IELTS documentation before booking.