March 2023, w/ Linh (sister)
~7 hrs, climbing mountains
coffee from carriage attendant
changed how I see her
SKILLME360 is your complete learning ecosystem for English mastery, academic success, and professional growth.
Learn from the best instructors globally
Speak with confidence in real scenarios
Track your progress with advanced analytics
Achieve your goals anywhere in the world
Between us we have spent 24 years in classrooms across Vietnam, Thailand, Mexico, Cambodia, Saudi Arabia, the UK and the US — and we currently teach at eight universities and English centres in Ho Chi Minh City. SKILLME360 is what we built for those rooms. Every lesson, rubric and dashboard was tested on our own students first.
Fifteen years teaching university English and running communication workshops for multinationals. In Ho Chi Minh City he teaches at HCMC Open University, UEH, Vietnam Aviation Academy, FPT Education and NIIE. He builds the curriculum, assessment and instructor training behind SKILLME360.
Nine years teaching across the US, Thailand, Mexico, Cambodia and Vietnam. He lectured the academic English course at UMT and teaches IELTS and university-pathway English at YOLA, after VUS. He shapes how SKILLME360 sequences lessons, feedback and assessment.
Every kind of English learner gets a dedicated experience built for their goals.
Book instructor-led English workshops for your team and track company-wide progress.
Full Grade 1–12 English curriculum with pre-built interactive lessons and activities.
100 self-teaching lessons for parents to use at home with young children.
IELTS self-study prep with optional one-on-one instructor support.
Join schools, companies, parents, and students across Vietnam building their future with SKILLME360.
Consistency today, success tomorrow.
20 structured speaking lessons built for Vietnamese learners. Explanations, native-quality examples, and interactive exercises.
Five full tests in the Cambridge format. Pick an exercise, listen once, then check your answers and read why each one is right.
Answer real exam questions out loud, record yourself, and keep the ones you're proud of. Each week, send your best clips to Lionel over Zalo for voice feedback.
Five questions across Part 1, 2 and 3. About 8 minutes.
Timed Reading and Listening tests, marked the second you finish. Every score becomes a dot on your band chart, so you can watch it climb week by week.
Deep-dive reference material organised by skill area. Everything the lessons touch on, with room to explore further at your own pace.
How words assemble into phrases, and phrases into meaning. Start with the noun phrase, which sits at the centre of nearly every sentence you'll ever build.
Five short lessons teach the machinery of fluent English — rhythm, chunking, pausing, stress and intonation — then the Reading Studio puts you behind a microphone: model audio, guided overlays, a full teleprompter and unlimited recorded attempts.
English is stress-timed, Vietnamese is syllable-timed — this single difference drives everything. Find the beat, compress the rest.
Fluent readers speak in chunks, not words. Learn where grammar draws the boundaries, then draw them yourself.
Two pause lengths, one law: pauses only fall on chunk boundaries. Build a pause map and hear it read back.
Content words up, function words down — and one sentence with six meanings, depending on where you aim the light.
Falls, rises, list tunes and question tags. English listeners hear your melody before your words.
A warm-up narrative. Listen with guided overlays, read it from the teleprompter, record and compare against the model voice.
An IELTS Part 2 style monologue. Storyteller rhythm: longer chunks, a list tune and one power pause.
Academic register with pronunciation hints on the hard vocabulary. Keep the beat steady as the sentences grow.
The full cabin-crew safety briefing — the longest script in the course. Formal register, repeating chunk patterns, built for the teleprompter.
Final call at gate twelve. Flight codes, row numbers and passenger names — where rhythm usually collapses. Slow the numbers, stress them.
A duty manager greets a tour group. Keep the announcement structure but let the voice smile — warmth without losing the beat.
Delays, apologies and alternative services. Neutral, dignified delivery — the information does the work, the rhythm keeps it calm.
A guide opens a gallery walk: housekeeping points, dates and names, then a three-thousand-year list tune.
Headline rhythm and hard stops. Read it like the autocue is live — a rise-and-hold story list, then straight to the main story.
Ten lessons that build an IELTS essay from the ground up — the sentence, then the paragraph, then the essay, then the two tasks. Written for Vietnamese candidates, with the transfer errors that cost you Band 6.5 named and fixed as we go.
Both tasks, the four criteria, and what the examiner actually does with your script in the ten minutes they spend on it. Start here — the rest of the course assumes this lesson.
Subject-verb-object, the four sentence types, fragments, run-ons — and the six Vietnamese transfer errors that appear in almost every script we mark.
One idea per paragraph, a topic sentence that commits, and P·E·E·L to develop it. Plus why your linking words are hurting your Coherence score rather than helping it.
Task types, the instruction words, and how to find what is actually being asked before you write a single word.
Two sentences in, two sentences out. Paraphrase without copying, state a position, and finish without a new idea.
Agree/disagree, discuss both views, two-part questions. The shape each one demands and how examiners tell them apart.
Precision over decoration. Collocation, register, and why the words you memorised from a band-9 list are marking you down.
Overview first, then selection. What to include, what to leave out, and the language of change.
The two task types nobody practises. Passive voice, sequencing, and describing change across time.
The last three minutes. A fixed checking order that finds more marks than another paragraph would.
Writing is the lowest-scoring of the four IELTS papers almost everywhere in the world, and Vietnam is no exception. The average Vietnamese candidate scores roughly half a band lower in Writing than in Reading. That gap is not a language problem. It is a structure problem, and structure is teachable.
"Most candidates lose marks for things they already know how to do. They know the article rule. They know a paragraph needs one idea. Under time pressure they stop doing it. This course is about making the right move automatic, so pressure can't take it away from you."
Nine stages. This first lesson is orientation — no essay writing yet. You need to know what the paper is and how it is judged before you can aim at it.
You are here — getting oriented.
Task 1 and Task 2 — the times, the word counts, and why one is worth double.
Task Response, Coherence, Lexical Resource, Grammar. 25% each, and each one is winnable.
What the examiner does in the ten minutes they spend on your script.
What costs the most marks — and the fix for each.
Four paragraphs, forty minutes. The frame everything else hangs on.
The habits that transfer from Vietnamese into English essays.
Ten questions. Prove the lesson landed.
Self-check, badge, and what Lesson 2 does.
One hour, no break, no extra paper time. You decide how to spend the sixty minutes — and most candidates spend them wrong.
Academic: describe a graph, chart, table, map or process. General Training: write a letter.
An essay responding to an argument, a problem, or a point of view. Same for Academic and General Training.
150 and 250 are not targets. They are the line below which you are penalised before anyone judges your English. Write 240 words in Task 2 and you are marked down on Task Response no matter how good the argument is.
Aim for 270 to 290. That is a twenty-word cushion above the line — enough to survive a miscount, not enough to drag you into the rambling that wrecks your grammar score. Very long essays are not rewarded. There is no band descriptor anywhere that says "wrote a lot".
Five minutes of planning feels like five wasted minutes. It is the highest-return five minutes on the paper: it is where Coherence is won.
Your Writing band is the average of four separate scores. Each one is judged on its own, from its own descriptor. This matters more than it sounds: a Band 8 vocabulary cannot rescue a Band 5 Task Response. The average just drags to 6.5.
Did you answer the question that was asked — all parts of it — with a clear position you hold from start to finish, developed with real support?
Can the reader follow you effortlessly? Logical paragraphs, one idea each, and linking that guides rather than decorates.
Range and accuracy of vocabulary. Precision, collocation, and register — not how rare your words are.
Variety of structures and how often you get them right. Both halves matter — range without accuracy scores no better than accuracy without range.
A typical Vietnamese script at 6.5 does not score 6.5 across the board. It scores unevenly, and the shape of the unevenness is remarkably consistent.
Average: 6.25, rounded to 6.5. The vocabulary work most candidates keep doing is polishing the one criterion that is already fine. The marks are sitting in Task Response, Coherence and Grammar, and all three are structural. That is what this course goes after.
Each sentence below describes a problem in a real script. Decide which of the four criteria takes the hit.
An examiner has your Task 2 essay for roughly six to eight minutes. They are not hunting for beauty. They are matching what is in front of them against four descriptors they know by heart.
Straight through, no pen. They are asking one question: did this person answer the question? By the end of the first read, Task Response is roughly settled.
Now they mark errors and note structures. Grammar and Lexical Resource are decided here. Repeated errors count once as a pattern, not fifteen times.
They look at the essay as an object. Four clear paragraphs? Or one grey block? Coherence is heavily influenced by what the page looks like before a word is read.
One per criterion, averaged, rounded. No discussion, no benefit of the doubt.
If the examiner reaches the end of the first read unsure what you think, Task Response is capped at 5. State your position in the introduction, in plain words.
Four visually distinct paragraphs signal control before anything is read. Indent or leave a blank line. Never write one continuous block.
Dropping articles fifteen times is one problem, marked once. Fixing that single habit moves Grammar a whole band. Fixing one sentence moves nothing.
In order of how much they cost. Number one costs more than the other four combined.
The essay discusses the topic instead of answering the question. Or it agrees in paragraph two and hedges in paragraph three. The examiner finishes not knowing what you think.
"In this day and age", "It is a double-edged sword", "Since the dawn of civilisation". Examiners have read these thousands of times. They are excluded from the word count in practice and they mark you down for register.
"Technology has many benefits for education. It helps students a lot. It is very useful." Three sentences, one unsupported idea, zero development.
Under 250 in Task 2 is an automatic penalty on Task Response, applied before quality is considered.
Two ideas, three ideas, five ideas in one block. The reader loses the thread and Coherence drops.
Question: Some people believe university education should be free for everyone. To what extent do you agree or disagree?
"University education is a controversial topic in modern society. Some people think it should be free, while others disagree. This essay will discuss both sides of this issue."
"Whether taxpayers should fund university tuition is contested in most developed economies. I largely agree that it should be free, because the public return on a graduate workforce outweighs the individual cost — though I would restrict this to publicly funded institutions."
Not a template for the words — a frame for the thinking. Every Task 2 essay in this course sits on this shape, whatever the question type.
Paraphrase the question. State your position.
Your strongest reason. Topic sentence, explanation, one concrete example, link back to the position.
Your second reason, or the counter-argument you concede and then answer. Same internal shape.
Restate the position in new words. One forward-looking sentence. No new ideas, ever.
Candidates aiming high often add a third body paragraph, reasoning that more ideas means more development. It does the opposite.
Each body gets ~70 words. That is a topic sentence, two thin claims, and no room for an example. Three underdeveloped ideas.
Each body gets ~100 words. Topic sentence, explanation, concrete example, link back. Two ideas, both finished.
These are not sloppiness. They are Vietnamese working correctly, in the wrong language. Naming them is most of the fix, because you cannot proofread for an error you don't know exists.
This is a genuine Band 6 paragraph from a Vietnamese candidate, lightly anonymised. Read it, then reveal the marked version.
"According to me, government should invest more money in public transport. In Hanoi, many people uses motorbike every day and it cause serious pollution. Last year the city introduce new bus route but it is not enough. Besides, moreover, the traffic jam is very serious problem."
"According to me, government should invest more money in public transport. In Hanoi, many people uses motorbike every day and it cause serious pollution. Last year the city introduce new bus route but it is not enough. Besides, moreover, the traffic jam is very serious problem."
"In my view, the government should invest more heavily in public transport. In Hanoi, hundreds of thousands of commuters use motorbikes daily, and this causes serious air pollution. The city introduced new bus routes last year, but capacity remains well short of demand. Congestion, meanwhile, is worsening rather than easing."
No time limit. Answer honestly — a wrong answer here is worth more than a lucky one.
No judgement — just a record of where you are starting from.
Worth double. 40 minutes, 270–290 words, five of those minutes spent planning.
Your vocabulary is probably fine. The marks are in Task Response, Coherence and Grammar.
The examiner settles Task Response on a single fast read. Don't make them hunt.
Two ideas developed beats four mentioned. Depth is the score.
Articles, plurals, tense, "according to me", buried position, stacked markers.
One habit corrected moves a whole band. One sentence corrected moves nothing.
You know what the paper is, how it is judged, and what the frame looks like. Now we build the thing that goes inside it — starting with the smallest unit that can be right or wrong.
Two of your four criteria — Grammatical Range and Grammatical Accuracy — are decided at this level. Not at the essay level, not at the paragraph level. Sentence by sentence, clause by clause.
"Candidates ask me for advanced structures. Then I read four sentences in a row that are all subject-verb-object, seven words long, and I understand why they're stuck at 6. Range isn't about difficulty. It's about not writing the same shape twice."
You are here.
What every English sentence has to have, and what Vietnamese lets you drop.
Simple, compound, complex, compound-complex. Range lives here.
The sentence that isn't one, and the three ways it happens.
Two sentences wearing one coat.
Articles, plurals, tense, agreement, prepositions, countability — dismantled one at a time.
Turning four flat sentences into a Band 7 paragraph without adding an idea.
Twelve questions.
Self-check, badge, next lesson.
An English sentence needs a subject and a finite verb. Not "usually". Always. Vietnamese does not have this requirement, and that single difference produces more errors in your writing than any other.
If the listener can recover it, Vietnamese drops it. English cannot. Three specific gaps cause almost all of the damage.
Grammatical Range means using more than one of these. Not exotic grammar. These four, mixed.
"Many students study abroad. It is expensive. Their parents pay a lot of money. Some students do not return home. This is a problem for Vietnam."
"Many Vietnamese students now study abroad, but the cost is considerable, and it usually falls on parents rather than on the students themselves. Because a significant proportion never return, this expenditure represents a net loss of talent for Vietnam."
A fragment has a capital letter and a full stop but no independent clause. It happens three ways, and once you know the three, you can find them all.
"Because the government reduced funding for public universities."
A complete thought that starts with because, although, while, if, when is not a sentence. It needs a main clause to attach to.
"Fees rose sharply because the government reduced funding for public universities."
"Many young people moving to the cities every year."
An -ing form is not a finite verb. On its own it can't run a sentence.
"Many young people are moving to the cities every year." — or — "Many young people move to the cities every year."
"…which harms the environment. For example, air pollution in major cities."
The most common fragment in IELTS scripts. "For example" does not license dropping the verb.
"For example, air pollution in major cities has reached hazardous levels."
Tap your verdict on each one.
The opposite failure. Two independent clauses joined with nothing, or with a comma that isn't strong enough to hold them.
"Public transport is underfunded, congestion continues to worsen."
Both halves are complete sentences. A comma cannot join them. This is the most common punctuation error in IELTS Writing worldwide.
"Public transport is underfunded. Congestion continues to worsen."
"Public transport is underfunded, so congestion continues to worsen."
"Public transport is underfunded; congestion continues to worsen."
"Because public transport is underfunded, congestion continues to worsen."
Vietnamese has no articles. English forces a choice in front of almost every noun. Here is the whole decision as three questions — run them in order and you will be right the large majority of the time.
Each sentence has exactly one error from the six. Tap the word that is wrong.
Watch the same content go from Band 6 to Band 7.5 in three moves. No new ideas. No harder vocabulary.
"Cities are growing quickly. This causes housing shortages. Prices increase. Young people cannot buy homes."
"Because cities are growing quickly, housing shortages have emerged. Prices increase. Young people cannot buy homes."
"Because cities are growing quickly, housing shortages have emerged, and prices have risen sharply. Young people cannot buy homes."
"Because cities are growing quickly, housing shortages have emerged, and prices have risen sharply — a trend which has put ownership beyond the reach of most young professionals."
Three flat sentences. Pick the move at each step and watch the sentence assemble.
Sentence types, fragments, splices, and the six transfer errors.
Always. Vietnamese lets you drop the subject, the "be", and "there is". English does not.
Range means variety, not difficulty. Complex sentences are where 6 becomes 7.
Stranded dependent clause, the -ing trap, the added-on example.
Nor are therefore, moreover, furthermore. A comma won't hold them.
General? Identifiable? Any one will do? That's the tree.
Articles, plural -s, third-person -s, past marking, countability, prepositions.
You can build a correct sentence and vary its shape. Now we stack them into the unit the examiner actually reads.
This is the unit the examiner reads as a whole. A sentence can be perfect and still land in a paragraph that scores 5, because the paragraph is where Coherence is won or lost — and Coherence is a quarter of your Writing band.
"I can usually predict a Coherence score from across the room, before I read a word. Four clean blocks with white space between them looks like a 7. One grey wall looks like a 5. What the page looks like is telling me something true about how the writer thinks."
You are here.
The rule that fixes half of all Coherence problems.
The sentence that commits — and the one that doesn't.
Point, Explain, Example, Link. How to fill 100 words with one idea.
Why "Firstly, Secondly, Moreover" is hurting you, and what to do instead.
Pronouns, this/these, and synonyms that hold a paragraph together invisibly.
A Band 6 paragraph rebuilt into a Band 7.5 one, live.
Twelve questions.
Self-check, badge, next lesson.
This one rule fixes roughly half of every Coherence problem we see. Not "one topic" — one idea. A body paragraph makes a single claim and spends its whole length supporting that claim. When the claim changes, the paragraph ends.
"Online learning is convenient because students can study anywhere. It is also cheaper than traditional courses. However, it requires self-discipline, and many students struggle without a teacher present. Universities should therefore invest in better platforms."
"The primary advantage of online learning is its flexibility. Because lessons are recorded and accessible on demand, students can fit study around work and family rather than the reverse. A nurse on rotating shifts, for instance, can complete a module at midnight or at dawn — an option a fixed timetable simply cannot offer."
After you write a body paragraph, ask: can I summarise this whole paragraph in one sentence that starts "This paragraph argues that…"? If you need "and" to finish that sentence, you have two paragraphs pretending to be one.
The first sentence of a body paragraph tells the reader the single claim the paragraph will prove. A weak one announces a topic. A strong one states a position on that topic.
"Another point is about the environment."
"There are many advantages of technology."
"Now I will discuss education."
"Heavy industry remains the largest single source of urban air pollution."
"Technology's greatest benefit to education is reach: it puts expert teaching within range of students who would never otherwise meet it."
Four moves that turn one claim into a full 100-word paragraph. This is the structure inside every body paragraph in the rest of the course.
Your topic sentence. The claim, committed.
Why is the claim true? The mechanism, in one or two sentences.
One concrete instance. A named thing, a number, a real situation. This is the sentence Vietnamese scripts most often skip.
Tie the example back to the claim, or forward to your position. One sentence.
Vietnamese scripts reliably do P and the first E. They assert a claim and explain it. Then they stop, or repeat the claim in new words, and the paragraph runs out of substance at 60 words.
"Public transport reduces congestion. When more people use buses and trains, fewer cars are on the road. This means less traffic. Therefore public transport is good for cities."
"Public transport reduces congestion. When more commuters travel by bus and train, fewer cars occupy the road. Bogotá's bus rapid transit system, for example, moves 2.4 million passengers a day on dedicated lanes, and average car journey times across the city fell measurably in the years after it opened. A single well-run network can therefore take pressure off an entire road system."
Claim: "Working from home improves productivity for many roles." Pick the sentence that does each job.
This surprises every candidate. You were taught that "Firstly, Secondly, Moreover, In conclusion" shows cohesion. At Band 6 it does. Above Band 6, mechanical linking is explicitly penalised — the descriptor says so.
"…uses a range of cohesive devices appropriately although there may be some under-/over-use."
"…uses cohesive devices effectively, but cohesion within and/or between sentences may be faulty or mechanical."
"Firstly, remote work saves time. Secondly, it reduces costs. Moreover, it is flexible. In addition, it is better for the environment. Therefore, it is beneficial."
"Remote work saves the hours most employees lose to commuting. Those recovered hours tend to go back into the job or into rest, and both raise output. The same shift also cuts office costs and road traffic — benefits that reinforce rather than compete with the productivity gain."
Real cohesion connects one idea to the next through meaning. Three techniques carry almost all of it, and none of them is a connector at the start of a sentence.
End one sentence on an idea, open the next by pointing to it.
Name a thing once, then refer to it with pronouns and "the".
When you do use a connector, bury it inside the sentence, not at the front.
These are the devices that hold a paragraph together without the reader noticing. Overuse the obvious connectors, underuse these, and that is exactly the Band 6 profile.
Everything from this lesson, applied to one real Band 6 paragraph. Question: Some people think children should start school as early as possible. Do you agree?
"Firstly, starting school early is good. Children can learn many thing when they are young. Secondly, it is good for working parents. Moreover, young children learn language fast. In addition, they make friends. Therefore early school is beneficial."
"The strongest case for early schooling rests on language. A child's capacity to absorb a new language peaks well before the age of seven and declines steadily afterwards, so the years before formal school are the ones least able to be recovered later. A five-year-old placed in a bilingual classroom will typically reach fluency that an adult, studying far harder, never quite matches. Beginning school early therefore captures a window that closes on its own."
One idea, topic sentences, P·E·E·L, and real cohesion.
The one-sentence test: if you need "and" to summarise it, it's two paragraphs.
Make a claim, don't announce a topic.
Point, Explain, Example, Link. The second E is where the marks live.
The descriptor says so. One sentence-initial connector per paragraph, maximum.
this / it / the / synonyms carry cohesion invisibly.
Run it on every body paragraph before you move on.
Sentence, then paragraph. You now have the two units every essay is built from. From Lesson 4, we assemble them into a full Task 2 answer — starting with the thing most candidates skip: reading the question properly.
Every examiner has marked it: a fluent, well-organised essay scoring Band 5.5 on Task Response because it answers a question the paper never asked. The candidate saw the topic — technology, education, the environment — and wrote their prepared essay about the topic. IELTS never asks about a topic. It asks a specific question inside one.
Reading the question is not the thing you do before the exam starts. It is the first scored act of the exam. Five careful minutes here are worth more than any vocabulary you memorised.
Every Task 2 question on every paper is one of five types. The type decides your essay's shape before you have a single idea. Learn to name the type in ten seconds.
The topic sentence of the question is scenery. The instruction line is the contract. Three phrases do most of the damage:
Most questions carry limiters — quiet words that shrink the territory: young people, in cities, developing countries, at primary school. Write outside the fence and your ideas stop counting.
Underline the limiters before you plan. If a body paragraph doesn't touch every fence post, it isn't answering this question — it's answering an easier cousin of it.
You've named the type, honoured the instruction words, and fenced the limits. Now the plan — on the question paper, in note form, in five minutes:
Four failure patterns account for nearly every Task Response penalty in Vietnamese scripts:
For each question: name the type, then pick what a complete answer must contain. The decoder gives instant feedback — take your time, this is the whole lesson in one exercise.
No time limit. A wrong answer here teaches more than a lucky one — read every explanation.
No judgement — just a record of where you are starting from.
Opinion, discussion, adv/disadv, problem/solution, two-part — name it in ten seconds.
Instruction words are orders. "Outweigh" demands a verdict; "discuss both" means both.
Limiters shrink the territory. Ideas outside them stop counting.
Decode, position, two ideas with examples, final check — then never re-decide.
You now read a Task 2 question the way an examiner does: type first, contract second, fence third, plan fourth. Lesson 5 builds the frame around your answer — an introduction in two sentences and a conclusion that knows when to stop.
The introduction and conclusion are the smallest paragraphs in your essay and the ones candidates spend the most wasted time on. Here is the whole secret: the introduction is two sentences with two jobs, and the conclusion is two sentences with one job. Everything longer is time stolen from the paragraphs that actually earn marks.
An examiner reads your introduction in eight seconds. In those eight seconds they form a hypothesis about your band. The rest of the essay mostly confirms it. Two clean sentences beat five impressive ones every single time.
Job one: show you understood the question, by restating it in your own words. Job two: state your position (or, for a discussion essay, your direction). That's the entire introduction.
Copying the question costs you: copied strings are subtracted from your word count and read as Band-5 behaviour. But paraphrase has one law — the meaning, including every limiter, must survive. Three safe tools:
Sentence two commits. The commitment has to be one you can defend for 270 words — which is why extreme positions are usually a trap and measured ones are a gift.
The conclusion has one job: confirm that the position promised in the introduction was delivered. Restate the position in fresh words, gesture at your reasons, stop. Two sentences.
Three candidates answered the same question. Read each introduction, diagnose it, and pick the fix. The surgery table gives instant feedback.
Memorised templates get discounted; frames get filled. The difference: a frame fixes the jobs, and every word is chosen fresh for this question.
If two of your practice essays contain the same sentence, it isn't a frame any more — it's a template, and the examiner has read it before.
No time limit. A wrong answer here teaches more than a lucky one — read every explanation.
No judgement — just a record of where you are starting from.
Sentence one restates in fresh words; sentence two commits to a position.
Common synonyms, word forms, structure swaps — with every fence post standing.
Measured beats absolute. Build the distinction in and the body plans itself.
Confirm in fresh words, compress the reasons, stop. No new ideas, no nutshells.
Two sentences that restate and commit; two that confirm and stop. The frame is done in under six minutes, forever. Lesson 6 fills it — the three essay shapes of Task 2, built paragraph by paragraph.
You can read the question, frame the answer, and build a paragraph. This lesson assembles those parts into the three full essays Task 2 actually asks for: the opinion essay, the discussion essay, and the two-part answer. Same bricks every time — different building.
Band 6 candidates have ideas. Band 7 candidates have a shape their ideas live in. When the shape is decided by the question type, the essay half-writes itself.
For agree/disagree, you have two legal shapes. Both are four paragraphs. Pick by asking: can I fill two paragraphs arguing one way?
"Discuss both views and give your own opinion" is the shape Vietnamese candidates most often get wrong, in one specific way: the view they disagree with gets a weak, two-sentence paragraph. The examiner reads that as failing to discuss both views.
Attribution is the trick. "Supporters argue" lets you write the other side at full strength without ever owning it. The stronger you make their case, the more impressive your rebuttal.
"Why is this happening? Is it a positive or negative development?" Two direct questions. The shape is mechanical, which makes it the safest question on the paper once you see it:
One move above all separates 6 from 7 in Task Response: showing you can hold the opposing idea, grant what is true in it, and still win. Three steps, one sentence pattern:
With the shape fixed, you need exactly two body-paragraph ideas. The selection test, in order:
One question, four decisions: shape, intro, Body 1, Body 2. Choose well and watch a complete essay plan assemble below your choices.
No time limit. A wrong answer here teaches more than a lucky one — read every explanation.
No judgement — just a record of where you are starting from.
Four paragraphs, always. The question type picks the building.
"Supporters argue…" writes the other side at full strength without owning it.
Grant the strongest true version, pivot, win with evidence — the Band 7 move.
Fence, mechanism, example, different in kind. Two strong beats three weak.
Opinion, discussion, two-part — each now has a fixed shape, a place for your position, and the concede-and-refute move inside it. Lesson 7 sharpens the material itself: vocabulary chosen for precision rather than decoration.
Here is the Lexical Resource secret nobody sells: examiners are not counting rare words. The descriptor rewards precision — the word that means exactly what you mean — and naturalness — words in the company they actually keep. A band-9 word in the wrong place scores below a band-6 word in the right one.
When a script says "the plethora of conveyances engendered vehicular congestion", I don't think Band 8. I think: this candidate has a list, and the list is writing the essay.
Three ways memorised vocabulary loses marks:
Collocation is which words live together. English decided these pairings centuries ago and takes them personally. The high-frequency ones for IELTS essays:
Register is temperature control. IELTS essays are formal — but formal means neutral and precise, not Victorian. The swaps that matter:
Band 6 scripts shout: "Technology destroys communication skills." Band 7+ scripts calibrate: "Heavy technology use may weaken face-to-face skills, particularly among teenagers." The toolkit:
Hedging is not weakness. It is the sound of somebody who has thought about how true their sentence is. Examiners are trained to hear exactly that.
Eight sentences, each with one word doing the wrong job — imprecise, mis-collocated, or wrong-temperature. Pick the precise replacement.
The last habit: vocabulary grows by topic family, with collocations attached, because IELTS topics repeat. Three families cover half of all papers:
No time limit. A wrong answer here teaches more than a lucky one — read every explanation.
No judgement — just a record of where you are starting from.
The exact word beats the rare word. Evenness beats spikes.
Make/do/take, assigned adjectives, verb+preposition — the pair is the word.
Neutral and precise. No kids, no contractions — and no "utilise" either.
Modal dial, frequency dial, scope tags: the grammar of careful claims.
Precision, collocation, register, hedging, families: your vocabulary now works for the essay instead of posing in it. Part 3 changes task: the 20-minute Task 1 report, starting with graphs and charts — and the single sentence worth more than any other in it.
Task 1 is not a small essay. It is a report: 150 words, 20 minutes, and a rule that surprises everyone — no opinions, no explanations, no causes. The chart shows car ownership rising; you report that it rose. Why it rose belongs to Task 2, and writing it here costs marks.
Task 1 is the most learnable 20 minutes in IELTS. It has a fixed shape, a fixed grammar, and a single sentence that carries half the score. Most candidates never learn which sentence.
The band descriptors are explicit: without a clear overview, Task Achievement cannot pass Band 5. The overview answers one question — standing back, what are the one or two biggest things happening? — and follows one law: no numbers.
The data-listing report — "In 2000 X was 5%. In 2005 X was 9%. In 2010 X was 15%…" — is the most common Task 1 failure. The chart has maybe twenty numbers; a Band 7 report uses six, chosen as evidence for patterns:
Task 1 grammar is a small machine with four parts:
Most charts compare categories, and comparison language is where accuracy slips. The reliable set:
A described chart, four decisions: introduction, overview, Body 1, Body 2. Choose well and the complete 150-word report assembles below.
No time limit. A wrong answer here teaches more than a lucky one — read every explanation.
No judgement — just a record of where you are starting from.
Intro, overview, two grouped bodies. No conclusion, no causes, no opinions.
One sentence, no numbers, starts with "Overall," — it decides the band.
Rose vs fell. Landmarks as evidence for patterns you named.
Ten verbs, six adverbs, noun forms, proportions — the whole machine.
Introduction, overview, two grouped bodies — with the overview written first and no conclusion anywhere. Lesson 9 covers the two Task 1 types nobody practises until they appear on the paper: maps and processes.
Maps and processes appear on roughly one paper in five, and most candidates meet them for the first time in the exam room. Yet they are the most formulaic Task 1 types of all — no trends, no data selection dilemmas, just two grammar machines: the passive voice and sequencing language. Learn the machines and these become the easiest twenty minutes on the paper.
A map question is a gift wearing a disguise. There are no numbers to mis-read, and the whole report runs on about eight verbs — provided you can put them in the passive.
Two maps, two dates. Your job: report the changes. The actor is unknown (who demolished the factory? nobody says), which is exactly what the passive is for:
A process diagram — how bricks are made, how honey is produced — is a chain of stages. The report walks the chain once, in order, in the present passive:
Maps carry dates, and the dates dictate the tense. Misreading them is the highest-frequency error in map reports:
A town map in 1995 and today, described in text below. Build the report: pick the overview, then order the process of one transformation, then choose the tense-correct detail sentences.
Candidates fear maps because they practised graphs. After this lesson you have practised maps. The fear was the only hard part.
No time limit. A wrong answer here teaches more than a lucky one — read every explanation.
No judgement — just a record of where you are starting from.
Eight passive verbs, compass points, "on the site of the former…"
Present passive, sequenced: first, then, once, at the final stage.
Past-past, past-now, now-future: the dates make the tense decision once.
Character of change for maps; stage count, start, end for processes.
Passive voice, sequencers, location phrases, and the date-pair tense rule: both rare Task 1 types are now the formulaic twenty minutes they always secretly were. One lesson remains — the three minutes at the end of the exam that recover more marks than any paragraph.
Here is a trade every candidate is offered and almost every candidate refuses: three minutes of checking, in exchange for ten to fifteen small errors removed. Grammar and Lexical accuracy are marked partly on error density — how often the reading is interrupted. Removing ten interruptions moves that needle more than any final paragraph could.
Vietnamese candidates lose most of their Grammar marks to five error types they can all correct when shown them. Under pressure, the writing hand makes the error and the reading eye never comes back for it. This lesson builds the coming back.
Five passes, fixed order, highest-frequency errors first. Each pass, your eye hunts one thing only — thirty seconds per pass across the whole script:
Take your last three practice essays and mark every error. You will find something both embarrassing and wonderful: it is mostly the same three errors, again and again. That short list is worth more than any grammar book, because it is your grammar book.
Band 7 is not written by people who know more grammar. It is written by people who stopped making their own three favourite mistakes.
Below is a genuine Band 6 paragraph containing eight errors — articles, endings, tense, collocation. Tap each error you find. The gym counts your catches and shows anything you missed.
Errors removed and a minute still on the clock? One upgrade pass — not rewriting, just three surgical swaps anywhere they fit:
A complete Band 6 short answer, thirteen errors seeded across every pass type. Run the five passes in order and catch them all — this is the exam-day rehearsal.
Ten lessons compress to this. Read it the night before the exam and never again:
Structure under pressure — that was the whole course. The candidate who does ordinary things in the right order beats the candidate who does impressive things in a panic. Go and be the first one.
No time limit. A wrong answer here teaches more than a lucky one — read every explanation.
No judgement — just a record of where you are starting from.
Articles → endings → tense → pairs → your list. Thirty seconds each.
Your three favourite errors, hunted until they graduate.
Precise verb, anchored "this", hedged claim — three swaps, then hands off.
Structure under pressure: ordinary things, right order, every time.
Ten lessons: the sentence, the paragraph, the question, the frame, the shapes, the words, the report, the maps, and the three minutes that protect all of it. The Writing course is finished — everything from here is practice essays, run through the workspace, checked with the five passes. Band by band, the structure does the lifting.
20 structured lessons built specifically for Vietnamese IELTS candidates. Each lesson follows the proven sequence: clear explanation, real examples, then guided practice.
What examiners ask, what they listen for, and how to answer naturally. The foundation every Part 1 answer is built on.
Hometown, family, daily life — the foundational topics every Part 1 starts with. Master these and you're armed for 70% of questions.
The "Do you like X?" formula, comparative grammar, and how to handle topics that genuinely bore you. 4 of every 10 Part 1 questions live here.
The grammar pillar of Part 1. Tense mastery for Vietnamese speakers — past tense, present perfect, future forms, and time-stamping detail.
The length lesson. Four extension moves — Example, Contrast, Consequence, Anecdote — that turn 8-second answers into 30-second ones, without padding.
The two-minute long turn. The prep ritual, the I·D·H·R framework, the 8 cue card categories, and the pacing markers that turn two minutes from terrifying into manageable.
The first Part 2 category — and the one Vietnamese candidates underperform on most. The trait-listing trap, the S·R·M Story framework, and how to handle Vietnamese kinship in English without awkwardness.
The second Part 2 category. The tourist-brochure trap that caps Vietnamese candidates at Band 5-6 — and Sensory Anchoring, the framework that replaces "very beautiful, very famous" with a place the examiner can actually picture.
The third Part 2 category. The chronological play-by-play trap that caps Vietnamese candidates at Band 5-6 — and Peak Moment Zoom, the framework that replaces 20 minutes of "then this, then that" with 30 seconds of the moment that actually mattered.
The fourth and final Part 2 category. The product-description trap that caps Vietnamese candidates at Band 5-6 — and Significance Unwrap, the framework that replaces "it is made of X, it is Y" with the three layers of meaning the object actually carries.
The first Part 3 lesson. The register shift from descriptive Part 2 to abstract Part 3 — and the P·E·E framework (Position · Evidence · Extension) for opinion questions. The section where most Vietnamese candidates lose Band 7, with the structural fix.
The second Part 3 question type. The X·Y·Z framework (eXtract the dimension · compare Y values · Zoom into one telling detail). Plus the Listing Trap failure mode and the "what's actually interesting" reframe move that lifts answers from Band 6 to Band 7.
The third Part 3 question type — the "why" question. The C·M·E framework (Cause · Mechanism · Effect), the Jump Trap failure mode, and the second-order effect that lifts answers from Band 6 to Band 8. The mechanism is the missing middle most candidates skip.
The fourth and final Part 3 question type — the "what will happen" question. The I·F·B framework (If-statement · First-order · Beyond), the Hedge Paralysis failure mode, and the non-obvious projection that lifts answers from Band 6 to Band 8. Commit to a grounded future, don't refuse the question.
The Part 3 capstone. Real questions don't come labelled — the examiner mixes types within a single thread. The framework-selector for reading unlabelled questions, the blended questions that need two frameworks at once, and switching fluently on the fly. Where P·E·E, X·Y·Z, C·M·E, and I·F·B become one system.
The first Core Skills lesson. The frameworks gave you what to say; now we sharpen how you say it. Pronunciation isn't accent elimination — it's finding the music English carries: punching the right syllables, linking words into a flow, and letting your pitch move for meaning. The S·F·M framework, built for Vietnamese speakers escaping syllable-timed, tonal habits.
The second Core Skills lesson. Fluency isn't speed — it's never letting the stream stop, and keeping it connected. The G·E·M framework: glide through gaps instead of stalling, extend every answer so you never dry up, and maneuver around words you don't have instead of freezing. Built for the Vietnamese learner's freeze-and-hunt habit.
The third Core Skills lesson. A wide vocabulary isn't rare, showy words — it's precision, variety, and natural fit. The P·F·N framework: upgrade vague words to precise ones, build topic word-families so you never repeat, and use collocations and idiom that sound natural rather than translated. Built for the Vietnamese learner's "good / nice / very" plateau.
The fourth and final Core Skills lesson. Grammar is scored on two halves — range and accuracy — and learners over-think both. The B·U·F framework: build correct simple sentences first, upgrade some to a few high-value complex structures, and fix the handful of errors that stack up fastest. Reach for grammar the way you reach for a precise word — never cram.
Complete 14-minute simulation — Parts 1, 2 and 3 under exam conditions.
In the next forty-five minutes, you will master the most important framework for the first four minutes of your IELTS Speaking test — the framework that turns nervous, short answers into confident responses that examiners reward.
"You don't need perfect English to score Band 7 in Part 1. You need a structure your brain can fall back on when you're nervous. That's exactly what this lesson gives you."
Nine stages, designed to take you from understanding Part 1 to performing it. Each stage builds on the last. By the end, the A + R + D method will feel automatic.
You are here — getting oriented.
What actually happens in those 4 minutes. The timeline, the rules, the rhythm.
What examiners write down. Fluency, vocabulary, grammar, pronunciation — each one explained.
What costs candidates the most marks — and how to fix each one immediately.
The framework. We'll build it together, piece by piece.
See the method applied to five real Part 1 topics.
The pronunciation issues examiners hear most often — and how to fix them.
Five exercises building from recognition to full production.
Self-assessment, your stats, and the path to Lesson 2.
This is not a textbook. You'll learn by doing. Here's what to expect.
Each screen focuses on a single idea. No long pages, no overwhelm. Click through at your own pace.
Every few screens you'll be asked to click, choose, drag, or type. The doing is where the learning happens.
The dots at the top show your progress. Click any completed stage to revisit it.
Soft tones confirm your answers. You can turn them off anytime using the sound icon in the header.
Picture this: you've been waiting outside for twenty minutes. The examiner opens the door and invites you in. Your hands are slightly sweaty. Your mind is racing through everything you memorised.
The examiner smiles. They say good morning. They check your ID and ask you to confirm your name. This part is friendly, but it's not casual — they're already listening. From the moment you say "yes, that's correct," they have begun forming an impression.
Then comes the line you've been waiting for: "Now in this first part, I'd like to ask you some questions about yourself."
Part 1 has begun.
Part 1 lasts four to five minutes total. Inside those minutes, there is a predictable rhythm. Once you see it, you'll never feel lost again.
This is the single most useful number in this lesson. Most candidates either undershoot it or massively overshoot it. Hit it consistently and you sound prepared, not nervous.
While you speak, the examiner is making quiet notes on a scoring sheet. Here is what is going through their head — and what they are writing down.
Notice what they are not doing. They are not waiting for big mistakes to punish you. They are looking for moments of natural English — a well-used idiom, a complex sentence structure, a confident shift between tenses.
Every time you do something well, a tick goes on their sheet. Your job is to give them as many opportunities to tick as possible — and the A + R + D method does exactly that.
Two quick questions before we move on. Take a moment — there is no time pressure here.
Your speaking band is not a single number that the examiner pulls from thin air. It is the average of four separate scores — each one watching a different part of your English. Understand them, and you can target your practice with precision.
Can you keep going without long pauses, and does your speech flow logically?
How wide is your vocabulary, and do you use words accurately and naturally?
Do you use a variety of structures, and how accurate are they?
Can the examiner understand you easily — sounds, stress, intonation?
This is the pillar that decides whether you sound like a confident speaker or a struggling one — often within the first thirty seconds. Two skills inside one score.
Speaking at a comfortable pace, without long unnatural pauses while you search for words. Crucially, fluency is not about speaking fast — it is about speaking continuously.
Your ideas connect logically. One sentence leads naturally to the next. The examiner can follow your thought without effort.
This pillar is about more than just knowing words. It is about using the right word at the right moment — and showing the examiner that your vocabulary extends beyond textbook basics.
Range, precision, naturalness, and the ability to paraphrase when you don't know a word. Hit all four and you are firmly in the 7+ territory.
"My city is very big and there are many many people. The food is good and the weather is nice."
"My city is incredibly densely populated — it can feel overwhelming at first. But the street food scene is something else, and the climate is fairly mild year-round."
The examiner is not just listening for whether your grammar is correct. They are listening for variety. Using ten different grammar structures, even with a small mistake or two, scores higher than using two perfect structures.
This is the rule that flips most Vietnamese candidates' approach. You don't need flawless grammar — you need visible variety. Show the examiner you can use complex structures, even imperfectly, and the score rises.
"I've lived here for about three years."
"If I had more free time, I'd probably take up photography."
"I used to play football every weekend when I was younger."
"That might be why so many people enjoy it."
"It's much busier than it was five years ago."
"My grandmother, who lives in Đà Lạt, taught me to cook."
This is the most misunderstood pillar. You do not need a British or American accent. The examiner does not care if you sound Vietnamese. They care about one thing: can you be understood without effort?
Pronunciation isn't only about individual sounds. It's about how those sounds combine into words, how words combine into sentences, and how your voice moves up and down to carry meaning.
The basic phonemes. For Vietnamese speakers, the common issues are final consonants (-s, -ed, -t, -k), the V/W distinction, and the th sound. We'll cover these in detail in Stage 7.
English has stressed syllables — say PHO-to, not pho-TO. Vietnamese is tonal, not stress-based, so this is a real shift. Saying "DEvelopment" instead of "deVELopment" can confuse native listeners.
Your voice rises and falls to carry meaning. Questions go up at the end. Surprises rise sharply. Boring lists stay flat. Examiners notice intonation immediately — it makes you sound natural even when individual words are imperfect.
Drag each habit on the left into the pillar it belongs to on the right. The next button unlocks once all six are placed correctly.
By far the most common mistake Vietnamese candidates make. The examiner asks a question. You answer in three words. There is silence. The examiner moves on. Your band score quietly drops.
Same question. Same candidate. But this time, instead of stopping, they extend with the A + R + D framework you'll learn fully in Stage 5. Notice how the answer feels confident, not rehearsed.
The opposite mistake. You know you should extend your answer — so you do, and do, and keep going. Two minutes later you're still talking, the examiner has subtly looked at the clock twice, and you've used up time meant for three different questions.
Good speakers know when to stop. They build a clear answer with a beginning, middle, and end — and the end signals to the examiner that they are ready for the next question.
This one is harder to see, easier to fall into, and most dangerous of all. Many candidates memorise full answers for predicted topics. The examiner spots this almost immediately — and the penalty is severe.
Three answers below. For each, identify which of the three mistakes the candidate is making. Both questions must be answered correctly to move on.
This is the heart of the lesson. A simple formula that works on every Part 1 question you'll ever face. Three letters. Three sentences. Around fifteen to twenty seconds. Once it becomes automatic, your nerves stop mattering.
Watch the formula in action. Take a moment with this example — every Part 1 answer you give in your test should sound like this.
One short sentence. No "Let me think." No long pause. The Answer respects the examiner's question — it shows you understood and you're not stalling. Get this right and the rest flows naturally.
Direct, confident, one sentence. Often eight words or fewer.
This is where most candidates stop. They give an Answer and trail off. The Reason is what turns a one-word reply into a real answer. It tells the examiner you can think, not just respond.
One full sentence explaining why your Answer is true for you. Personal, specific, ideally using a complex grammar structure.
The Detail is what makes you sound human, not robotic. It's a specific example, a memory, a small story. Examiners love this part — because it's impossible to fake from a textbook.
One concrete, specific thing. A memory, a number, a name, a place, a story. Anything textbook-free.
"last week," "when I was twelve," "every Saturday"
"this little café in District 1," "my grandmother's village in Đà Lạt"
"my dad," "a colleague at work," "this friend of mine from university"
"for about three years," "two or three times a week," "the last ten years"
A real Part 1 question. You write each piece, one at a time. The system checks for the basics — length, structure, grammar markers — and shows you your assembled answer at the end.
The single most common Part 1 topic. Almost every IELTS Speaking test opens with this one. Let's see the A + R + D method applied to two real questions.
"I am a student. I study at university."
I'm a student at the moment — I'm in my third year of an international business programme, which I find genuinely fascinating because no two lectures cover the same kind of problem. I picked it specifically because I want to work in cross-border trade after I graduate next summer.
"Because I like it. It is interesting."
Honestly, it was a bit of a gradual decision. I'd been interested in how Vietnamese companies expand into other markets for a long time, and this programme seemed to cover exactly that. My uncle runs a small export business and watching him struggle with international contracts was probably what sparked the whole idea.
The second-most-common Part 1 topic. This one is dangerous — candidates fall into tourist-brochure language, which sounds memorised instantly.
"My hometown is Đà Nẵng. It is a beautiful coastal city in central Vietnam with a long and rich history…"
I'm from Đà Nẵng, on the coast in central Vietnam. It's the kind of city that's growing incredibly fast — what used to be a sleepy seaside town has become one of the busiest in the country. My family has lived there for three generations, so I've watched the change happen first-hand.
"I like the food. The food is very delicious and famous."
Probably the food, to be honest. Đà Nẵng has its own version of almost every Vietnamese dish, and most of them are spicier than what you'd find in the north. I genuinely miss mì Quảng when I'm away — it's a noodle dish you can only really get right at home.
A favourite of examiners because it reveals personality. The danger here is candidates listing hobbies like a CV — instead of telling small stories about them.
"In my free time, I like to read books, watch movies, listen to music, and spend time with my family and friends."
Most weekends I'll go for a long ride on my motorbike. It's the only time during the week when I'm not checking my phone every few minutes, and that mental break has become important for me. A couple of weeks ago I rode out to Củ Chi just to see somewhere new, and I came back feeling completely refreshed.
Actually, yes — I started learning film photography about six months ago. I was getting tired of the way phone photos all look the same, and shooting film forces you to slow down and actually think about each shot. I bought an old Olympus camera at a market in District 1, and developing my first roll was honestly one of the more exciting things I've done this year.
Food questions are golden because they reward specific details. Use a real dish name, a real memory, a real opinion — and the answer comes alive.
"No, I don't. I prefer to eat outside because it is convenient."
Not really, to be honest — I live two minutes from one of the best phở places in District 3, and it's almost cheaper to eat out than to buy the ingredients. The one thing I do make at home is breakfast — I've been a coffee snob for years and no shop quite gets it the way I want.
It changes, but right now it's probably bún bò Huế. There's something about the lemongrass and the spice level that you just don't get with phở — it's more punchy, more aggressive in flavour. There's a place on Lê Văn Sỹ street that does it the proper Huế way, and I've been going there at least once a week for the past three months.
A modern Part 1 staple. The trap here is sounding like a tech magazine review. The fix is making it about you — your habits, your frustrations, your one specific moment.
"I use my phone every day. It is very useful for me."
Honestly, more than I'd like to admit — my screen time report says about six hours a day, which is embarrassing when I think about how much else I could be doing with that time. I tried switching to a "dumb phone" for a week last month, and the first two days were genuinely uncomfortable — that's when I realised how dependent I'd become.
Mostly messaging and music, with way too much scrolling thrown in. Zalo is basically how my family stays in touch — my mother sends voice notes faster than most people type — so the phone is genuinely the easiest way to keep up with everyone. Beyond that, Spotify probably gets the most use — I've been deep into Vietnamese indie this year, especially Ngọt and Vũ.
Notice what every strong answer has in common: a specific name, place, time, or number. Phở on Lê Văn Sỹ. A motorbike ride to Củ Chi. Ngọt and Vũ on Spotify. Six hours of screen time. These tiny anchors are what stop your answer from sounding generic — and they're exactly what examiners reward.
Every language leaves a fingerprint on how its speakers produce English. Vietnamese has a few patterns examiners recognise instantly. The good news: most of them are quick to fix — and fixing just two or three is the difference between Band 6 and Band 7 in pronunciation.
Words ending in -s, -ed, -t, -k, -d, -th often get dropped or softened. "asked" becomes "ask"; "works" becomes "work".
Vietnamese has no /v/ sound, so "very" often sounds like "wery," and "work" can sound like "vork" when over-corrected.
Vietnamese doesn't conjugate verbs — so "-ed" endings disappear easily. "I worked yesterday" becomes "I work yesterday."
Vietnamese is tonal, not stressed. English needs syllable emphasis: PHO-to, not pho-TO. Getting this wrong can confuse listeners completely.
This is the single most common Vietnamese pattern examiners hear. The endings of English words simply disappear. Once you start listening for it, you'll hear it everywhere.
Vietnamese words rarely end with consonant clusters. When you carry that habit into English, words like asked, walked, worked lose their -ed. Plural and possessive -s endings get dropped. Past tense disappears.
For one full day, exaggerate every word ending you speak. Yes, even at home, even in Vietnamese-mixed sentences. You will feel ridiculous. After 24 hours, the muscle memory shifts — and you can ease back into a natural rhythm with the endings preserved.
Vietnamese has no /v/ sound. Vietnamese has no /w/ sound either — but Vietnamese speakers often substitute one for the other, which creates the classic confusion: "vork" for work, or "wery good" for very good.
The /v/ is made by gently biting your lower lip and letting the air vibrate through. You'll feel the buzz. Try saying "vvvvvery" — feel your top teeth touching your lip.
The /w/ is lips only — no teeth involved. Like blowing out a candle softly. Try saying "wwwwwork" — your lips should pucker, not bite.
Look in a mirror while practising. If your top teeth aren't touching your bottom lip for /v/ — it's not a /v/. If your lips aren't rounded for /w/ — it's not a /w/. Watch the mirror, train the muscle, the sound follows.
Vietnamese doesn't conjugate verbs. To express the past, you add words like đã or hôm qua — the verb itself stays the same. In English, the verb has to change, and that -ed at the end is critical.
When you say "I work yesterday" instead of "I worked yesterday," two scoring pillars suffer at once. Pronunciation drops because the consonant cluster is missing. Grammar drops because the past tense is missing. One small habit, two penalties.
English has three different sounds for the -ed ending — and getting them right makes you sound dramatically more confident.
For one week, read aloud for 5 minutes a day from any English news article. Force yourself to land every -ed ending. Don't worry about which sound it is yet — just make sure the consonant exists. Your brain learns the patterns naturally after a few days.
Vietnamese is a tonal language — pitch carries meaning at the word level. English is a stressed language — emphasis carries meaning at the syllable and sentence level. This shift is one of the most important pronunciation upgrades you can make.
Most English words of two syllables or more have one stressed syllable. Get it wrong and even simple words become unrecognisable. PHO-to and pho-TO sound like different words to a native ear.
English voices rise and fall to carry meaning. A flat, robotic delivery is one of the biggest giveaways of memorised speech — and one of the easiest things to fix once you notice it.
Examiners are forgiving of accent. They are not forgiving of monotone delivery. If your voice carries music — rises at questions, falls at conclusions, lifts at interesting words — you score Band 7 in pronunciation even with a clearly Vietnamese accent. The music is more powerful than the sounds.
A real Part 1 answer, broken into three sentences. Your job: identify which is the Answer, which is the Reason, and which is the Detail. Click the matching letter under each sentence.
Three answers below. Each is incomplete in a different way. Identify what each is missing — Answer, Reason, or Detail.
"Yes, I do. I grew up there and my parents still live there."
"I really enjoy reading because it lets my mind switch off after work."
"Mainly because I love trying food I've never had before — last year in Hội An I had cao lầu for the first time."
Below are two real Part 1 questions with the A already written for you. Your job: type your own Reason and Detail to complete each answer.
Given (A): Most of the time, yes.
Given (A): Definitely alone.
Five real Part 1 questions. Type a complete A + R + D answer for each. Aim for 25–35 words. The counter will tell you if you're in range.
The whole point of speaking practice is speaking. Take the five answers you just wrote and read each one aloud three times. First slowly, then at normal pace, then as if you're in the test room.
Here's how you did across the five exercises. These results are saved for the lesson — you can come back and improve them anytime.
Your performance across the practice arena suggests you have absorbed the A + R + D framework well. Move into Stage 9 to complete the lesson.
Before we celebrate, a moment of honesty. For each statement below, drag the slider to where you are right now. There is no judgement here — just a record of your starting point.
Forty-five minutes of focused work. Here is everything you covered in this single lesson — and what it represents for the rest of your IELTS journey.
Lesson 1 lays the foundation. The next nineteen lessons build the house. Here's what's coming, and why each lesson matters.
You'll build vocabulary banks for the most common Part 1 topics — hometown, family, work, hobbies — and practise extending short answers without ever sounding rehearsed.
The famous cue card. You'll learn to plan in 60 seconds, structure 2 minutes of speech, and never run out of things to say. Topics: describing people, places, events, and objects.
Abstract questions, opinions, speculation about the future. This is where Band 7+ candidates separate themselves — and where The Lionel Method scales beautifully.
Deep dives on pronunciation for Vietnamese speakers, building fluency, expanding vocabulary range, mastering complex grammar, and a full mock test under exam conditions.
"You've done the hardest part — you've started, and you've finished a whole lesson. From here, every lesson is just one more brick in the wall. Twenty bricks. That's all this is."
You have officially completed Part 1 Foundations. This badge is now part of your learner profile. One down, nineteen to go.
In Lesson 1 you learned the framework — the A + R + D method that works on any Part 1 question. In Lesson 2, you learn the topics themselves. The questions Vietnamese candidates face most often, and how to answer each one with the kind of detail examiners reward.
"Knowing the framework is one thing. Having the words ready when the examiner asks about your family in Vietnamese is another. This lesson closes that gap."
Before we dive into topics, a thirty-second refresher on what you mastered in Lesson 1. If any of these feels shaky, jump back — there is no shame in revisiting the foundation.
Every Part 1 answer needs an Answer, a Reason, and a Detail. Three sentences. 15–20 seconds.
Short answers, rambling, and memorisation. Each one signals the examiner to lower your fluency band.
Fluency, Lexical Resource, Grammar, Pronunciation. Each carries equal weight.
Final consonants, V/W swap, missing tense markers, word stress, monotone delivery.
Lesson 2 is denser than Lesson 1 — less framework, more content. Here's what's coming.
You are here.
What makes Part 1 questions personal, why this is good for you, and the three categories you'll face.
The single most common Part 1 topic. Vocabulary, common questions, the tourist-brochure trap.
Generational vocabulary, Vietnamese family complexity, story templates.
Time expressions, habit vocabulary, present-simple precision.
80+ phrases organised by Band level, with audio. Bookmark this — you'll come back to it.
Vietnamese phrases that fail when translated literally — and their natural English equivalents.
Six exercises focused specifically on personal topics.
Self-assessment, stats, badge, and what's coming in Lesson 3.
There is a reason every Part 1 question is about your life. The examiner isn't curious about you — they need a controlled context to test your English. Personal questions give them that control.
Personal questions are easier to understand than abstract ones — that's the gift. But they are also harder to fake. You can memorise an answer about climate change. You cannot memorise an answer about your own mother that doesn't sound rehearsed.
Every personal question is about something you know more than the examiner does. You've lived in your hometown longer than they have. You know your family. You don't need to research anything.
You can describe your father in Vietnamese in five seconds with rich detail. Doing it in English requires specific vocabulary you may not have. That's what this lesson gives you.
The more truthful your answer, the more natural it sounds. The more natural it sounds, the more fluency marks you score. Truth is not just ethical — it's a Band 7 weapon.
Every personal Part 1 question fits into one of three buckets. Once you can recognise which bucket a question belongs to, you know exactly what vocabulary to reach for.
Where you live, your hometown, your house or apartment, your neighbourhood.
Your family, your friends, your colleagues, the people in your life.
Your daily life, your habits, what you do on weekends, your relationship with time.
This is a point most IELTS books miss. Vietnamese culture is rich in exactly the kind of detail Part 1 examiners reward. You just need to know how to channel it into English.
Many Vietnamese families have three or four generations under one roof. That's instant material for any "family" question — and Western examiners find it genuinely interesting.
Where Western candidates struggle to talk about "what they ate yesterday," you have phở bò for breakfast, cơm tấm for lunch, and a specific bún chả place you visit every Friday.
Tết, the Mid-Autumn Festival, ancestor commemorations — these are emotionally rich topics most candidates worldwide cannot match.
From cà phê sữa đá at street stalls to motorbike commutes through Saigon traffic — your daily life is naturally interesting if you learn to describe it precisely.
Across hundreds of personal Part 1 answers I've seen scored, the same things separate Band 6 from Band 7. Here are the four that matter most.
"District 3" beats "the city." "My grandmother Hà" beats "an older relative." "Last Tết" beats "during a holiday." Every proper noun signals a real life behind the answer.
"I've been living there for three years" scores higher than "I live there." Present perfect signals grammar range AND specific commitment to the timeline.
"District 3 is busier than where I grew up" earns more than a flat description. Comparative grammar is hard for many candidates, so using it well stands out.
"Which I genuinely love" or "to be honest, it can be exhausting" — these tiny editorial flashes show personality and confidence. They are the single biggest differentiator at Band 7+.
Two questions before we move on. Take your time.
Of every Part 1 topic, hometown shows up in roughly seventy percent of all IELTS Speaking tests. The examiner almost always opens here. Master this one topic and you've handled the first two minutes of your test before you even sit down.
Hometown is not just where you were born. The examiner usually means where you grew up or where you currently live — whichever you have most to say about. Pick the one that gives you richer material, then commit to it for the whole question.
After fifteen years of watching IELTS Speaking tests, I can tell you that roughly six questions cover ninety percent of hometown conversations. Memorise the questions, not the answers — then practice your A + R + D structure on each one.
This is the single most common mistake on hometown questions. The candidate has practised describing their city for tourists — and they forget that the examiner is not a tourist. They are a fluency examiner.
"My hometown is Hà Nội, a beautiful capital city located in northern Vietnam. It has a rich and ancient history dating back over a thousand years, with many historical sites such as the Temple of Literature, the One Pillar Pagoda, and the Old Quarter. The local cuisine is world-renowned, particularly phở and bún chả, which have attracted visitors from all around the globe..."
"I'm from Hà Nội — I've lived in the same neighbourhood near Hồ Tây for most of my life. It's funny, my parents bought our apartment when the area was completely quiet, and now there's a coffee shop on every corner. I sort of miss the quietness, but the food got significantly better, so I can't really complain."
Twenty-four phrases organised by Band level. Click any phrase to hear it. These are exactly the kinds of words that take you from Band 6 to Band 7 — natural, specific, varied.
The framework applied to four real hometown questions. Click each card to reveal the model answer with colour-coded A, R, and D.
I'm from Hà Nội, in the north of Vietnam. It's the kind of city where everything happens at street level — the cafés, the food stalls, the conversations spill out onto the pavement. I've lived in the same neighbourhood near West Lake for about twenty years, which probably explains why I can't imagine living anywhere else.
Honestly, it's been transformed beyond recognition. When I was growing up, the area around our apartment was mostly quiet residential streets, and now it's full of cafés, gyms, and co-working spaces. My parents complain about it every Sunday at lunch, but I secretly love that everything is so accessible now.
Probably the food scene, if I had to pick one thing. Hà Nội takes its food incredibly seriously — there are restaurants that have been doing the same dish for three generations, and they're often better than the fancy places. There's a tiny bún chả place on Lê Văn Hưu street that I go to every other Saturday — it's been there since the seventies.
It depends on what they're looking for, but for most people — yes. Hà Nội has this rare combination of low cost of living, strong food culture, and walkability that you really don't get in many capital cities. A friend of mine from London moved here last year and within three months she was telling me she'd never go back.
Fill in this skeleton with your own answers. By the end you'll have a personal hometown answer you can adapt to any of the six common questions. Save it, practise it, but never memorise it word-for-word.
After hometown, family is the second-most-common Part 1 topic. And for Vietnamese candidates, it is also the topic where you have the most natural advantage — because Vietnamese family life is genuinely richer in detail than what most Western examiners are used to hearing.
In a typical Western family answer, the candidate has two parents, maybe a sibling, and the family lives in different cities. The answer is over in fifteen seconds because there is not much to say. Vietnamese families have more layers, more relationships, more shared spaces — and the examiner finds that genuinely interesting.
Slightly more variety than hometown, but the patterns repeat. Master the structure and any family question becomes manageable.
Vietnamese has more precise family terms than English. Where Vietnamese distinguishes between bác, chú, cậu, English just says "uncle." Here's how to talk about your family without losing the texture.
Beyond who is in your family, the examiner wants to hear what they're like and how you relate to them. These are the two vocabulary banks most Vietnamese candidates lack — and they're the easiest to fix.
Four real family questions with strong model answers. Click to reveal each one.
We're a pretty typical three-generation household. My parents, my grandmother on my mother's side, and my younger brother all live in the same apartment in District 5 — which sounds chaotic but actually works really well most of the time. My grandmother does all the cooking, and we have dinner together every single evening, which is non-negotiable in our family.
Probably my grandmother, oddly enough. My parents have always been busy with work, but my grandmother basically raised me — she has this incredibly patient way of listening to everything I say without judging. Even now, at 78, she's the first person I call when something goes wrong at work, before my parents.
I'm definitely my mother's daughter when it comes to temperament. She's that classic Vietnamese mum — appears soft-spoken but is genuinely tough underneath, and I've inherited exactly that combination. Last year my colleagues at work nicknamed me "the quiet storm," and when I told my mother, she laughed for about five minutes.
Tết is the big one for us — it's basically the equivalent of Christmas, New Year, and a family reunion all rolled into one. Everyone in my extended family travels back to our grandparents' house in Bắc Ninh for a week, even relatives who live abroad. My favourite moment is the night before Tết — there must be twenty of us in one tiny kitchen, all making bánh chưng together.
Same approach as the hometown template — fill in the blanks with your real life. The result is a personal answer you can flex onto any family question the examiner asks.
After hometown and family, daily life is the third pillar of Part 1 personal questions. It seems easy — "what do you do every day?" — but it's the topic where most Vietnamese candidates lose marks they didn't even realise were available. The reason is almost always grammar.
Daily life questions are secretly a tense test. The examiner is checking whether you can move smoothly between present simple (habits), present continuous (current activities), and frequency adverbs (sometimes, often, never). Get this right and you score Band 7+ on grammar within the first answer.
"I wake up at six." Use this for things you do regularly. The verb stays in its base form.
"At the moment I'm trying to wake up earlier." Use this for what's currently happening or a recent change in habit.
"I used to stay up until 2am." A Band 7 structure that signals you understand how habits change over time.
"I occasionally skip breakfast." Adverbs like occasionally, usually, rarely add precision and variety.
Vietnamese candidates often use the same three time phrases over and over — "every day," "in the morning," "after work." The fix is a vocabulary upgrade. Twelve time expressions, organised by what they let you say.
Same approach as hometown and family — six questions cover almost everything the examiner will ask about your daily routine. Each one has a hidden trap.
Four real daily-life questions with strong model answers. Notice how each one mixes present simple, frequency adverbs, and specific time anchors.
My weekdays are pretty structured, more by habit than design. I'm at my desk by 8:30 most mornings — give or take a few minutes — and I try to do the hardest work before lunch, because I know my focus drops after that. My one ritual is a cà phê sữa đá from the same stall on Pasteur street, religiously, at 9am.
Definitely a morning person, though I haven't always been. I used to stay up until two or three in the morning back in university, but at some point my body just gave up on that — now I'm in bed by ten without fail. There's something about being awake at 6am with a quiet city outside the window that I've come to genuinely enjoy.
My weekends look completely different from my weekdays, which is the whole point. Saturdays I keep open — usually a slow brunch with friends in District 2, then a long walk somewhere — but Sundays are non-negotiable family lunch. My grandmother cooks for ten of us every single Sunday, and I genuinely look forward to it all week.
Quite a lot, actually — more than I realised until you asked. A year ago I was working from home five days a week and going to the gym at 9pm; now I'm in the office most days and the gym has moved to before work entirely. It sounds small, but it's flipped my whole evening — I'm in bed at ten now instead of finishing dinner at ten.
Same approach as before — fill in your real life. The result is a personal answer you can adapt to any of the six daily-life questions.
This is the stage to bookmark. Over 80 phrases organised by topic and Band level, every one clickable to hear. The mistake most candidates make is trying to memorise the whole bank. Don't. Pick five per topic — the ones that feel like things you would actually say — and learn those properly.
Click every phrase you find interesting. Hearing the rhythm matters more than reading it.
From each band, choose the phrases that sound like something the real you would say. Quality over coverage.
Repeat each one three times until it feels natural in your mouth. That's when it becomes yours.
Don't try to absorb this in one sitting. Use it as a reference — return between practice sessions.
Twenty-one phrases for describing cities, neighbourhoods, and the place you live. Useful for hometown questions, "where do you live," and any question about location.
Twenty-one phrases for describing the people closest to you. Mix personality words with relationship language for maximum range.
Twenty-one phrases for describing your routines, weekends, and the rhythm of your week. Includes the high-value frequency adverbs that lift fluency scoring instantly.
The smallest topic but the highest-value. These are the phrases that signal personality to the examiner — the editorial flashes that separate Band 6 from Band 7. Twenty-one of them, ready to use.
You do not need to climb from Band 5.5 to Band 7.5 in one go. Pick one phrase from the next band up and use it three times this week. When it feels natural, pick another. That is how vocabulary actually moves — one phrase at a time, until your brain reaches for the new word automatically.
Every Vietnamese candidate, no matter how advanced, mentally translates from Vietnamese to English at some point during the test. The problem is that some Vietnamese phrases translate beautifully, while others — when translated word-for-word — produce English that no native speaker would ever say. This stage shows you the most common traps and their fixes.
A wrong word is a small mistake. A directly translated phrase is a bigger one — because it doesn't just sound foreign, it actively confuses the examiner. They have to pause and decode what you meant. Each pause costs you fluency marks.
When you're nervous, your brain reaches for Vietnamese first and then translates. It feels like the "right" way — but the structure is wrong.
The patterns are predictable. Examiners trained on Vietnamese candidates hear the same mistakes hundreds of times — and notice them every time.
Most translation traps have a one-line fix. Once you know the natural English equivalent, you stop reaching for the literal version automatically.
Fixing five of these moves you from "clearly Vietnamese" to "comfortable English speaker" without any new vocabulary. Pure structural change.
These five appear in roughly every other Vietnamese IELTS Speaking test. Master these and you've already eliminated half the translation damage.
These five are subtler. They don't break English completely — they just make your speech sound slightly off, which the examiner registers as a fluency issue even if they can't quite name it.
Four sentences below. Each contains exactly one translation trap from this stage. Click the part of the sentence that's wrong, then read the explanation. All four must be answered to unlock the next screen.
Knowing the traps is one thing. Stopping yourself from falling into them under exam pressure is another. Here is the only method I've seen work — a habit-forming approach over seven days.
Don't try to fix all ten at once. Choose the three from this lesson that feel most like patterns you use. Write them down somewhere visible.
For two full days, every time you speak any English at all, listen for your three traps. When you catch one, stop mid-sentence and rephrase. It will feel embarrassing. Do it anyway.
Record yourself answering five Part 1 questions each day. Listen back. Mark every trap you slipped into. Re-record those specific sentences.
By now your three target traps should feel automatic to avoid. Move on to three new ones from the list. Repeat the cycle. In a month you've eliminated every translation trap from your speech.
You will always translate from Vietnamese to some degree, no matter how advanced you become. Fluent bilinguals do it too. The goal isn't to stop translating — the goal is to catch the bad translations before they leave your mouth. That catching gets faster with practice, until eventually it happens in milliseconds and feels automatic. That's the real definition of fluency.
Five Part 1 questions. Three categories. Drop each question into the right pile. Remember from Stage 2: Places, People, or Routines — every personal question belongs to one of these three.
Four pairs of phrases. In each pair, click the one that's Band 7+ (the more natural, more native-sounding option). Pay attention to specificity and idiomatic language.
Four sentences with translation traps from Stage 7. Type the corrected version in the field below each one. Don't overthink it — just write the natural English equivalent.
Five real personal-topic questions. Type your full A + R + D answer for each. Pull from the vocabulary banks if you need to. Aim for 25–35 words and at least one Band 7 phrase per answer.
Five answers written. Reading them once is not enough. Read each answer aloud three times — slow, normal, exam pace — then tick what you noticed.
Here's how you did across the five exercises. Same rule as before — these are saved for the lesson and you can revisit anytime.
Your performance across the personal-topics arena shows where the framework has settled and where it still needs reps. Move into Stage 9 to complete the lesson.
Lesson 1 was about the framework. Lesson 2 was about the topics — places, people, routines. For each statement below, drag the slider to where you genuinely are right now. No judgement. Just a snapshot.
Another forty-five minutes invested. Forty-six screens worked through. Here's what that adds up to.
Lesson 1 was the framework. Lesson 2 covered personal topics. Lesson 3 takes you into the opinion territory — and this is where things get interesting.
Every Part 1 test has at least three "do you like..." questions. They sound innocent, but they're a goldmine — opinion questions let you show personality, comparative grammar, and idiomatic expressions all in one breath. Lesson 3 shows you how.
The structure that turns a yes/no question into a Band 7 answer in 15 seconds.
Comparative grammar without sounding like a textbook — and the phrases examiners notice.
What to say when you genuinely don't like something the examiner is asking about.
30+ phrases for expressing preferences with conviction and nuance.
"Two lessons down. You've got the framework, you've got the topics. From here on, every lesson sharpens a specific edge. Keep coming back. The students who finish Lesson 5 have already changed by the time they reach Lesson 10. You're closer than you think."
You now own the framework and the foundational topic vocabulary. Two badges on your shelf. Eighteen lessons to go.
Lesson 1 gave you the framework. Lesson 2 gave you the foundational topics. Lesson 3 is where things start to get interesting — because opinion questions are where Vietnamese candidates either soar or stall. They are the most frequent type in Part 1, and also the type most often mishandled.
"I've sat in on hundreds of Speaking tests. The single biggest difference between Band 6 and Band 7 isn't grammar or vocabulary — it's how the candidate handles 'do you like.' This lesson fixes that."
Two lessons of foundation behind you. Before we layer opinions on top, a quick reminder of what's already in your toolkit.
From Lesson 1. Every Part 1 answer needs an Answer, a Reason, and a Detail.
From Lesson 2. Places, People, Routines — every personal question fits one of these buckets.
From Lesson 2. 80+ phrases by Band level. Still bookmarked, still useful.
From Lesson 2. The 10 most damaging Vietnamese-to-English literal translations.
Same 9-stage structure, new territory.
You are here.
Why "do you like" questions are secretly hard, and what examiners actually want from your answer.
Lionel's signature 4-part framework. Drag-and-build interactive.
The grammar that makes preferences sound natural — "rather than", "I prefer X to Y", and the trap of "more better."
How to say "I don't really like this" without sounding rude or bland.
80+ phrases across 4 topics, all clickable for audio.
What to say when the examiner asks about something you genuinely don't care about. Five rescue strategies.
Six exercises specifically on opinion questions.
Self-assessment, stats, badge, and what's coming in Lesson 4.
On the surface, "do you like coffee?" is the easiest possible question. You either do or you don't. That simplicity is the trap. Because the question is binary, candidates answer it binarily — "yes I do" — and lose marks they didn't realise were on the table.
The examiner doesn't care whether you like coffee. They care whether you can express, justify, and qualify an opinion — three skills bundled into one innocent-sounding question. A good answer demonstrates all three. A weak answer just hits the first one.
Take a clear position. "Yes" or "no" works — but better candidates lean further. "Honestly, I'm obsessed."
Give a reason that goes beyond "because it's good." Real reasons feel real. Generic ones feel rehearsed.
Add nuance. "Most of the time, anyway." "Apart from the milk-heavy ones." Qualifiers signal sophistication.
Connect to a specific moment, place, or memory. This is the A + R + D method applied to opinions.
Examiners ask opinion questions in four distinct patterns. Each one demands a slightly different response shape. Spot the pattern, pick the right tool.
"Do you like coffee?" · "Do you enjoy reading?" · "Are you interested in music?"
"Do you prefer tea or coffee?" · "Books or movies?" · "City or countryside?"
"How often do you exercise?" · "How often do you read?" · "How often do you eat out?"
"What do you think of social media?" · "What's your opinion on online learning?"
If Lesson 1's "3 killer mistakes" were about Part 1 generally, these three are specific to opinions. Each one quietly drops a candidate from Band 7 to Band 6.
The candidate answers the question and stops. Three words, all of which were given to them in the question. Zero language demonstration. This is the single most common opinion mistake.
"I like it because it's good." "I enjoy it because it's interesting." These aren't reasons — they're restatements of "I like it." The examiner registers them as filler.
The candidate pretends to love something they're indifferent about, because they've been told positive answers score better. Examiners detect this instantly — the language goes generic and the voice flattens.
Vietnamese culture has a distinct relationship with expressing strong opinions — and this is worth being aware of before you walk into the test. It's not a flaw. It's a calibration issue.
Soften opinions to maintain harmony. "Cũng được" ("it's okay too") and "tuỳ thôi" ("it depends, really") are the default settings. Direct strong opinions can feel impolite.
Western examiners expect a clear position. "It depends" without then specifying what it depends on reads as evasive or under-confident, not polite.
Two questions before we move to the formula in Stage 3.
After watching thousands of opinion answers, I've boiled the strongest ones down to four moves. They happen in roughly this order, though Band 7+ candidates blur the boundaries. Master the four pieces separately first, then learn to flow between them.
Take a clear stand. Lean further than yes/no — use intensity vocabulary.
Narrow the topic. What kind specifically? Which version?
Give a real, specific reason — not "it's good."
Qualify with a counter or specific detail. This is where Band 7 lives.
A real opinion question, broken into the four moves. Each piece colour-coded so you can see how they stack.
"Honestly, I'm obsessed with it."
"Mostly Vietnamese indie at the moment — bands like Ngọt and Vũ."
"The lyrics feel honest in a way that international pop usually doesn't, and the production has improved so much in the last few years."
"That said, I'll happily put on classical when I need to focus — it's basically the only genre I can work to."
The first move of the formula. Instead of "yes I do," reach for intensity. Each phrase signals a different level of enthusiasm — pick the one that's true.
A real opinion question. Below are the four pieces, scrambled. Drag each piece into its correct P · T · R · N slot. The pieces self-validate as you drop them.
Three real opinion questions across food, travel, and technology. Each model answer follows the P · T · R · N structure. Click to reveal.
Honestly, I've grown to love it. Specifically Vietnamese home cooking — the slow stuff, like bún bò Huế or phở. It started as a way to save money after I moved out of my parents' place, but now I find it genuinely meditative — chopping, simmering, getting the broth right. I'm useless at anything western, though — give me a pasta dish and I panic.
I'm absolutely a fan, yes. Mostly domestic travel within Vietnam — Đà Lạt, Hội An, the central coast. There's a whole side of the country that most people skip, and I find it more rewarding than the typical tourist routes — slower pace, better food, fewer crowds. International travel feels appealing in theory, but the logistics genuinely exhaust me — so I'm in no rush.
I have genuinely mixed feelings about it. I use Instagram and a little bit of TikTok, but I've completely given up on Facebook. The good moments — staying in touch with friends who've moved abroad, discovering new food places — are real. But the time it eats is also real, and I'm not always sure it's worth it. I've actually been trying to cut back lately, mostly by deleting the apps from my phone on weekends.
Fill in this skeleton with any topic of your choice — a hobby, a kind of food, a music genre. The template walks you through P · T · R · N step by step.
Pattern 2 from Stage 2 — the "X or Y?" question — is secretly a grammar test. Examiners use comparative questions to check whether you can handle a specific set of structures. Get those structures right and your grammar band jumps. Get them wrong and the answer fluency is irrelevant.
Comparative questions check three specific grammar skills: comparative adjectives ("better," "more interesting"), preference verbs ("I prefer," "I'd rather"), and contrast connectors ("whereas," "while," "on the other hand"). Most Vietnamese candidates get one of these right but never all three in the same answer.
"X is more interesting than Y" — the basic structure. Get this wrong and the whole answer collapses.
"I prefer X to Y" / "I'd rather X than Y" — Band 7 candidates use both, not just "I like X more."
"Whereas," "while," "on the other hand" — these signal sophistication. Vietnamese candidates often default to "but" only.
"I lean more towards X" or "I'd probably pick X" — softening preferences signals you can hold complexity.
English comparative adjectives have two forms — and Vietnamese speakers often mix them up. The rule is simple: short adjectives take "-er", long adjectives take "more". Once it clicks, you stop hesitating.
Vietnamese speakers default to "I like X more than Y." It works, but it's the Band 5 version. Here are the four structures that move you to Band 7 — pick two and rotate between them in your test.
"I prefer reading to watching TV — books leave more room for your imagination."
"I'd rather cook at home than eat out, mostly because I can control what goes into the food."
"I'd much sooner spend a Saturday wandering around District 3 than sit in a shopping mall."
"I lean more towards coffee, although tea has its moments in the afternoon."
A great preference answer doesn't just say "I prefer X." It acknowledges Y in the same breath. Contrast connectors are the words that let you do this gracefully. Most Vietnamese candidates use only one — "but" — over and over. Here are five more.
"Coffee energises me, whereas tea calms me down."
"While books require focus, podcasts I can listen to anywhere."
"District 1 has the buzz; on the other hand, District 2 is more livable."
"I usually drink coffee. That said, I'll switch to tea after 4pm."
"I lean towards quiet places, although I can handle a busy café for short bursts."
"Unlike most of my friends, I find watching films genuinely tiring."
Three blanks, one full sentence. Pick the right option for each slot. The sentence reads aloud when complete — and you'll see if the grammar holds together.
"I prefer coffee ___ tea ___ it gives me ___ energy in the morning."
"I prefer coffee to tea because it gives me more energy in the morning." Three Band 7 grammar features stacked: preference structure, causal connector, comparative.
Three real comparative questions with model Band 7+ answers. Notice the structures from this stage stacking up — preference verb + comparative + contrast connector, all in 25 seconds.
"Honestly, I lean more towards the city — at least for now. I grew up in a quiet town outside Đà Nẵng, and while it was peaceful, everything closed at 8pm and I was bored by sixteen. Saigon, on the other hand, has the kind of energy I find addictive — although I'd probably retire somewhere quieter."
"I'd much sooner travel alone, to be honest. Groups mean compromise — what to eat, when to start the day, how long to spend somewhere — whereas alone I can completely follow my own pace. That said, I did a week in Tokyo with two friends last year and it was genuinely one of the best trips I've ever had, so I won't say it's a rule."
"I prefer books to films, although I know I'm in the minority on that. Films are quicker and more social, sure, but books leave more room for your imagination — and they don't get worse with bad acting. Unlike most of my friends, I'd happily spend a Saturday afternoon with a novel rather than at the cinema."
Notice the pattern across all three answers: preference verb at the start, contrast connector in the middle, comparative adjective somewhere along the way. Three Band 7 grammar features per answer — and they sound like normal speech, not a textbook drill. That's the goal.
Most IELTS courses spend ninety percent of their time on positive answers. They teach you how to enthuse, praise, agree. Almost no time goes into the harder skill — how to express dislike or disagreement in confident, natural English. For Vietnamese speakers, this is where most band points are quietly lost.
Vietnamese candidates face a real cultural problem here. Saying "I don't like it" feels too direct — it cuts against politeness norms. But saying "yes, I like it" when you don't sounds rehearsed, generic, and the examiner detects it instantly. You need a third option — a way to disagree that is genuine, polite, and idiomatic in English.
Grammatically correct, but flat. The examiner hears a dead end. No language demonstration, no personality, no follow-up material.
The voice tightens. The vocabulary goes generic. The examiner registers the flatness and quietly drops your fluency band by half a point.
Confident dislike, with a graceful counter that shows perspective. Examiner hears nuance, sophistication, and genuine voice — all in one breath.
Native English speakers rarely express disagreement bluntly. They almost always soften the disagreement first — and then deliver it. Three moves, in this order: concede something positive, then pivot, then land your real opinion.
Acknowledge the appeal — even one small thing. This buys you permission to disagree.
Signal the turn. One word or phrase that prepares the examiner for your real view.
Deliver your real opinion — confidently. The hedge is what makes the dislike feel polite, not the dislike itself.
"I can see why people love shopping malls — that said, for me, they're genuinely exhausting and I avoid them when I can."
Twenty phrases for expressing disagreement, dislike, and skepticism in English without sounding rude. Pick three from each category and rotate them — using one per test is enough to register sophistication.
The hedge-then-disagree formula handles dislike. But the highest-scoring opinion answers go one step further — they hold two contradictory truths at once. The candidate likes something AND has a problem with it. Loves it AND finds it exhausting. This kind of nuance signals Band 8+ thinking.
"I adore Hà Nội — although I have to admit, I find the traffic genuinely exhausting after a long day."
Why it works: The "love" is sincere, the "exhausted" is also sincere — and both can be true. Examiner hears emotional intelligence, not contradiction.
"I love the idea of waking up at five and going for a run — in practice, I've never once actually done it."
Why it works: Acknowledges the gap between aspiration and reality. This is a deeply human thing to say, and examiners love it.
"I used to be obsessed with travel, but I've come to appreciate staying still — there's a lot to be said for knowing a city deeply rather than ten cities shallowly."
Why it works: Shows time, growth, and a considered position. "Used to" + "come to appreciate" is a Band 8 grammar stack.
"It depends entirely on the day — I can love a busy café in the morning and find the same café unbearable by four o'clock."
Why it works: Refuses the binary the examiner offered. But instead of being evasive, it justifies why with specifics. Confident hedging at its best.
Three opinion questions where the candidate doesn't love the topic — but answers strongly anyway. Notice how each one concedes, pivots, and lands.
"To be fair, I can see the appeal — there's something genuinely therapeutic about wandering through a mall on a quiet weekday. But honestly, the whole experience leaves me cold most of the time. The crowds, the air conditioning, the music — it adds up to something exhausting rather than fun. I'd much sooner pick up what I need online and spend that time doing literally anything else."
"In fairness, I have a lot of respect for it — my grandmother used to play đàn tranh at home and I grew up around it. Even so, I'd be lying if I said I actively listen to it now. I find it beautiful at weddings or in cafés, but I've never really warmed to putting it on by choice. It's a bit like reading classics — I appreciate them more than I enjoy them."
"It's complicated, honestly — I'm on Instagram most evenings, and yet I'd happily delete it tomorrow if I could. There's something to be said for staying connected with friends who live abroad, especially the ones I'd otherwise lose touch with. But the time it eats — and the way it makes me compare my life to strangers' — those parts I find genuinely tedious."
By the time you walk into your test, change how you think about opinion questions. The examiner is not asking whether you like something. They are asking whether you can express any opinion — positive, negative, or mixed — in fluent, natural English. Disagreement, done well, is just as Band 7 as enthusiasm. Done with hedging and nuance, it can be more.
Lesson 2's bank gave you 80+ phrases for personal topics. This one gives you 80+ phrases for expressing opinions about specific topics. Organised by the four most common Part 1 opinion areas: food, activities, places, and abstract concepts. Every phrase is clickable for audio.
Lesson 2's bank was general. This one gives you the exact phrases for the topics that come up most in Part 1 opinion questions.
Each topic has love, lukewarm, and dislike phrases. Pick the one that's actually true for you — and have all three ready in case the examiner pushes back.
Every phrase here is designed to combine with the P · T · R · N formula from Stage 3 and the hedge-then-disagree pattern from Stage 5.
Same rule as before — pick five per topic, learn those properly. Don't try to absorb all 80 in one session.
Twenty-one phrases for talking about food preferences. Useful for any "do you like..." question about cuisine, cooking, restaurants, or drinks. Includes both general food expressions and Vietnamese-specific food phrases.
Twenty-one phrases for talking about hobbies, sports, exercise, reading, gaming, and free-time activities. Every Part 1 test has at least one question in this territory.
Twenty-one phrases for talking about cities, countries, holidays, and travel preferences. Useful for "do you like travelling," "do you prefer cities or countryside," and any place-comparison question.
The trickiest category — opinions on topics that don't have an obvious "yes/no" answer. Social media, technology, change, learning, time. Twenty-one phrases for sounding intelligent about big topics without overcommitting.
Eighty phrases is too many to memorise. Here's what to do instead: open this bank once a week. Pick five phrases that genuinely sound like something you'd say. Practice each one out loud three times. The next week, pick five more. By the time you walk into your test, ten or fifteen phrases will feel like yours — and that's far more useful than memorising eighty that don't.
Every candidate hits this moment in the test. The examiner asks about something you have no opinion on, no experience with, and no interest in. "Do you like flowers?" "What do you think of traditional art?" Your mind goes blank. You stall. You panic. The fluency band drops half a point in three seconds.
Here's what most candidates miss. The boring question isn't a test of your interest — it's a test of your resourcefulness. Examiners deliberately ask about random, everyday topics to see whether you can generate language about anything, not just topics you've prepared for. The candidate who handles the boring question well demonstrates exactly the skill IELTS Speaking is designed to measure.
Two ways to generate language about a topic you don't care about. Both are honest — neither asks you to fake enthusiasm.
Use honest distance as your starting point.
"Honestly, it's not something I [verb] much, mostly because [real reason]."
Borrow opinion through a person you know.
"It's not really my world, but my [person] is properly into it — they [specific behaviour]."
Two more ways out. These work best when you have some opinion but not a strong one — when the honest answer is "it depends."
"Depends on..." but with specifics.
"It depends entirely on [specific context]. In [context A] I [reaction], whereas in [context B] I [opposite reaction]."
Shift from object to your changing relationship with it.
"I used to [past relationship], but [recent change] — now I [current state]."
The most powerful strategy of the five, and the one most candidates never think to use. When you don't have a personal opinion, talk about your culture's opinion. Vietnamese culture is genuinely interesting to Western examiners — and your perspective on it is something only you can offer.
Reframe through Vietnamese context.
"In Vietnam, [topic] is actually quite [adjective] — [cultural detail]. Personally I [your relationship], but I think the cultural context matters."
Four boring topic questions below. For each one, pick the rescue strategy you'd use. There's no single right answer — any of the five can work — but each question has a strategy that's particularly well-suited to it. See if your instincts match.
You will get a boring topic in your test. It's almost guaranteed. The candidates who score Band 7 aren't the ones who got lucky with interesting topics — they're the ones who had a strategy ready before they walked in. Pick two of these five strategies and rehearse them on five different boring topics this week. By test day, they'll be reflexes.
An opinion answer broken into four sentences. Your job: tag each one as P (Position), T (Type), R (Reason), or N (Nuance). All four must be answered correctly to unlock the next exercise.
Three opinion answers below. Each one uses a different rescue strategy from Stage 7. Identify which one. Click the strategy number under each answer.
"I used to listen all the time growing up — my dad always had it on during breakfast. But since podcasts took over, I almost never tune in anymore. Now I'd rather pick what I'm listening to than have it picked for me."
"In Vietnam, tea is more of a social ritual than a drink — when guests visit, you always offer trà nóng. It's the first move of hospitality. Personally I drink it daily, but what I genuinely appreciate is the cultural function more than the taste itself."
"It's never been my world, but my uncle is properly obsessed — he disappears for entire weekends to a lake outside Cần Thơ. I've gone with him twice and honestly spent most of the time on my phone, but I can see why he finds it so peaceful."
Three Band 5 opinion answers below. Your job: rewrite each one to hit Band 7. Use at least one Band 7 phrase, one comparative or contrast connector, and aim for 20–30 words. Reveal a model answer when you want to compare.
Five opinion questions across the four patterns from Stage 2 (Direct, Comparative, Frequency, Evaluative). Type a full P·T·R·N or hedge-then-disagree answer for each. Aim for 25–35 words.
Five opinion answers written. Time to make them yours. Read each one aloud three times — slow, normal, exam pace — and tick what you notice.
Five exercises down. Here's how it landed. These results are saved for the lesson and you can return any time.
Your performance across the opinion arena suggests the framework has taken hold. Move into Stage 9 to complete the lesson.
Lesson 1 was the framework. Lesson 2 was the topics. Lesson 3 was the opinions. For each statement below, drag the slider to where you genuinely are right now. No judgement.
Another forty-five minutes invested. The opinion machinery is now in your toolkit. Here's what you actually covered.
Three lessons down. The fourth one is where time enters the picture. Past tense answers. Future plans. Present habits. Lesson 4 is the grammar pillar of Part 1 — and it's where most candidates lose marks they didn't even know were available.
Roughly half of every Part 1 test asks about time — "what did you do last weekend," "are you planning to travel this year," "have you always lived here." Vietnamese speakers struggle here because Vietnamese doesn't conjugate verbs by tense. Lesson 4 bridges that gap.
Why Vietnamese candidates drop "-ed" — and the drill that fixes it for life.
The tense Vietnamese has no equivalent for — "I've lived" vs "I live" vs "I lived."
"Will," "going to," present continuous, present simple — and when to use which.
The single most underused Band 7 move: anchoring every detail to a specific year, month, or moment.
"Three lessons in. Notice how each one has built on the last? Lesson 1 gave you the framework. Lesson 2 gave you topics. Lesson 3 gave you opinions. Lesson 4 gives you time. Every lesson sharpens something the previous one made possible. You're not just learning IELTS — you're learning how English itself works for someone with your specific Vietnamese starting point."
You now own the framework, the topics, and the opinion machinery. Three badges on your shelf. Seventeen lessons still to go — but the spine of Part 1 is in your hands.
Three lessons in, you have the framework, the topics, and the opinion machinery. Lesson 4 tackles the single biggest grammar gap in Part 1 — tense. Vietnamese has no verb conjugation. English forces every verb to declare itself in time. This lesson bridges that gap, properly and for the long term.
"I can tell within thirty seconds whether a candidate's grammar will hold up. The tell is always the same — do they drop their -ed endings, and do they use present perfect at all? Lesson 4 fixes both. Permanently, if you do the drills."
Three lessons of toolkit behind you. A brief recap before we layer time onto opinions.
From Lesson 1. The base structure of every Part 1 answer.
From Lesson 2. Places, People, Routines — plus the 80-phrase vocabulary bank.
From Lesson 3. The opinion-answer spine — Position, Type, Reason, Nuance.
From Lesson 3. The toolkit for boring topics — Admit, Borrow, Pivot, Past-to-present, Cultural reframe.
Same 9-stage structure, with a focus shift to grammar. The arena at the end will feel more like a tense workout than a creative writing exercise — and that's the point.
You are here.
Why tense is hard for Vietnamese speakers — and what the examiner is specifically listening for.
The -ed drop drill, irregular verbs, and the past-time-marker trap.
The tense Vietnamese has no equivalent for. Timeline visualizer plus the since-vs-for rule.
"Will," "going to," present continuous, and present simple — and when to use which.
80+ time phrases organised by tense, all clickable for audio.
The past-to-present comparison — Part 1's signature grammar showcase.
Six tense-specific exercises.
Self-assessment, badge, and Lesson 5 preview.
Before we drill any specific tense, you need to understand why tenses are hard for you in the first place. It's not a personal failing — it's a deep structural mismatch between Vietnamese and English. Once you see it, the drills make sense.
The verb đi never changes. Only the time word at the start shifts.
Three completely different verb forms. The time word is almost a courtesy — the verb already tells you when.
Examiners trained on Vietnamese candidates listen for four specific things. Hit them deliberately and your grammar band lifts within the first answer.
"I worked" not "I work." The most damaging Vietnamese drop. Worth a tense drill on its own.
"I've lived here for ten years." If this tense never shows up, the candidate is capped around Band 6.
"I'm going to," "I'll probably," "I'm thinking of" — not just "I will."
"Last March," "since 2018," "for about three years" — anchoring details in time signals control of the system.
You don't need to master twelve English tenses for IELTS Part 1. Six is enough. Here they are, mapped to when you'd actually use each one in a real Speaking answer.
Three specific mistakes that drop a candidate's grammar band by half a point. All three are fixable. None of them require a tense you haven't already met.
"Last weekend I watch a movie." The time word makes the meaning clear, so the candidate's brain skips the verb ending. The examiner registers it as a grammar slip every single time.
"I live here ten years" instead of "I've lived here for ten years." Vietnamese has no present perfect equivalent, so the candidate's brain reaches for present simple every time. This is the single biggest reason candidates plateau at Band 6.
"Tomorrow I will go to the gym. Next year I will study abroad. After that I will buy a house." Three "will"s in a row. Examiners hear monotony where they should hear range.
Two questions before we drill into past tense in Stage 3.
Of all the grammar slips Vietnamese candidates make, dropping the final -ed is the costliest. The reason is brutal — it doesn't just sound wrong, it actively confuses the tense. The examiner hears "I work yesterday" and has to mentally reconstruct what you meant. Each pause costs you fluency marks.
Three missing -eds. The time word "last weekend" carries the whole burden — and the examiner mentally corrects three verbs while you keep talking. Each correction costs.
Three landed -eds. Same sentence, same content, but the tense system holds. The examiner stops correcting and starts listening.
"Worked," "visited," and "loved" all end in -ed in writing — but they sound completely different. There are three -ed sounds in English. Once you can hear them apart, you can produce them. Click each example to hear it.
After p, k, f, s, sh, ch, x — the -ed sounds like "t"
After b, g, v, z, m, n, l, r, vowels — the -ed sounds like "d"
After t or d only — the -ed adds a whole new syllable
English has over 200 irregular verbs. You don't need all of them for IELTS Part 1. You need the twelve that show up in roughly every personal answer. Master these and you've handled 90% of past-tense moments in your test.
Vietnamese speakers often default to "in the past" or "before" when talking about past events. Both are weak. Examiners reward specific past time markers — they signal you have control over the timeline, not just the tense.
Four common Part 1 past-tense questions with Band 7+ model answers. Notice the -ed endings landing cleanly, the irregular verbs in place, and the specific time markers anchoring each story.
"It was a pretty slow weekend, honestly. On Saturday I met up with a couple of friends in District 2 for brunch — we ended up spending the whole afternoon there. Sunday I visited my grandmother in Bình Thạnh, and we cooked bún bò Huế together. Nothing exciting, but exactly what I needed."
"Probably my trip to Hội An back in 2022. I went with two friends just after Tết — we stayed in a small homestay near the old town. The thing I remember most is one evening when we were eating cao lầu at this tiny family restaurant, and it started raining heavily, and the owner just brought us tea and let us sit there for hours. It was one of those unplanned moments that made the whole trip."
"Honestly, mixed feelings. I liked the social side — my classmates and I spent almost all our free time together — but academically I found it pretty stifling, especially in high school. We memorised a lot more than we questioned. Back when I was seventeen I wanted to escape it desperately. Looking back though, those friendships have stuck."
"A couple of weeks ago, actually. My younger sister came back from studying abroad in Australia, and the whole family gathered at our parents' place for dinner. My grandmother cooked for about twelve of us and we stayed at the table for hours just talking. I realised halfway through that I hadn't felt that present in a long time."
Four sentences, each with one verb missing its past-tense ending. Click the word that should have been past-tense. The correct -ed will appear when you get it right.
"Last weekend I went to Đà Lạt and stayed at a small hotel, then I visit my old school."
"I work at a coffee shop in District 1 for two years before I changed careers and started studying design."
"My mother invited us over and cook a huge meal — we stayed until midnight."
"I graduated from university in 2020, moved to Saigon, and started my first job, which I finish just a few months ago."
Notice anything? The dropped verbs almost always come after another past verb that already has -ed. Your brain marks the tense once and then relaxes. The fix is mechanical: every verb in a past-tense answer needs its own marking, regardless of how many other past verbs are around it.
Present perfect is the single most diagnostic tense in IELTS Speaking. It's the tense that bridges past and present — and Vietnamese has no equivalent. Most Vietnamese candidates avoid it entirely, defaulting to present simple. That single avoidance caps their grammar band at Band 6, no matter how good the rest of their answer is.
In Vietnamese, you'd say "tôi sống ở đây mười năm rồi" — "I live here ten years already." The word rồi ("already") signals the continuity. Vietnamese candidates often translate this directly: "I live here ten years" — which sounds wrong in English. The English version requires a tense Vietnamese simply doesn't have: "I've lived here for ten years."
All three signal duration but the verb stays in present simple — which would be correct for habits, not for spans bridging past to now.
The verb itself carries the duration meaning. The examiner registers the present perfect within the first answer — and your grammar band lifts immediately.
Present perfect is best understood visually. Click each scenario below to see where it lives on the timeline. Notice how the actions in green all bridge from a past moment to now — that's the signature.
Present perfect uses one of two duration words — and Vietnamese speakers often mix them up. The rule is simple and worth learning by heart, because once it's automatic, half the present-perfect problem disappears.
Present perfect has three core uses in Part 1 Speaking. If you can recognise which use you need, you stop second-guessing the structure. These three cover roughly every Part 1 situation you'll encounter.
Started in the past, still true now.
"I've + [verb-ed/past participle] + for/since + [duration]"
Have you ever / I've never / I've X times.
"I've + [verb-ed/past participle] + [ever/never/X times/once/twice]"
Just happened, still matters now.
"I've + [just/already/recently/lately] + [verb-ed/past participle]"
Try to hit one present perfect in every long answer. Not every sentence — once is enough. The examiner registers it within the first 15 seconds and your grammar band shifts upward. The candidate who uses zero present perfects in five minutes of talking is the one stuck at Band 6.
Three common Part 1 questions where present perfect is the natural fit. Notice how it stacks with other tenses — past simple for specific moments, present simple for habits, present perfect for the bridge.
"I've been living in District 2 for about four years now — I moved there just after I started my current job back in 2021. Before that I'd spent most of my twenties bouncing between District 1 and District 3, so settling somewhere quieter has been a bit of a relief, honestly."
"Yes, a couple of times — I've been to Thailand and Singapore, both for short trips. The Thailand one was a family holiday back when I was seventeen, and Singapore was a work conference last year. Honestly though, I've never been to Europe, and that's the one I'd love to do next."
"Honestly, I've been getting into long walks lately — there's a route along the canal in District 3 that I've been doing two or three times a week since the start of the year. It started as just exercise but it's become the way I think through things — phones off, no podcasts, just walking. I've probably saved a fortune on therapy."
Four blanks to fill. Each one tests a piece of the present-perfect machinery — the auxiliary, the past participle, the duration word, and the time anchor. Get all four right and you've built a textbook-clean Band 7 sentence.
"I ___ ___ at this company ___ ___."
"I have worked at this company for three years." Present perfect + "for" + a duration. The exact structure examiners listen for in the first 30 seconds of any "how long have you" answer.
Vietnamese candidates almost always default to "will" for anything in the future. "I will go," "I will study," "I will travel." Three "will"s in twenty seconds. The examiner registers it as a single-tense answer — and your range mark drops accordingly. English has at least four different future forms, and each one means something subtly different.
Three uses of "will" in one answer. The candidate is reaching for the only future form they trust. The examiner hears monotony where they should hear range.
Three different futures, three different meanings. Present continuous for the firm plan, going to for the intention, will for the prediction. Same content, three Band 7 grammar moves stacked.
Each future form does a specific job. The trick is recognising what job the situation needs, then reaching for the matching form. Once you internalise this, picking the right future feels automatic.
Six scenarios below. For each one, pick the future form that fits best. Watch for context clues — words like "booked," "I think," "the train at six," and "I've decided" each signal a different form.
Native English speakers almost never say "I will" with full certainty. Real conversation is full of softening — "probably," "I'm planning to," "I might." Hedging your future forms makes your speech sound authentic and shows the examiner you can express degrees of certainty.
Three common future-themed Part 1 questions. Notice how the strongest answers move between multiple future forms within a single response — the move that signals Band 7+ range to any examiner.
"Saturday I'm having brunch with a friend in District 2 — that's already arranged, around eleven. After that I'm going to try to get some studying done at a café nearby, although honestly I'll probably get distracted within an hour. Sunday is more open — I might just spend the day reading at home."
"Honestly, I'm hoping to have moved into a more senior role by then — possibly heading a small product team. I'm thinking of doing an MBA at some point, although I haven't fully committed yet. And ideally I'd love to have travelled to at least one new country every year between now and then."
"Yes — I'm flying to Hội An at the end of next month, that one's booked. After that I'm going to spend Tết with my family up in Hà Nội, which is non-negotiable. And then there's a half-formed plan for a trip to Japan in the spring — but who knows, that one might get pushed depending on work."
Aim for at least two different future forms in any answer about future plans. The single biggest signal that you control the tense system is moving between forms based on the actual nature of each plan. A booked trip gets present continuous. An intention gets "going to." A prediction gets "will." A possibility gets "might." Mixing them isn't showing off — it's accuracy.
Every tense in Lesson 4 comes with its own family of time markers — the words that tell the examiner exactly when something happened. "Last March," "since 2018," "at the moment," "in a few weeks." Without these markers, even correct tenses sound vague. With them, your answers carry weight.
"Last March" beats "before" every time. Examiners reward the candidate who places events on a real calendar, not in a vague past.
"Since" needs present perfect. "Yesterday" forces past simple. The marker and the verb form must agree — get the marker right and the tense often follows.
Listen to any native speaker for ten seconds — they're constantly time-stamping. "The other day," "around then," "lately." This is what naturalness sounds like.
Stage 3 covered 18 past markers. Stage 4 covered "since" and "for." This stage adds 84 more across all four time families — your full time-marking toolkit.
Phrases that anchor an action in current time — used with present simple ("I work here") or present continuous ("I'm working on a project at the moment"). Twenty-one phrases organised by routine, right-now, and recent.
The phrases that signal present perfect territory — the tense that bridges past and now. These are the marker words examiners listen for specifically, because using them correctly means you're operating in the Vietnamese-doesn't-have-this zone of English grammar.
You already met 18 past markers in Stage 3.4 — these are the additional 21 you'll reach for in real Part 1 answers. Past markers are the most common time markers Vietnamese candidates use, so refining beyond "yesterday" is what separates Band 6 from Band 7.
Phrases that anchor an action in coming time — used with whichever future form fits the situation. Twenty-one phrases across "soon," "scheduled future," and "distant future."
Don't try to memorise all 84 phrases. Pick one phrase from each family — one present, one perfect, one past, one future — and learn those four properly. By the end of the week, you'll have four reliable markers across all four tenses. That's enough to land any Part 1 time question. The other 80 are reference material for when you want to expand your range later.
"How has your hometown changed since you were young?" "Has technology changed how people communicate?" "Do you think children's lives have changed compared to your parents' generation?" These are the "what's changed" questions — and they appear in roughly every other Part 1 test. They are the single question type designed to test whether you can move between past, present perfect, and present in one cohesive answer.
One question forces three tenses to show up. The candidate has to use past simple to describe what was, present perfect to describe what has shifted, and present simple to describe what is. There's no way to dodge the grammar. Either you can move between tenses or you can't — and the examiner gets the answer within thirty seconds.
Every "what's changed" answer follows the same three-move shape. Master the shape and the tenses fall into place automatically — because each move requires a specific tense to make sense.
Set the past scene.
Describe the transformation.
Land in the present.
"Growing up, my hometown was a sleepy market town — everyone knew each other and most shops shut by eight. Since then, it's grown enormously and a whole new generation has moved in. These days it feels more like a small city than a town."
Once you have the Then-Shift-Now shape, you need the right verbs to describe the transformation. "Changed" gets you to Band 6. The verbs below get you to Band 7. Twenty-one phrases organised by direction of change.
An "what's changed" question below. Build your answer by picking one phrase for each of the three moves. Watch the answer assemble in real time — and notice how the three tenses lock together when you get the picks right.
[ Then — pick the past simple option ] [ Shift — pick the present perfect option ] [ Now — pick the present simple option ]
Past simple → present perfect → present simple. The exact grammar showcase examiners listen for in any "what's changed" question. This is what Band 7+ sounds like on tense.
Three of the most common "what's changed" questions from real Part 1 tests. Each model answer hits all three tenses in the right places. Click each to reveal.
"Growing up in Đà Nẵng, the city was much smaller and quieter — there were maybe five tall buildings and you could ride a motorbike from one end to the other in fifteen minutes. Since then, tourism has absolutely transformed the place — new bridges, beach resorts, an entire skyline that has appeared in the last decade. These days, it feels like a completely different city — more cosmopolitan, but I miss the quieter version, honestly."
"My grandparents' generation ate almost entirely from local markets — whatever was in season, prepared at home. Things have shifted a lot since then — international food has become everyday, and apps like QuickBite have changed how people eat completely. Now most of my friends order in three or four times a week, although traditional cooking is having a real comeback, which I'm glad about."
"For my parents' generation, work was pretty stable — you got a job after university and stayed there for decades. That model has all but disappeared for my age group — most of my friends have switched jobs three or four times by thirty, and remote work has become standard. These days, stability looks different — it's less about one company and more about your own skill base."
Any question with the words changed, transformed, different, evolved, used to be is a Then-Shift-Now question in disguise. The candidate who hears those signals reaches for three tenses on autopilot. The candidate who doesn't stays in one tense and caps their grammar band. Train your ear to spot the signal.
Four sentences, each with one wrong verb form. Click the word that breaks the tense. Each pick is graded — the correct verb form appears once you find the error.
"I grew up in Hà Nội, but I moved to Saigon five years ago and I live here ever since."
"Last weekend my family and I had a get-together and we go to my grandmother's place where she cooked for everyone."
"I study English since I was in primary school, although I only started taking it seriously last year."
"I work in District 1 and love the energy there. Next Friday I have a meeting in Đà Nẵng, so I will fly there in the morning — already booked."
Four scenarios. For each one, pick which tense fits the situation best. Watch for context clues — "last year," "since 2018," "at the moment," "next Friday" each signal a different tense.
Three Band 5 answers below. Each one has tense problems — dropped -ed, missing present perfect, single future form. Your job: rewrite each one using the tense moves from Lessons 4. Reveal model answers to compare when you're done.
Five questions, each targeting a specific tense move. Build a Band 7 answer for each — aim for 25–35 words. The hints under each question tell you which tense to hit.
Five answers written. Time to speak them. The tense moves are only worth anything if your mouth can produce them under pressure — so the drill below is specifically for tense delivery.
Five tense-specific exercises done. Here's how it landed. The arena scores are saved for the lesson — you can return any time.
Your performance across the tense arena suggests the tense system is settling. Move into Stage 9 to complete the lesson.
Tense is the most technical lesson so far — and the one that most often needs repeat passes. Slide each statement to where you genuinely are. No judgement. This is just for you.
Another forty-five minutes invested in the toughest grammar lesson of the program. Here's what you covered.
Four lessons in. You now have the framework, the topics, opinions, and tense. The fifth lesson is where your answers get longer. Most Vietnamese candidates give answers that are too short — three sentences and then silence. Lesson 5 fixes that. Not by adding fluff. By teaching you the four moves that naturally extend an answer to its proper length.
Part 1 answers should land at 25-40 seconds. Most Vietnamese candidates run dry at 8-12 seconds. The gap isn't vocabulary — it's not knowing how to keep talking after the direct answer. Lesson 5 gives you four reliable extension moves: the example, the contrast, the consequence, and the personal anecdote.
"For instance...", "Take last Tuesday...", "A perfect example is..." — extending by naming something specific.
"Whereas my brother...", "Unlike most of my friends...", "In Vietnam it's different from..." — extending by comparison.
"That's actually why I...", "Because of that, I now...", "It's the reason I ended up..." — extending by causation.
"There was a moment when...", "I remember once...", "Just the other week..." — extending by story.
"Four lessons in, and you've crossed the hardest grammar threshold most Vietnamese candidates ever face. The -ed drill, the present perfect bridge, the four futures — these are the moves that separate Band 6 from Band 7. The remaining lessons polish what you already have. Honestly, you've done the hard part."
Four badges on your shelf. You now hold the tense system that most Vietnamese candidates never properly close. Sixteen lessons still to go, but the grammar pillar of Part 1 is built.
Four lessons in. You have the framework, the topics, opinions, and tense. Your answers are grammatically clean and full of Band 7 phrases. So why do most Vietnamese candidates still cap at Band 6 in mock tests? Because their answers are too short. Lesson 5 is the length lesson — and it's the most under-taught skill in IELTS Speaking.
"I can give you the most beautiful eight-second answer in the world — perfect grammar, Band 8 vocabulary, native-sounding intonation — and you'll still cap at Band 6.5. Length isn't padding. Length is what gives the examiner enough surface area to measure your skills. Without length, none of your other work shows up."
Four lessons of toolkit. A quick reminder before we layer length onto everything.
From Lesson 1. The base structure of every Part 1 answer.
From Lesson 3. The opinion-answer spine.
From Lesson 4. Past simple, present perfect, four future forms, and Then-Shift-Now.
From Lesson 3. Tools for any boring topic you face.
Same 9-stage shape. New focus: the four reliable moves that turn a short answer into a confident long one.
You are here.
Why your answers are running out at 10 seconds — and what the ideal length actually is.
Example, Contrast, Consequence, Anecdote — the E · C · C · A formula.
The simplest extension. Interactive builder + the "for instance" toolkit.
Three more extension moves, each with its own mini-builder.
80+ phrases organised by which extension move they serve.
How to combine two or three extension moves in one answer — the Band 7.5+ move.
Six extension-specific exercises.
Self-assessment, stats, badge, and Lesson 6 preview.
An IELTS examiner gives you four marks — Fluency, Lexical Resource, Grammar, and Pronunciation. Every one of these is measured by how much you say. If your answer lasts ten seconds, the examiner has ten seconds of material to score on. If it lasts thirty seconds, they have three times the evidence — and three times the room to lift your band.
A candidate who finishes in eight seconds doesn't just give the examiner less material — they signal three other problems: not confident enough to keep going, not fluent enough to talk at length, and not sure what else to say. None of those things might be true about you. But that's what the eight-second answer communicates.
One question, two answers. Same content, same opinion. The first one runs dry too early. The second one lands in the Band 7 zone — not because the candidate knows more, but because they had four extension moves ready.
The answer is grammatically correct. The vocabulary is fine. But the examiner has barely heard anything to score on. The next question arrives immediately — and the band lifts that could have happened didn't.
Same opening sentence as the short version — but then three extension moves get added: Example (the specific book), Contrast (vs scrolling), and Consequence (improved sleep). The examiner now has three times the surface area to hear your range.
Vietnamese candidates run out of things to say in roughly four distinct ways. Each one feels like a unique problem in the moment, but they're really the same problem dressed up differently — the candidate doesn't have a default "what to say next" move ready. The next stage gives you exactly that.
The candidate answers the question literally and stops. "Yes I do." "No I don't." "It's fine." The wall hits in three words and there's no plan for what comes after.
The candidate adds one reason and then stops, thinking the answer is complete. "Yes, because it's relaxing." The reason felt like the second sentence the answer needed — but it was actually the second of five.
Out of nothing else to say, the candidate paraphrases the question back at the examiner. "Reading is important. Reading helps people." This is filler. The examiner registers it as filler.
The candidate buys time with stalling phrases — "I think... maybe... it depends... actually..." Two or three of these in a row, then a real sentence, then more stalling. The examiner counts every "I think" as silence.
If you watched a hundred Band 7 Part 1 answers back to back, you'd notice they all have the same rough shape. Not the same content — the same architecture. Four moves, in roughly this order, with subtle variation in how many show up.
Two questions before we move to the four moves in Stage 3.
Four extension moves. Memorable initials. You don't need to use all four in every answer — pick one, sometimes two, and develop. The point of having all four ready is that at least one of them always fits, no matter what topic the examiner throws at you.
Give a specific instance. The easiest extension — almost any answer can take one.
Compare to something different — a person, a time, a situation.
Explain what your situation has led to — what changed, what you do now because of it.
Tell a specific micro-story from your life. The most powerful — and the rarest in Vietnamese candidates' answers.
A demonstration. The candidate uses all four extension moves in a single answer — which you'd never actually do in your test, but it shows the moves clearly. Real test answers use one or two. Pick one for now and master it.
"Yes, I try to — I run three or four times a week, usually in the morning before work."
"For instance, I went out at six this morning along the canal in District 3 — about forty minutes."
"Unlike my brother, who's obsessed with the gym, I really only enjoy outdoor exercise — gyms feel suffocating to me."
"That's actually why I've shifted to morning runs — my energy through the day is noticeably better since I started."
"There was one morning during Tết when I ran along an empty street at sunrise — felt like I had the whole city to myself for ten minutes."
Not every extension move fits every question equally well. Some questions practically beg for an Example. Others naturally want a Contrast or an Anecdote. Here's the rough pairing — once you internalise this, picking the right move becomes automatic.
Three specific mistakes that turn good extension moves into score-lowering ones. All three are common. None are obvious from the inside — which is exactly why I'm flagging them before you start using the moves.
The candidate adds words but no new content. "I really like reading, and reading is really good, and I read a lot of books, and books are great." The answer got longer but the surface area for scoring didn't.
The candidate invents a story to extend, and the examiner can hear the fabrication immediately — the details get fuzzy, the voice tightens, the timing goes off. Forced anecdotes cost more than they buy.
The candidate, eager to show range, tries to land all four moves in one answer. The result feels like a checklist being read aloud. Examiners notice the artificiality and discount the answer.
A note on culture, before we drill into the moves. Vietnamese conversation styles are partly responsible for the short-answer problem — and naming this makes the fix feel less like personality change and more like switching modes.
Conversational politeness rewards brevity. Going on too long can seem self-centered. Answering exactly what was asked is considered respectful. Extending unprompted can feel boastful.
Brevity reads as evasive or unable. Examiners expect candidates to elaborate by default, not because they're showing off but because the test format demands it. Length is required, not optional.
Two questions before we go deep on the Example move in Stage 4.
Of the four E · C · C · A moves, Example is the one to master first. It works on roughly 80% of Part 1 questions. It requires no clever grammar. It demands no invented stories. It just asks you to swap one general statement for one specific instance — and that swap alone can add 10-15 seconds of natural speech to your answer.
Examples generate language automatically. As soon as you pick a specific thing — a real place, a real time, a real food — your brain produces detail without effort. You're not thinking about what to say; you're describing something that exists. Specificity is its own engine.
The candidate is talking about reading, not from reading. The examiner hears no real content — just statements that could apply to anyone, anywhere.
Same answer, with one specific example bolted on. Notice how much detail came for free once the book got named — the title, the author, the topic, the pace. None of that required clever vocabulary.
Not every example is the same kind. Depending on the question, you'll reach for one of four specific types. The good news: you only need one of them at a time. The better news: every Part 1 question fits at least one of these types comfortably.
A specific book, song, place, restaurant, app, person. Naming something is the simplest possible upgrade from "things I like" to "this specific thing."
Something that happened to you this week, this month, this year. Anchors a general claim in a real recent timeframe. Especially powerful with past simple verbs.
A specific recurring habit — what you typically do, how often, at what time, with whom. Makes a vague "I do this" land as a real practice.
The opposite move — when the question is too narrow, broaden it by listing one or two specific kinds. Often pairs with "especially" or "particularly."
The phrase that introduces your example matters. Native speakers rarely say "for example" — they use one of about a dozen softer, more natural variants. Each one signals a slightly different shade. Pick three or four and rotate them.
Pick a question. Then build the example — choose the type, pick the connecting phrase, and fill in the specifics. Watch the answer assemble in real time. The full sentence reads naturally when all four pieces fit together.
"Yes, honestly I love reading. [connecting phrase] [specific detail]"
Base answer + connector + specific detail. The whole thing takes about 4 seconds longer than the base answer alone — and the examiner has 4× the material to score on. Pure surface area.
Three real Part 1 questions with model answers that use the Example move well. Each one uses a different example type — see how the move flexes to fit the question.
"Yes — probably one or two a week, mostly at home. The most recent one was a Korean thriller called 'The Wailing' — I'd been meaning to watch it for ages and finally got around to it last Sunday. Honestly I lean more towards East Asian cinema these days; Hollywood feels predictable by comparison."
"I lean more towards home, actually. On any given weekday I'll cook something simple — bún or a quick stir-fry — and we tend to eat around seven, with whoever's home that night. Restaurants are more of a Friday-night thing for me, when I want a break from washing up."
"Honestly, it depends what kind of thing. Just the other day I forgot a colleague's name halfway through introducing him — properly embarrassing. But ask me what I ate for dinner three weeks ago and I could probably tell you. So I'd say my memory works in patches, not evenly."
One real example beats five generic statements. If you can't think of an example, that's a signal that you're being too abstract — narrow the topic until you can. "What about food?" is hard to answer. "What did I have for breakfast?" is easy. The trick is mentally narrowing big questions down to small ones, then giving the small one's answer as your example.
If Example is the safest move, Contrast, Consequence, and Anecdote are the lifting moves — the ones that take your answer from Band 7 to Band 7.5+. Each one does something Example alone can't. Each one is its own small craft. The next four screens build each of them, one at a time.
Compares your situation to something different — a different person, a different era, a different version of yourself.
Connects your reason to a real outcome — what your habit has produced, what changed because of it.
Tells a small story — a specific moment with a beginning, middle, and small landing.
Contrast extends your answer by pointing at something different. The classic pattern: "I do X. Unlike Y, who does Z." Three flavours — contrast with people, contrast with times, contrast with situations. Pick one, deploy, get 10-12 extra seconds of natural speech.
Pick a flavour, then pick a phrase to slot in. See the full sentence assemble.
"Yes — I run three or four times a week. [contrast]"
Consequence extends your answer by completing a cause-and-effect chain. You did X. As a result, Y. Three flavours — what you changed, what improved, what you now do differently. The Band 7 signal here is the connector — "that's why," "as a result," "because of that" all force the examiner to hear a logical sequence.
Pick a flavour, then pick the consequence to slot in.
"Honestly, much better than I used to. [consequence]"
Anecdote extends your answer by telling a tiny story. Not a movie plot — just a moment, with a real time, place, and small landing. Three flavours — the funny moment, the unexpected moment, the meaningful moment. This is the most powerful extension move when it's real. Forced anecdotes cost more than they buy, so only deploy when you actually have a moment ready.
Pick a flavour, then pick the story opener.
"I try to, though it's harder now than it used to be. [anecdote]"
One question. Three answers. Each one uses a different extension move on the same base. Notice how the move you pick shifts the entire feel of the answer — Contrast feels analytical, Consequence feels deliberate, Anecdote feels human.
"Yes, more than I used to — I try to get a walk in every evening, even if it's just twenty minutes."
"Unlike most of my colleagues, who spend their evenings on screens, I find that even a short walk along the canal puts me in a much better mood — it's the closest thing I have to a daily ritual."
"That's actually why my sleep has improved so much in the last year — those evening walks force my brain to slow down before I get home, and the rest of the night just flows better."
"There was one evening last month when I got caught in a sudden rainstorm halfway around the loop — I ended up sheltering under a kiosk with a stranger for twenty minutes, and we just talked. Small thing, but I remember it weekly."
If you want the answer to sound thoughtful → Contrast. If you want it to sound logical → Consequence. If you want it to sound human → Anecdote. None is universally better. The one you pick should match the tone the question seems to invite — and on most Part 1 days, Example + one of these three is plenty.
Two diagnostic questions. Pick the extension move that fits each one best.
You have the four moves. Now you need the vocabulary that signals each one. The bank below organises 84 extension phrases — 21 for each move — across three strength bands. Every phrase is clickable for audio. Pick three from each move and rotate.
Twelve phrases total, three per move. That's enough to feel natural without sounding rehearsed.
Every pill plays at slower-than-natural speed so you can hear the stress and rhythm clearly.
One Direct phrase, one Natural phrase, one Sophisticated phrase. Variety signals range to the examiner.
Using the same opener twice in one test is fine. Three times signals you're operating from a small stock — and examiners notice.
You already met 18 "for instance" phrases in Stage 4. Here are three more bands — naming things specifically, broadening with categories, and pointing at recent moments. Twenty-one in total for this move alone.
Contrast phrases come in three shapes — comparing to people, comparing to past selves, and comparing to alternative situations. The strongest of the three is usually the past-self contrast, because it doubles as a tense showcase.
Consequence phrases connect cause and effect. Three flavours — explaining a change you made, describing an improvement you noticed, or naming a new habit that resulted. All three signal logical reasoning to the examiner.
Anecdote phrases open a tiny story. Three flavours — opening with a specific time, opening with a memory, or opening with the unexpected nature of the moment. Each kind invites a different mood from the listener.
Eighty-four phrases is a lot. Don't try to learn them all. Pick three per move, write them out, and use them in every practice session for the next two weeks. By the end of week two they'll feel automatic — and that's all you need for a real test. The rest of the bank is reference material for when you want to expand your range further.
A single extension move gets you to Band 7. Two well-chosen moves, stacked together, push you into Band 7.5+ territory. But stacking has rules — pick the wrong two and the answer sounds rehearsed. Pick the right two and the examiner hears genuine range. This stage teaches the difference.
A candidate at Band 7 reaches reliably for one extension per answer. A candidate at Band 7.5+ instinctively layers two — usually Example + one other move — because they've stopped thinking about it. The moves flow naturally because the candidate practised them as a sequence, not as separate options.
Three or more reads as performance. The examiner notices you running through a list. Stop at two.
Specific instances ground your answer in reality. Once you have something specific, the second move (Contrast, Consequence, or Anecdote) can connect to it. Order matters.
If the second move doesn't connect to the first one, don't force it. A clean one-move answer beats a forced two-move one every time.
The four E·C·C·A moves give you six possible pairs. Three of them are natural stacks — they flow together as if they were one move. Three of them are forced — they sound stitched. Here's the map.
The most reliable pair. You name something specific, then explain what it led to. Cause-and-effect tied to a real instance — easiest to deliver naturally.
You name something specific, then contrast with something different. Works especially well for preference questions — "I like X. Unlike Y..."
You name a general thing, then tell a specific story about it. Pairs feel intimate — works well for the more personal Part 1 questions.
Both are abstract moves. Without a specific Example grounding them, they float — the answer feels theoretical rather than personal.
Tense shifts conflict — Contrast usually wants present or past habitual, Anecdote wants past simple. Hard to land smoothly without explicit signposting.
Order is hard. If Consequence comes first, the Anecdote feels redundant. If Anecdote comes first, the Consequence sounds tacked on. Either way, the join shows.
Five questions. For each one, pick which two-move stack fits best. Watch for what the question is really asking — questions invite different combinations naturally.
Three Part 1 questions, each answered with a different two-move stack. Notice how the second move connects directly to the specific thing named in the Example — that's the join that makes stacks sound natural rather than forced.
"Yes, more than I used to. Right now I'm halfway through 'The Mountains Sing' by Nguyễn Phan Quế Mai — about thirty pages a night for the last two weeks. That's actually why my sleep has improved so much — reading slows my brain down in a way nothing else does."
"Mostly Vietnamese indie at the moment — especially Ngọt and Vũ — there's something about their lyrics that lands more for me than anything in English. Unlike most of my friends, who are still really into K-pop, I find myself drawn back to Vietnamese music more and more."
"My go-to is a small place called The Workshop on Ngô Đức Kế — a third-wave coffee spot above a leather shop, you wouldn't find it unless someone told you about it. Just last week I went with a friend who was visiting from Đà Nẵng, and we ended up staying nearly four hours — they kept refilling our water and the staff never rushed us once."
Before you move into the Practice Arena, here's the rule for stacking in your real test — and one final check to lock the principle in.
In a real Part 1 test, aim for one move per answer 80% of the time, two moves per answer 20% of the time. Save your stacks for the questions that genuinely invite them — typically the third or fourth question on a topic, where you've built some context and the second move can connect naturally. Don't stack the opening question of a topic, and don't stack two answers in a row. The strategic deployment of stacks is what makes them feel earned rather than performed.
Five Part 1 answers below. For each one, identify which extension move (E, C-contrast, C-consequence, or A) was used after the base. Wrong picks show you which move you mistook it for.
Four Band 6 answers that ran dry too early. Read each one, then pick which extension move would best fix it. The right pick gets you to Band 7 length.
Three Band 5 answers with weak or missing extensions. Your job: rewrite each one with a proper extension move. Aim for 25-35 words. Reveal the model answer when you're done.
Five questions. Each one specifies which extension move to use. Build a Band 7 answer of 25-35 words. The fifth is a stack — Example + your choice of second move.
Extension moves are only real once your mouth produces them under pressure. The drill below trains the rhythm — the small pause before the connector, the slight stress on the named thing, the dropping intonation that signals the answer is landing.
Five extension-specific exercises done. Here's how it landed.
Your performance across the extension arena suggests the E·C·C·A moves are settling. Move into Stage 9 to complete the lesson.
Extension is more art than mechanics — knowing the moves is one thing, deploying them naturally is another. Slide each statement to where you genuinely are. No judgement.
Forty-five minutes invested in the lesson that fixes the single biggest Vietnamese-candidate problem in Part 1 — running dry too early.
Five lessons of Part 1 are done. The framework, the topics, opinions, tense, length — that's the full Part 1 toolkit. Now we move to Part 2: the two-minute long turn, where the examiner hands you a card and you talk solo for two full minutes. Different beast entirely.
Part 2 is where most Vietnamese candidates either thrive or collapse. The two-minute long turn rewards structure and punishes hesitation. Lesson 6 introduces the cue card framework — the prep ritual, the four-section structure, and the pacing markers that get you to two full minutes without padding.
What to write down on the examiner's paper, in what order, with what shorthand — so the two minutes feel rehearsed rather than improvised.
Introduce · Develop · Highlight · Reflect. Roughly 30 seconds each. The architecture that turns two minutes from terrifying into manageable.
Person, Place, Object, Event, Experience, Skill, Plan, Idea. Every Part 2 prompt fits one of these — and each has its own micro-template.
"Initially..." "Then..." "What stood out was..." "Looking back..." — connectors that signal you're moving through your prep, not running out.
"You've now completed all five Part 1 lessons. The candidate who's worked properly through these five has a Part 1 toolkit most native speakers couldn't articulate. You won't run dry, you won't drop your -eds, you won't hedge into nothing. That's a real shift. Take a day off, then come back for the long turn."
Five badges. Part 1 in full. You now hold the complete Part 1 toolkit — framework, topics, opinions, tense, and length. The opening eleven minutes of your IELTS Speaking test should now feel like home ground.
Five lessons of Part 1 are behind you. You can now answer almost any short question well. Now the test shifts. The examiner stops asking questions and hands you a card. You get one minute to think. Then you talk for two full minutes, alone. No follow-up questions, no rescuing, no examiner nodding to keep you going. This is Part 2, and it requires a different toolkit.
"Part 2 is the section where most Vietnamese candidates either break out into Band 7 or collapse to Band 5.5. The difference between the two isn't English ability — they have roughly the same vocabulary. The difference is whether they have a system for the two minutes. Lesson 6 gives you the system."
Part 1 and Part 2 are different sports played by the same body. The mental shift between them is what catches most candidates out — they bring Part 1 habits into Part 2 and the answers run out at 45 seconds.
Same 9-stage shape as Part 1 lessons. New focus: building the system that turns a 2-minute long turn from terrifying into manageable.
You are here.
What to write down on the prep paper, in what order, with what shorthand.
Introduce · Develop · Highlight · Reflect — the 4-section structure that fills 2 minutes naturally.
Every Part 2 prompt fits one of eight categories. Each has its own micro-template.
80+ Part 2 connector phrases — "Initially..." "Then..." "What stood out was..." — that signal you're moving through your plan.
A 2-minute model answer annotated section by section, showing the prep paper that produced it.
The 4 ways Part 2 collapses — and the specific fix for each.
Six Part 2-specific exercises.
Self-assessment, badge, and Lesson 7 preview (Describing People).
Most Vietnamese candidates waste their prep minute. They stare at the card, panic about content, and start writing words randomly with 20 seconds left. By the time the timer goes off, their paper has three disconnected words on it — and their 2-minute talk has no spine. The prep minute, used well, is the single largest predictor of Part 2 success.
Sixty seconds, split into four windows. Each window has a single job. Do these in order — even if you finish a window early, don't move to the next prematurely. The order is the ritual.
Read the card twice. Decide the specific thing you'll talk about. One concrete subject, not a category. If the card says "describe a memorable journey," you decide: "the train to Đà Lạt with my sister, March 2023." That decision is the entire content of Window 1.
Write four section markers on your paper: I · D · H · R. Under each, write 2-3 words — never full sentences. These are anchors, not content. The four section letters take 5 seconds to write; the 2-3 words per section take ~15 seconds.
Go back through I·D·H·R and add one specific detail to each — a name, a time, a sensory detail, a quote. These details are what stop the answer from sounding generic. "sister" becomes "sister — Linh". "train" becomes "train — 7am, sleeper carriage".
Eye-scan the paper once. Mentally walk through each section. If a section feels thin (fewer than 30 seconds of talk), add one more detail there. If a section feels too packed, mentally pick which detail you'll drop. Then put the pencil down before the examiner does.
Here's a real prep paper for a real cue card, after 60 seconds of using the ritual. Notice how little is on it. Notice the section markers down the left. Notice the bullet shorthand. This is what your paper should look like at the 60-second mark.
Describe a memorable journey you have taken.
You should say:
and explain why it was memorable.
Sixty seconds doesn't leave room for full words. The shorthand below is what serious candidates use — a small toolkit of symbols and abbreviations that lets you fit twice as much meaning onto the page. Practise these and the prep window stops feeling rushed.
Two questions before we move to the I·D·H·R framework in Stage 3.
Four sections. Two minutes. Roughly 30 seconds each. The framework that turns the Part 2 long turn from a wall of empty time into four manageable chunks. Once you internalise these four sections, you stop watching the clock — you watch your structure instead.
The first 30 seconds. Your job: name the subject, plant the basic facts, and orient the listener. Don't rush past it — but don't get stuck here either. Three sentences is plenty.
The second 30 seconds. Your job: build out the context. Walk through what happened, in what order, with what circumstances around it. This is where most candidates either expand naturally or stall — having two or three concrete facts on your prep paper is what saves you.
The third 30 seconds. Your job: zoom in on the single most vivid moment. This is the section that separates Band 6 answers from Band 7. Sensory detail. Specifics. The bit a listener could picture. Don't summarise — paint.
The final 30 seconds. Your job: step back and explain why this matters. Why is it memorable? What did it change about you? This is where most cue cards explicitly ask "and explain why..." — and where you can either land a real insight or limp out with a generic "it was very memorable."
Two questions to lock in the four sections before we move on.
Part 2 cue cards look infinitely varied. They're not. Every prompt — past, present, and future — fits one of eight categories. Once you can recognise the category, you know which I·D·H·R template fits, which content angle to take, and which trap to avoid. Recognition is half the work.
Person, Place, and Object are the easiest to start with — they're all concrete nouns you can picture and describe. The I·D·H·R template flexes slightly for each one. Pick a specific instance, anchor it in your real life, and let the structure do the work.
Event, Experience, and Skill all involve telling a story — they have a temporal arc. The I·D·H·R template feels most natural here because narrative naturally builds toward a moment and then reflects on it. These are also the most common cue cards in the exam, so master these and you're covered for most test days.
Plan and Idea are the trickiest two. They don't have a natural temporal arc — Plan is future-facing, Idea is conceptual. The I·D·H·R template still works but needs adapting. For both, the Highlight section has to be a concrete instance even though the topic is abstract — that's the move that prevents your answer from floating.
No matter which of the eight categories you face, Highlight must be concrete and specific. A real moment with sensory detail. An actual scene that happened or could happen. Even abstract Plan and Idea cards need a concrete Highlight — that's what stops your 2-minute answer from floating in generalities and gives the examiner something to remember.
Six real Part 2 cue cards. For each one, pick the category. Recognition speed is what matters in the test — a 5-second category-spot saves 30 seconds of prep panic later.
Describe a time when you helped a stranger.
Describe a piece of equipment you use frequently.
Describe a place you would like to visit in the future.
Describe a piece of advice that has stayed with you.
Describe a celebration or festival in your country.
Describe something you would like to be better at.
Pacing markers are the connector phrases that signal to the examiner — and to yourself — where you are in your answer. "Initially..." tells the examiner you're starting Develop. "What stood out was..." tells them you're entering Highlight. These small words do enormous work: they hold the structure together for the listener and remind you where you should be heading next.
The examiner hears "looking back..." and knows you're entering Reflect. Without markers, they have to guess what section you're in — and guessing penalises your fluency score.
When you say "initially," your brain commits to Develop content next. The marker forces you to follow your prep paper rather than drift wherever the thought takes you.
"What I remember most vividly..." is three seconds of natural speech that gives your brain time to retrieve the next detail. Markers function as fluent pause-fillers.
A candidate who uses 8-10 different markers across their 2 minutes signals lexical range. One who relies on "and then..." three times signals the opposite.
Twenty phrases for the first 30 seconds. Three sub-uses: opening the answer, naming the subject, and setting up the why-this-one hook. Pick three favourites from across the bands — one direct, one natural, one sophisticated.
Twenty phrases for the second 30 seconds. Three sub-uses: temporal sequencing, context-laying, and detail-stacking. This section needs the strongest connectors because it's the longest stretch of straight content before Highlight.
Twenty phrases for the third 30 seconds — the section where Band 6 splits from Band 7. Three sub-uses: zooming in on the moment, painting sensory detail, and capturing what was said or felt. These are the most important markers in the entire lesson — they're the verbal cues that the examiner is about to hear your best material.
Twenty phrases for the final 30 seconds. Three sub-uses: stepping back to look at the meaning, explaining what changed, and landing the answer cleanly. Most candidates underweight Reflect — these phrases let you fill the section confidently.
Across all four sections, aim for roughly three markers per section — one direct, one natural, one sophisticated. That's twelve markers across two minutes — one every ten seconds — which feels natural without sounding rehearsed. Don't try to memorise all 80. Pick twelve, write them on a card, practise them for a week, and they'll become reflexes.
Five lessons of theory, fifteen screens of framework, eighty pacing markers. Time to see it all assembled into one real answer. Over the next four screens, you'll see: the cue card, the 60-second prep paper that produced the answer, and the full 2-minute answer split into its four I·D·H·R sections — every marker tagged, every move explained.
Describe a memorable journey you have taken.
You should say:
and explain why it was memorable.
This is what was on the prep paper when the timer went off. Twenty-three words total. Eight specific anchor points. Section letters down the left. No full sentences — just enough to keep the structure intact while the brain produces the actual language live.
Below is the entire answer, split into its four sections. Each marker is highlighted in the colour of its section. Each annotation tells you which move is being made. Read it through once normally, then click "Show annotations" to see the architecture.
"The journey I'd like to talk about is a train trip I took from Saigon up to Đà Lạt with my sister Linh, back in March 2023. It's the one that immediately came to mind because it was the first proper trip we'd taken together as adults — both of us moved out of our parents' place years ago, and the chances to actually be together as siblings rather than just family had basically dried up. So when she suggested the train instead of flying, I jumped at it."
"We started off by boarding the train around 7am at Sài Gòn station — a sleeper carriage with about four other people in our compartment, all reading or already half-asleep. The context is that the route takes about seven hours total — slow but properly scenic, climbing up through the mountains the whole way. On top of that, for the first few hours we mostly just talked — catching up on her work, my work, the kind of stuff you never quite say over phone calls. The carriage attendant kept coming around with these tiny paper cups of coffee."
"But the moment I keep coming back to was about six in the morning when we were going through Định Quán. I can still picture the sun just starting to come up, and outside the window there was this thick mist sitting over the pine forests, with the train climbing slowly through it. I remember Linh actually whispered 'oh my god' under her breath — we'd both stopped talking — and the attendant came back with two more coffees without saying anything. Like he'd seen this view a thousand times and knew what it did to people."
"Looking back, what made it memorable wasn't really the scenery, honestly — Vietnam has a lot of beautiful scenery — it was the fact that we'd been technically close as siblings our whole lives but had never actually chosen to spend that kind of time together as adults until that trip. Since then we've made a point of doing at least one trip together a year. So yeah, that's the one that comes to mind. Worth more than the seven-hour ticket price, frankly."
The answer wasn't lyrical, didn't use rare vocabulary, and had at least one stumble. So what put it in Band 7 territory? Six specific moves — listed below. Each one is something you've already met across Lessons 1–6.
11 markers across 2 minutes meant the examiner heard the I·D·H·R bones throughout. They never had to guess what section the candidate was in.
"Mist sitting over the pine forests," "the train climbing slowly through it," "Linh actually whispered 'oh my god' under her breath" — the examiner can picture this moment. That's the band-lifter.
"We'd been technically close as siblings our whole lives but had never chosen to spend time together as adults" — that's a real reflection, not "it was a very memorable journey."
Past simple ("we boarded"), past continuous ("we were going through"), present perfect ("we've made a point of"), and a casual present ("Vietnam has a lot of beautiful scenery") — handled cleanly throughout.
"Honestly," "basically," "kind of stuff," "frankly" — these tiny conversational fillers signal genuine fluency rather than a rehearsed script. Examiners trained to spot the difference instantly hear it.
"Worth more than the seven-hour ticket price, frankly" — a small, dry, personal close. Examiners remember answers that end with personality. They don't remember "and that's the journey I wanted to tell you about."
Look back through the six moves. Every one is a structural or register move, not a vocabulary one. The answer would still have hit Band 7 if you'd swapped "mist" for "fog" or "pines" for "trees." The structure carried it. Most Vietnamese candidates lose Band 7 not because their vocabulary is weak but because they have no structural backbone. Lesson 6 fixes that.
Two diagnostic questions before we move into Stage 7 (common failures).
Watch a thousand Vietnamese candidates do Part 2 and the failures cluster into four distinct patterns. Same root cause behind each — no usable system to lean on — but four different symptoms in the actual test. Knowing which pattern you're prone to is half the fix. Most candidates own one or two of these, not all four.
Talks confidently for 30-45 seconds, then visibly runs out of content. Stares at the prep paper. Says "um" three times. Tries to restart from a different angle. Examiner has to keep the silence with their face.
Talks for the full two minutes but reads as a flat list of facts. No section structure audible. No vivid moment, no reflection — just a chronological recital of what happened. Sounds like reading bullet points aloud.
The opposite problem. Delivers an obviously memorised piece — too smooth, too formal, too "British." Examiner can hear the rehearsal. Vocabulary suddenly jumps two bands above the rest of the test. Confidence drops in the follow-up question.
Misreads the cue card subtly in the prep minute — talks about a journey instead of their most memorable journey, or describes an object instead of explaining why they like it. The answer is technically wrong for the prompt, even though every sentence is fluent.
The most common Vietnamese Part 2 failure. The candidate has roughly 40 seconds of content prepared. They deliver it well. Then the silence comes — and the silence is what costs them the band. The fix isn't more vocabulary. It's more Reflect content during the prep minute.
"I would like to talk about my best friend. Her name is Hương. We have been friends for about ten years since we met in high school. She is a very kind person and she always helps me when I have problems... (pause) ... um, she also likes the same things as me, like... reading books and... watching films... (pause) ... yeah, she is a very good friend to me."
A subtler failure than the stall — and harder to self-diagnose because the candidate does reach the two-minute mark. But the answer reads like an ordered list of facts. The examiner hears a sequence of events with no peak, no perspective, no shape. Band 6 ceiling regardless of vocabulary.
"I want to describe a trip to Đà Nẵng. I went last summer with three of my friends. We flew from Saigon. We arrived in the afternoon. We checked into our hotel. We went to the beach. We had dinner at a seafood restaurant. The next day we went to Marble Mountain. We took many photos. The third day we went to Hội An. We came back to Saigon on the fourth day. It was a good trip."
The IELTS prep industry's gift to examiners: candidates who memorise entire Part 2 templates ahead of time and recite them verbatim. Examiners are trained to catch this — vocabulary that suddenly jumps two bands above the rest of the test, formality that doesn't match Part 1, and an unnatural flow with no real hesitation. Counterintuitively, this fails harder than the 45-Second Stall.
"Today I would like to enlighten you about a truly remarkable individual who has profoundly impacted my existence — my beloved grandmother. She possesses an unparalleled wisdom that has been cultivated through decades of life experience. Her benevolent nature radiates throughout our entire family, fostering an environment of unconditional love and unwavering support..."
The cruellest failure — because the candidate often delivers a fluent, well-structured 2-minute answer, but it's about the wrong thing. They misread the cue card subtly in the prep minute and the misreading propagates through the whole talk. The fix is in Window 1 of prep.
Cue card: "Describe a piece of clothing you wore for a special occasion."
Answer: "I'd like to talk about my favourite jacket. It's a black leather jacket I bought in District 1 about three years ago. I wear it almost every day in cooler weather..."
(Two minutes of detailed jacket description, but the cue card asked about a special-occasion garment specifically.)
Don't try to fix all four. Re-read the four failures and identify which one is yours — the one you can feel happening in mock tests. That's the one to drill. The other three are still useful to know about, but applying one specific fix consistently beats half-fixing four patterns at once. One pattern, one fix, two weeks of practice — and Part 2 stops being the section you dread.
Five real Part 2 answer fragments. For each one, identify which of the four collapse patterns the candidate fell into. Recognition is the first step toward catching the pattern in your own answers.
Five Part 2 sentence fragments. For each, identify which I·D·H·R section it belongs to based on the marker and content. This sharpens your ear for the architecture you'll be deploying in your own answers.
Three cue cards. For each one, write what you'd put on your prep paper in the 60-second window. Two to three shorthand lines per section — no full sentences. The "Reveal model" button shows a working example.
The Highlight section is what splits Band 6 from Band 7. Three cue cards below — write only the Highlight section for each one (about 60-80 words). Force yourself to use a zoom-in marker, one sensory detail, and one human moment. The 30-second section that decides your band.
Pick one of the three cue cards you prepped in Exercise 3. Set a timer for exactly two minutes. Stand up if you can. Deliver the full I·D·H·R answer aloud, using only the prep paper you wrote earlier as your reference. This is the closest thing to a real test you can do alone.
Five Part 2-specific exercises done. Here's how it landed.
Your performance across the Part 2 arena suggests the I·D·H·R system is settling. Move into Stage 9 to complete the lesson.
Part 2 is the section most candidates rate themselves either too high or too low on. Be honest. The point isn't to feel good — it's to know which screens to revisit when you do the morning drill this week.
Forty-five minutes invested in the lesson that turns Part 2 from the section most Vietnamese candidates dread into the section where they can make up real ground.
Lesson 6 gave you the universal system: prep ritual, I·D·H·R framework, 8 categories, 80 markers. Lessons 7-10 take that system and apply it lovingly to each of the four most common Part 2 categories — starting with the one that catches Vietnamese candidates out most often: Describing People.
Person cue cards are deceptively hard. Vietnamese candidates default to trait-listing ("he is kind, smart, helpful, friendly...") because Vietnamese culture rewards describing virtues. English examiners reward showing the person through specific stories. Lesson 7 fixes the translation problem.
Why "she is kind, smart, helpful, friendly" sounds Band 5 even with perfect grammar — and how to convert each trait into a 15-second story instead.
Specific moment · Reveals trait · Moment landing. The micro-formula for embedding character into narrative — every trait you'd list becomes a 30-second mini-anecdote instead.
How to handle anh / chị / cô / bác / chú in English without sounding awkward, and 40+ character description phrases that go beyond "kind" and "smart."
A Highlight section built around a single moment that captures who the person is — laughing too loudly at her own joke, refusing to take a compliment, knowing exactly when to leave the room.
"You've now built the structural backbone for the entire 2-minute long turn. From here on, Part 2 is just specific applications of the same I·D·H·R system you already own. Lessons 7-10 are going to feel familiar fast — same shape, new content. Take a day off, do Lionel's drill for a week, then come back ready for Part 2 category one."
Six badges. Part 1 in full, plus the universal Part 2 framework. The first eleven minutes of the test are now home ground, and the two-minute long turn has a system you can lean on.
Look at any year of IELTS Speaking cue cards and one category appears more than any other: People. Describe a family member. Describe a person you admire. Describe someone who has influenced you. The chance you'll get a Person card is high — and the chance Vietnamese candidates underperform on it is even higher. Not because the English is harder. Because a deep cultural translation problem sits at the heart of it.
"I've watched maybe four thousand Vietnamese candidates do Person cue cards. The grammar is fine, the vocabulary is fine — but ninety percent of them list virtues for two minutes. 'She is kind, she is smart, she is helpful, she is hardworking, she is loving...' In Vietnamese culture, that's honouring someone. In English IELTS, it's a Band 5 ceiling no matter how clean the grammar is. Lesson 7 fixes the translation."
Two candidates. Same grandmother. Same trait they want to convey — kindness. Watch what happens when one lists the trait and the other shows it through a single specific moment. Both candidates have the same vocabulary. Only one is at Band 7.
Grammar perfect. Vocabulary correct. But the examiner has no idea who this grandmother actually is. Every adjective applies to anyone's grandmother. Nothing is hers.
No adjectives. No "she is kind." Just one moment — broken tea cup, bathroom floor, ten silent minutes — and a landing line that names what the moment showed. The examiner now knows exactly who this grandmother is.
Same 9-stage shape as every other lesson. New focus: applying the I·D·H·R framework specifically to Person cue cards, and learning the S·R·M micro-formula that lives inside the Highlight section.
You are here.
The cultural diagnosis. Why virtue-listing is honoured in Vietnamese and penalised in English IELTS. The 4 trait-listing failure modes.
The fourth core formula. Specific moment · Reveals trait · Moment landing. Lives inside the Highlight section.
How the universal Part 2 framework adapts specifically for Person cue cards.
80+ Vietnamese-flavoured character phrases beyond "kind / smart / helpful."
40+ phrasings for anh / chị / cô / chú / bác without sounding awkward.
The Bác Hùng answer — annotated 2-minute Person card with the prep paper that produced it.
Six Person-specific exercises.
Self-assessment, badge, and Lesson 8 preview (Describing Places).
If you've ever delivered a beautiful Vietnamese description of a family member, listing their virtues with real warmth and respect, and then watched your IELTS examiner's pen stop moving — this stage is for you. The collapse isn't a language failure. It's a cultural mode mismatch that English IELTS specifically penalises.
Listing someone's virtues is a sign of respect. "Bà em rất hiền, đảm đang, thương con cháu lắm" — gentle, hardworking, loves her grandchildren — sounds warm and reverent. Direct virtue-naming is how Vietnamese honour the people we describe.
Listing virtues sounds Band 5 even with perfect grammar. The examiner hears generic adjectives that could describe anyone's grandmother and concludes the candidate can't paint a real person. Showing through a specific moment is what carries the band.
The Speaking band descriptors reward specificity, fluency over extended speech, and idiomatic flexibility. A trait list fails all three. Adjectives are generic. They run out fast (you can list six in twenty seconds). And "she is kind" uses the most basic possible grammar pattern repeated five times. The examiner trains to listen for the moment the candidate says "there was one time when..." — and panics quietly when it never comes.
Watch enough Vietnamese candidates do Person cue cards and the trait-listing problem appears in four distinct forms. Each one feels different in the moment. Each one has the same root cause — virtue-naming instead of scene-painting — but recognising your specific shape makes the fix more targeted.
The candidate lists five or six adjectives in a row without development. The pile collapses fast because adjectives don't generate further language.
The candidate keeps circling back to the same one or two virtues, rephrasing them differently each time. Sounds like they're saying more but they're not.
The candidate uses praise phrases that apply to anyone. The compliment could fit a stranger's grandmother as well as their own.
The candidate defines the person only by their role rather than by what makes them themselves. They become a category, not a character.
This is a translation guide, not a judgement. Vietnamese conversational norms are doing exactly what they should be doing inside Vietnamese culture. English IELTS norms reward something different. The point of this screen is to make the contrast visible so you can switch consciously.
Describes a person by listing their good qualities. The list itself is the respect.
Describes a person by painting one specific moment that shows who they are. The moment carries the respect.
Whenever you catch yourself about to list a virtue — "she is kind", "he is hardworking", "they are loyal" — pause and ask: what's one specific moment when I saw that? The answer to that question is the English IELTS sentence. The original virtue claim is just the Vietnamese mode equivalent.
It's not that you can never use the adjective. It's that the adjective alone isn't enough. "She was kind — once at Tết, when I was eight, she sat with me on the bathroom floor for ten minutes after I broke her tea cup..." The adjective opens, the scene carries.
Candidates sometimes resist this stage. "But it sounds disrespectful to talk about my grandmother that way." Worth addressing directly: nothing in Lesson 7 asks you to disrespect anyone. It asks you to switch modes for two minutes, then switch back. Here's how Lionel frames it.
"Outside the test, you describe your grandmother in Vietnamese, the way Vietnamese describes grandmothers — with virtue, with reverence, with the warmth that Vietnamese culture has built into its conversation. Nothing about Lesson 7 changes that. Inside the test — for two specific minutes — you switch into English IELTS mode. You paint one moment. You let the moment carry the respect that virtue-listing would carry in Vietnamese. Then you switch back. It's the same grandmother. It's the same love. You're just using a different brush for the next two minutes."
Before the test, you're chatting with the friend who drove you there in Vietnamese. "Bà em hiền lắm..." — virtue mode, completely natural.
You walk into the room, the test begins. For eleven minutes — including the two minutes of Part 2 — you're in English IELTS mode. Specific scenes. Named moments. Picturable detail.
You walk out of the room. Your friend asks how it went. You're back in Vietnamese mode, virtue and warmth, the way you always speak.
Two minutes of switched mode. Then back to who you actually are. Nothing more is being asked of you.
When you paint a real moment about your grandmother — broken tea cup, ten silent minutes on the bathroom floor — you're not being less reverent than the Vietnamese virtue list. You're being more reverent, by a different convention. The examiner hears a candidate who knows their grandmother well enough to remember a specific Tết from when they were eight. That depth of knowledge is the respect. The IELTS test just measures it in scenes instead of adjectives.
Two questions before we move to the S·R·M framework in Stage 3.
Three letters. One micro-formula. Every virtue you'd list in Vietnamese mode becomes a 15-30 second story in English IELTS mode. S·R·M sits inside the Highlight section of the I·D·H·R framework you learned in Lesson 6 — it doesn't replace it, it powers it. This is the fourth core formula in The Lionel Method.
The first letter does the heavy lifting. A specific moment is one concrete scene — not a habit, not a general behaviour, not a typical Sunday. One specific event. Three anchors make it specific: named time, named place, named action.
The middle letter is where most candidates trip. R doesn't mean "and then add the adjective at the end." R means: the moment itself shows the character — without you having to name it. The examiner watches the scene and concludes the trait independently. Your job is to pick the right scene, not to translate it.
"My uncle is very generous. He always shares his money. He once gave me money when I needed it. He is one of the most generous people I know."
"My uncle Hùng worked night shifts at a packaging factory for six years to put my cousin Linh through university. He never mentioned the money. The first I heard about it was at her graduation — she thanked him in her speech and he just looked at his hands."
"If your S·R·M ends with 'and that's why she is kind,' you haven't done R right. The scene should make the examiner conclude 'kind' on their own. If you have to name the trait, the trait wasn't in the moment."
The third letter is the small line at the end of the story — one sentence that names what the moment showed you. Not the trait itself. The meaning the moment carried. The landing is what transforms an anecdote into reflection. It's also what makes the Highlight section actually land rather than trailing off.
Pick a trait. Pick a specific moment that proves it. Pick the action that reveals the trait. Pick a landing line. Watch the full Highlight assemble in real time. Three of the four picks shape a different complete answer, so try each trait at least once.
[specific moment — pick the trait first] [revealing action] [landing line]
~30 seconds of Highlight built from three picks. Notice how the trait gets shown without ever being named directly — the moment carries it. This is what lifts a Person Highlight from Band 6 to Band 7.
Two questions before we move to how the People I·D·H·R micro-template works in Stage 4.
The universal Part 2 framework from Lesson 6 still applies. But each of the four sections has Person-specific moves — small adjustments that turn a generic I·D·H·R answer into one that actually shows who a person is. Stages 4-6 of Lesson 7 walk through those adjustments section by section, starting with the overview here.
Name + relationship + one anchoring fact. The hardest move is starting with the name, not the role.
Background, role in your life, how often you see them. Without trait-listing. Context not virtue.
The S·R·M Story lives here. One specific moment that shows who they are.
What this specific person taught you. Not generic "kindness" — what they specifically shifted.
Most Vietnamese candidates open with the role: "I'd like to talk about my best friend." The IELTS Band 7 move is to open with the name, then explain the relationship. The difference sounds small. It isn't. Naming the person tells the examiner you know them as a specific human, not a category.
Develop is where the trait-listing trap is most likely to ambush you. Thirty seconds to fill, no story yet, brain reaches for adjectives. Resist. Develop builds context — background, role, frequency, dynamic. Save the trait-showing for the Highlight section, where it belongs.
"Bà Hoa lived through the war, raised five children mostly on her own when ông nội was working away from home, and somehow still found time to teach herself to read at the local pagoda in her thirties. She lives in the same house in Cần Thơ where my mum and her siblings grew up — three concrete rooms with a small mango tree out the back. I see her maybe four times a year now since I moved to Saigon for work, but we talk on the phone most Sunday evenings — usually about whether I'm eating enough."
Notice what's missing: no "she is kind," no "she is hardworking," no "she is wise." All three traits are there — visible through the war detail, the five children, the self-taught reading, the Sunday calls about eating. But the Develop section never names them. It lets the context speak.
This is where Lesson 7 connects directly to Lesson 6. The Highlight section of a Person cue card is the S·R·M Story you built in Stage 3. Same architecture, just applied. The pacing markers from Lesson 6 still open the section ("What I remember most vividly was..."), and S·R·M fills the 30 seconds that follow.
The marker tells the examiner the band-deciding section has started. The S (Tết when I was eight, broken tea cup, hiding in the bathroom) makes the moment specific and picturable. The R (she sat next to me for ten minutes without a word) reveals kindness through action without naming it. The M (choosing the child over the object) lands the meaning. Thirty seconds of band-7 content from a four-letter formula.
Most candidates Reflect lazily on Person cards. "She taught me to be kind." "He showed me the value of family." Generic claims that could apply to anyone's grandmother or anyone's friend. The Band 7 move is making Reflect about what this specific person shifted in you — concretely, with a real change you can name.
Lead with the name in Introduce. Build context, not virtues, in Develop. Use S·R·M in Highlight. Make Reflect about a specific shift this person caused. Four small moves applied across the I·D·H·R framework you already own — and a Person cue card answer goes from generic to genuinely about the person.
Most candidates lose their Person card not because their vocabulary is weak, but because their character vocabulary is adjectival. "Kind, smart, hardworking, friendly." This bank gives you eighty character phrases that aren't adjectives — they describe people through action, anchor, voice, and meaning. Pick three from each category and you've replaced the entire trait-listing reflex with something the examiner actually rewards.
Phrases that anchor the person in your life. "The type to remember everyone's coffee order." Used in Develop, after introducing the person.
Phrases that show character through what someone actually does. "Wouldn't dream of leaving someone to pay alone." Used in Develop and Highlight.
Phrases that capture how someone moves through the world. "Speaks Vietnamese with a Huế accent you can hear from three streets away." Used anywhere.
Phrases for naming what this person has shifted in you. "I think of him every time I have to make a hard call." Used in Reflect specifically.
Twenty phrases that anchor the person as a specific human in your life. Pick three that match the person you're describing — one direct, one natural, one sophisticated. Used in Develop, right after the name and relationship clarifier.
Twenty phrases that show character through behaviour rather than naming traits. These are the ones that do the heaviest lifting in your Highlight section — they convert "she is kind" into "she'd never let you finish a drink without offering you another." The action is the trait.
Twenty phrases that capture how someone moves through the world — voice, presence, the small things that make them recognisable. These work especially well in the opening lines of Introduce, when you're sketching the person quickly. The examiner forms an immediate mental image.
Twenty phrases for the final 30 seconds. These are the lines that name what this specific person has shifted in you — without saying "she taught me to be kind." Most candidates lose Band 7 in Reflect by generalising. These phrases lock you into specifics.
Across the four categories — anchor, action, voice & manner, reflect — pick three favourites per category. Twelve total phrases that you can deploy on any Person cue card. Don't try to memorise eighty. Pick twelve, write them on a card, practise them for a week, and they'll become reflexes. Twelve phrases × four categories = a full Band 7 Person card vocabulary.
Vietnamese kinship is denser than English. anh, chị, em, cô, dì, chú, bác, ông, bà all encode information English collapses into "uncle / aunt / older brother / grandparent." Direct translation makes Person card answers feel awkward — too literal in one direction, too vague in the other. This stage gives you a small set of natural English phrasings that handle the problem cleanly.
"I would like to talk about my father's older brother's wife's youngest sister, who is..."
"I would like to talk about my cô, she is very..."
Six bridge constructions that handle every Vietnamese kinship term in natural English. The pattern is always: name + English-style clarifier, where the clarifier carries the Vietnamese information without the literal translation.
Audio-clickable phrasings for every major kinship category. Pick three to five favourites — enough to handle the family member you're most likely to describe in your real test. All deliberately natural-sounding, not literal translations.
From Stage 4 you already know to open with the name, not the role. For kinship specifically, the pattern is: Vietnamese name + dash + English clarifier. This handles the kinship problem in one move and signals to the examiner that the person is a real human, not a category.
The examiner gets three pieces of information in one sentence — the person's name, their relationship to you, and one concrete fact that anchors them in the world. That's a complete Introduce in roughly ten seconds, leaving twenty seconds of your Introduce section for the "why I picked this person" hook. Compare to "I'd like to talk about my uncle" — which gives the examiner nothing for the same word count.
Two questions before we move to the full worked example in Stage 7.
Six stages of theory — the trait-listing trap, the cultural diagnosis, the S·R·M framework, the People I·D·H·R micro-template, character vocabulary, and Vietnamese kinship handling. Time to see all six built into a single Band 7 answer. Over the next four screens, you'll see the cue card, the prep paper, the full annotated 2-minute answer, and the six structural moves that lifted it.
Describe a family member who has influenced you.
You should say:
and explain how they have influenced you.
What was on the prep paper when the timer ran out. Twenty-six words. Eight specific anchor points. Section letters down the left. The S·R·M micro-formula visible in the Highlight notes — specific moment, revealing action, landing line. The brain produces the live English; the paper just keeps the structure intact.
Four sections. Each marker colored by the section it belongs to. Toggle annotations to see the move each marker makes. Click "Hear it spoken" to listen to the full answer. Notice the S·R·M structure inside the Highlight section — it's three sentences, but each one is doing the specific job from Stage 3.
"I'd like to talk about Bác Hùng — my dad's older brother, who's lived in Vũng Tàu since I was small and worked as a fisherman for most of his life before retiring a few years ago. I picked him because when I think about who actually taught me anything that's stuck, it's him — more than any teacher I had at school, honestly."
For some context, we visited Bác Hùng and his family every summer when I was between about five and twelve — my parents would drop me off at the start of the school holidays and pick me up six or seven weeks later. Across those years, he taught me how to swim, how to fish off the rocks at Bãi Sau, and basically how to read the sea — when it was safe, when it wasn't. He never made a thing of any of it. He had this slow, deliberate way of explaining things — he'd just show you, then watch while you tried, then say one quiet sentence."
The moment I keep coming back to was one morning at Đầm Sen pool, when I was about six — I was terrified of the deep end and frozen at the edge. He stood in the water for what must have been forty minutes while I worked up the nerve to jump — never once told me to hurry, never once said anything actually. That was the first time I really understood what patient teaching looks like — it's giving someone the time they need, not the time you have."
Looking back, what he shifted in me wasn't an idea — it was a default behaviour. I notice now, whenever I'm teaching anyone anything — a junior at work, my younger cousin trying to ride a motorbike — I catch myself trying to wait the way he waited. So yeah, Bác Hùng. Quiet teacher of basically everything important I know."
No rare vocabulary. No memorised script. Six structural and register moves drawn from across Lesson 7 — each one something you've already met. Listed below.
"I'd like to talk about Bác Hùng — my dad's older brother" instead of "I'd like to talk about my uncle." Five words longer, light-years more specific.
"Bác Hùng — my dad's older brother" uses the bridge pattern from Stage 6 — Vietnamese name plus English clarifier. Sounds natural, not over-translated and not under-translated.
"He taught me how to swim, how to fish off the rocks at Bãi Sau, how to read the sea." No "he is patient, kind, wise." The patience and kindness are visible in the Develop section without being named.
One morning at Đầm Sen (S). Stood in the water for 40 minutes without rushing (R). "Giving someone the time they need, not the time you have" (M). Three sentences, full S·R·M, the band-deciding 30 seconds.
"What he shifted in me wasn't an idea — it was a default behaviour" + "I catch myself trying to wait the way he waited." Specific, observable change. Not generic "he taught me patience."
"So yeah, Bác Hùng. Quiet teacher of basically everything important I know." Small, dry, personal. Closes the answer with character, not closure formula.
No rare vocabulary. No British idioms. No memorised "topic vocabulary list." Every move came from structural choices that Lesson 7 taught you — and the answer hits Band 7 because of the structure, not the words. Most Vietnamese candidates lose Band 7 on Person cards because their vocabulary is adjectival, not because their grammar is weak. Lesson 7 fixes that — and the Bác Hùng answer is the proof.
Two diagnostic questions before we move to the Practice Arena.
Five real Person card excerpts. For each one, identify which of the four trait-listing failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four traits. For each, write a 3-line S·R·M conversion: the Specific moment, the Revealing action, the Moment landing. Don't try to be polished — get the shape right. Reveal each model after your attempt to see one possible version.
Three Person cue cards. For each, write the 60-second prep paper — 2-3 shorthand lines per I·D·H·R section. Notice the Highlight section needs S·R·M notes specifically. The "Reveal model" button shows a working example with the Vietnamese-flavoured detail you've been learning.
The band-deciding 30 seconds. Three Person cue cards — write only the Highlight section for each one (60-90 words). Force yourself to use the S·R·M architecture: zoom-in marker → S → R → M. No virtue-listing.
Pick one of the three cue cards you prepped in Exercise 3. Set a timer for exactly two minutes. Deliver the full Person card answer aloud, using only the prep paper you wrote as your reference. Three rounds — slow pass, natural pace, exam pressure.
Five Person-card exercises done. Here's how it landed.
Your performance across the Person card arena suggests the S·R·M system is taking root. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 7. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Person-card drill this week.
Forty-five minutes invested in the lesson that fixes the single most common Band 6 ceiling Vietnamese candidates hit on Part 2 — the trait-listing trap.
You've now done one full Part 2 category lesson. Lessons 8, 9, and 10 follow the same architecture — universal I·D·H·R framework plus a new category-specific Highlight tool. Up next: Places. The cue card category where most candidates default to tourist-brochure language and the band drops accordingly.
Place cue cards are deceptively hard. Vietnamese candidates default to "very beautiful, very famous, many tourists visit, very enjoyable" — tourist-brochure language that signals Band 5 instantly. Lesson 8 replaces that with sensory anchoring, the second category-specific Highlight tool in your toolkit.
Why "Hội An is very beautiful, has many tourists, and is very famous" caps at Band 5 — and how to replace the brochure language with a candidate's actual lived experience of the place.
The Highlight tool for Places. Sight + sound + smell + temperature + texture — pick two senses, deploy them concretely, and the place becomes picturable in the examiner's head.
80+ Vietnamese-flavoured place descriptors — Đà Lạt pine fog, the Mekong dawn, Hà Nội's old quarter at 5am, the back streets of Hội An after the lanterns go out.
Place Reflects need to be personal — not "I love Hội An, it's very beautiful." Lesson 8 teaches the personal Reflect pattern: what this place specifically did to you, that no other place could have done.
"Person cards are the hardest category for Vietnamese candidates and you just finished the lesson on them. Everything from here is going to feel familiar — same I·D·H·R framework, just different Highlight tools. Take a day off. Do Lionel's Person-card drill for a week. Then come back for Places when the trait-listing reflex has actually started fading."
Seven badges. The full Part 1 toolkit, the universal Part 2 framework, and the first Part 2 category deep dive. The trait-listing trap has a name now, a diagnostic, and a fix.
Person cards (Lesson 7) made Vietnamese candidates list virtues. Place cards do something subtly worse — they make every candidate sound like the back cover of a tourist brochure. "Hội An is very beautiful, has many tourists, is very famous, and very enjoyable." The IELTS examiner has heard exactly that answer about Hội An three hundred times this year. They cap it at Band 5-6 every time, regardless of grammar.
"Person cards collapse because Vietnamese candidates list virtues. Place cards collapse because Vietnamese candidates list adjectives the tourism board put on a billboard. 'Very beautiful, very famous, many tourists.' Generic doesn't get you Band 7. Specific does. Specific means the smell of pine fog in Đà Lạt at six in the morning, not 'Đà Lạt has nice weather.' That's what this lesson teaches — how to make a place feel like your place, not the tourist board's."
Two candidates. Both choosing Đà Lạt as their cue card answer. Both with the same vocabulary range. Watch what happens when one defaults to tourist-brochure language and the other anchors the place in one specific sensory moment. Same Đà Lạt. Only one is at Band 7.
Grammar perfect. Vocabulary correct. Could be lifted word-for-word from a tourism board pamphlet. The examiner has no idea what this candidate's Đà Lạt actually looks like.
No adjectives. No "very beautiful, very famous." Just one morning — pine fog, cold cup, condensed milk — and a landing line that names what the place did. The examiner now has a picturable scene of this candidate's actual Đà Lạt.
Same 9-stage shape as every other lesson. New focus: applying the I·D·H·R framework to Place cards specifically, plus learning Sensory Anchoring — the second category-specific Highlight tool in your growing toolkit.
You are here.
The 4 brochure-language failure modes — and why they cap at Band 5-6.
The second category-specific Highlight tool. Five senses, pick two for any Place Highlight. Plus the interactive Sensory Anchor Builder.
How the universal Part 2 framework adapts specifically for Place cue cards.
80+ Vietnamese-flavoured place phrases beyond "very beautiful / very famous."
When to use diacritics, how to anchor unfamiliar places, what to do with internationally known places.
The Đà Lạt morning answer — annotated 2-minute Place card with the prep paper that produced it.
Six Place-specific exercises.
Self-assessment, badge, and Lesson 9 preview (Describing Events).
If you've described Hội An, Đà Lạt, Hà Nội or Hạ Long Bay in IELTS practice and felt your answer turning into a leaflet — this stage is for you. The collapse isn't a vocabulary problem. It's a cultural pattern where every Vietnamese candidate has been trained, often unintentionally, to describe their country's places with the same six adjectives the tourism board uses.
School textbooks. Tourism board billboards. National identity messaging. "Vietnam, the hidden charm", "very beautiful, very famous". Vietnamese candidates absorb this language for years before they ever sit IELTS — and reach for it under pressure because it's there.
The examiner has heard the same five adjectives about Hội An from thirty different candidates this month. Generic praise is the opposite of lexical resource. There's nothing in your answer that an examiner can mark as specific, idiomatic, or extended. Band 5-6 ceiling automatic.
This is the key shift. Place cards aren't asking you to be a tour guide selling Vietnam to a foreigner. They're asking you to describe a place that matters to you specifically — what it sounds like when you're there at 6am, what it smells like in the rain, which corner of it you keep coming back to. The brochure language fails because it could apply to anyone visiting once. Your place can't.
Listen to enough Place card answers and the failures cluster into four distinct patterns. Same root cause behind each — defaulting to tourism-board language instead of your lived experience — but each has a different surface symptom. Recognising your specific shape makes the fix more targeted.
Five or six generic praise-adjectives in a row. "Very beautiful, very famous, very interesting, very enjoyable, very nice." Pure tourism-board copy with no specificity.
Facts about the place that read like a Wikipedia summary. Population, history, UNESCO status. The candidate becomes an unpaid tour guide for their own country instead of describing somewhere they actually know.
A flat list of attractions the candidate visited. No focus, no peak moment — just a sequence of stops on a tourist itinerary. Sounds like reading captions off Instagram posts.
The place becomes a stand-in for the country itself. The candidate describes it as proof Vietnam is great, with patriotic language that could be lifted from a speech. The actual place disappears under the politics.
A translation guide between the two registers. The Vietnamese tourism-board mode isn't wrong inside its context — it's the right register for selling the country. It's the wrong register for an IELTS Speaking test, which rewards lived specificity. Same place described in both modes makes the difference visible.
Describes a place by its reputation — what guidebooks say, what tourists are told. The marketing copy is the description.
Describes the place by what you specifically heard, saw, smelled, felt the last time you were actually there. One concrete moment.
Whenever you catch yourself reaching for a brochure adjective — "beautiful," "famous," "charming," "stunning" — pause and ask: what did I actually hear, see, or smell the last time I was there? The answer to that question is the IELTS sentence. The adjective is what the tourism board put on the billboard.
This doesn't mean the adjective is forbidden. It means the adjective alone is empty — it needs a specific sensory anchor to mean anything. "Hội An is beautiful — especially the back streets around 10pm, when the tour buses have left and you can hear the river." The adjective opens, the sensory anchor carries.
Sometimes candidates resist this stage too. "But Hội An is beautiful and famous." Worth addressing directly: nothing in Lesson 8 says the brochure isn't true. It says the brochure isn't the right register for a Band 7 IELTS answer. The fix is a brush change, not a reality change.
"Vietnam is beautiful. Hội An is famous. The brochure isn't lying — it's just selling the country to people who've never been. You've been. You've stood on a particular street at a particular time. You've heard a particular sound. You've smelled the river. The IELTS test isn't asking you to sell Vietnam — it's asking you to describe one place you actually know. That's a completely different brush. Smaller, finer, more personal. Two minutes of your Hội An, not the brochure's Hội An."
The brochure has a wide-angle lens. "Beautiful ancient town, peaceful, famous for lanterns." Everyone visiting Hội An can confirm all of that — which is the problem.
Your brush has a narrow lens. "The back streets behind the lantern market, around 10pm, after the tour buses have left, the river running properly, lantern shops dimming one by one." Only someone who has stood on that specific corner at that specific time could say it.
Same Hội An. Different brush. The Band 7 examiner is rewarding the narrow lens, not the wide one. Pull in tight, find the specific moment, paint that.
When you describe Hội An through one specific moment — the back streets at 10pm, the dimming lanterns, the river running — you're not being less proud of Vietnam than the brochure version. You're being more proud, in a different way. The examiner hears a candidate who knows their country well enough to have particular streets they love at particular times. That depth of knowledge is the patriotism. The IELTS test just rewards it in sensory detail instead of in adjectives.
Two questions before we move to the Sensory Anchoring framework in Stage 3.
Five senses. Pick two. Build the Highlight from there. That's the entire framework. Sensory Anchoring is the second category-specific Highlight tool in your toolkit — same role for Place cards that S·R·M plays for Person cards. It sits inside the Highlight section of I·D·H·R, and replaces brochure adjectives with what you actually heard, saw, smelled, felt, or touched.
If you only learn one combination, learn this one. Sight and sound are the two most natural senses to deploy under exam pressure — they're what the brain reaches for first when describing a real place. Used together, they cover roughly 80% of Highlight situations. Used alone with one well-chosen detail each, they're enough.
Sight and sound are what every candidate reaches for. Smell and temperature are what very few do — which is exactly why they lift the band. A single well-chosen smell or temperature detail signals to the examiner that you're not reciting a textbook, you're describing a place you've actually been. These are the senses that buy Band 7.
"If you remember nothing else from Stage 3, remember this: add one smell. Candidates who include one specific smell in their Place Highlight jump half a band almost automatically. It's the single most under-used sensory anchor in IELTS Speaking."
Texture is the least-used sense and the most under-rated. It's also the easiest to combine with the other four. This screen covers texture as a standalone and then shows you how to combine two senses cleanly — the move that takes a Highlight from "good" to "Band 7."
One sense is good. Two senses, deployed in the same sentence, is the move that lifts the Highlight from Band 6 to Band 7. The combination tells the examiner you're inside the scene, not describing it from outside.
"The river was running properly — the kind of quiet you don't get in Saigon — and the smell was still incense from the temples, but quieter than during the day."
"The fog was sitting on the pines below us, and the air was cold enough that my hands were wrapped around the cup just to feel something warm."
"The smell was wet pine bark and condensed milk, and the wooden steps under me were slightly damp from the morning fog."
Pick two senses. Deploy them in the same Highlight. Don't try for all five. Two senses concretely deployed beats five vague mentions. The hardest part of Sensory Anchoring isn't choosing senses — it's resisting the urge to pile on a third, a fourth, a fifth. Discipline yourself to two and let those two do their work.
Four picks. Pick a place. Pick two senses. Pick the specific detail for each sense. Pick a landing line. Watch the full Highlight assemble in real time. Each place has different sense options — try at least two different places to see the range.
[pick the place first]
~30 seconds of Highlight built from four picks. Notice how the senses do all the work — the place is anchored through what you heard, saw, smelled, or felt, never through generic adjectives. This is what lifts a Place Highlight from Band 6 to Band 7.
Two questions before we move to how the Places I·D·H·R micro-template works in Stage 4.
The universal Part 2 framework from Lesson 6 still applies. The universal Person-specific moves from Lesson 7 don't. Place cards have their own four Place-specific moves — small adjustments to each section that turn a generic I·D·H·R answer into a Band 7 Place description. Stage 4 walks through those adjustments section by section.
Name + location + one anchoring fact. The hardest move is naming a specific corner of the place, not the place itself.
How you came to know the place, frequency, what kind of place it is in your life. Not guidebook facts.
Sensory Anchoring lives here. Two senses, one specific moment.
What this place specifically does to you. Not "I love it" — the precise effect this place has that no other place could give you.
Most candidates open with the city: "I'd like to talk about Hà Nội." The IELTS Band 7 move is to open with a specific corner of the place — a particular street, a particular time of day, a particular kind of moment. The difference does the same work as opening with a person's name in Lesson 7: it signals that you know this place as a specific somewhere, not a category.
Develop is where the Tour Guide Recitation lurks. Thirty seconds to fill, no specific moment yet, brain reaches for guidebook facts. Resist. Develop builds your relationship with the place — when you first went, how often, what role it plays in your life. Save the guidebook for someone who hasn't been.
"I first went up to Đà Lạt with my cousin Mai when I was about seventeen — it was the kind of trip you take when you've just finished high school and want to feel like you've gone somewhere different. Since then I've been back probably six or seven times, almost always between November and February, which is when the mornings get properly cold. I usually go for two or three days, mostly on my own these days, with a notebook and no real plan. It's become the place I go when Saigon starts to feel too much."
Notice what's missing: no "Đà Lạt is famous for its flowers," no "it became a city in 1893," no UNESCO talk. Three facts about this candidate's Đà Lạt instead — when they first went, the rhythm of the visits, what the place does for them. The brochure version of Đà Lạt is invisible. The candidate's version is everywhere.
This is where Stage 3 plugs in. The Highlight section of a Place cue card is the Sensory Anchor you built in the Stage 3 Builder. Same architecture: opening with a Highlight marker from Lesson 6, then two senses deployed in one specific moment, then a landing line that names what the place did to you. Thirty seconds. Band-deciding.
The marker tells the examiner the band-deciding section has started. The specific moment (early morning, the steps outside one coffee shop on Đường Hồ Tùng Mậu) puts the listener inside the scene. The two senses — temperature (cold air, hands wrapped around the cup) and smell (pine bark and condensed milk) — paint the place from the inside. The landing names the effect (nothing in Saigon ever feels like this) without using a single brochure adjective. Thirty seconds of band-7 content from the framework you already built.
Most candidates Reflect lazily on Place cards. "I love it." "It is very special to me." "I will visit again." Statements that fit any place anywhere. The Band 7 move is making Reflect about a specific effect this place has on you — something that only this corner of this town, in this rhythm of your life, can do.
Name the corner in Introduce. Tell your story, not the guidebook's, in Develop. Use Sensory Anchoring in Highlight. Make Reflect about a this-place-only effect. Four small moves applied across the I·D·H·R framework you already own — and a Place card answer goes from generic-Band-6 to specific-Band-7.
The same architecture as the L7 character vocab bank, applied to places. Eighty Vietnamese-flavoured Place phrases that aren't brochure adjectives. Pick three from each category and you've replaced "very beautiful, very famous" with vocabulary the examiner actually rewards.
Phrases for naming the specific corner of a place rather than the place itself. "The back streets behind the lantern market." Used in Introduce.
Ready-made sensory anchors you can drop straight into a Highlight. "The kind of quiet you don't get in Saigon." Used in Highlight.
Phrases that describe your relationship with the place — how often, what season, what role. Used in Develop.
Phrases that name the specific effect this place has on you. "The kind of city where you remember what slow feels like." Used in Reflect.
Twenty phrases for naming the specific corner of a place rather than the place itself. These are your Introduce-section openers. Pick the structure that fits the place you've chosen — back street, rooftop, alley, particular morning, particular bench.
Twenty ready-made sensory anchors you can drop straight into a Highlight. These cover the five senses across Vietnamese-specific contexts. Pick the two senses that fit your moment from Stage 3, and you'll find your wording here.
Twenty phrases for your Develop section — describing the pattern of your relationship with the place. How often you go, what season, what brings you back, what role the place plays. These let you fill Develop with your story instead of the guidebook's.
Twenty phrases for your Reflect section. These name the specific effect this place has on you — what makes it irreplaceable. Most candidates lose Band 7 in Reflect by saying "I love it" or "I will visit again." These phrases lock you into specifics.
Across the four categories — corner, sensory, rhythm, this-place-only — pick three favourites per category. Twelve phrases total that you can deploy on any Place cue card. Don't try to memorise eighty. Pick twelve, rehearse for a week, and they'll become reflexes. Twelve phrases × four categories = a full Band 7 Place card vocabulary.
Lesson 7 had the kinship problem — Vietnamese terms denser than English. Lesson 8 has the place-name problem — Vietnamese places ranging from internationally famous to known only inside one province. Examiners have heard of Hà Nội. Many haven't heard of Cần Thơ. None will have heard of Bãi Sau. The Band 7 candidate knows which place needs anchoring and which doesn't, and handles each appropriately without breaking flow.
"I'd like to talk about Hà Nội, the capital of Vietnam, which is located in the north of the country and has a population of approximately 8 million people, and is famous for..."
"I'd like to talk about Bãi Sau, where I used to swim every summer..."
For any tier-2 or tier-3 Vietnamese place, the same 3-piece pattern works. Vietnamese name + English clarifier + locator detail. Three pieces, one sentence, the place is placed. Same architecture as the L7 kinship bridge — different content, identical move.
Direction + distance: "about three hours south of Saigon" or "an hour north of Hà Nội." This is the most efficient locator — gives the examiner a mental map in five words.
Geographic feature: "in the Mekong delta," "in the Red River Delta," "in the central highlands." Useful when the place is best known for its region.
Elevation or coast: "1,500 metres above Saigon," "on the coast east of Saigon." Specific and concrete — works particularly well for hill towns and coastal places.
Audio-clickable phrasings for every major Vietnamese place a candidate is likely to choose. Each is a working anchor — Vietnamese name with the English clarifier ready to deploy. Use them as templates: keep the structure, swap in the place you actually know best.
A small but band-relevant question: do you say "Hà Nội" with full tones, or do you anglicise it to "Hanoi" with English stress? The answer is small but consistent: say it the Vietnamese way naturally, but don't make a thing of it. The examiner isn't testing your pronunciation of Vietnamese — but consistent, comfortable Vietnamese place-name pronunciation signals an authentic register.
Two questions before we move to the worked example in Stage 7.
Six stages of theory — the tourist-brochure trap, the cultural diagnosis, the Sensory Anchoring framework, the Places I·D·H·R micro-template, place vocabulary, and Vietnamese place-name anchoring. Time to see all six built into a single Band 7 answer. Over the next four screens, you'll see the cue card, the prep paper, the full annotated 2-minute answer, and the six structural moves that lifted it.
Describe a place you like to visit.
You should say:
and explain why you like this place.
What was on the prep paper when the timer ran out. Twenty-eight words. Eight specific anchor points. Section letters down the left. The S·A micro-formula visible in the Highlight notes — two senses + landing line. The brain produces the live English; the paper just keeps the structure intact.
Four sections. Each marker colored by the section it belongs to. Toggle annotations to see the move each marker makes. Click "Hear it spoken" to listen to the full answer. Notice the S·A structure inside the Highlight section — temperature and smell, deployed in one specific morning, with a landing line that doesn't say "beautiful" once.
"I'd like to talk about a particular coffee shop on Đường Hồ Tùng Mậu, in Đà Lạt — a hill town about 1,500 metres above Saigon — which has somehow become the place I keep going back to whenever the city stops working for me. I picked this place because out of everywhere I could have talked about, this one specific corner is the version of Vietnam I most look forward to seeing every year."
I first went up to Đà Lạt with my cousin Mai when I was about seventeen — the kind of trip you take when you've just finished high school and want to feel like you've gone somewhere different. Since then I've been back probably six or seven times, almost always between November and February, which is when the mornings get properly cold. I usually go for two or three days, mostly on my own these days, with a notebook and no real plan. It's become the place I go when Saigon starts to feel too much."
What I keep coming back to is one early morning, sitting on the wooden steps outside that coffee shop — fog still everywhere, no traffic yet. The air was cold enough that my hands were wrapped around the cup just to feel something warm — and the smell was wet pine bark and condensed milk coffee, the kind of smell you only get above 1,500 metres. And I remember thinking — nothing in Saigon ever feels like this."
Looking back, what Đà Lạt does for me isn't really about the place — it's that it's the only city in Vietnam where I can sit in silence for an hour without it feeling like a waste of time. Saigon doesn't allow that. Đà Lạt seems to almost demand it. So yeah, that one coffee shop on Đường Hồ Tùng Mậu. Quiet teacher of how slow is meant to feel."
No rare vocabulary. No memorised script. Six structural and register moves drawn from across Lesson 8 — each one something you've already met. Listed below.
"I'd like to talk about a particular coffee shop on Đường Hồ Tùng Mậu" instead of "I'd like to talk about Đà Lạt." Five words longer, light-years more specific.
"Đà Lạt — a hill town about 1,500 metres above Saigon" uses the 3-piece anchor pattern. Vietnamese name + English clarifier + locator detail. Examiner has a mental map in five seconds.
"First went at seventeen / been back six or seven times / November to February / two or three days alone with a notebook." Four rhythm-of-relationship facts. No "Đà Lạt is famous for its flowers." Develop section reads like the candidate's actual life, not Wikipedia.
One specific morning. Two senses — temperature (hands wrapped around the cup) and smell (wet pine bark + condensed milk). Landing line ("nothing in Saigon ever feels like this"). The band-deciding 30 seconds, with not a single "beautiful" or "famous."
"It's the only city in Vietnam where I can sit in silence for an hour without it feeling like a waste of time. Saigon doesn't allow that. Đà Lạt seems to almost demand it." Specific contrast with home. No "I love it." The this-place-only effect, named precisely.
"So yeah, that one coffee shop on Đường Hồ Tùng Mậu. Quiet teacher of how slow is meant to feel." Small, dry, personal. Closes the answer with character — not "I will visit again."
No "beautiful." No "famous." No "many tourists." No Wikipedia facts about when Đà Lạt was founded. Every move came from structural choices that Lesson 8 taught you — and the answer hits Band 7 because of the structure, not the words. Most Vietnamese candidates lose Band 7 on Place cards because their vocabulary is brochure-shaped, not because their grammar is weak. Lesson 8 fixes that — and the Đà Lạt answer is the proof.
Two diagnostic questions before we move to the Practice Arena.
Five real Place card excerpts. For each one, identify which of the four brochure-language failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four brochure adjectives. For each, write a 2-line sensory-anchored Highlight — pick two senses, deploy them in one specific moment. Don't try to be polished — get the shape right. Reveal each model after your attempt.
Three Place cue cards. For each, write the 60-second prep paper — 2-3 shorthand lines per I·D·H·R section. The Highlight section needs two sense notes specifically. The "Reveal model" button shows a Vietnamese-flavoured working example.
The band-deciding 30 seconds. Three Place cue cards — write only the Highlight section for each one (60-90 words). Force yourself to use the Sensory Anchoring architecture: zoom-in marker → specific moment → two senses → landing line. No brochure adjectives.
Pick one of the three cue cards you prepped in Exercise 3. Set a timer for exactly two minutes. Deliver the full Place card answer aloud, using only the prep paper you wrote as your reference. Three rounds — slow pass, natural pace, exam pressure.
Five Place-card exercises done. Here's how it landed.
Your performance across the Place card arena suggests the S·A system is taking root. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 8. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Place-card drill this week.
Forty-five minutes invested in the lesson that fixes the second most common Band 6 ceiling Vietnamese candidates hit on Part 2 — the tourist-brochure trap.
Two Part 2 categories done. Two more to go. Lesson 9 is the third Part 2 category — Events. Same I·D·H·R framework, same 9-stage shape, third category-specific Highlight tool. Up next: the Peak Moment Zoom — the framework that prevents Event cards from becoming chronological play-by-plays.
Event cue cards are the third Part 2 category, and they have their own specific failure mode — the chronological play-by-play. Vietnamese candidates tend to narrate events minute by minute: "First, we arrived at the wedding. Then we sat down. Then the bride came in. Then..." Lesson 9 replaces that with Peak Moment Zoom, the third category-specific Highlight tool in your toolkit.
Why narrating an event minute by minute caps at Band 5-6 — and how the Peak Moment Zoom replaces 20 minutes of "then this, then that" with 30 seconds of the moment that actually mattered.
The Highlight tool for Events. Pick the one peak moment, zoom into it tightly, dilate time across 30 seconds. The rest of the event becomes context — your peak moment becomes the answer.
80+ Vietnamese-flavoured event phrases — Tết family gatherings, weddings at a friend's restaurant, university graduations, Lunar New Year midnight, the first concert someone took you to.
Event Reflects need to name the shift — what was true before this event that wasn't true after. Lesson 9 teaches the before/after Reflect pattern: the specific way this event left you different.
"Two Part 2 categories down. Same I·D·H·R framework, two different category-specific Highlight tools — S·R·M for People, S·A for Places. Two more to go. Events come next, and they have the easiest fix of any of the four categories. Take a day off, do Lionel's Place-card drill for a week, and the brochure reflex will be gone. Then Events."
Eight badges. The full Part 1 toolkit, the universal Part 2 framework, and the first two Part 2 categories — People and Places. The brochure trap has a name now, a diagnostic, and a fix.
Person cards made Vietnamese candidates list virtues. Place cards made them sound like brochures. Event cards do something different again — they make every candidate narrate the event minute by minute, like an itinerary. "First we arrived. Then we sat down. Then they cut the cake. Then we ate. Then we took photos. Then we left." Twenty things happened in your two minutes. None of them peaked. Band 5-6 ceiling, every time.
"Person cards collapse on virtues. Place cards collapse on brochure adjectives. Event cards collapse on chronology. Vietnamese candidates narrate every wedding from 'first the bride arrived' to 'then we went home,' with equal weight on every step. The examiner ends up with twenty events and not one of them mattered. Band 7 needs the opposite — one moment that mattered, dilated across thirty seconds. The rest of the event becomes background. Your peak moment becomes the answer."
Two candidates. Both describing a cousin's wedding. Same vocabulary range, same grammar. Watch what happens when one narrates chronologically and the other zooms into one specific peak moment. Same wedding. Only one is at Band 7.
Grammar perfect. Vocabulary correct. Could be the itinerary printed in the wedding programme. The examiner has no idea why this wedding mattered to this candidate.
No itinerary. No "then this, then that." Just ten seconds of silence — magnified into thirty seconds of speech, with one specific person doing one specific thing. The examiner is inside the wedding hall, watching Uncle Tuấn stop talking.
Same 9-stage shape as Lessons 7 and 8. New focus: applying I·D·H·R to Event cards specifically, plus learning Peak Moment Zoom — the third category-specific Highlight tool in your growing toolkit.
You are here.
The 4 chronological-narration failure modes — and why they cap at Band 5-6.
The third category-specific Highlight tool. Pick the peak, magnify it, zoom out into meaning. Plus the interactive Peak Moment Builder.
How the universal Part 2 framework adapts specifically for Event cue cards.
80+ Vietnamese-flavoured event phrases beyond "very nice / very enjoyable."
Tết, weddings, đám giỗ (death anniversaries), lễ hội (festivals) — how to anchor each cleanly without over-explaining.
The cousin's wedding answer — annotated 2-minute Event card with the prep paper that produced it.
Five Event-specific exercises.
Self-assessment, badge, and Lesson 10 preview (Describing Objects).
If you've described a wedding, a Tết gathering, a birthday party, or a graduation in IELTS practice and felt your answer turning into an itinerary — this stage is for you. The collapse isn't a vocabulary problem. It's a cultural pattern where Vietnamese candidates have been trained, often unintentionally, to narrate events with equal weight on every step — like a school recital.
Vietnamese school storytelling. Family retellings of Tết and weddings, where the social purpose is including everyone — every cousin's outfit, every dish, every speech. "First we did X. Then we did Y." That's how events are remembered out loud at home. Vietnamese candidates carry the same shape into English.
Twenty events given equal weight = no event mattered. The examiner ends up with the itinerary but never the story. Band 7 specifically rewards structure with shape — a peak, a valley, a turning point. Flat chronology has no shape. The grade caps automatically.
This is the key shift. Event cards aren't asking you to recount the whole event from arrival to departure. They're asking you to describe an event in a way that lets the examiner feel why this specific event mattered to you specifically. The chronological version fails because it could be the wedding programme — facts about what happened in order, without any moment carrying weight. Your version can't.
Listen to enough Event card answers and the failures cluster into four distinct patterns. Same root cause behind each — defaulting to chronological narration instead of peak-zooming — but each has a different surface symptom. Recognising your specific shape makes the fix more targeted.
Equal-weight chronological narration. Every step gets one sentence. "First we did X. Then we did Y. Then we did Z." No moment matters more than any other. Pure programme-of-events recitation.
The event becomes a roll call of who was there. Names, relationships, what each person was wearing or doing. The actual event disappears under the cast. Sounds like reading the seating chart.
Especially common at weddings and Tết. The candidate lists every dish, drink, and course. Phở, gỏi cuốn, cá kho tộ, bánh chưng, tré. The event becomes a food log instead of a memory.
After the chronology, the candidate adds a generic emotional close. "It was a very nice day." "I had a great time." "I will remember it forever." Equal-weight chronology + content-free emotion = nothing the examiner can grade as Band 7.
A translation guide between the two registers. Vietnamese chronological narration isn't wrong inside its context — it's the right register for family retellings, where the social purpose is including everyone. It's the wrong register for an IELTS Speaking test, which rewards peak-zooming. Same event described in both modes makes the difference visible.
Describes an event by the order things happened. Equal weight on each step. The timeline is the story.
Picks the one moment that actually mattered. Slows time down inside it. The rest of the event becomes background.
Whenever you catch yourself reaching for "first... then... then... then..." — pause and ask: which moment of this event do I actually still see in my head when I think about it? The answer to that question is the IELTS sentence. The chronology is what the wedding programme had.
This doesn't mean chronology is forbidden. It means chronology alone is empty — it needs one specific moment that's been zoomed into. "At my cousin's wedding — particularly the ten seconds before she walked in — my uncle Tuấn just stopped talking..." The event opens, the peak moment carries.
Sometimes candidates resist this stage too. "But all those things at the wedding really happened." Worth addressing directly: nothing in Lesson 9 says the chronology isn't true. It says the chronology isn't the right register for a Band 7 IELTS answer. The fix is a focal-length change, not a reality change.
"Of course the whole wedding happened. The tea ceremony, the speeches, the food, the photos — all of it. The chronology isn't a lie, it's just the wrong tool. The IELTS test isn't asking you to recount the wedding. It's asking you to describe the wedding in a way that lets me feel why it mattered to you. That's a completely different camera. Wide-angle becomes telephoto. Less event, more moment. Two minutes of your peak, not the programme's chronology."
Chronological narration has a wide-angle lens. "First X, then Y, then Z, then W, then V." Everyone at the wedding can confirm all of it — which is the problem.
Your peak zoom has a telephoto lens. "Ten seconds of silence right before she walked in. Uncle Tuấn — who'd been talking the whole morning about how she'd never settle down — just stopped, mid-sentence." Only someone who was standing next to Uncle Tuấn at that specific moment could say it.
Same wedding. Different camera. The Band 7 examiner is rewarding the telephoto, not the wide-angle. Pull in tight. Pick the moment. Magnify it.
When you describe a wedding through one specific moment — Uncle Tuấn going quiet, your mum's shaking hands, the ten seconds of silence before the music — you're not honouring the event less than the chronological version. You're honouring it more, in a different way. The examiner hears a candidate who remembers the wedding well enough to have a specific peak — which is what real memory looks like. That peak is the respect. The IELTS test just rewards it in zoom instead of in chronology.
Two questions before we move to the Peak Moment Zoom framework in Stage 3.
Three moves. Pick the peak. Magnify the moment. Zoom out into meaning. That's the entire framework. Peak Moment Zoom (P·M·Z) is the third category-specific Highlight tool — same role for Event cards that S·R·M plays for People and S·A plays for Places. It sits inside the Highlight section of I·D·H·R, and replaces equal-weight chronology with one peak moment, dilated across 30 seconds.
The first move is selection. Out of everything that happened at the event, you pick one moment to zoom into. The wrong moment kills the answer. The right moment makes the rest of the framework do its work. There are three reliable tests for picking the right peak — they all converge on the same kind of moment.
The second move is the technical one. Once you've picked your peak, you have to magnify it — slow real time down inside your speech. Ten seconds of real time should expand into about 30 seconds of speech. This is the Time-Dilation Technique, and it's the single biggest band-lifter in Lesson 9.
"If you remember nothing else from Stage 3, remember this: name one specific person doing one specific small thing inside the peak. That single move dilates time more than any other technique. The bride walked in — that's coverage. Uncle Tuấn stopped talking mid-sentence — that's a magnified moment. The examiner can see Uncle Tuấn. That's Band 7."
The third move is the landing. Once you've picked the peak and magnified it across 25 seconds of speech, you need 5 seconds of zoom-out — one sentence that names what the moment did. Not the event's meaning. The peak moment's specific shift. This is what closes the Highlight and prepares the Reflect section.
One full Peak Moment Zoom in one block — Pick, Magnify, Zoom. Notice how the three moves layer: Pick is one line of setup, Magnify is two or three lines of dilation with a specific person and contrast, Zoom is one closing line.
"What I keep coming back to is the ten seconds right before my cousin walked in. My uncle Tuấn — who'd been talking the whole morning about how she'd never settle down — he just stopped, mid-sentence. You could hear the air conditioning. He didn't speak again until she was halfway down the aisle. That was when I realised this wasn't just a wedding to him — it was something else entirely."
Pick one peak. Magnify it across most of the Highlight. Zoom out in one closing sentence. The hardest discipline is the proportions — most candidates magnify too little (they jump straight to the zoom-out) or magnify too much (they never zoom out, the Highlight bleeds into Reflect). Aim for roughly 25 seconds of magnification + 5 seconds of zoom-out.
Four picks. Pick an event. Pick the peak (10 seconds of real time). Pick what magnifies it (the specific person doing the specific thing). Pick the zoom-out. Watch the full P·M·Z Highlight assemble in real time. Each event has different peak options — try at least two different events to see the range.
[pick the event first]
~30 seconds of Highlight built from four picks. Notice how the three moves layer: Pick sets up the moment, Magnify slows time inside it (specific person + specific small thing + contrast), Zoom out closes with the shift. This is what lifts an Event Highlight from Band 6 to Band 7.
Two questions before we move to the Events I·D·H·R micro-template in Stage 4.
The universal Part 2 framework from Lesson 6 still applies. The People-specific moves from Lesson 7 and the Place-specific moves from Lesson 8 don't. Event cards have their own four Event-specific moves — small adjustments to each section that turn a generic I·D·H·R answer into a Band 7 Event description. Stage 4 walks through those adjustments section by section.
Event name + when + one anchor. The hardest move is naming the specific instance of the event, not the category of event.
Three or four context moves to set up the peak — not the full chronology. Background sketch only.
Peak Moment Zoom lives here. Pick · Magnify · Zoom out.
The before/after shift. What was true about you before this event that wasn't true after.
Most candidates open with the category: "I'd like to talk about a wedding." The Band 7 move is to open with a specific instance of the event — whose, when, where. The same kind of specificity move as L7 (name the person) and L8 (name the corner), applied to events.
Develop is where the Itinerary March lurks. Thirty seconds to fill, peak not yet arrived, brain reaches for chronology. Resist. Develop builds just enough context for the Highlight peak to land — not the full timeline. Three or four background sentences, no more. Save the chronology for the wedding programme.
"The wedding was on a Saturday in August, at her husband Tuấn-Anh's family's seafood restaurant in Vũng Tàu — about three hours from Saigon by bus. The whole of my mum's side of the family came down, including my uncle Tuấn — Linh's father — who'd been famously sceptical of the whole match for about a year and a half. The morning of the wedding he was still telling anyone who'd listen that Linh was too stubborn to ever actually go through with it. The part I keep going back to, though, is what happened in the ten seconds before she walked into the reception hall."
Notice what's missing: no chronology. No "first the tea ceremony, then the speeches." Three context facts (who, where, the family tension) + one forward-pointing setup line. Develop is doing exactly what Develop is meant to do — making the peak land harder. The full timeline is invisible. The candidate's relationship to this specific wedding is everywhere.
This is where Stage 3 plugs in. The Highlight section of an Event cue card is the Peak Moment Zoom you built in the Stage 3 Builder. Same architecture: opening with a Highlight marker from Lesson 6, then Pick the peak, Magnify it, Zoom out. Thirty seconds. Band-deciding.
The marker tells the examiner the band-deciding section has started. The Pick names the peak — 10 seconds of real time. The Magnify does the heaviest lifting: Uncle Tuấn (specific person), stopped mid-sentence (specific small thing), morning sceptic vs sudden silence (contrast). The Zoom out closes with the shift — "this wasn't just a wedding to him" — naming what those ten seconds revealed. Thirty seconds of Band 7 content from the framework you already built.
Most candidates Reflect lazily on Event cards. "It was a great day." "I will remember it forever." Statements that fit any event anywhere. The Band 7 move is making Reflect about a specific before/after shift — what was true about you, your family, or your understanding before this event that wasn't true after.
Name the specific instance in Introduce. Use Develop for setup, not narration. Use Peak Moment Zoom in Highlight. Make Reflect about a before/after shift. Four small moves applied across the I·D·H·R framework you already own — and an Event card answer goes from generic-Band-6 to specific-Band-7.
Same architecture as the L7 character bank and the L8 place bank, applied to events. Eighty Vietnamese-flavoured Event phrases that aren't "very nice" or "I will remember forever." Pick three from each category and you've replaced the chronological-narration reflex with vocabulary the examiner actually rewards.
Phrases for naming the specific instance of an event rather than the category. "My cousin Linh's wedding two summers ago." Used in Introduce.
Ready-made openers and pivots for the peak-moment Highlight. "What I keep coming back to is the ten seconds when..." Used in Highlight.
Phrases that magnify a moment — specific-person constructions, contrast openers, sensory drop-ins. The body of the Magnify move.
Phrases for the Reflect section. "Before that day I'd always thought... After that day I started noticing..." The shift pattern.
Twenty phrases for naming the specific instance of an event rather than the category. These are your Introduce-section openers. Pick the structure that fits your specific event — whose wedding, which Tết, what year of graduation, which đám giỗ.
Twenty ready-made openers and pivots for the peak-moment Highlight. These are the phrases that say "I'm zooming in now" — Lesson 6's markers, retuned for events. Pick three favourites and drill them until they're reflexive.
Twenty phrases for your Magnify section — the body of the Highlight. These are the highest-leverage phrases in the entire bank because they're what dilates time inside a peak moment. Specific-person constructions, contrast openers, sensory drop-ins, and pause-detail combinations.
Twenty phrases for your Reflect section. These name the specific before/after shift — what was true about you, your family, or your understanding before the event that wasn't true after. Most candidates lose Band 7 in Reflect by saying "I will remember forever." These phrases lock you into the shift pattern.
Across the four categories — instance-naming, peak-moment, time-dilation, before/after shift — pick three favourites per category. Twelve phrases total that you can deploy on any Event cue card. Don't try to memorise eighty. Pick twelve, rehearse for a week, and they'll become reflexes. Twelve phrases × four categories = a full Band 7 Event card vocabulary.
Lesson 7 had the kinship problem. Lesson 8 had the place-name problem. Lesson 9 has the Vietnamese-event problem — events ranging from internationally known to specifically Vietnamese cultural practices the examiner may never have encountered. Examiners have heard of Tết and weddings. Many haven't heard of đám giỗ. None will have heard of đầy tháng or thôi nôi. The Band 7 candidate knows which event needs anchoring and which doesn't, and handles each cleanly.
"I'd like to talk about Tết, which is the Vietnamese New Year, a major traditional festival in Vietnam, celebrated according to the lunar calendar, lasting several days, marked by family gatherings, special foods like bánh chưng..."
"I'd like to talk about my niece's đầy tháng, when the whole family came together..."
Three constructions — one for each tier that needs anchoring. Each is a sentence-level template you can plug any Vietnamese event into. Light anchor for tier-2, full anchor for tier-3, anchor + function for tier-4. The patterns are identical in structure to the kinship bridges of L7 and the place-name anchors of L8 — same architecture, new content.
Name + literal translation: "Tết — the Vietnamese Lunar New Year." Cleanest for tier-2 events. Examiner gets the category in five words.
Name + cultural function: "Đám giỗ — the annual death-anniversary ceremony." Best for tier-3 events. Gives the examiner what the event is for, not just what it's called.
Name + function + sense of scale: "Đầy tháng — when a baby turns one month old, marking the first time they can be properly introduced to the extended family." Best for tier-4 events. The Band 8 anchor — gives the examiner the function and a feel for how big a deal it is in the culture.
Audio-clickable phrasings for every major Vietnamese event category a candidate is likely to choose. Each is a working anchor — Vietnamese name with the English clarifier ready to deploy. Use them as templates: keep the structure, swap in the specific event you actually attended.
A small but band-relevant question: do you say "Tết" with the full tone, or do you anglicise it? Answer: say it the Vietnamese way naturally, but don't make a thing of it. The examiner isn't testing your pronunciation of Vietnamese — but consistent, comfortable Vietnamese event pronunciation signals an authentic register. Move on quickly.
Two questions before we move to the worked example in Stage 7.
Six stages of theory — the chronological-narration trap, the cultural diagnosis, the Peak Moment Zoom framework, the Events I·D·H·R micro-template, event vocabulary, and Vietnamese event anchoring. Time to see all six built into a single Band 7 answer. Over the next four screens, you'll see the cue card, the prep paper, the full annotated 2-minute answer, and the six structural moves that lifted it.
Describe a memorable event in your life.
You should say:
and explain why this event is memorable.
What was on the prep paper when the timer ran out. Thirty-two words. Eight specific anchor points. Section letters down the left. The P·M·Z micro-formula visible in the Highlight notes — Pick / Magnify / Zoom. The brain produces the live English; the paper just keeps the structure intact.
Four sections. Each marker colored by the section it belongs to. Toggle annotations to see the move each marker makes. Click "Hear it spoken" to listen to the full answer. Notice the P·M·Z structure inside the Highlight section — Pick the peak, Magnify with Uncle Tuấn, Zoom out into the shift. No chronology anywhere in the answer.
"I'd like to talk about my older cousin Linh's wedding, two summers ago, in her husband's family's seafood restaurant in Vũng Tàu — which I keep coming back to in my head for reasons I didn't fully understand at the time. I picked this event because out of every family event I've been to, this is the one that I find myself describing whenever anyone asks about Vietnamese weddings — and it's never the parts you'd expect."
"The wedding was on a Saturday in August, at her husband Tuấn-Anh's family's restaurant on the coast — about three hours from Saigon by bus. The whole of my mum's side of the family came down, including my uncle Tuấn — Linh's father — who'd been famously sceptical of the whole match for about a year and a half. The morning of the wedding he was still telling anyone who'd listen that Linh was too stubborn to ever actually go through with it. The part I keep going back to, though, is what happened in the ten seconds before she walked into the reception hall."
What I keep coming back to is those ten seconds right before Linh walked in. My uncle Tuấn — who'd been telling stories all morning about how stubborn she was, how she'd never go through with it — he just stopped, mid-sentence. You could hear the air conditioning. He didn't say another word until she was halfway down the aisle. That was when I realised this wasn't just a wedding to him — it was something else entirely."
Looking back, before that wedding I'd always thought of my Uncle Tuấn as a hard man — gruff, sceptical, hard to read. After those ten seconds of silence I started noticing him completely differently. I think that whole afternoon, I was the only one in the family who saw him cry. So yeah, Linh's wedding. The day I learned that some of the most important moments in family life are about ten seconds long and easy to miss."
No rare vocabulary. No memorised script. Six structural and register moves drawn from across Lesson 9 — each one something you've already met. Listed below.
"I'd like to talk about my older cousin Linh's wedding, two summers ago, in her husband's family's seafood restaurant in Vũng Tàu" instead of "I'd like to talk about a wedding." Twelve words longer, light-years more specific.
Three context facts (Saturday in August, the restaurant, Uncle Tuấn's scepticism) + one forward-pointing setup line. No "first the tea ceremony, then the speeches, then the photos." Develop did exactly what Develop is meant to do — make the peak land harder.
One peak (10 seconds before Linh walked in). Magnified with one specific person doing one specific small thing (Uncle Tuấn stopping mid-sentence + air-con audible + 30 seconds of silence). One zoom-out line ("this wasn't just a wedding to him"). The band-deciding 30 seconds, with not a single "memorable" or "I will remember forever."
"Before that wedding I'd thought of my Uncle Tuấn as a hard man... After those ten seconds I started noticing him completely differently." Specific shift named. No "I will remember it forever." The exact pattern Stage 4 R-move teaches.
"Just stopped, mid-sentence." "You could hear the air conditioning." "Halfway down the aisle." Three Band 7-8 time-dilation phrases from Stage 5 — deployed inside the Magnify section to slow real time down.
"So yeah, Linh's wedding. The day I learned that some of the most important moments in family life are about ten seconds long and easy to miss." Small, dry, personal. Closes the answer with character — not "I will remember it forever."
No "memorable." No "amazing." No "I will remember forever." No "everyone enjoyed the food." No chronology of the wedding. Every move came from structural choices that Lesson 9 taught you — and the answer hits Band 7 because of the structure, not the words. Most Vietnamese candidates lose Band 7 on Event cards because their structure is chronological, not because their grammar is weak. Lesson 9 fixes that — and Linh's wedding answer is the proof.
Two diagnostic questions before we move to the Practice Arena.
Five real Event card excerpts. For each one, identify which of the four chronological-narration failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four chronological narrations. For each, write a 2-3 line Peak Moment Zoom — pick one 10-second peak, magnify it with one specific person, zoom out into meaning. Don't try to be polished — get the shape right. Reveal each model after your attempt.
Three Event cue cards. For each, write the 60-second prep paper — 2-3 shorthand lines per I·D·H·R section. The Highlight section needs three P·M·Z notes specifically (Pick / Magnify / Zoom). The "Reveal model" button shows a Vietnamese-flavoured working example.
The band-deciding 30 seconds. Three Event cue cards — write only the Highlight section for each one (60-90 words). Force yourself to use the Peak Moment Zoom architecture: peak-moment marker → Pick (10-sec moment) → Magnify (specific person + small thing) → Zoom out (one-sentence shift). No chronology.
Pick one of the three cue cards you prepped in Exercise 3. Set a timer for exactly two minutes. Deliver the full Event card answer aloud, using only the prep paper you wrote as your reference. Three rounds — slow pass, natural pace, exam pressure.
Five Event-card exercises done. Here's how it landed.
Your performance across the Event card arena suggests the P·M·Z system is taking root. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 9. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Event-card drill this week.
Forty-five minutes invested in the lesson that fixes the third most common Band 6 ceiling Vietnamese candidates hit on Part 2 — the chronological-play-by-play trap.
Three Part 2 categories done. One more to go. Lesson 10 is the fourth and final Part 2 category — Objects. Same I·D·H·R framework, same 9-stage shape, fourth and final category-specific Highlight tool. Up next: Significance Unwrap — the framework that prevents Object cards from becoming product-description recitations.
Object cue cards are the fourth Part 2 category, and they have their own specific failure mode — the product-description trap. Vietnamese candidates tend to describe objects like online listings: "It is made of leather. It is brown. It is from my mother..." Lesson 10 replaces that with Significance Unwrap, the fourth category-specific Highlight tool. It's the lesson that completes your Part 2 toolkit.
Why describing the object's features minute by minute caps at Band 5-6 — and how Significance Unwrap replaces "it is X and Y and Z" with the layers of meaning the object carries.
The Highlight tool for Objects. Three layers of meaning — surface (what it is), middle (who gave it / when you got it), deep (what it really represents). The unwrap moves outward from object → person → meaning.
80+ Vietnamese-flavoured object phrases — heirloom rings, áo dài, family photos, grandmother's bowl, that one specific phone you can't replace.
Object Reflects need to name what the object anchors — what part of yourself, your past, or your relationships it makes visible that nothing else does. Lesson 10 teaches the object-as-anchor Reflect pattern.
"Three Part 2 categories down — People, Places, and Events. Same I·D·H·R framework, three different category-specific Highlight tools — S·R·M, S·A, and P·M·Z. One more Part 2 category to go. Objects is the fourth and most quietly powerful — the cue card you can prep for at the kitchen table, holding the actual object in your hand. Take a day off, do Lionel's Event-card drill for a week, and the chronological reflex will be gone. Then Objects — and then your full Part 2 toolkit is complete."
Nine badges. The full Part 1 toolkit, the universal Part 2 framework, and the first three Part 2 categories — People, Places, and Events. The chronological-narration reflex has a name now, a diagnostic, and a fix.
People cards made Vietnamese candidates list virtues. Place cards made them sound like brochures. Event cards made them narrate chronologically. Object cards do something different again — they make every candidate produce an online product listing. "It is made of leather. It is brown. It is small. My mother gave it to me. It is very meaningful to me." Five sentences of feature description, zero meaning surfaced. Band 5-6 ceiling, every time.
"Person cards collapse on virtues. Place cards collapse on brochure adjectives. Event cards collapse on chronology. Object cards collapse on product-spec language. Vietnamese candidates describe a precious family heirloom the way they'd write a Marketly listing — material, colour, size, who gave it. The actual significance never gets unwrapped. Band 7 needs the opposite — one specific object, one specific human story, and one layer underneath that nothing else in your life carries the same way. The object is just the wrapping paper. The meaning is what you're actually describing."
Two candidates. Both describing the same family heirloom — a silver bracelet that's been in the family for three generations. Same vocabulary range, same grammar. Watch what happens when one describes the bracelet as a product and the other unwraps its significance. Same bracelet. Only one is at Band 7.
Grammar perfect. Vocabulary correct. Could be the listing on a second-hand jewellery site. The examiner has no idea why this bracelet matters to this candidate any more than any other bracelet would.
One sentence of Surface (thin silver bracelet). Three sentences of Story (worn 60 years, mother's two-year silence, the 25th-birthday handover). Two sentences of Significance (chain of three generations of women, the bracelet as a position not an object). The bracelet becomes a vehicle for everything underneath it.
Same 9-stage shape as Lessons 7, 8, and 9. New focus: applying I·D·H·R to Object cards specifically, plus learning Significance Unwrap — the fourth and final category-specific Highlight tool. After this lesson, your Part 2 toolkit is complete.
You are here.
The 4 product-description failure modes — and why they cap at Band 5-6.
The fourth and final category-specific Highlight tool. Three layers: Surface → Story → Significance. Plus the interactive Significance Builder.
How the universal Part 2 framework adapts specifically for Object cue cards.
80+ Vietnamese-flavoured object phrases beyond "made of X" and "very meaningful."
Áo dài, lì xì, gia phả, bàn thờ — how to anchor each cleanly without over-explaining the cultural context.
Bà nội's silver bracelet — annotated 2-minute Object card with the prep paper that produced it.
Five Object-specific exercises.
Self-assessment, badge, and Lesson 11 preview (Part 3 — abstract discussion).
If you've described a piece of jewellery, a phone, a notebook, or a family heirloom in IELTS practice and felt your answer turning into a product spec — this stage is for you. The collapse isn't a vocabulary problem. It's a cultural pattern where Vietnamese candidates have been trained, often unintentionally, to describe objects through their physical features rather than their significance.
School English curriculum. "What colour is it? What is it made of? Where is it from?" The same five questions every Vietnamese student has answered about every object since primary school. Combined with the daily reality of Marketly listings, Lazada descriptions, and TikTok product reviews — the product-spec template is the most familiar way to talk about objects in any language.
Product descriptions are interchangeable. Material, colour, size — every silver bracelet has them, which means none of those facts say anything about this bracelet specifically. Band 7 rewards what makes this object in this candidate's life irreplaceable — not what makes the object identifiable to a stranger. The descriptors specifically want significance, not spec.
This is the key shift. Object cards aren't asking you to describe the object as if you were listing it for sale. They're asking you to describe an object in a way that lets the examiner feel why this specific object matters to you specifically. The product version fails because it could be the Marketly listing — facts about the bracelet, the watch, the áo dài, that anyone selling one would write. Your version can't.
Listen to enough Object card answers and the failures cluster into four distinct patterns. Same root cause behind each — defaulting to product-spec language instead of unwrapping significance — but each has a different surface symptom. Recognising your specific shape makes the fix more targeted.
Material, colour, size, shape, age. Pure product spec. Every sentence describes a different physical feature. No human story, no significance — just the listing copy. The dominant failure mode for Vietnamese candidates.
A roll call of how the object travelled through the family. Names without stories. The chain of ownership gets named, but the human meaning attached to each handover never surfaces. Sounds like reading a will.
Especially common with heirlooms and gifts. The candidate describes how they look after the object instead of what it means. Storage, cleaning, conditions of use — pure maintenance log. The meaning hides behind the care routine.
After the product description, the candidate adds a generic emotional close. "It is very precious to me." "It is very meaningful." "I love it very much." "I will keep it forever." Feature list + content-free emotion = nothing the examiner can grade as Band 7.
A translation guide between the two registers. Vietnamese product-description language isn't wrong inside its context — it's the right register for buying, selling, or showing-and-telling. It's the wrong register for an IELTS Speaking test, which rewards unwrapping significance. Same object described in both modes makes the difference visible. Here's a small embroidered handkerchief from bà ngoại, in both registers.
Describes the object through its physical features and origin chain. Surface coverage only. The wrapping paper is the description.
Picks the layers underneath. Brief surface, then unwraps the human story, then the meaning underneath. The object is just the door.
Whenever you catch yourself describing the material, the colour, or the size — pause and ask: what's underneath this surface? Who made it? Who carried it? Who gave it to whom, and what did that handover mean? The answer to that question is the IELTS sentence. The product description is what the Marketly seller had.
This doesn't mean physical detail is forbidden. It means surface detail alone is empty — it needs one specific story attached. "A small yellow handkerchief embroidered with red flowers — bà ngoại's, made the year before she met my grandfather." The surface opens, the human path follows.
Sometimes candidates resist this stage too. "But the examiner asked me to describe the object." Worth addressing directly: nothing in Lesson 10 says the description isn't allowed. It says the description alone is incomplete. The fix is a layer-change, not a content-change.
"Of course the bracelet is made of silver. The handkerchief is yellow with red flowers. The áo dài is made of silk. None of that is wrong. It's just the wrapping paper. The IELTS test isn't asking you to describe the wrapping. It's asking you to describe the gift inside — the human story, the meaning, what this specific object carries that nothing else in your life carries the same way. Less surface, more unwrap. One sentence for what it is. The rest of the time for what it's for."
Product description lives on the outer layer. "It is silver. It is small. It is old. It is from my grandmother." Anyone holding the bracelet could describe it the same way — which is the problem.
Your significance unwrap moves inward through the layers. "Bà nội wore it every single day for sixty years. My mother kept it untouched for two years after she passed. On my twenty-fifth birthday she put it on my wrist without saying a word." Only the person who lived through that handover can say it. The bracelet's specifications are unchanged. The candidate's relationship to the bracelet is everything.
Same object. Different layer. The Band 7 examiner is rewarding the inner layers, not the outer wrap. Peel back. Unwrap. Get to what it actually carries.
When you describe a bracelet through three generations of women, or a handkerchief through the year before your grandmother met your grandfather, or an áo dài through the morning your mother stopped fitting into hers — you're not honouring the object less than the feature-list version. You're honouring it more, in a different way. The examiner hears a candidate who knows the object well enough to have a peeled-back understanding of it — which is what real possession looks like. That depth is the respect. The IELTS test just rewards it in layers instead of in specs.
Two questions before we move to the Significance Unwrap framework in Stage 3.
Three layers. Surface. Story. Significance. That's the entire framework. Significance Unwrap (S·U) is the fourth and final category-specific Highlight tool — same role for Object cards that S·R·M plays for People, S·A plays for Places, and P·M·Z plays for Events. It sits inside the Highlight section of I·D·H·R, and replaces feature-listing with three concentric layers of meaning, peeled back in order.
The first layer is the briefest. One sentence describing what the object physically is. The whole Surface layer should be smaller than the Story layer and the Significance layer combined. The discipline is brevity — give the examiner just enough physical detail to picture the object, then immediately move on.
The second layer is where the human story lives. Who made it, who carried it, who gave it to whom. This layer is where the object stops being a product and starts being a vehicle. Two to three sentences. Not the full family history — just the specific handover or origin that makes this object yours specifically.
"If you remember nothing else from Stage 3, remember this: name the specific moment the object changed hands. Not "she gave it to me." But "she put it on my wrist without saying a word." Or "she left it on the kitchen table on the morning she moved out." The moment of handover is where the meaning lives. Frame it like a scene, not like a transaction."
The third and deepest layer. After the Surface (5 seconds) and the Story (15-20 seconds), you have 5-10 seconds left for the Significance — one or two sentences naming what the object actually carries. Not "it is very meaningful." But what specifically makes this object irreplaceable in your life.
One full Significance Unwrap in one block — Surface, Story, Significance. Notice how the three layers layer: Surface is one line, Story is the bulk of the Highlight with the human handover, Significance is one closing redefining line.
"It's a thin silver bracelet with a small engraving worn almost smooth on the inside — bà nội's. She wore it every single day for sixty years. When she passed away, my mother kept it in her dresser for two years, untouched. On my twenty-fifth birthday she put it on my wrist without saying a word. The engraving is my great-grandmother's name. It doesn't feel like an object. It feels like a position I'm holding for the next person."
One sentence Surface. Two to three sentences Story. One to two sentences Significance. The hardest discipline is the proportions — most candidates spend most of the Highlight on Surface (Feature List mode) and reach Significance with five seconds left, forced into generic sentiment. Plan the layers proportionally: 5 sec Surface + 15-20 sec Story + 5-10 sec Significance = 30 sec Highlight.
Four picks. Pick an object. Pick the Surface (one sentence). Pick the Story (the human handover). Pick the Significance (the deepest layer). Watch the full S·U Highlight assemble in real time. Each object has different layer options — try at least two different objects to see the range.
[pick the object first]
~30 seconds of Highlight built from four picks. Notice how the three layers layer: Surface sets up the physical (5 sec), Story unwraps the human path (15-20 sec), Significance closes with what it actually carries (5-10 sec). This is what lifts an Object Highlight from Band 6 to Band 7.
Two questions before we move to the Objects I·D·H·R micro-template in Stage 4.
The universal Part 2 framework from Lesson 6 still applies. The People-specific, Place-specific, and Event-specific moves don't. Object cards have their own four Object-specific moves — small adjustments to each section that turn a generic I·D·H·R answer into a Band 7 Object description. Stage 4 walks through those adjustments section by section.
Name the specific object + whose it was + one anchoring fact. Not "a piece of jewellery" — "my grandmother's silver bracelet."
Context for the object — but not product description. When you got it. How it came into your life. Your relationship with it now.
Significance Unwrap lives here. Surface · Story · Significance.
The object-as-anchor. What this specific object anchors that nothing else in your life does.
Most candidates open with the category: "I'd like to talk about a piece of jewellery." The Band 7 move is to open with a specific object — whose, what kind, with one anchoring detail. The same kind of specificity move as L7-9: name what makes this object yours specifically.
Develop is where the Feature List lurks. Thirty seconds to fill, Highlight not yet reached, brain reaches for material/colour/size. Resist. Develop builds context — when the object entered your life, how it came to you, what part of your routine it now lives in. Two or three context lines, no product spec.
"I got it on my 25th birthday — three years ago — though it had been in our family for at least three generations before that. It lives on my left wrist now, basically permanently. I take it off to swim and to shower, and that's it. It's become part of how I dress in the morning without thinking — and increasingly, part of how I think about the women in my family without saying it out loud. What I find myself coming back to, though, is the morning my mother actually put it on me."
Notice what's missing: no material/colour/size. No "it is silver, it is small." Three context facts (when, where it lives, the relationship-building forward line) + one pointing-forward setup. Develop did exactly what Develop is meant to do — built context for the Highlight. The product spec is invisible. The candidate's relationship with this specific bracelet is everywhere.
This is where Stage 3 plugs in. The Highlight section of an Object cue card is the Significance Unwrap you built in the Stage 3 Builder. Same architecture: opening with a Highlight marker from Lesson 6, then Surface (5 sec), Story (15-20 sec), Significance (5-10 sec). Thirty seconds. Band-deciding.
The marker tells the examiner the band-deciding section has started. The Surface takes one sentence — thin silver bracelet, engraving worn smooth. The Story does the heaviest lifting: bà nội wearing it 60 years, mother's two-year silence, the wordless 25th-birthday handover, the great-grandmother's engraving. The Significance closes with the redefining line — "a position I'm holding for the next person" — which names what the bracelet actually carries. Thirty seconds of Band 7 content from the framework you already built.
Most candidates Reflect lazily on Object cards. "It is very precious." "I will keep it forever." Statements that fit any object anywhere. The Band 7 move is making Reflect about what the object anchors — what part of yourself, your past, or your relationships this specific object makes visible that nothing else in your life does.
Name the specific object in Introduce. Use Develop for context, not catalogue. Use Significance Unwrap in Highlight. Make Reflect about what the object anchors. Four small moves applied across the I·D·H·R framework you already own — and an Object card answer goes from generic-Band-6 to specific-Band-7.
Same architecture as the L7 character bank, the L8 place bank, and the L9 event bank — applied to objects. Eighty Vietnamese-flavoured Object phrases that aren't "very precious" or "I will keep it forever." Pick three from each category and you've replaced the product-description reflex with vocabulary the examiner actually rewards.
Phrases for naming the specific object rather than the category. "Bà nội's silver bracelet" not "a piece of jewellery." Used in Introduce.
Ready-made one-sentence physical descriptions that pivot straight into Story. "A thin silver bracelet with an engraving worn almost smooth on the inside — bà nội's." Used in Highlight Layer 1.
Phrases for naming the specific moment the object changed hands. "She put it on my wrist without saying a word." Used in Highlight Layer 2 (Story).
Phrases for the deepest layer. "It feels like a position I'm holding for the next person." Used in Highlight Layer 3 (Significance) and Reflect.
Twenty phrases for naming the specific object rather than the category. These are your Introduce openers. The structure: whose it is + what kind + one distinguishing detail. Pick the structure that fits the object you've chosen — bracelet, watch, photograph, handkerchief, áo dài.
Twenty ready-made one-sentence physical descriptions. Each one packs the description into a single phrase and immediately pivots into Story. Use them as your Surface layer in the Highlight — five seconds, one sentence, then move on.
Twenty phrases for the Story layer — the specific moment the object changed hands. These are the highest-leverage phrases in the entire bank because they replace "she gave it to me" with a frameable scene. Pick the structure that fits your specific handover.
Twenty phrases for the Significance layer and the Reflect section. These name what the object specifically anchors — what nothing else in your life carries the same way. Most candidates lose Band 7 here by saying "very precious." These phrases lock you into specifics.
Across the four categories — specific-object naming, surface, handover-moment, object-as-anchor — pick three favourites per category. Twelve phrases total that you can deploy on any Object cue card. Don't try to memorise eighty. Pick twelve, rehearse for a week, and they'll become reflexes. Twelve phrases × four categories = a full Band 7 Object card vocabulary.
Lessons 7, 8, and 9 each had their own anchoring problem — kinship terms, place names, event types. Lesson 10 has the Vietnamese-object problem: objects ranging from universally recognisable (watches, photographs) to highly culturally specific (lư hương, mâm cúng, trầu cau). Examiners have seen watches. None will have seen a lư hương. The Band 7 candidate knows which object needs anchoring and which doesn't, and handles each cleanly.
"I'd like to talk about a watch, which is a small mechanical device you wear on your wrist for telling the time, made up of a face, hands, and a strap, which I inherited from..."
"I'd like to talk about my grandmother's lư hương, which has been on our family altar for generations..."
Three constructions — one for each tier that needs anchoring. Each is a sentence-level template you can plug any Vietnamese object into. Light anchor for tier-2, full anchor for tier-3, anchor + ritual function for tier-4. The patterns are identical in structure to the kinship bridges of L7, the place-name anchors of L8, and the event anchors of L9 — same architecture, new content.
Name + literal translation: "Áo dài — the traditional Vietnamese dress." Cleanest for tier-2 objects. Examiner gets the category in five words.
Name + cultural function: "Gia phả — the handwritten genealogy book Vietnamese families keep." Best for tier-3 objects. Gives the examiner what the object does in the culture.
Name + ritual function + cultural weight: "Lư hương — the small bronze incense burner at the centre of every Vietnamese family altar, where you light incense to your ancestors." Best for tier-4 objects. The Band 8 anchor — gives the examiner the function and a feel for the object's role in daily Vietnamese life.
Audio-clickable phrasings for every major Vietnamese object category a candidate is likely to choose. Each is a working anchor — Vietnamese name with the English clarifier ready to deploy. Use them as templates: keep the structure, swap in the specific object you actually own.
A small but band-relevant question: do you say "áo dài" with the full tones, or do you anglicise it? Answer: say it the Vietnamese way naturally, but don't make a thing of it. The examiner isn't testing your pronunciation of Vietnamese — but consistent, comfortable Vietnamese object pronunciation signals an authentic register. Move on quickly.
Two questions before we move to the worked example in Stage 7.
Six stages of theory — the product-description trap, the cultural diagnosis, the Significance Unwrap framework, the Objects I·D·H·R micro-template, object vocabulary, and Vietnamese object anchoring. Time to see all six built into a single Band 7 answer. Over the next four screens, you'll see the cue card, the prep paper, the full annotated 2-minute answer, and the six structural moves that lifted it.
Describe an object that is important to you.
You should say:
and explain why this object is important to you.
What was on the prep paper when the timer ran out. Thirty-four words. Eight specific anchor points. Section letters down the left. The S·U micro-formula visible in the Highlight notes — Surface / Story / Significance. The brain produces the live English; the paper just keeps the structure intact.
Four sections. Each marker colored by the section it belongs to. Toggle annotations to see the move each marker makes. Click "Hear it spoken" to listen to the full answer. Notice the S·U structure inside the Highlight section — Surface (5 sec) → Story (20 sec) → Significance (5 sec). No product description anywhere in the answer.
"I'd like to talk about a thin silver bracelet that belonged to my paternal grandmother — bà nội — with my great-grandmother's name engraved on the inside. I picked this one because out of everything I could have inherited from my family, this is the object I think about most — and the one I've worked the hardest to actually understand."
"I got it on my 25th birthday — three years ago next April — though it had been in our family for at least three generations before that. It lives on my left wrist now, basically permanently. I take it off to swim and to shower, and that's it. It's become part of how I dress in the morning without thinking — and increasingly, part of how I think about the women in my family without saying it out loud. What I find myself coming back to, though, is the bracelet itself — what it is, where it's been, and what it's still doing in our family."
What I keep coming back to is the bracelet itself. A thin silver bracelet with an engraving worn almost smooth on the inside — bà nội's. She wore it every single day for sixty years. When she passed, my mother kept it in her dresser, untouched, for two years. On my 25th birthday she put it on my wrist without saying a word. The engraving on the inside is my great-grandmother's name. It doesn't feel like an object. It feels like a position I'm holding for the next person."
Looking back, what the bracelet actually anchors is a version of bà nội I never knew when she was alive — the daily-wearing, never-removing version who wore it through three wars and sixty years of normal mornings. If I lost it tomorrow, that version of her would have nothing left to attach to in my daily life — and I think that's why I take it off as little as I can manage. So yeah, that's what the bracelet actually is. Three generations of women, and a fourth I haven't met yet."
No rare vocabulary. No memorised script. Six structural and register moves drawn from across Lesson 10 — each one something you've already met. Listed below.
"A thin silver bracelet that belonged to my paternal grandmother — bà nội — with my great-grandmother's name engraved on the inside" instead of "a piece of jewellery." Twelve words longer, light-years more specific.
25th birthday three years ago + lives on left wrist + part of dressing without thinking + part of how she thinks about the women in her family. Four context lines + one forward-pointing setup. No material, colour, size. Develop did exactly what Develop is meant to do — built context for the Highlight.
One sentence Surface (thin silver, engraving worn smooth). Three sentences Story (60 years daily wear, mother's two-year silence, wordless 25th-birthday handover, great-grandmother's engraving). One sentence Significance ("a position I'm holding for the next person"). The band-deciding 30 seconds, with not a single "made of" or "I will keep forever."
"On my 25th birthday she put it on my wrist without saying a word." One specific moment of object transfer, framed as a scene. The Story layer's highest-leverage sentence — what Lionel calls the "specific moment the object changed hands." Frame it like a scene, not a transaction.
"What the bracelet actually anchors is a version of bà nội I never knew when she was alive..." Specific anchor named. No "I will keep it forever." The exact pattern Stage 4 R-move teaches — what would disappear from her life if the object did.
"So yeah, that's what the bracelet actually is. Three generations of women, and a fourth I haven't met yet." Small, quiet, specific. Closes the answer with character — not "I will treasure it always."
No "made of." No "very precious." No "I will keep it forever." No material/colour/size catalogue. No "this is a very important Vietnamese tradition." Every move came from structural choices that Lesson 10 taught you — and the answer hits Band 7 because of the structure, not the words. Most Vietnamese candidates lose Band 7 on Object cards because their structure is product-spec, not because their grammar is weak. Lesson 10 fixes that — and bà nội's silver bracelet is the proof.
Two diagnostic questions before we move to the Practice Arena.
Five real Object card excerpts. For each one, identify which of the four product-description failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four feature-list excerpts. For each, write a 3-line Significance Unwrap — one sentence of Surface (what it physically is), 2-3 sentences of Story (the human handover), and one line of Significance (what it actually carries). Don't try to be polished — get the shape right. Reveal each model after your attempt.
Three Object cue cards. For each, write the 60-second prep paper — 2-3 shorthand lines per I·D·H·R section. The Highlight section needs three S·U notes specifically (Surface / Story / Significance, marked with 🔹 📖 💎). The "Reveal model" button shows a Vietnamese-flavoured working example.
The band-deciding 30 seconds. Three Object cue cards — write only the Highlight section for each one (60-90 words). Force yourself to use the Significance Unwrap architecture: peak-moment marker → Surface (one sentence) → Story (2-3 sentences with the human handover) → Significance (one closing redefining line). No product description.
Pick one of the three cue cards you prepped in Exercise 3. Hold the actual object in your hand if you have it — Object cards reward physical proximity. Set a timer for exactly two minutes. Deliver the full Object card answer aloud, using only the prep paper you wrote as your reference. Three rounds — slow pass, natural pace, exam pressure.
Five Object-card exercises done. Here's how it landed.
Your performance across the Object card arena suggests the S·U system is taking root. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 10. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Object-card drill this week.
Forty-five minutes invested in the lesson that fixes the fourth and final most common Band 6 ceiling Vietnamese candidates hit on Part 2 — the product-description trap.
Ten lessons done. Part 1 toolkit complete. Part 2 toolkit complete. The full middle band of the test is now structurally yours. Lesson 11 begins Part 3 — the abstract discussion that follows your 2-minute Part 2 answer. Different register, different mental gear. Where Part 2 is concrete and personal ("describe X"), Part 3 is abstract and societal ("what do you think about X in general"). The first Part 3 lesson teaches the framework for the whole module.
Part 3 is the section where most Vietnamese candidates lose Band 7. After your 2-minute Part 2 answer, the examiner pivots into 4-5 minutes of abstract questions on the same topic. Where Part 2 asked you to describe a specific person, place, event, or object, Part 3 asks you what you think about people, places, events, or objects in general. The shift catches most candidates flat-footed. Lesson 11 teaches the framework that makes it manageable.
Concrete-specific ("my grandmother's bracelet") to abstract-societal ("how have attitudes toward family heirlooms changed in Vietnamese society?"). Why this is the section that catches Band 6 candidates and why even Band 7 candidates often slip back.
Opinion · Compare-and-contrast · Cause-effect · Speculation. Four distinct question shapes, each with its own answer architecture. Recognising the question type in two seconds is the band-jump.
Position · Evidence · Extension. The first Part 3 micro-framework — used for opinion questions specifically. State your position cleanly, support it with one concrete example, extend with one broader implication. The skeleton that prevents drift.
Where Part 2 candidates lean on Vietnamese cultural specifics, Part 3 demands generalisation across cultures. How to use a Vietnamese example without making the whole answer feel parochial — and when to deliberately invoke a Vietnamese contrast as a Band 8 move.
"Halfway. Take a moment. The full Part 1 toolkit, the universal Part 2 framework, and all four Part 2 categories — People, Places, Events, Objects. Four category-specific Highlight tools: S·R·M, S·A, P·M·Z, S·U. Four Vietnamese cultural-anchoring layers. Four canonical worked examples — Bác Hùng, Đà Lạt, Linh's wedding, bà nội's bracelet. That's the entire middle band of the IELTS Speaking test, structurally complete in your hands. Part 3 is a different register but the same person doing the talking. Do Lionel's Object-card drill for a week, take a real day off, and then we start Part 3."
Ten badges. The Part 1 toolkit, the universal Part 2 framework, and the full set of four Part 2 categories — People, Places, Events, and Objects. The product-description reflex has a name now, a diagnostic, and a fix. Halfway through the program, and structurally through the middle band of the IELTS Speaking test.
The section of the IELTS Speaking test where most Vietnamese candidates quietly lose Band 7. After your 2-minute Part 2 answer about a specific person, place, event, or object — the examiner pivots into 4-5 minutes of abstract follow-up questions on the broader topic. The register changes completely. The mental gear changes completely. Most candidates either drift back into Part 2 mode (personal anecdotes) or collapse into one-sentence yes/no answers. Both cap at Band 6. The structural fix is in the next nine stages.
"Part 3 is where the examiner asks you to think, not just describe. The questions move from concrete to abstract, from personal to societal, from narrative to argumentative. Vietnamese candidates have been trained for years to describe well — the Lesson 10 candidate could describe bà nội's bracelet beautifully. But when the examiner asks 'how have attitudes toward family heirlooms changed in recent generations?' — the same candidate often goes back to bà nội's bracelet, because that's the version of the question their brain already has an answer for. The fix isn't more vocabulary. It's a register shift: state a position, support it with one example, extend it with one broader implication. Three sentences, one minute, Band 7. The P·E·E framework is the skeleton."
Two candidates. Both being asked the same Part 3 question right after Part 2 about bà nội's bracelet. Same vocabulary range, same grammar, same Vietnamese background. Watch what happens when one stays in Part 2 mode (personal anecdote) and the other shifts into Part 3 mode (abstract position with evidence). Same question. Only one is at Band 7.
"How have attitudes toward family heirlooms changed in recent generations in your country?"
The candidate has stayed in Part 2 mode — recapping their own family story. The examiner asked about generational attitudes in the country; the answer is about one bracelet in one family. Grammar is fine. Register is wrong. Examiner cannot grade abstract thinking from a personal anecdote.
Position-Evidence-Extension visible. Sentence 1 is the position. Sentences 2-3 are the evidence (a generational contrast — older generation vs candidate's generation). Sentence 4 is the extension (counter-trend among a specific demographic). Abstract pattern + counter-trend = Band 7+. The candidate's own family doesn't appear once — they could, but here the candidate has chosen to generalise instead, which is the correct Part 3 move.
Same 9-stage shape as Lessons 7-10, applied to Part 3 instead of Part 2. New module focus: abstract discussion register, the P·E·E framework, and the four Part 3 question types. After this lesson, the foundational Part 3 framework is yours, ready to specialise in Lessons 12-15.
You are here.
The 4 Part 3 failure modes — Personal-Anecdote Drift, Yes/No Collapse, Confused Position, Opinion-Without-Evidence — and why they cap at Band 5-6.
The first Part 3 micro-framework. Position · Evidence · Extension. Three sentences, one minute, Band 7. Plus the interactive Opinion Builder.
Opinion · Compare-and-Contrast · Cause-and-Effect · Speculation. Four distinct question shapes, each with its own answer architecture.
80+ Part 3-appropriate phrases — none of them "I think" or "in my opinion." The position-stating, evidence-introducing, and extension-launching phrases that signal Band 7.
How to use a Vietnamese example without making the answer feel parochial — and when to deliberately invoke a Vietnamese contrast as a Band 8 move.
Five Part 3 questions answered to Band 7 — building on the bà nội's bracelet Part 2 from Lesson 10. Annotated answers with P·E·E structure visible.
Five Part 3-specific exercises.
Self-assessment, badge, and Lesson 12 preview (Compare-and-Contrast questions).
If you've been asked an abstract Part 3 question and felt your brain reaching for the Part 2 story you just finished telling — this stage is for you. The collapse isn't a vocabulary problem and it isn't an opinions problem. It's a register-switching problem: Vietnamese candidates have been trained for years to describe specific people, places, events, and objects, with very little practice generating abstract societal positions in English on demand.
Vietnamese school English curriculum. Years of "describe your family / your hometown / your favourite festival / your most important object." Combined with the Lessons 1-10 of this very program — all of which trained you to describe brilliantly. The descriptive register is the most familiar way to talk about anything in English, and when Part 3 hits, the brain defaults to what it already knows how to do.
Part 3 is graded specifically on abstract thinking, position-taking, and the ability to extend an idea beyond personal experience. A candidate who describes their own family's bracelet for the third time when asked about "generational attitudes" demonstrates excellent descriptive English — which is graded in Part 2. In Part 3, the descriptors specifically want generalisation, hypothesis, comparison, and speculation.
This is the key shift. Part 2 asked you to make the examiner see a specific person, place, event, or object in your life. Part 3 asks you to make the examiner see how you think about that category of thing across society, generations, and cultures. The personal story is no longer the answer — it's at most one piece of evidence inside a larger argument. Your job stops being narrator and starts being analyst.
Listen to enough Part 3 answers and the failures cluster into four distinct patterns. Same root cause behind each — staying in descriptive mode when the question wants argumentative mode — but each has a different surface symptom. Recognising your specific shape makes the fix more targeted.
The Part 2 reflex. The candidate hears the abstract question and immediately reaches for the specific story they just finished telling. "In my family..." / "I have a friend who..." / "When I was younger..." The story replaces the position. The dominant failure mode for candidates fresh out of Part 2.
The opposite failure mode. The candidate panics at the abstraction, freezes, and gives a one-sentence answer hoping the examiner moves on. "Yes, I think so." "No, not really." "Yes, definitely." The collapse signals "I don't know how to extend this." Most Part 3 questions expect 3-4 sentences minimum.
The candidate tries to be balanced and ends up contradicting themselves. "Attitudes have changed a lot. But also they haven't really changed. Maybe some have changed. But it depends." The position becomes invisible. Balance is a Band 7 move; incoherence is a Band 5 collapse. The two look similar from the inside, different from the outside.
The candidate states a clear position — and then repeats it three different ways instead of supporting it. "Family heirlooms are important. They are meaningful. They represent the family. People should keep them. So they're important." The position is fine. The evidence is missing. Part 3 specifically rewards P + E together, not P alone.
A translation guide between the two registers. Descriptive English is what Lessons 1-10 built — and it's exactly the right register for Part 2. It's the wrong register for Part 3, which rewards argumentative English instead. Same Vietnamese candidate answering the same Part 3 question in both modes makes the difference visible. Here's "Why do you think some people keep objects that have no practical use?" in both modes.
Reaches for a specific object, a specific person, a specific moment. Tells what happened. Lets the story imply the answer. Right register for Part 2, wrong register for Part 3.
Opens with a position. Supports it with an example (which can be personal, but functions as evidence inside an argument). Closes with an extension or implication. Same Vietnamese candidate, different gear.
Whenever you catch yourself opening a Part 3 answer with "in my family" or "I have a friend who" — pause and ask: what's the position underneath the story? The story is fine as evidence, but it can't be the entire answer. The argument has to come first, then the story serves the argument.
This doesn't mean personal examples are forbidden in Part 3. It means personal examples alone are incomplete — they need a stated position above them and an extension below them. "Position. Then evidence — which can include your bracelet. Then extension." That's the architecture.
Sometimes candidates resist this stage too. "But my Part 2 answer worked so well — why can't I just keep going with that?" Worth addressing directly: Lesson 11 doesn't say your Part 2 stories are wrong. It says they no longer belong at the top of the answer. The fix is a position-change, not a content-change.
"Of course your bà nội's bracelet matters. Of course your family's stories are good evidence. None of that is wrong. But the moment the examiner moves into Part 3, the candidate's lived experience stops being the answer and starts being the evidence. State a position first. Use your bracelet to support it. Then extend with one broader implication. The bracelet hasn't disappeared — it's just stopped being the headline. Argumentative English wears the story inside the position, not instead of it. Three sentences, one minute, Band 7."
Descriptive mode lives in narrative. "My grandmother had a bracelet. She wore it for sixty years. She gave it to my mother. My mother gave it to me." The story is the answer. Right for Part 2, wrong for Part 3.
Argumentative mode opens with a claim and uses narrative as support. "I think people keep practically useless objects because those objects do work that has nothing to do with utility. The bracelet that's been in my family for three generations doesn't function as jewellery — it functions as a small material record of who we've been." The position frames the story. Same content, different shape, different band.
Same candidate. Same Vietnamese background. Same bracelet. The Band 7 examiner is rewarding the argument that contains the story, not the story alone. Position first. Story inside. Extension after.
When you take your grandmother's bracelet and use it as evidence inside a broader claim about why families keep practically useless objects — you're not honouring it less than the personal-anecdote version. You're honouring it more, in a different way. The examiner hears a candidate who can see their own life as part of a larger pattern — which is what real thinking looks like. That stepping-back is the respect. The IELTS test just rewards it in arguments instead of in stories.
Two questions before we move to the P·E·E framework in Stage 3.
Three sentences. Position. Evidence. Extension. That's the entire framework. P·E·E is the first Part 3 micro-framework — used specifically for opinion questions (the most common Part 3 type, roughly 40% of the questions you'll face). Same role for Part 3 that I·D·H·R played for Part 2: a skeletal structure that keeps your answer cohesive while still leaving room for everything you actually want to say.
The first layer is the most important. One sentence stating your clear claim about the question. Not "I think it depends" — but "I think attitudes have shifted quite significantly over the last twenty years." The Position has to be takeable — the examiner has to be able to point at it and say "that's the candidate's claim." Hedge a little (Part 3 rewards nuance), but commit.
The second layer is where your Part 2 stories finally come back — but in a new role. Where Part 2 made the story the answer, Part 3 makes the story the evidence. One concrete example that supports your Position. Vietnamese-specific examples are welcome here; in fact, they're often the strongest evidence because they're things only you can speak to with first-person credibility. One to two sentences. Not the whole story — just the relevant slice.
"If you remember nothing else from Stage 3, remember this: the Vietnamese example becomes more powerful in Part 3, not less. In Part 2 the bracelet was the whole story. In Part 3 the bracelet is a one-sentence proof inside a societal argument — and the contrast between the personal and the societal is where Band 7 lives. Use your life as evidence for a position that's bigger than your life."
The third layer is the Band 7 differentiator. After Position and Evidence, you have 10-15 seconds left for one Extension sentence — a counter-trend, an implication, a paradox, or a cross-cultural observation. Most candidates stop at P + E and wonder why they're stuck at Band 6. The Extension is the move that signals "this candidate can think beyond the obvious answer."
One full P·E·E in one block — Position, Evidence, Extension. Notice how the three layers layer: Position is one sentence stating the claim cleanly, Evidence is one to two sentences supporting it with a specific instance, Extension is one closing sentence complicating or extending the picture.
"I think attitudes have shifted quite significantly in Vietnam over the last twenty years — for older generations, an heirloom was a position-holding object that came with responsibility, but for my generation, heirlooms feel increasingly optional. You can see this in even very small objects — the silver bracelet in my family has been carrying three generations of women silently for decades, and none of us would have thought of it as 'an heirloom' in any official sense. That said, there's actually a counter-trend among urban professionals in their thirties, who are starting to ask their parents specifically for heirlooms again — for reasons their grandparents wouldn't have recognised."
One sentence Position. One to two sentences Evidence. One sentence Extension. Total: 45-60 seconds, three to four sentences. The hardest discipline is adding the Extension — most candidates stop at P+E and wonder why they're stuck at Band 6. 10 sec Position + 20-25 sec Evidence + 10-15 sec Extension = 45-60 sec Part 3 answer.
Four picks. Pick a Part 3 question. Pick a Position. Pick the Evidence. Pick the Extension. Watch the full P·E·E answer assemble in real time. Each question has different layer options — try at least two different questions to see how the same architecture handles different topics.
[pick a Part 3 question first]
~45-60 seconds of Part 3 answer built from four picks. Notice how the three layers layer: Position takes a stake (10 sec), Evidence supports it with one instance (20-25 sec), Extension complicates with a counter-trend or implication (10-15 sec). This is what lifts a Part 3 answer from Band 6 to Band 7.
Two questions before we move to the four Part 3 question types in Stage 4.
P·E·E from Stage 3 is the foundational structure for opinion questions — the most common Part 3 type. But Part 3 actually has four distinct question types, each requiring a slightly different answer shape. Recognising the question type in the first two seconds is the band-jump: it tells you which architecture to deploy before you've even said your first word. Lessons 12-15 deep-dive each type. This stage gives you the recognition map.
"Do you think...?" / "What's your view on...?" / "How important is...?" The most common Part 3 type.
"What are the differences between...?" / "How is X different from Y?" Two-sided structural questions.
"Why do you think...?" / "What causes...?" / "What are the effects of...?" Mechanism questions.
"What if...?" / "How will X change in the future?" / "Imagine if..." Hypothesis questions.
The most common Part 3 question type. Roughly 40% of the questions you'll face. Recognisable by the verbs "think," "view," "agree," "important," "should." The architecture is P·E·E — the framework you built in Stage 3. This screen gives you the recognition patterns and a small bank of opener phrases.
Roughly 25% of Part 3 questions. Recognisable by "differences," "compare," "how is X different from Y?" Requires a two-sided answer with a clear basis of comparison. Full architecture in Lesson 12 (X·Y·Z framework — eXtract dimension, compare Y values, Zoom into one telling detail). This screen gives you the recognition pattern and a one-paragraph preview.
"I think the most interesting difference is in how each generation decides what's worth keeping (X — the dimension). My grandmother's generation kept objects almost automatically — anything that had survived was assumed to be worth preserving (Y₁). My generation, by contrast, has to actively choose, because we have access to so much more stuff that the default isn't 'keep' anymore — it's 'discard unless there's a reason' (Y₂). You can see this in something as small as old photographs — my grandmother kept every single one in shoeboxes; I keep maybe 5% of mine, all of them deliberately, all of them tagged (Z — the zoom)."
Two more question types to recognise. Cause-Effect (~20%) asks for the mechanism behind something. Speculation (~15%) asks you to imagine an alternative or future. Both have their own architectures — C·M·E and I·F·B — taught in full in Lessons 13 and 14. Recognition patterns below.
Lesson 12 will deep-dive Compare-and-Contrast (X·Y·Z). Lesson 13 will deep-dive Cause-and-Effect (C·M·E). Lesson 14 will deep-dive Speculation (I·F·B). Lesson 15 will teach the meta-skill of recognising hybrid questions that combine multiple types. For now, the recognition map alone is worth a lot — it tells you which architecture to deploy in two seconds.
Four Part 3 questions. Identify the question type and which framework to deploy. Two seconds per question — the same speed you'll need in the real exam.
Same architecture as the L7-10 vocabulary banks, applied to Part 3 opinion construction. Eighty Part 3-appropriate phrases that aren't "I think" or "in my opinion." Pick three from each category and you've replaced the descriptive register with the argumentative phrases the examiner actually rewards.
Phrases for opening with a clear claim. Not "I think" — "I'd argue," "My sense is," "What I find most interesting is." The Part 3 opener bank.
Phrases that connect your example to your Position. "You can see this in..." / "What I notice in my own family is..." / "Take an object like..."
Phrases that signal the Band 7 closer. "That said..." / "What's actually counterintuitive is..." / "Which I think implies..."
Phrases that signal nuance without losing the position. "For the most part..." / "In some ways more than others..." / "Roughly speaking..."
Twenty phrases for opening with a clear claim. Avoid the bare "I think" — varied openers signal a candidate with range. Pick three favourites and use them across the 4-5 Part 3 questions you'll face, rotating so you don't sound repetitive.
Twenty phrases for connecting your example to your Position. The connector tells the examiner "this is the evidence move now." Pick three favourites and use them when introducing your specific example.
Twenty phrases for launching the Band 7 closer. Counter-trends, implications, paradoxes, cross-cultural observations. These are the highest-leverage phrases in the entire bank because they signal "this candidate can think beyond the obvious answer." Pick three favourites and drill them.
Twenty phrases for signalling nuance without losing your position. Hedge once, take a clear stake, and move on. Two hedges in one sentence usually means you've stopped taking a position at all (Confused Position trap from Stage 2). Use these strategically.
Across the four categories — Position, Evidence, Extension, Hedging — pick three favourites per category. Twelve phrases total that you can deploy on any Part 3 opinion question. Don't try to memorise eighty. Pick twelve, rehearse for a week, and they'll become reflexes. Twelve phrases × four categories = a full Band 7 Part 3 vocabulary.
Lessons 7-10 each had a Vietnamese cultural anchoring problem — kinship terms, place names, event names, object names. Lesson 11 has a different version: the Vietnamese examples problem. Part 3 questions usually ask about "society," "people in general," "your country" — and the candidate has to decide whether to invoke a Vietnamese example, when, and how much. The wrong move makes the answer feel parochial. The right move makes it land as Band 8.
"In Vietnam we have this tradition called đám giỗ, which is a death-anniversary ceremony. And we also have bàn thờ, which is the family altar. And gia phả, which is the genealogy book. And tảo mộ, which is the grave-cleaning visit..."
"People generally value family heritage. Across cultures, in different societies, family traditions matter. People keep these things. Yes, generally everyone does this."
Three constructions for inserting a Vietnamese example into a Part 3 answer at the right level of specificity. The candidate can choose which pattern to deploy depending on whether the question frame is country-specific, societal, or comparative.
Vietnamese specific only: "In Vietnam we keep heirlooms." Functional for "in your country" questions. Stops at Band 6 because no abstract claim sits above it.
Position + Vietnamese as Evidence: "People keep useless objects because of memory — like the bracelet in my family." The Band 7 default. Vietnamese example does its work as supporting instance.
Position + Vietnamese Evidence + cross-cultural Extension: "People keep useless objects because of memory — like the bracelet in my family — and this might be more pronounced in Vietnam than elsewhere because of our recent history." The Band 8 move. Vietnamese specific used twice: once as evidence, once as cross-cultural contrast.
Five Part 3-appropriate Vietnamese deployments, each tied to a specific kind of claim. Each is audio-clickable and structured as a working sentence. Notice none of them is a culture lesson — each is a Vietnamese specific deployed inside a societal-level argument.
Five common traps when deploying Vietnamese examples in Part 3. The pattern matters: candidates either over-Vietnamise (Vietnam culture lecture, no abstract claim) or under-Vietnamise (purely generic, no first-person grounding). The Band 7 fix is keeping Vietnam as one piece of evidence inside a wider argument.
Two questions before we move to the worked example in Stage 7.
The candidate from Lesson 10 has just finished their 2-minute Part 2 about bà nội's silver bracelet. The examiner now pivots into Part 3 — five abstract follow-up questions over the next 4-5 minutes. You'll see all five answered to Band 7. Q1 and Q5 are Opinion questions using full P·E·E (this lesson). Q2-Q4 preview the three other Part 3 frameworks that Lessons 12, 13, and 14 deep-dive — Cause-Effect, Compare-Contrast, and Speculation. Same candidate. Same Vietnamese voice. Five distinct architectures.
The candidate's Part 2 answer was about "a thin silver bracelet that belonged to my paternal grandmother — bà nội — with my great-grandmother's name engraved on the inside," using the Significance Unwrap framework. They closed on the landing line: "Three generations of women, and a fourth I haven't met yet."
Type: Opinion. Architecture: P·E·E (this lesson). Full annotated answer next.
Type: Cause-Effect. Architecture: C·M·E preview (Lesson 13).
Type: Compare-Contrast. Architecture: X·Y·Z preview (Lesson 12).
Type: Speculation. Architecture: I·F·B preview (Lesson 14).
Type: Opinion. Architecture: P·E·E (this lesson, concise version).
The first Part 3 question. Opinion type. Full P·E·E architecture. Toggle annotations to see the move each marker makes. Click "Hear it spoken" to listen to the full answer. Notice the three tagged sections — Position (pink), Evidence (purple), Extension (teal).
"How have attitudes toward family heirlooms changed in recent generations in your country?"
"I think attitudes have shifted quite significantly in Vietnam over the last twenty years — for older generations, an heirloom was a position-holding object that came with responsibility, but for my generation, heirlooms feel increasingly optional.
You can see this in even very small objects — the silver bracelet in my family has been carrying three generations of women silently for decades, and none of us would have thought of it as 'an heirloom' in any official sense. For my grandmother it was just 'something you wear'; for my mother it became 'something you're meant to keep'; for me, it's somewhere between both — which I think tells you something about how the category itself has shifted."
That said, there's actually a counter-trend worth mentioning — urban professionals in their thirties are starting to ask their parents for heirlooms again, for reasons their grandparents wouldn't have recognised."
Two more questions. Each uses a different framework — preview-style, since Lessons 13 (C·M·E) and 12 (X·Y·Z) deep-dive them. Notice how the bracelet doesn't reappear; the candidate has moved into broader Vietnamese-flavoured examples instead.
Examiner: "Why do you think some people keep objects that have no practical use?"
"(Cause) I think people keep practically useless objects because those objects do work that has nothing to do with utility — they hold memory, anchor identity, and extend relationships beyond the people who're gone. (Mechanism) The way this works, I think, is that physical contact does something for memory that photographs and stories can't — you can put your hand on the same surface another person put their hand on, fifty years apart, and that physical link does cognitive work most people don't notice they're using. (Effect) The effect is that these supposedly useless objects end up doing the heavy lifting of family continuity in ways that more useful objects can't — which is why families that lose everything else still go back for these specific things first."
Examiner: "What are the differences between how older and younger generations value heirlooms?"
"(eXtract the dimension) I think the most interesting difference isn't whether each generation values heirlooms, but in how each one decides what's worth keeping. (Y values) My grandmother's generation kept objects almost automatically — anything that had survived was assumed to be worth preserving. My generation, by contrast, has to actively choose, because we have access to so much more stuff that the default isn't 'keep' anymore — it's 'discard unless there's a reason.' (Zoom) You can see this in something as small as old photographs — my grandmother kept every single one in shoeboxes; I keep maybe five percent of mine, all of them deliberately, all of them tagged. The objects haven't changed; the relationship to them has."
Notice the framework switch. Q2 uses C·M·E (Cause → Mechanism → Effect) because the question asks "why" — a mechanism question. Q3 uses X·Y·Z (eXtract dimension → Y values across the dimension → Zoom into telling detail) because the question asks for differences — a structural comparison. Same candidate, two different architectures, both deployed cleanly in under a minute each.
The last two questions. Q4 deploys the I·F·B framework (Lesson 14 preview). Q5 returns to P·E·E for a concise second Opinion answer. By the end of these five questions, the candidate has demonstrated all four Part 3 frameworks — without ever defaulting to "in my family I have a silver bracelet" beyond Q1.
Examiner: "How will attitudes toward physical objects change in the next twenty years?"
"(If-statement) If current trends continue, my sense is that physical objects will become more emotionally specialised, not less — they'll do less utility work and more memory work. (First-order) The first-order consequence we're already seeing is the revival of physical formats people thought were dying — vinyl, printed photographs, paper books — chosen specifically because they're not digital, by people who could choose either. (Beyond first-order) The more interesting second-order effect, I think, is that objects will become more deliberate — fewer per household, but each one carrying more meaning. Which might amplify the inequality between families with long material histories and those without — heirlooms will matter more, precisely because most things won't."
Examiner: "Should family heirlooms ever be donated to museums?"
"(Position) I'd say yes, but only when the object has cultural significance beyond the family itself, and when the family no longer has someone willing to actively maintain it. (Evidence) Take an object like a pre-1954 Vietnamese áo dài — that has wider cultural value because it's a physical record of a clothing tradition that's evolved significantly, and a museum can give it context a private family can't. (Extension) That said, the museum-versus-family binary is actually unhelpful — what most families need is something in between, and digital archives are starting to fill that gap surprisingly well."
Five questions, four frameworks, ~4 minutes of Part 3. The bracelet appeared in exactly one answer (Q1, as Evidence) — and then the candidate moved to other Vietnamese-flavoured examples: photographs in Q3, vinyl in Q4, áo dài in Q5. Each answer used a calibrated number of Vietnamese specifics (1-2 per answer). Each followed its framework cleanly. This is what Part 3 sounds like at Band 7+, with the bracelet from Lesson 10 used as supporting evidence rather than the entire answer.
Two diagnostic questions before we move to the Practice Arena.
Five real Part 3 answer excerpts. For each one, identify which of the four Part 3 failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four Part 2 drift excerpts. For each, write a 3-line P·E·E — one sentence Position, 1-2 sentences Evidence (the personal story can stay, but as Evidence not as the answer), one sentence Extension. Reveal each model after your attempt.
Three Part 3 questions. For each, identify the question type first, then either write a P·E·E (if it's Opinion) or note which framework would apply (X·Y·Z, C·M·E, I·F·B). Recognition first, deployment second — same skill drilled in Stage 4 but now applied to writing.
Three P+E answers — Position and Evidence are already written. Your job is to add the Extension sentence — the Band 7 differentiator. Pick a counter-trend, an implication, a paradox, or a cross-cultural observation. One sentence each. This is the move that lifts a Part 3 answer from Band 6 to Band 7.
Pick one Part 3 question — any of the four from the Opinion Builder in Stage 3 (heirlooms, useless objects, digital age, museums), or one from Exercise 2. Set a timer for 50 seconds. Deliver the P·E·E answer aloud. Three rounds — slow pass, natural pace, exam pressure.
Five Part 3-specific exercises done. Here's how it landed.
Your performance across the Part 3 arena suggests the P·E·E framework is taking root. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 11. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Part 3 drill this week.
Forty-five minutes invested in the lesson that fixes the most common Band 6 ceiling Vietnamese candidates hit on Part 3 — the Personal-Anecdote Drift.
Eleven lessons done. Part 3 foundation in place. Lesson 12 deep-dives the second Part 3 question type — Compare-and-Contrast (~25% of Part 3 questions). Different mental move from Opinion: where P·E·E takes a single stake, X·Y·Z holds two positions in tension along one specific dimension.
"What are the differences between...?" / "How is X different from Y?" / "Compare X in your country with elsewhere." Twenty-five percent of Part 3 questions ask for a comparison — and most candidates list unconnected differences instead of comparing along one specific dimension. Lesson 12 teaches the X·Y·Z architecture that lifts a Compare-Contrast answer from Band 6 to Band 7.
eXtract one specific dimension of comparison. Compare the Y-values across that dimension. Zoom into one telling detail that makes the contrast vivid. One axis, deeply explored.
The dominant failure mode for Compare-Contrast: listing five unconnected differences (X is fast, Y is slow, X is new, Y is old, X is expensive, Y is cheap...). The Band 7 move is one dimension explored deeply, not five touched lightly.
Most Vietnamese-relevant Compare-Contrast questions land on generational or urban/rural axes. Lesson 12 will give you the specific phrasings for these two recurring patterns.
The Band 8 move on Compare-Contrast questions: reframe the comparison the examiner asked for into a more interesting comparison the candidate noticed. "I think the more interesting difference isn't X vs Y, but actually..."
"Part 3 foundation is in. The Personal-Anecdote Drift has a name now — and the P·E·E framework is the fix. One Position. One Evidence. One Extension. Three sentences, one minute, Band 7. Do the drill for a week — one Part 3 answer per day with the recording on — and the Extension sentence will start surfacing automatically. After that, Lessons 12, 13, 14, 15 just teach you three more architectures to swap in: X·Y·Z, C·M·E, I·F·B. The hard mental gear-change is the one you just did."
Eleven badges. The full Part 1 toolkit, the universal Part 2 framework, the full set of four Part 2 categories — and now the Part 3 foundation with the P·E·E framework. The Personal-Anecdote Drift has a name, a diagnostic, and a fix. Lessons 12-15 deep-dive the three remaining Part 3 frameworks.
The second Part 3 question type. After P·E·E for Opinion questions in Lesson 11, this lesson teaches X·Y·Z for Compare-and-Contrast questions — roughly 25% of Part 3 questions. Where P·E·E asked you to take a single stake, X·Y·Z asks you to hold two positions in tension along one specific dimension. The Band 7 candidate explores one axis deeply. The Band 5-6 candidate lists five differences and never picks an axis at all.
"Compare-and-Contrast questions look easy because they hand you the structure: X versus Y. The trap is in how candidates fill the structure. Most candidates list five surface differences — older people stay home, younger people go out; older people speak Vietnamese, younger people speak English; older people respect tradition, younger people don't. Five axes, all touched lightly, none of them said anything interesting. The Band 7 move is one axis, said deeply. Pick the comparison the examiner didn't quite ask for — the one underneath the question — and explore it for the full minute. Lesson 12 teaches the X·Y·Z framework that makes this automatic, and the reframe move that turns Band 7 into Band 8."
Two candidates. Same Compare-Contrast Part 3 question. Same Vietnamese background, same vocabulary range, same grammar. Watch what happens when one lists five surface differences (Listing Trap) and the other picks one specific dimension and explores it (X·Y·Z). Same question. Only one is at Band 7.
"How do Vietnamese weddings today differ from weddings of your grandparents' generation?"
Five differences listed, none of them said anything specific. Location, clothing, cost, helpers — four separate axes touched lightly. The examiner cannot grade analytical thinking from a list. Grammar is fine; structure is wrong.
X·Y·Z architecture visible. Sentence 1 picks the dimension ("whose investment they demonstrate") and reframes the question. Sentences 2-3 give the Y-values across that dimension. Sentence 4 zooms into one specific detail (photo arrangement). One axis, deeply explored.
Same 9-stage shape as Lessons 7-11, applied to Compare-and-Contrast questions specifically. Focus: the X·Y·Z framework, the Listing Trap failure mode, the Vietnamese comparison patterns (generational, urban/rural, diaspora/domestic), and the Band 8 reframe move.
You are here.
The 4 Compare-Contrast failure modes — Listing Trap, Polarity Collapse, Vague-Axis Drift, One-Sided Compare — and the cultural reasons each one shows up.
The second Part 3 micro-framework. eXtract the dimension · compare Y values · Zoom into telling detail. Plus the interactive Comparison Builder.
The four recurring axes in Part 3 Vietnamese questions — generational, urban/rural, past/present, diaspora/domestic. Pre-loaded dimensions for each.
80+ Compare-Contrast-appropriate phrases — dimension-naming, Y-value-contrasting, zoom-launching, reframe-opening. The L12 phrase bank.
The Band 8 move: "what's actually interesting in this comparison isn't X vs Y, but..." Turning the examiner's question into a more interesting question — without dodging.
Five Compare-Contrast questions answered to Band 7 — same Vietnamese candidate, five X·Y·Z architectures across five different dimensions.
Five Compare-Contrast-specific exercises.
Self-assessment, badge, and Lesson 13 preview (Cause-and-Effect questions · C·M·E).
If you've ever been asked a Compare-Contrast Part 3 question and felt your brain reaching for a quick list of three or four differences — this stage is for you. The collapse isn't a vocabulary problem and it isn't a thinking problem. It's a discipline problem: listing five surface differences feels comprehensive, while committing to one axis feels risky. Vietnamese candidates default to listing precisely because it feels like the safer move. Lesson 12 is about why it actually isn't.
Vietnamese essay structure (and a lot of school-tested English) rewards breadth: identify five points, give one or two sentences each, conclude. The student who covers five angles feels more comprehensive than the one who explores one. In Part 3, that instinct inverts — five surface points look unfocused, while one deeply-explored axis looks like analytical thinking.
The Compare-Contrast descriptors specifically reward the candidate who can identify the most interesting comparison underneath the question and explore it analytically. A list of five differences demonstrates English fluency at the sentence level. An axis chosen and explored demonstrates analytical thinking at the discourse level — which is what Band 7 is actually grading.
From the examiner's side, a list of five differences sounds like the candidate has noticed many surface contrasts but doesn't know which one matters. An axis chosen and explored sounds like the candidate has done the harder mental work of asking "what's the actual difference underneath all these surface ones?" — which is the move that distinguishes a Band 7 thinker from a Band 6 lister.
The good news: picking an axis is a skill, not a vocabulary acquisition. Once you've practised picking axes on three or four Compare-Contrast questions, the move starts becoming automatic — and almost any axis you commit to, explored deeply, will outperform any list of five surface differences.
Listen to enough Compare-Contrast Part 3 answers and the failures cluster into four distinct patterns. Same root cause behind each — failing to commit to one axis — but each has a different surface symptom. Recognising your specific shape makes the fix more targeted.
The dominant default. The candidate lists 4-6 unconnected surface differences without committing to any axis. Each difference gets one short sentence, none gets explored. From the outside it looks like good coverage; from the examiner's side it sounds like the candidate hasn't decided what the comparison is actually about.
The candidate flattens the comparison into a moral binary: one side is implicitly better. "Old way is good, new way is bad" (or the reverse). Nuance disappears; the comparison becomes an evaluation. The examiner can't grade analytical thinking when the candidate has implicitly already decided who's winning.
The candidate uses the word "different" repeatedly without ever naming what's different. "They are very different. The difference is big. Things have changed a lot. It's not the same." The examiner hears a candidate gesturing at a comparison without committing to one. The L12-specific cousin of L11's Confused Position.
The candidate describes one side of the comparison thoroughly and then runs out of time before doing the other. 45 seconds on the old way, 10 seconds on the new way (or vice versa). The comparison fails because both sides aren't actually present in the answer. Common when the candidate has more first-hand experience of one side than the other.
Same Vietnamese candidate, same Compare-Contrast Part 3 question, two structural registers. Here's "What's the difference between how older and younger Vietnamese view marriage?" in both modes. Notice the listing version has more sentences and feels more comprehensive — and yet the axis version is the one that lifts the band.
Lists 5-6 surface differences. Each gets one sentence. Feels comprehensive. Sounds unfocused. The examiner hears coverage without analysis.
Picks one specific dimension. Holds two values in tension across that dimension. Zooms into one telling detail. One axis, deeply explored.
Whenever you catch yourself starting a Compare-Contrast answer with two short sentences contrasting surface features — pause and ask: what's the axis underneath these surface differences? The surface differences (age at marriage, location, family size) are real, but they're symptoms of a deeper axis (what marriage is for). Pick the axis. Explore it. Zoom in. Stop.
This doesn't mean surface differences are forbidden — they're useful as Y-values once you've named the axis. Wedding photographs as a detail is more concrete than any of the listed differences in the Listing version. The shift is in what the candidate does with the surface details, not in whether they appear at all.
Sometimes candidates resist this stage too. "But the question literally asked me how things are different — surely listing differences is what's being asked for?" Worth addressing directly: Lesson 12 doesn't say the question is wrong. It says the question hands you a structure, not an axis. The axis is your job.
"Compare-Contrast questions hand you the structure but never the axis. 'How are X and Y different?' — that's the structure. The axis is what you bring to the question. The candidate who picks an interesting axis demonstrates analytical thinking; the candidate who lists five surface differences demonstrates English fluency. The descriptors specifically reward the first. Your job in Part 3 Compare-Contrast isn't to inventory the differences — it's to identify the one comparison underneath the surface that's actually interesting, and explore it. Picking the axis is the work. Once you've picked one, almost any of them will outperform the list of five."
Imagine the examiner asks: "What's the difference between how children play in cities vs in rural Vietnam?" Five surface axes are available: location, equipment, supervision, social grouping, screen time. The Listing candidate touches all five lightly. The X·Y·Z candidate picks one and stays.
If you pick "supervision," for instance, you get to explore: urban children play under direct adult supervision (parents, helpers, after-school staff); rural children play in shifting peer groups with much looser adult presence. Then you zoom: the same eight-year-old in Saigon has someone tracking her at all times; her cousin in Cần Thơ can disappear for an afternoon and nobody worries. Same age, same family, different supervision structure. One axis, two-sided, telling detail. Band 7 from one good axis pick.
When you pick one specific dimension and explore it for the full minute — instead of touching five lightly — you're not narrowing the question. You're respecting it. You're saying: this comparison deserves a real argument, not a checklist. The examiner hears a candidate who can see the comparison the question was actually asking and chose to engage with it. The choosing is the respect. Lesson 11 was about respecting the abstraction. Lesson 12 is about respecting the axis. Same move, different framework.
Two questions before we move to the X·Y·Z framework in Stage 3.
Three moves. eXtract the dimension. Compare the Y-values. Zoom into one telling detail. That's the entire framework. X·Y·Z is the second Part 3 micro-framework — used specifically for Compare-and-Contrast questions (~25% of Part 3). Same role for Compare-Contrast that P·E·E played for Opinion: a skeletal structure that forces you to pick one axis and explore it instead of listing five lightly.
The first move is the most important — and the one most candidates skip. Before comparing anything, name the single dimension you're going to compare along. Not "they're different in many ways" — but "the most interesting difference is in how each generation decides what's worth keeping." One axis, named explicitly, so the examiner knows exactly what comparison is coming.
Once the dimension is named, give the value on each side. This is the actual comparison — one sentence for X's value on the dimension, one for Y's. The discipline: both sides must be present, both must be specific, and both must be values on the same axis you just named. This is where One-Sided Compare gets fixed.
"The test of a good Y-comparison is simple: could you swap the two halves and have it still make sense as a contrast? If your X-value is 'the families' event' and your Y-value is 'the couple's event,' yes — they're two values on one axis. If your X-value is 'held at home' and your Y-value is 'more expensive,' no — those are two different axes wearing a contrast connector. Both values, one axis. That's the whole discipline of the Y move."
The third move is the Band 7 differentiator. After the dimension and the two Y-values, you have 10-15 seconds for one concrete zoom — a small, specific, sensory detail that makes the abstract contrast vivid. The Zoom is to X·Y·Z what the Extension was to P·E·E: the move that separates a competent answer from a memorable one.
One full X·Y·Z in one block — eXtract, Y-values, Zoom. Notice how the three moves layer: the dimension names the axis (10 sec), the Y-values populate both sides of it (20-25 sec), the Zoom makes it physical and lands a closing line (10-15 sec).
"I think the most interesting difference is in whose event the wedding actually is. For my grandparents' generation, a wedding belonged to the two families — both sides contributed labour and visibly invested in each other's commitment; for my generation, it increasingly belongs to the couple, with the families paying but mostly receding into the audience. You can see this in something as small as who arranges the photos on the wall the morning after — in my grandmother's case, both grandmothers; in my cousin's case last year, just the couple and their wedding planner. Same wedding, different protagonist."
One sentence eXtract. Two sentences Y-values. One sentence Zoom. Total: 45-60 seconds, four to five sentences. The hardest discipline is the eXtract — committing to one axis when listing five feels safer. 10 sec eXtract + 20-25 sec Y-values + 10-15 sec Zoom = 45-60 sec Compare-Contrast answer.
Four picks. Pick a Compare-Contrast question. Pick the dimension (X). Pick the two-sided Y-values. Pick the Zoom. Watch the full X·Y·Z answer assemble in real time. Each question offers different dimensions — try at least two to see how the same architecture handles different topics.
[pick a Compare-Contrast question first]
~45-60 seconds of Compare-Contrast answer built from four picks. Notice how the three moves layer: eXtract names one axis (10 sec), Y-values populate both sides (20-25 sec), Zoom lands a telling detail and a closing line (10-15 sec). One axis, deeply explored — not five touched lightly.
Two questions before we move to the Vietnamese Compare-Contrast patterns in Stage 4.
Vietnamese Part 3 Compare-Contrast questions cluster around four recurring comparison patterns. If you have a strong dimension ready for each one, you'll almost never be caught without an axis to pick. This stage gives you the four patterns and the best dimensions to deploy on each — so the eXtract move becomes instant recognition rather than on-the-spot invention.
"How do older and younger generations differ on...?" The single most common Vietnamese Compare-Contrast pattern — because Vietnam's compressed modernisation makes generational gaps unusually visible.
"How does X differ between cities and the countryside?" Vietnam's sharp urban-rural divergence makes this a frequent and rich pattern.
"How has X changed over the years?" Looks like Generational but isn't — this is about societal change over time, not about people of different ages coexisting now.
"How does X in Vietnam differ from other countries?" The cross-cultural pattern — where the L11 Vietnamese-examples calibration becomes directly useful.
The most common Vietnamese Compare-Contrast pattern. Recognisable by "older and younger," "your generation vs your parents'," "young people today." The strongest default axis: what the thing is FOR — because the deepest generational differences are almost always about changing purpose, not changing surface.
The second most common pattern. Recognisable by "cities vs countryside," "Hà Nội vs rural areas," "big cities vs small towns." The strongest default axis: the pace and structure of time — because the deepest urban-rural difference in Vietnam isn't wealth or facilities, it's how time itself is organised.
Two more recurring axes. Past/Present looks like Generational but is subtly different — it's about societal change over time rather than coexisting age groups. Domestic/Foreign is the cross-cultural pattern, where the Lesson 11 Vietnamese-examples calibration becomes directly useful.
Notice that Past/Present's "gained vs lost" axis is specifically designed to defuse the Polarity Collapse failure mode — by requiring you to name both a gain and a loss, it forces a genuine comparison instead of a moral verdict. And Domestic/Foreign's "one specific cultural value" axis is designed to defuse the National Pride Speech — by forcing one named, specific value rather than general cultural praise.
Four Compare-Contrast questions. Identify which of the four Vietnamese patterns each one is — which tells you the pre-loaded default axis to deploy. Two seconds per question.
Same architecture as the L7-11 vocabulary banks, applied to Compare-and-Contrast construction. Eighty phrases that aren't "they are very different" or "on the other hand." Pick three from each category and you've replaced the listing reflex with the axis-naming, contrasting, and zooming phrases that signal Band 7.
Phrases for the eXtract move — naming the axis. "The most interesting difference is in..." / "The real contrast is around..." The signpost that says "I picked an axis."
Phrases for the Y-values move — holding two sides in tension. "Whereas for my generation..." / "By contrast..." / "The older pattern was... the newer pattern is..."
Phrases for the Zoom move — landing the telling detail. "You can see this in something as small as..." / "It shows up in one detail: ..."
Phrases for the Band 8 reframe (Stage 6) — turning the question into a better one. "What's actually interesting isn't X vs Y, but..." / "The more revealing comparison is..."
Twenty phrases for the eXtract move — naming the single axis you'll compare along. These are your openers. The signpost tells the examiner "I've picked an axis," which immediately separates you from the lister. Pick three favourites and rotate them.
Twenty phrases for the Y-values move — holding two sides in tension across one axis. The connector makes the two-sidedness audible and keeps both values on the same dimension. Pick three favourites and use them to join your X-value and Y-value cleanly.
Twenty phrases for the Zoom move — launching into the one telling detail that makes the contrast vivid. These are the highest-leverage phrases in the bank because they signal the Band 7 differentiator. Pick three favourites and drill them until they're reflexive.
Twenty phrases for the Band 8 reframe move — turning the comparison the examiner asked for into a more interesting one (the full move is built in Stage 6). These signal the highest-level Compare-Contrast thinking: not just answering the question, but improving it. Use sparingly — one well-placed reframe per Part 3 set.
Across the four categories — dimension-naming, contrast-connector, zoom-launching, reframe-opening — pick three favourites per category. Twelve phrases total that you can deploy on any Compare-Contrast question. The reframe phrases are the most advanced — keep one ready, but only deploy it when you genuinely see a better comparison. Twelve phrases × four categories = a full Band 7 Compare-Contrast vocabulary.
The Band 8 move on Compare-Contrast questions. After the eXtract-Y-Zoom architecture is solid, the highest-level candidates do one more thing: they notice when the comparison the examiner asked for isn't the most interesting one available — and they offer a better axis. The key is that this is improving the question, not dodging it. The difference between the two is the whole skill.
Examiner: "How are city and rural weddings different?" Candidate: "Well, I don't really know about rural weddings. But I think all weddings are about love. Love is the same everywhere. So maybe they're not so different."
Examiner: "How are city and rural weddings different?" Candidate: "The obvious difference is scale and cost — but I think the more revealing comparison is in who the wedding is performed for. A rural wedding is performed for the community that will hold the couple accountable; a city wedding is increasingly performed for an audience that will mostly never see the couple again."
A reframe that works always has two parts. First, acknowledge the obvious comparison — this proves you understood the question and aren't avoiding it. Then, offer the sharper axis — and crucially, deliver a full X·Y·Z on the new axis. A reframe that names a better comparison but doesn't then explore it is just a fancier form of dodging.
The phrase template: "The obvious difference is [the expected axis] — but I think the more revealing comparison is [your sharper axis]." Then run the full eXtract-Y-Zoom on the sharper axis. Acknowledge, pivot, deliver.
The reframe is powerful but not free. Deployed at the right moment, it signals Band 8 insight. Deployed reflexively on every question, it starts to sound like you can't answer questions on their own terms. Here's the calibration — when the reframe earns its keep, and when a clean straight X·Y·Z is the better move.
Five worked reframes, each tied to a common Compare-Contrast question. Each follows the acknowledge-pivot-deliver structure. Audio-clickable. Notice how each one names the obvious axis first, then offers a sharper one — and the sharper axis is always more about meaning than about surface features.
The single most important distinction in this stage. A reframe and a dodge can use almost the same opening words — "well, I think the real question is..." — but they go in opposite directions. Five paired examples, dodge vs reframe, so you can hear the difference and stay on the right side of the line.
Uses reframe language to escape a comparison the candidate can't make. Lands on something vague and universal. Never delivers a real contrast.
Acknowledges the obvious axis, names a sharper one, and delivers a full X·Y·Z on it. Lands on something specific and two-sided. Improves the comparison.
Before you reframe, ask yourself one question: am I moving toward a more specific comparison, or away from a comparison I can't make? If you're moving toward something sharper and you can deliver a full X·Y·Z on it — reframe. If you're moving toward something vaguer because you don't know the topic — don't reframe; just attempt a straight X·Y·Z on the obvious axis, even if it's imperfect. A flawed straight answer beats a polished dodge every time.
Two questions before we move to the worked example in Stage 7.
The same Vietnamese candidate from Lessons 10-11. They've done a Part 2 about bà nội's bracelet and the Part 3 Opinion set in Lesson 11. Now the examiner is running a Compare-Contrast-heavy Part 3. Five questions, each demanding an X·Y·Z — and one of them gets the Band 8 reframe. Watch how the candidate picks a different axis for each, never lists, and lands a closing line every time.
Topic thread: family, tradition, and change in Vietnam. The examiner moves through five Compare-Contrast questions, each comparing across a different pairing — generational, urban/rural, past/present, domestic/foreign.
Pattern: Generational. Axis: whose event it is. Full annotated X·Y·Z next.
Pattern: Urban/Rural. Axis: unsupervised time.
Pattern: Past/Present. Axis: what's gained vs what's lost.
Pattern: Domestic/Foreign. Axis: one specific cultural value.
Pattern: Urban/Rural — but answered with the Band 8 reframe.
The first Compare-Contrast question. Generational pattern. Full X·Y·Z architecture. Toggle annotations to see the move each marker makes. Click "Hear it spoken" to listen. Notice the three tagged sections — eXtract (rose), Y-values (purple), Zoom (amber).
"How do weddings today differ from your grandparents' generation?"
I think the most interesting difference is in whose event the wedding actually is — the two families', or the couple's.
For my grandparents' generation, a wedding belonged to the two families — both sides contributed labour, both visibly invested in each other's commitment, and the couple were almost the occasion rather than the owners of it. For my generation, by contrast, it increasingly belongs to the couple — the families still pay, but they mostly recede into the audience, and the day is built around the couple's taste and story.
You can see it in something as small as who arranges the photos on the wall the morning after — in my grandmother's case, both grandmothers did it together; in my cousin's case last year, just the couple and their wedding planner. Same wedding, different protagonist.
Two more questions, two more axes. Each picks a different dimension and runs a clean X·Y·Z. Notice the closing lines — each one uses the "Same [surface], different [axis]" template.
Examiner: "How is childhood different in cities versus the countryside?"
(eXtract) I think the real contrast is in how much unsupervised time a childhood actually contains. (Y-values) An urban childhood now contains almost none — there's always an adult tracking the schedule and the location, between school, tutoring, and supervised play. A rural childhood still contains long stretches where no adult knows exactly where the child is, and nobody treats that as a problem. (Zoom) The same eight-year-old in Saigon has someone accounting for her every hour; her cousin in Cần Thơ can vanish for a whole afternoon and turn up at dinner. Same age, different amount of childhood that nobody is watching."
Examiner: "How has the way families celebrate Tết changed over the years?"
(eXtract) Rather than say it's simply better or worse, I'd frame it as what's been gained against what's been lost. (Y-values) What's gained is reach — families scattered across the country, or abroad, can now all be present by video, which was impossible for my grandparents. What's lost is a kind of enforced slowness — the days of preparation, the shared cooking, the fact that you couldn't be anywhere else, which is exactly what made the holiday feel separate from ordinary time. (Zoom) You can hold both in one image: we can video-call relatives in three countries on the first morning of Tết now, but almost nobody makes bánh chưng from scratch overnight anymore — it's bought, sealed, ready. Same holiday, more reach and less ritual."
Notice Q3 defuses Polarity Collapse. The question invites "Tết is worse now" — the classic moralising trap. The "gained vs lost" axis forces the candidate to name both a real gain (reach) and a real loss (enforced slowness), which produces a genuine comparison instead of a verdict. The axis itself prevented the failure mode.
The last two questions. Q4 runs a clean Domestic/Foreign X·Y·Z on one specific cultural value. Q5 demonstrates the Band 8 reframe — acknowledging the obvious urban/rural axis, then offering a sharper cross-cutting one and delivering on it fully.
Examiner: "How does the parent-child relationship differ in Vietnam versus the West?"
(eXtract) Rather than say the cultures are just different, I'd point to one specific value: whether adult independence is the goal of parenting or a side effect of it. (Y-values) In a lot of Western parenting, raising a child who eventually leaves and lives fully separately is the explicit success condition — independence is the point. In Vietnam, independence happens, but it's rarely the goal; the goal is a relationship that stays mutually obligated for life, where the adult child remains woven into the parents' daily existence. (Zoom) You see it in one question: a Western parent often asks 'when will you move out?'; a Vietnamese parent more often assumes 'when you marry, we'll work out how the household expands.' Same love, different default about distance."
Examiner: "How do city and rural attitudes to marriage differ?"
(Acknowledge) The obvious answer is timing and independence — city people marry later and choose more freely, rural people marry earlier and with more family involvement. (Pivot to sharper axis) But honestly, I'm not sure city versus rural is the sharpest divide anymore — I think the real split is between families who've stayed in one place for generations and families who've moved, wherever they now live. (Deliver X·Y·Z on the new axis) A rooted family — rural or urban — still treats marriage as something the wider family has a stake in, because everyone will keep living near each other. A mobile family treats it as the couple's private decision, because the relatives are scattered anyway. You see it in who gets consulted before a proposal: in a rooted family, several aunts; in a mobile one, maybe a phone call afterward. Same country, but the dividing line runs through movement, not through the map."
Five questions, five distinct axes, zero lists. Q1 generational (whose event), Q2 urban/rural (unsupervised time), Q3 past/present (gained vs lost), Q4 domestic/foreign (independence as goal vs side effect), Q5 the reframe (movement, not map). Each landed a "Same X, different Y" closing line. The reframe in Q5 acknowledged the obvious axis, offered a sharper cross-cutting one, and delivered a full X·Y·Z on it — the textbook Band 8 move. This is what a Compare-Contrast-heavy Part 3 sounds like at the top of the band.
Two diagnostic questions before we move to the Practice Arena.
Five real Compare-Contrast answer excerpts. For each one, identify which of the four failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four listing excerpts. For each, write a 3-line X·Y·Z — name one dimension (eXtract), give both Y-values on that axis, land one zoom with a closing line. Reveal each model after your attempt.
Three Compare-Contrast questions. For each, identify the Vietnamese pattern (Generational / Urban-Rural / Past-Present / Domestic-Foreign), then write just the eXtract sentence — the one dimension you'd commit to. Recognition plus axis-naming, the two-second skill from Stage 4.
Three X·Y answers — eXtract and Y-values are already written. Your job is to add the Zoom — one small concrete detail that holds the whole contrast, plus a "Same [surface], different [axis]" closing line. This is the Band 7 differentiator. One sentence of detail, one short closing line.
Pick one Compare-Contrast question — any from the Comparison Builder in Stage 3, or from Exercise 2. Set a timer for 50 seconds. Deliver the full X·Y·Z aloud: name one axis, populate both sides, land a zoom with a closing line. Three rounds — slow pass, natural pace, exam pressure.
Five Compare-Contrast exercises done. Here's how it landed.
Your performance across the Compare-Contrast arena suggests the X·Y·Z framework is taking root. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 12. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Compare-Contrast drill this week.
Forty-five minutes invested in the lesson that fixes the most common Band 6 ceiling on Compare-Contrast questions — the Listing Trap.
Twelve lessons done. Two Part 3 frameworks in hand — P·E·E for Opinion, X·Y·Z for Compare-Contrast. Lesson 13 deep-dives the third Part 3 question type — Cause-and-Effect (~20% of Part 3). The "why" question. Different mental move again: where X·Y·Z held two things in tension, C·M·E traces a chain from cause through mechanism to effect.
"Why do you think...?" / "What causes...?" / "What are the effects of...?" Twenty percent of Part 3 questions ask for a mechanism — and most candidates either name a cause and stop, or jump straight to effects without explaining how. Lesson 13 teaches the C·M·E architecture: name the Cause, explain the Mechanism in the middle, name the Effect. The mechanism is where most candidates skip, and it's exactly where Band 7 lives.
Cause · Mechanism · Effect. Name the specific cause, explain how it produces the effect (the chain in the middle most candidates skip), then name the specific outcome with one concrete instance.
The dominant Cause-Effect failure mode: jumping from cause straight to effect with no mechanism. "Phones cause loneliness." How? The missing middle is the band-decider.
When to commit to one deep causal chain vs when to acknowledge multiple causes — and how to avoid the "many reasons" list that mirrors L12's Listing Trap.
The Band 8 move on Cause-Effect: naming the non-obvious downstream consequence. Not just the first effect, but the effect of the effect.
"Two Part 3 frameworks down. P·E·E for opinions, X·Y·Z for comparisons. The Listing Trap has a name now, and the one-axis discipline is the fix — one dimension, both sides, one zoom, one closing line. Do the drill for a week, and watch the second axis stop sneaking in. Then Lesson 13 adds the third framework: C·M·E for the 'why' questions. After that it's just Speculation and the hybrid questions, and the whole Part 3 toolkit is complete. You're past the halfway point of Part 3 already."
Twelve badges. The full Part 1 and Part 2 toolkits, plus two of the four Part 3 frameworks — P·E·E for Opinion, X·Y·Z for Compare-Contrast. The Listing Trap has a name, a diagnostic, and a fix. Lessons 13-15 finish the Part 3 module.
The third Part 3 question type. After P·E·E for Opinion and X·Y·Z for Compare-Contrast, this lesson teaches C·M·E for Cause-and-Effect questions — roughly 20% of Part 3, and the most common single word in Part 3 is "why." Where X·Y·Z held two things in tension, C·M·E traces a chain: name the Cause, explain the Mechanism, name the Effect. The Band 5-6 candidate names a cause and jumps straight to an effect. The Band 7 candidate explains the mechanism in the middle — the how that connects them.
"Cause-and-Effect questions are the ones that start with 'why' or 'what causes' or 'what are the effects of.' They look easy — you name a cause, you name an effect, done. And that's exactly the trap. 'Phones cause loneliness' isn't an answer; it's a headline. The examiner is waiting for the part where you explain how a phone produces loneliness — the mechanism in the middle. That missing middle is the entire difference between Band 6 and Band 7 on these questions. Lesson 13 teaches you to never skip it — and then teaches you the Band 8 move of naming the second-order effect, the consequence of the consequence."
Two candidates. Same Cause-Effect Part 3 question. Same Vietnamese background, same vocabulary, same grammar. Watch what happens when one jumps from cause straight to effect (Jump Trap) and the other explains the mechanism in the middle (C·M·E). Same question. Only one is at Band 7.
"Why do you think young people in Vietnam are choosing to marry later than previous generations?"
Three causes named, each jumped straight to the same effect with "so." Nothing in between explains how the cause produces the effect. It's also secretly a Listing Trap — three shallow causal jumps instead of one explained chain. The examiner cannot grade reasoning from a list of "because" statements.
C·M·E architecture visible. Sentence 1 names one specific cause. Sentence 2 is the mechanism — the "how" that the Jump Trap skips entirely. Sentence 3 names the effect and then pushes to the second-order effect (a new life stage), which is the Band 8 reach. One chain, fully explained.
Same 9-stage shape as Lessons 7-12, applied to Cause-and-Effect questions. Focus: the C·M·E framework, the Jump Trap failure mode, the Vietnamese causal patterns, and the second-order effect Band 8 move.
You are here.
The 4 Cause-Effect failure modes — Jump Trap, Cause Listing, Single-Link Stop, Correlation Confusion — and why each one shows up.
The third Part 3 micro-framework. Cause · Mechanism · Effect. Plus the interactive Causal-Chain Builder.
The recurring "why" themes in Vietnamese Part 3 — modernisation, urbanisation, technology, economic change — with pre-loaded mechanisms.
80+ Cause-Effect phrases — cause-naming, mechanism-explaining, effect-stating, second-order-launching.
The Band 8 move: naming the consequence of the consequence — the non-obvious downstream effect most candidates never reach.
Five Cause-Effect questions answered to Band 7 — same Vietnamese candidate, five C·M·E chains.
Five Cause-Effect-specific exercises.
Self-assessment, badge, and Lesson 14 preview (Speculation questions · I·F·B).
If you've ever answered a "why" question by naming a cause, saying "so," and naming an effect — and then felt like you'd finished — this stage is for you. The mechanism-skip isn't a vocabulary problem or a thinking problem. It's a habit problem: in everyday speech, the mechanism is usually obvious and we leave it out. "I'm tired because I didn't sleep" needs no mechanism. But in Part 3, the mechanism is the answer — and the habit of leaving it out is exactly what caps candidates at Band 6.
In normal conversation, cause and effect are usually close enough that the mechanism is assumed. "I was late because of traffic" — nobody needs the chain explained. Everyday English (and everyday Vietnamese) trains us to drop the middle. Part 3 inverts this: the examiner already knows the cause and the effect; they're testing whether you can articulate the connection between them.
The Cause-Effect descriptors specifically reward the candidate who can explain how a cause produces an effect — that's reasoning made visible. Naming a cause and an effect demonstrates that you know two facts. Explaining the mechanism between them demonstrates that you understand a process. Band 7 is graded on the second.
From the examiner's side, "phones cause loneliness" sounds like a slogan the candidate has heard and repeated. "Phones create loneliness because they replace the small, low-effort social contact — the chat at the shop, the nod to a neighbour — with frictionless alternatives, so the easy social muscles slowly atrophy" sounds like a candidate who has actually thought about the process. Same cause, same effect. The mechanism is the entire difference.
The good news: the mechanism is a skill, not knowledge. You don't need to know more about phones or loneliness — you need to build the habit of asking "but how exactly?" after every cause you name, and then answering it for a sentence before moving to the effect.
Listen to enough Cause-Effect Part 3 answers and the failures cluster into four patterns. The root cause behind most of them is the same — skipping the mechanism — but each has a distinct surface symptom. Recognising your shape makes the fix targeted.
The dominant default. The candidate names a cause, says "so" or "because," and jumps straight to the effect — with nothing in between explaining how the one produces the other. The mechanism, which is the actual answer, is entirely absent.
The L12 Listing Trap in causal clothing. Instead of explaining one causal chain deeply, the candidate lists several causes shallowly — "because of A, and B, and also C, and D." Breadth replaces depth; no single mechanism gets explained. Often signalled by "there are many reasons."
The candidate explains one cause-mechanism-effect link competently — and then stops, when the interesting answer was one link further down the chain. They reach the first-order effect and treat it as the destination, missing the second-order effect where Band 8 lives.
The candidate mistakes "happens at the same time" for "causes." Two things that co-occur get linked causally without a real mechanism — because the mechanism would expose that the link isn't actually causal. The missing mechanism here isn't laziness; it's that there isn't one.
Same Vietnamese candidate, same Cause-Effect question, two registers. Here's "Why do you think fewer young people want to work in agriculture?" in both modes. The headline version names causes and effects; the explanation version inserts the mechanism. Notice the headline version is actually longer — and yet the explanation version lifts the band.
Names several causes, jumps each to an effect with "so." Feels complete. Sounds like repeated slogans. The examiner hears facts, not reasoning.
Names one cause, explains how it actually produces the effect, then reaches a second-order consequence. One chain, fully traced.
Whenever you catch yourself saying "X, so Y" in a Cause-Effect answer — stop and insert the mechanism between them. "X — and the way that actually produces Y is... — so Y." The mechanism sentence is the one doing the band-lifting work. The causes and effects are the scaffolding; the mechanism is the building.
And once you've explained one mechanism well, don't stop at the first effect — ask "and what does that effect then cause?" That second question takes you to the second-order effect, which is where Band 8 lives. One cause, one mechanism, one first-order effect, one second-order effect: that's a complete Band 7-8 Cause-Effect answer.
Some candidates resist this stage. "But I did answer — I gave the cause and the effect. Isn't that the whole question?" Worth addressing directly: a "why" question isn't asking you to name a cause. It's asking you to explain a process. The cause is the starting point of the answer, not the answer itself.
"When the examiner asks 'why,' they're not asking you to point at a cause — they could guess the cause themselves. They're asking you to walk them through the machinery: how does this cause actually turn into this effect? The candidate who explains the mechanism is doing the thing the question was really testing. Naming the cause is the easy 20% everyone can do. The mechanism is the 80% that separates the bands. So when you hear 'why,' don't reach for a cause and relax — reach for a cause and then immediately ask yourself 'how does that actually work?' The answer to that is what you say out loud."
Imagine the examiner asks: "Why do people in cities have fewer children?" The Jump Trap candidate says "because it's expensive, so they have fewer." The C·M·E candidate inserts the machinery: "the cost isn't really the direct cause — the real mechanism is that in a city, raising a child well requires buying things that used to be free in a village, like space, childcare, and supervised time, so each child represents a much larger and more visible trade-off against the parents' own goals, which makes people stop at one."
Notice the mechanism sentence does the work: it explains how urban living converts into fewer children, through the specific machinery of cost-visibility and trade-offs. That's the sentence the examiner is grading. The mechanism is where the reasoning becomes visible.
When you explain how a cause becomes an effect — instead of just asserting that it does — you're respecting the question. You're treating "why" as the real intellectual request it is, not as a prompt to recite a cause. The examiner hears a candidate who understands processes, not just facts. The explaining is the respect. Lesson 11 was the abstraction, Lesson 12 was the axis — Lesson 13 is the mechanism. Same move, different framework.
Two questions before we move to the C·M·E framework in Stage 3.
Three moves. Name the Cause. Explain the Mechanism. Name the Effect — then reach for the second-order effect. That's the framework. C·M·E is the third Part 3 micro-framework, used for Cause-and-Effect questions (~20% of Part 3). Same role for "why" questions that P·E·E played for opinions and X·Y·Z for comparisons: a skeleton that forces you to insert the mechanism instead of jumping straight to the effect.
The first move is deceptively simple: name one specific cause — not a list. The discipline here mirrors the X·Y·Z eXtract: commit to a single causal chain rather than spreading across many. One well-explained cause beats five named ones, every time. Pick the cause you can best explain the mechanism for.
The heart of the framework. This is the move the Jump Trap skips entirely — and the one the examiner is actually grading. Two sentences explaining how the cause produces the effect. Not that it does — how. This is where the reasoning becomes visible and the band gets decided.
"The test of a real mechanism is simple: could the listener have said it themselves before you spoke? If your 'mechanism' is just 'because it's expensive,' they already knew that — it's not a mechanism, it's a restatement of the cause. A real mechanism tells them something they hadn't assembled yet: the steps in between, the non-obvious link, the 'it's not X, it's actually Y.' If they nod and think 'huh, I hadn't connected those,' you've found the mechanism."
The third move names the effect — and then, for Band 8, pushes one step further to the second-order effect. After the cause and the mechanism, you have 10-15 seconds. Use the first few to name the obvious effect, then the rest to reach the non-obvious consequence of that effect. The second-order reach is to C·M·E what the Zoom was to X·Y·Z: the move that lifts the answer.
One full C·M·E in one block — Cause, Mechanism, Effect-plus-second-order. Notice how the moves layer: the cause names one specific driver (8-10 sec), the mechanism explains the machinery in two sentences (25-30 sec), the effect names the result and reaches past it (10-15 sec).
"I think the main driver is the rising cost of setting up an independent household in cities. The way that actually delays marriage is subtle — it's not that people can't afford a wedding, it's that marriage in Vietnam still implies a stable home and often supporting parents, so young people feel they need years of financial runway before they can take that on responsibly. The effect isn't just later marriage — it's that a whole new life stage has opened up that didn't exist for my parents: a decade of independent adulthood before family obligations begin, which is reshaping everything from how people save to what they think their twenties are for."
One sentence Cause. Two sentences Mechanism. One sentence Effect-plus-second-order. Total: 45-60 seconds, four to five sentences. The hardest discipline is staying in the Mechanism — resisting the jump to the effect. 10 sec Cause + 25-30 sec Mechanism + 10-15 sec Effect = 45-60 sec Cause-Effect answer.
Four picks. Pick a Cause-Effect question. Pick the cause (C). Pick the mechanism (M). Pick the effect + second-order (E). Watch the full C·M·E answer assemble in real time. Each question offers different causal chains — try at least two to see how the same architecture handles different "why" questions.
[pick a Cause-Effect question first]
~45-60 seconds of Cause-Effect answer built from four picks. Notice how the moves layer: the cause names one driver (10 sec), the mechanism explains the machinery in two sentences (25-30 sec), the effect names the result and reaches to the second-order consequence (10-15 sec). One chain, fully traced — never a jump.
Two questions before we move to the Vietnamese causal patterns in Stage 4.
Almost every Vietnamese Part 3 "why" question traces back to one of four big societal engines. If you have a ready mechanism for each, you'll rarely be caught jumping. This stage gives you the four engines and the best mechanism to deploy for each — so the Mechanism move becomes pattern-recall rather than on-the-spot invention.
"Why has X changed in recent decades?" The biggest engine — Vietnam compressed a century of change into a generation, so almost any "why has this changed" traces back here.
"Why do people move to / leave / behave differently in cities?" The mass movement from countryside to city drives a huge share of behaviour-change questions.
"Why has technology changed how people do X?" Phones, internet, and social media as causal engines — but the mechanism is rarely the device itself.
"Why do people work / spend / save differently now?" Rising incomes, new aspirations, and shifting job markets as the engine.
The biggest causal engine in Vietnamese Part 3. Recognisable by "in recent decades," "compared to the past," "as society develops." The core mechanism: Vietnam compressed change that took the West a century into a single generation, so the gap between what people grew up expecting and what now exists is unusually wide — and that gap is the engine behind most "why has this changed" questions.
The second engine. Recognisable by "moving to cities," "urban life," "leaving the countryside." The core mechanism: things that were free and ambient in a village — space, childcare, supervision, community, food from your own land — all become things you have to buy in a city. That single conversion drives an enormous range of behaviour change.
Two more engines. Technology's mechanism is rarely the device itself — it's the friction the device removes. Economic change's mechanism is the new trade-offs that rising options create. Both reward candidates who look past the obvious cause to the machinery underneath.
Notice the shared move across all four engines: the obvious cause is never the mechanism. "Technology" isn't a mechanism — "friction removed" is. "The economy" isn't a mechanism — "new trade-offs appear" is. "The city" isn't a mechanism — "free things become bought" is. In every case, the engine names the cause, but the band-lifting mechanism is one layer underneath. When you recognise the engine, reach past it to the pre-loaded mechanism.
Four Cause-Effect questions. Identify which of the four engines each one runs on — which tells you the pre-loaded mechanism to reach for. Two seconds per question.
Same architecture as the L7-12 vocabulary banks, applied to Cause-and-Effect construction. Eighty phrases that aren't "because of" or "so." Pick three from each category and you've replaced the jump reflex with the cause-naming, mechanism-explaining, and effect-reaching phrases that signal Band 7.
Phrases for the C move — committing to one specific cause. "I think the main driver is..." / "The biggest single factor is..." The signal that says "I'm choosing one cause to explain."
Phrases for the M move — the band-decider. "The way that actually works is..." / "What happens is..." / "It's not X, it's actually Y..." The signal that you're explaining the process.
Phrases for the first-order E — naming the direct result cleanly. "The result is..." / "Which leads to..." / "The immediate effect is..."
Phrases for the Band 8 reach (Stage 6) — the consequence of the consequence. "The deeper effect is..." / "And the self-reinforcing part is..." / "Which then causes..."
Twenty phrases for the C move — committing to one specific cause. These are your openers. The signal tells the examiner "I've chosen one cause to explain deeply," which separates you from the lister. Pick three favourites and rotate them.
Twenty phrases for the M move — the band-decider. These signal "I'm now explaining the process," which is exactly the move the examiner is listening for. The highest-leverage phrases in the whole bank. Pick three favourites and drill them until they're reflexive — they're your insurance against the Jump Trap.
Twenty phrases for the first-order E move — naming the direct result cleanly before you reach for the second-order effect. Keep these brief; they're the setup for the Band 8 reach, not the destination. Pick three favourites.
Twenty phrases for the Band 8 reach — the consequence of the consequence (the full move is built in Stage 6). These signal the highest-level Cause-Effect thinking: not just the obvious result, but the effect of the effect. Use one per Cause-Effect answer where you can see the downstream consequence.
Across the four categories — cause-naming, mechanism-explaining, effect-stating, second-order-launching — pick three favourites per category. Twelve phrases total that you can deploy on any Cause-Effect question. The mechanism phrases are your insurance against the Jump Trap; the second-order phrases are your reach for Band 8. Twelve phrases × four categories = a full Band 7-8 Cause-Effect vocabulary.
The Band 8 move on Cause-Effect questions. After the C·M·E chain is solid, the highest-level candidates do one more thing: they don't stop at the first effect — they ask "and what does that then cause?" The second-order effect is the consequence of the consequence, the downstream result most candidates never reach. It's the Single-Link Stop fix turned into a positive skill.
Q: "Why is remote work growing?" A: "Because technology makes it possible — the mechanism is that good internet and video tools mean physical presence is no longer required for most knowledge work. So more people work from home now."
...So more people work from home — but the deeper effect is that the office stops being where careers are quietly built, which disadvantages younger workers who used to learn by being in the room, so we may end up with a generation that's much harder to mentor and promote.
The whole move comes down to asking one more question after you name the first effect: "and what does that effect then cause?" The first effect is usually the obvious one the question half-implied. The second-order effect is the one that shows you've thought about the system, not just the single link. It's almost always more interesting — and it's almost always the part the examiner remembers.
The transition is simple: name the first effect, then pivot with "but the deeper effect is..." / "and what that in turn causes is..." / "the second-order consequence is..." One sentence of first-order, one sentence of second-order, and the answer lands at the top of the band.
Second-order effects come in three recognisable shapes. Knowing the shapes means you can reach for one quickly instead of inventing from scratch. When you've named your first effect, scan these three and grab whichever fits.
You don't need all four shapes — pick the one or two that come most naturally and keep them ready. The loop and the paradox are the most impressive because they're the most surprising, but the new-stage shape is the easiest to reach for and works on almost any "why has society changed" question. When in doubt, ask: does my first effect loop back, flip around, land elsewhere, or open something new?
Five worked second-order effects, each tied to a common Cause-Effect question. Each names the first effect briefly, then reaches the second-order one. Audio-clickable. Notice the shape of each — loop, paradox, displaced cost, or new stage.
The second-order effect is powerful, but it has a failure mode of its own: reaching for a dramatic downstream consequence that isn't actually plausible. A grounded second-order effect signals systems thinking; an exaggerated one ("so society will collapse") signals the opposite. Here's where the reach earns its keep and where it tips into overreach.
Reaches for the most extreme downstream consequence, skipping the plausible intermediate ones. Sounds alarmist rather than analytical. The examiner hears exaggeration, not insight.
Reaches exactly one link past the first effect, to something genuinely plausible and specific. Stays measured. The examiner hears someone who can trace a system without exaggerating it.
The discipline is simple: reach exactly one link past the first effect, and make sure that link is plausible enough that a thoughtful listener would nod. Don't chain three speculative consequences together to reach a dramatic endpoint — that's not depth, it's a slippery slope. One grounded second-order effect beats three escalating ones. If your second-order effect needs the word "completely," "collapse," or "disappear," you've probably overreached — pull back to the measured, specific consequence one step down.
Two questions before we move to the worked example in Stage 7.
The same Vietnamese candidate from Lessons 10-12. They've done their Part 2 and the Opinion and Compare-Contrast Part 3 sets. Now the examiner runs a Cause-Effect-heavy Part 3 — five "why" questions. Watch how the candidate names one cause, explains the mechanism, and reaches a second-order effect every time — never jumping, never listing.
Topic thread: social and generational change in Vietnam. The examiner moves through five Cause-Effect questions, each running on one of the four causal engines from Stage 4 — modernisation, urbanisation, technology, economic change.
Engine: Economic change. Full annotated C·M·E next.
Engine: Urbanisation. Mechanism = free contact becomes effortful.
Engine: Technology. Mechanism = friction removed from novelty.
Engine: Modernisation. Mechanism = lost the conditions that made them work.
Engine: Economic + Urbanisation, with a self-reinforcing loop second-order effect.
The first Cause-Effect question. Economic-change engine. Full C·M·E architecture. Toggle annotations to see each move. Click "Hear it spoken" to listen. Notice the three tagged sections — Cause (crimson), Mechanism (purple), Effect (teal).
"Why are young people in Vietnam marrying later than previous generations?"
I think the main driver is the rising cost of setting up an independent household in cities.
The way that actually delays marriage is subtle, though — it's not that people can't afford a wedding, it's that marriage in Vietnam still implies setting up a stable home and often supporting parents, so young people now feel they need years of financial runway before they can take that on responsibly.
The effect isn't just later marriage — it's that a whole new life stage has opened up that didn't exist for my parents: a decade of independent adulthood before family obligations begin, which is reshaping everything from how people save to what they think their twenties are for.
Two more questions, two more engines. Each names one cause, inserts the mechanism, and reaches a second-order effect. Watch the mechanism sentence in each — it's the part the Jump Trap would have skipped.
Examiner: "Why do people in cities seem to feel lonelier?"
(Cause) I think the main driver is the move to cities itself — specifically, what city living does to everyday contact. (Mechanism) The way that works is that a village gives you ambient, unchosen contact — you literally can't avoid people, you pass the same faces daily; a city makes every interaction chosen and effortful, so connection has to be deliberately scheduled, and anything that has to be scheduled can be postponed, and often is. (Effect + second-order) So the first effect is more isolation — but paradoxically it's isolation in a crowd, which is harder to recognise and admit than rural isolation ever was, so people often don't even name what they're feeling, which makes it harder to address."
Examiner: "Why do people find it harder to concentrate for long periods now?"
(Cause) The cause everyone names is phones, but the real driver is what phones removed — the friction that used to sit between us and novelty. (Mechanism) What happens is that fresh stimulation is now available instantly and endlessly, with zero effort, so the mind gets trained to expect a new hit of novelty every few seconds — and once it's trained that way, the slow, single-threaded focus that a book or a deep task demands starts to feel almost like deprivation. (Effect + second-order) So concentration gets harder — but the deeper effect is that the activities which used to build attention, like sustained reading, get abandoned precisely because they now feel hard, so the muscle weakens further for lack of use. It's self-reinforcing."
Notice the mechanism does the work in both. Q2's "free contact becomes effortful, scheduled things get postponed" and Q3's "the mind trained on novelty experiences focus as deprivation" are the sentences a Jump Trap answer would skip entirely — and they're exactly what lifts these to Band 7. Each also reaches a second-order effect: isolation-in-a-crowd that's hard to name (Q2), and the self-reinforcing attention loop (Q3).
The last two questions. Q4 runs a clean modernisation C·M·E. Q5 ends on the most advanced second-order shape — the self-reinforcing loop, where the effect worsens its own cause.
Examiner: "Why are traditional festivals celebrated less enthusiastically than before?"
(Cause) I'd trace it back to how fast Vietnam modernised — the festivals didn't change, but the world around them did. (Mechanism) The mechanism is that these festivals were tuned for a slow village world of fixed neighbours, shared preparation, and time that couldn't be spent any other way — and when you drop them into a fast urban world of scattered families and infinite alternatives, they lose the very conditions that made them feel effortless, so what used to be rhythm now feels like effort. (Effect + second-order) So they're observed more thinly — and the deeper effect is that they survive increasingly as performance rather than practice: the Tết photo gets posted, but the three days of shared cooking that gave the photo meaning quietly disappear."
Examiner: "Why do fewer young people want to work in agriculture?"
(Cause) I think the main driver is that farming has stopped reading as a path and started reading as a trap in young people's eyes. (Mechanism) The way that works is generational and visible: they watched their parents work brutally hard on small plots for incomes that never compounded into security, while classmates who left for factory or office work visibly pulled ahead — so agriculture now signals "the choice that guarantees you fall behind," regardless of whether that's strictly true. (Effect + second-order) So young people leave farming — and the self-reinforcing part is that the countryside loses exactly the educated young people who could have modernised it and made it pay, so the conditions that made farming unattractive get worse, not better, which pushes the next wave out even faster. It's a loop that feeds itself."
Five questions, five mechanisms, zero jumps. Q1 economic (new life stage), Q2 urbanisation (isolation in a crowd), Q3 technology (self-reinforcing attention loop), Q4 modernisation (performance not practice), Q5 the loop (farming exodus feeds itself). Each named one cause, inserted the mechanism the Jump Trap skips, and reached a second-order effect in one of the four shapes. Q5's self-reinforcing loop is the most advanced reach. This is what a Cause-Effect-heavy Part 3 sounds like at the top of the band.
Two diagnostic questions before we move to the Practice Arena.
Five real Cause-Effect answer excerpts. For each one, identify which of the four failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four jump excerpts. For each, write a 3-line C·M·E — name one cause, explain the mechanism (the missing middle), name the effect and reach one second-order step. Reveal each model after your attempt.
Three Cause-Effect questions. For each, identify the causal engine (Modernisation / Urbanisation / Technology / Economic), then write just the mechanism sentence — the "how" that connects cause to effect. The band-deciding move, drilled in isolation.
Three C·M answers with the first effect already named. Your job is to add the second-order effect — one plausible link further down the chain, in one of the four shapes (loop, paradox, displaced cost, new stage). One link only — remember the one-link rule, no slippery slopes.
Pick one Cause-Effect question — any from the Causal-Chain Builder in Stage 3, or from Exercise 2. Set a timer for 50 seconds. Deliver the full C·M·E aloud: name one cause, explain the mechanism for two sentences, name the effect and reach one second-order step. Three rounds — slow pass, natural pace, exam pressure.
Five Cause-Effect exercises done. Here's how it landed.
Your performance across the Cause-Effect arena suggests the C·M·E framework is taking root. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 13. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Cause-Effect drill this week.
Forty-five minutes invested in the lesson that fixes the most common Band 6 ceiling on "why" questions — the Jump Trap.
Thirteen lessons done. Three Part 3 frameworks in hand — P·E·E for Opinion, X·Y·Z for Compare-Contrast, C·M·E for Cause-Effect. Lesson 14 deep-dives the fourth and final Part 3 question type — Speculation (~15% of Part 3). The "what will happen" / "what would happen if" question. The mental move shifts from explaining the present to projecting a hypothetical forward.
"What will X look like in fifty years?" / "What would happen if...?" / "How might X change in the future?" Fifteen percent of Part 3 questions ask you to project forward — and most candidates either refuse to commit ("nobody can know the future") or make a flat one-line prediction with nothing behind it. Lesson 14 teaches the I·F·B architecture: a grounded If-statement, the First-order consequence, and the move Beyond it.
If-statement (a grounded premise, not a wild guess) · First-order consequence (the direct, near-certain result) · Beyond (the less obvious longer-term shift). It borrows the second-order reach from C·M·E and points it at the future.
The dominant Speculation failure mode: refusing to commit because "the future is uncertain." The Band 7 candidate commits to a grounded projection while signalling appropriate uncertainty — confidence and humility together.
How to anchor a future projection in a present-day trend so it sounds reasoned rather than invented. "If current trends continue..." done properly, with a real trend named.
The Band 8 move: projecting the consequence others miss — not "there will be more technology" but the subtle human or social shift that technology will produce.
"Three Part 3 frameworks down, one to go. P·E·E, X·Y·Z, C·M·E — opinions, comparisons, and causes. The Jump Trap has a name now, and the mechanism is the fix: never let a cause jump straight to an effect without explaining the how in between. Do the drill for a week and the 'but how exactly?' question will start firing on its own. Then Lesson 14 adds the last framework — I·F·B for the future questions — and after that, Lesson 15 teaches you to mix all four on the fly. The hardest conceptual work of Part 3 is now behind you."
Thirteen badges. The full Part 1 and Part 2 toolkits, plus three of the four Part 3 frameworks — P·E·E, X·Y·Z, and C·M·E. The Jump Trap has a name, a diagnostic, and a fix. Lessons 14-15 finish the Part 3 module.
The fourth and final Part 3 question type. After P·E·E for Opinion, X·Y·Z for Compare-Contrast, and C·M·E for Cause-Effect, this lesson teaches I·F·B for Speculation questions — roughly 15% of Part 3, the "what will happen" and "what would happen if" questions. The mental move shifts from explaining the present to projecting a hypothetical forward. The Band 5-6 candidate refuses to commit — "nobody can know the future." The Band 7 candidate commits to a grounded projection while signalling honest uncertainty: confidence and humility together.
"Speculation questions terrify people, because they feel like a trap — 'how could I possibly know what will happen in fifty years?' So candidates hedge into paralysis: 'it's hard to say, nobody can predict, maybe yes maybe no.' But the examiner isn't asking you to be right. They're asking you to reason about the future in a structured way. A grounded projection — 'if this present trend continues, then this follows, and beyond that, this subtler thing' — is exactly what they want. You anchor the guess in something real, you trace it forward, and you signal appropriate uncertainty without drowning in it. Lesson 14 teaches the I·F·B architecture that makes a future projection sound reasoned instead of either reckless or paralysed."
Two candidates. Same Speculation Part 3 question. Same Vietnamese background, same vocabulary, same grammar. Watch what happens when one hedges into paralysis (refusing to commit) and the other grounds a projection and traces it forward (I·F·B). Same question. Only one is at Band 7.
"How do you think family life in Vietnam will change over the next thirty years?"
Pure hedging, zero content. The candidate refuses the question out of fear of being "wrong" — but there's no wrong answer to grade, only the reasoning, which is absent. From the examiner's side, this sounds like someone who can't construct a hypothetical, not someone being appropriately cautious.
I·F·B architecture visible. Sentence 1 grounds the projection in a real present trend ("if current urbanisation continues"). Sentence 2 traces the direct first-order consequence. Sentence 3 reaches Beyond to the subtler human shift — and signals honest uncertainty ("I suspect") without collapsing into paralysis. Committed but humble.
Same 9-stage shape as Lessons 7-13, applied to Speculation questions. Focus: the I·F·B framework, the Hedge Paralysis failure mode, grounding the If-statement, and the non-obvious projection Band 8 move. This is the last of the four Part 3 frameworks.
You are here.
The 4 Speculation failure modes — Hedge Paralysis, Wild Guess, Flat Prediction, Present-Tense Drift — and why each one shows up.
The fourth Part 3 micro-framework. If-statement · First-order · Beyond. Plus the interactive Projection Builder.
How to anchor a future projection in a real present-day trend so it sounds reasoned. The Vietnamese trends most likely to appear, pre-loaded.
80+ Speculation phrases — projection-grounding, consequence-tracing, beyond-reaching, calibrated-hedging.
The Band 8 move: projecting the consequence others miss — the subtle human or social shift, not the obvious "more technology."
Five Speculation questions answered to Band 7 — same Vietnamese candidate, five I·F·B projections.
Five Speculation-specific exercises.
Self-assessment, badge, and Lesson 15 preview (Hybrid Part 3 — mixing all four frameworks).
If a "what will happen" question has ever made your mind go blank — or sent you into a spiral of "it's hard to say, nobody knows" — this stage is for you. The freeze isn't a vocabulary problem or a thinking problem. It's a fear problem: candidates believe a wrong prediction will be penalised, so they refuse to predict at all. But there is no wrong prediction to penalise — only present or absent reasoning. The fear is misdirected, and once you see that, the freeze melts.
Years of schooling trained you that questions have correct answers and wrong answers are marked down. A "what will happen in fifty years" question triggers that reflex — "I might be wrong" — and the safest-feeling move is to not commit. But IELTS Speculation questions have no answer key; the examiner can't mark a prediction wrong because the future hasn't happened. They can only grade whether you reasoned.
The Speculation descriptors reward the candidate who can construct a hypothetical and follow it through — "if this, then this, and beyond that, this." That's reasoning under uncertainty, which is a genuine cognitive skill. Refusing to guess demonstrates the opposite: an inability to build a hypothetical. The grounded guess scores; the refusal doesn't.
From the examiner's side, "nobody can know the future" sounds like a candidate who can't construct a hypothetical sentence — which is a real language-and-reasoning gap. "If urbanisation continues, then households will keep shrinking, and beyond that, elder care will have to be reinvented" sounds like a candidate who can build and extend a conditional. Same uncertainty about the actual future; completely different demonstration of skill.
The good news: this is the most purely psychological of all the Part 3 fixes. You already have the reasoning ability — you reason about hypotheticals all day ("if it rains, I'll take an umbrella"). The fix is permission: permission to commit to a grounded guess, knowing you can't be marked wrong, only marked absent.
Listen to enough Speculation Part 3 answers and the failures cluster into four patterns. Two come from too little commitment, two from the wrong kind of commitment. Recognising your shape makes the fix targeted.
The dominant default. The candidate refuses to commit to any projection, hiding behind "nobody can know." Hedging language fills the whole answer with no actual content underneath. The fear of being wrong produces an answer that can't be graded because there's nothing in it.
The opposite over-correction. The candidate commits boldly — but to a projection with no anchor in any present trend. A dramatic, ungrounded prediction that sounds like science fiction rather than reasoning. Commitment without grounding reads as recklessness, not insight.
The candidate makes one prediction — and stops. No grounding before it, no consequence traced after it. A bare one-line guess that answers the question literally but demonstrates no reasoning. The Speculation cousin of the Cause-Effect Jump Trap: the projection with nothing on either side.
The candidate is asked about the future but describes the present instead. They slide into "these days" and "nowadays," never actually projecting forward. The question asked "will," the answer keeps saying "is." Often the candidate doesn't notice they've answered a different question.
Same Vietnamese candidate, same Speculation question, two registers. Here's "How do you think people's working lives will change in the future?" in both modes. The refusal version is full of cautious language; the projection version commits to a grounded guess and traces it. Notice the refusal version is just as long — and yet only the projection lifts the band.
Fills the time with cautious language. Sounds responsible but commits to nothing. The examiner hears avoidance, not analysis.
Anchors in a present trend, commits to a direction, traces a first-order consequence, then reaches a subtler one — with honest uncertainty signalled lightly, not drowning the answer.
Whenever you feel the urge to say "nobody can know" — stop and convert it into a grounded conditional. "If [a real present trend] continues, then [a direct consequence], and beyond that [a subtler shift]." The hedge becomes the grounding ("if this continues" carries the uncertainty), and the projection carries the content. You're still being appropriately cautious — the "if" does that — but now there's something to grade.
And notice: a light touch of uncertainty actually strengthens a Band 7 projection. "I'd expect," "I suspect," "this seems likely" — these signal that you know you're projecting, not asserting fact. The skill isn't removing all hedging; it's hedging the right amount, in the right place, while still committing to content.
Some candidates resist this stage hardest of all. "But I genuinely don't know what will happen — isn't it dishonest to pretend I do?" Worth addressing directly: a grounded projection isn't a claim to know the future. It's a demonstration that you can reason forward from what's true now. Nobody is asking you to be a prophet.
"When the examiner asks 'what will happen,' they don't think you have a crystal ball — and they don't want you to pretend you do. They want to see you take something true about the present and reason forward from it carefully. That's not prediction; it's structured reasoning under uncertainty. The 'if' protects you — it says 'I'm reasoning from a premise, not claiming to know.' So you get to commit fully to the logic while staying honest about the uncertainty. Refusing to do this doesn't make you humble; it makes you look like you can't construct a conditional. Commit to the reasoning. Hedge the certainty. Those are two different things, and Band 7 candidates keep them separate."
Imagine the examiner asks: "Will people still read physical books in the future?" The Hedge Paralysis candidate says "it's hard to know." The I·F·B candidate reasons forward: "if the current revival of physical formats among people who grew up digital holds — and there are signs it will — then physical books won't disappear but will become a deliberate choice rather than a default, and beyond that, I'd guess reading a physical book becomes a small status signal, a way of saying 'I chose to slow down,' which is almost the opposite of what people predicted twenty years ago."
Notice the candidate never claims to know. Every move is conditional and hedged ("if... holds," "I'd guess"). But the reasoning is fully committed and fully traced. That's reasoning forward, not predicting.
When you take a real present trend and reason it forward — instead of refusing the question — you're respecting it. You're treating "what will happen" as the genuine invitation to think that it is, not as a trap to dodge. The examiner hears a candidate who can build and extend a hypothetical, which is exactly the skill being tested. The committing is the respect. Lesson 11 was the abstraction, 12 the axis, 13 the mechanism — Lesson 14 is the projection. The eighth and final member of the set. Same move, different framework.
Two questions before we move to the I·F·B framework in Stage 3.
Three moves. Ground the If-statement. Trace the First-order consequence. Reach Beyond. That's the framework — the fourth and final Part 3 micro-framework, used for Speculation questions (~15% of Part 3). Same role for "what will happen" questions that P·E·E played for opinions, X·Y·Z for comparisons, and C·M·E for causes: a skeleton that forces you to anchor, commit, and extend instead of freezing.
The first move is the one that defeats the freeze. Instead of guessing wildly or refusing to guess, you anchor your projection in a real present-day trend: "If [a trend that's genuinely happening now] continues..." This single move converts an impossible question ("predict the future") into an answerable one ("reason forward from what's already true").
Once the If is grounded, trace the direct consequence — the near-certain, obvious result that follows if the trend holds. Two sentences. This is the committed part: having anchored in a real trend, you now follow it forward without hedging every word. The "if" already did your hedging; here you commit to the logic.
"People get the proportions backwards. They hedge the first-order consequence to death — 'maybe possibly perhaps people might leave cities' — and then make the Beyond a wild unhedged leap. Flip it. Commit to the first-order, because it follows almost certainly from your grounded If. Save your hedging for the Beyond, where the uncertainty is genuinely higher. Confident on the near, humble on the far. That's the rhythm of a Band 7 projection."
The Band-lifting move. After the grounded If and the direct First-order consequence, you reach Beyond — to the subtler, less obvious shift that the first-order effect produces. This is the same second-order reach you learned in C·M·E, now pointed at the future. The Beyond is where the human or social insight lives, the part the wild guesser and the freezer both miss.
One full I·F·B in one block — grounded If, committed First-order, hedged Beyond. Notice how the moves layer: the If anchors in a real trend (10-12 sec), the First-order traces the direct result with light hedging (20-25 sec), the Beyond reaches the subtler human shift with honest uncertainty (12-15 sec).
"If the current trend toward smaller, more dispersed families continues — which seems likely given urbanisation — then the most direct change is that the multigenerational household becomes the exception rather than the norm within a generation, and care for elderly parents shifts from something ambient and shared to something deliberately arranged. But beyond that, I suspect the deeper change is emotional: filial duty won't disappear, it'll just have to be expressed in new ways — money and scheduled visits rather than daily presence — which may feel like both progress and loss at the same time."
One sentence If. Two sentences First-order. One sentence Beyond. Total: 45-60 seconds, four to five sentences. The hardest discipline is grounding the If instead of freezing — and remembering to hedge the Beyond, not the First-order. 10 sec If + 20-25 sec First-order + 12-15 sec Beyond = 45-60 sec Speculation answer.
Four picks. Pick a Speculation question. Pick the grounded If. Pick the First-order consequence. Pick the Beyond. Watch the full I·F·B answer assemble in real time. Each question offers different projections — try at least two to see how the same architecture handles different "what will happen" questions.
[pick a Speculation question first]
~45-60 seconds of Speculation answer built from four picks. Notice how the moves layer: the If anchors in a real trend (10 sec), the First-order traces the direct result (20-25 sec), the Beyond reaches the subtler human shift with honest uncertainty (12-15 sec). Grounded, committed, and humble — never frozen.
Two questions before we move to grounding the If-statement in Stage 4.
The If-statement only works if it's anchored in a real, recognisable trend. Most Vietnamese Speculation questions can be grounded in one of four big present-day trends. If you have these ready, you'll never freeze — you'll always have a real anchor to project forward from. This stage gives you the four trends and the projections they unlock.
"Will technology change how we...?" The most common Speculation anchor — the ongoing move of more life online, more automation, more frictionless digital contact.
"How will cities / where people live change?" The continuing flow into cities, plus the new ability to live and work anywhere.
"How will family / society change?" Vietnam's falling birth rate and ageing population — a powerful, well-documented anchor for family and social questions.
"How will attitudes / lifestyles change?" Rising incomes shifting priorities from survival toward meaning, experience, and individual choice.
The most common Speculation anchor. Recognisable in questions about technology, communication, work, shopping, learning. The grounding move: "if more of life keeps moving online, as it clearly is..." The key, just like in C·M·E, is to project the human consequence of the technology, not the technology itself.
The second anchor. Recognisable in questions about cities, housing, where people live, travel. The grounding move: "if people keep flowing into cities — but also gain the ability to work from anywhere..." The interesting tension is that two trends pull in opposite directions, which makes for rich projections.
Two more anchors. Ageing/Demographics is Vietnam's quietly powerful trend — a falling birth rate and rising lifespan that few candidates think to use, which makes it impressive. Prosperity/Values anchors questions about changing attitudes and lifestyles in rising incomes.
Notice the shared move across all four anchors: the Beyond is always human, never technological. Digitalisation's Beyond is about closeness, not devices. Ageing's Beyond is about how duty gets re-expressed, not about numbers. The technology, the demographics, the economics are the First-order; the Beyond is always what it does to how people feel, relate, and live. When you recognise the anchor, ground the If, trace the obvious First-order, then reach for the human Beyond.
Four Speculation questions. Identify which of the four trend anchors each one is best grounded in — which gives you a ready If-statement to project from. Two seconds per question.
Same architecture as the L7-13 vocabulary banks, applied to Speculation. Eighty phrases that aren't "it's hard to say" or "maybe yes maybe no." Pick three from each category and you've replaced the freeze reflex with the grounding, tracing, reaching, and calibrated-hedging phrases that signal Band 7.
Phrases for the I move — anchoring in a real trend. "If current trends continue..." / "Assuming the present shift toward X holds..." The signal that says "I'm grounding, not guessing."
Phrases for the F move — committing to the direct result. "The most direct change would be..." / "The obvious first effect is..." The signal that you're following the logic forward.
Phrases for the B move — the Band 8 reach. "But beyond that..." / "The subtler shift, I suspect, is..." The signal that you're pushing past the obvious to the deeper human consequence.
Phrases for honest uncertainty without paralysis. "I'd expect..." / "my guess is..." / "it may even..." The light touch that signals you know you're projecting — applied to the Beyond, not the whole answer.
Twenty phrases for the I move — anchoring in a real trend. These are your openers, and they're the cure for Hedge Paralysis: the moment you say "if current trends continue," you've committed to having something to say. Pick three favourites and rotate them.
Twenty phrases for the F move — committing to the direct consequence. These follow the grounded If and trace the obvious first result. Remember: commit here, lightly hedged. One "would" or "I'd expect" is enough — don't pile on the hedges. Pick three favourites.
Twenty phrases for the B move — the Band 8 reach. These pivot from the obvious first-order consequence to the subtler human shift beneath it. Use one per Speculation answer, right after the first-order. Pick three favourites.
Twenty phrases for honest uncertainty without paralysis. These are the light touch that signals "I know I'm projecting" — the antidote to both the Wild Guess (no hedge) and the Hedge Paralysis (all hedge). Sprinkle one or two, mostly on the Beyond. Pick three favourites.
Across the four categories — projection-grounding, consequence-tracing, beyond-reaching, calibrated-hedging — pick three favourites per category. Twelve phrases total that you can deploy on any Speculation question. The grounding phrases get you unfrozen; the hedging phrases keep you from over-committing. Twelve phrases × four categories = a full Band 7-8 Speculation vocabulary.
The Band 8 move on Speculation questions. After the I·F·B is solid, the highest-level candidates make the Beyond non-obvious — they project the consequence almost nobody else thinks to mention. Not "there will be more technology" but the subtle human or social shift that the technology produces. It's the Beyond move from Stage 3, sharpened to its most insightful form.
Q: "How will technology change communication?" A: "If digital communication keeps growing, then we'll message more and call less, and beyond that, people will use even more apps and devices to stay in touch in the future."
...then we'll message more and call less — but beyond that, I suspect the real shift is that constant low-effort contact quietly replaces the rarer, deeper kind, so we may end up more connected and less close at the same time, and loneliness may rise in the middle of more communication than ever.
The obvious projection extends the trend in a straight line: more technology, more cities, more change. The non-obvious projection asks a sharper question — "what will this do to how people feel, relate, or live that nobody expects?" It looks for the human consequence, the paradox, the thing that's almost the reverse of what the straight-line projection would predict. The technology is the First-order; the human shift is the non-obvious Beyond.
This is exactly the C·M·E second-order move from Lesson 13, now pointed at the future. There, you asked "and what does that effect then cause?" Here, you ask "and what will that change do to people that isn't obvious?" Same reach, future tense, aimed at the human.
Non-obvious projections come from three reliable places. When you've named your first-order consequence, scan these three and grab whichever fits — it's faster than inventing insight from scratch.
You don't need all four — pick the one or two that come most naturally and keep them ready. The paradox is the most memorable because it's the most surprising; the redefinition is the easiest to reach for and works on almost any "will X survive?" question; the displacement and new divide reveal the most systems thinking. When in doubt, ask: does my first-order change flip into its opposite, quietly redefine something, shift value elsewhere, or open a new gap?
Five worked Beyonds, each tied to a common Speculation question. Each names the first-order briefly, then reaches the non-obvious human shift. Audio-clickable. Notice the source of each — paradox, redefinition, displacement, or new divide.
The non-obvious projection is powerful, but it has its own failure mode — the same one C·M·E's second-order had: reaching for something dramatic that isn't actually plausible. A grounded non-obvious Beyond signals insight; a sci-fi leap signals the Wild Guess returning. Here's where the reach earns its keep and where it tips into recklessness.
Reaches for a sensational future that has no plausible path from the first-order. Sounds like science fiction, not reasoning. The examiner hears the Wild Guess sneaking back in through the Beyond.
Reaches a genuinely non-obvious human shift that still follows plausibly from the first-order. Surprising but defensible. The examiner hears insight, not sci-fi.
The discipline is the same as C·M·E's one-link rule, adapted: your non-obvious Beyond must have a plausible path from the first-order consequence. A listener should be able to follow the logic — "ah, yes, if that happens, then I can see how this follows." If your Beyond needs a leap nobody can trace (brain implants, language disappearing, society collapsing), you've slipped back into Wild Guess. Non-obvious doesn't mean impossible; it means "true but overlooked." Reach for the surprising-yet-defensible, not the sensational-yet-groundless.
Two questions before we move to the worked example in Stage 7.
The same Vietnamese candidate from Lessons 10-13. They've done their Part 2 and the Opinion, Compare-Contrast, and Cause-Effect Part 3 sets. Now the examiner runs a Speculation-heavy Part 3 — five "what will happen" questions. Watch how the candidate grounds an If, commits to a first-order, and reaches a non-obvious Beyond every time — never freezing, never leaping.
Topic thread: the future of life in Vietnam. The examiner moves through five Speculation questions, each best grounded in one of the four trend anchors from Stage 4 — digitalisation, urbanisation/mobility, ageing/demographics, prosperity/values.
Anchor: Ageing/Demographics. Full annotated I·F·B next.
Anchor: Digitalisation. Beyond = paradox of connection.
Anchor: Urbanisation/Mobility. Beyond = city redefined as luxury.
Anchor: Digitalisation + Mobility. Beyond = informal learning lost.
Anchor: Prosperity/Values. Beyond = paradox of choice.
The first Speculation question. Ageing/demographics anchor. Full I·F·B architecture. Toggle annotations to see each move. Click "Hear it spoken" to listen. Notice the three tagged sections — If-statement (magenta), First-order (purple), Beyond (amber).
"How do you think family life in Vietnam will change over the next thirty years?"
If the current trend toward smaller, more dispersed families continues — which seems likely given urbanisation —
then the most direct change is that the multigenerational household becomes the exception rather than the norm within a generation, and care for elderly parents shifts from something ambient and shared to something that has to be deliberately arranged, and often paid for.
But beyond that, I suspect the deeper change is emotional: filial duty won't disappear, it'll just have to be expressed in new ways — money and scheduled visits rather than daily presence — which may feel like both progress and loss at the same time.
Two more questions, two more anchors. Each grounds an If, commits to a first-order, and reaches a non-obvious Beyond. Watch the Beyond in each — it's always the subtler human shift, never just "more of the same."
Examiner: "How do you think the way people communicate will change?"
(If) If digital communication keeps replacing in-person contact, which it clearly has been, (First-order) then the obvious first change is that we'll be in touch with far more people, far more often — distance basically stops mattering, and staying connected becomes effortless and constant. (Beyond) But beyond that, I suspect the real shift is a paradox: constant low-effort contact may quietly crowd out the rarer, effortful kind that actually builds closeness — so we could end up more connected and less close at the same time, with loneliness rising in the middle of more communication than the world has ever had."
Examiner: "What do you think cities will look like fifty years from now?"
(If) If remote work keeps making where you live independent of where you earn, as it's started to, (First-order) then the direct change is that some people leave expensive cities for smaller towns, because the job that once chained them there no longer does, so the relentless one-way pull into the biggest cities may finally ease. (Beyond) But beyond that, my guess is the city gets redefined: it stops being the place you must be for work and becomes the place you choose to be for culture and energy — so living in a big city shifts from an economic necessity to something closer to a lifestyle luxury, which completely changes who lives there and why."
Notice the Beyond does the work in both. Q2's "more connected and less close" paradox and Q3's "city redefined from necessity to luxury" are the moves that lift these to Band 8 — the subtler human shifts a straight-line projection would miss. And in both, the hedging sits on the Beyond ("I suspect," "my guess is"), while the first-order is stated with confidence. The rhythm holds.
The last two questions. Q4 reaches a displacement Beyond (informal learning lost). Q5 ends on the paradox of choice — the same C·M·E mechanism from Lesson 13, now projected into the future.
Examiner: "How will people's working lives change in the future?"
(If) If the shift toward remote and flexible work keeps going, which seems very likely, (First-order) then the part I'm confident about is that the rigid nine-to-five in a fixed office breaks down, and work becomes something defined by output and projects rather than hours and presence. (Beyond) But beyond that, I suspect the subtler cost lands on younger workers: the office was quietly where careers got built through informal contact — overhearing, being mentored, absorbing the unwritten rules — and if that disappears, we may produce a generation that's technically capable but much harder to bring up to speed, so companies may have to reinvent on purpose what the office used to do by accident."
Examiner: "Do you think people's values and priorities will change in the future?"
(If) If incomes keep rising so that basic survival is assured for more and more people, which has been the long trend in Vietnam, (First-order) then the clear first change is that priorities shift from security toward meaning — people start chasing experiences, self-expression, and purpose rather than just stability, the way my generation already differs from my grandparents'. (Beyond) But beyond that, paradoxically, I'd expect this to bring more anxiety, not less: when survival isn't the question anymore, the question becomes 'am I living the right life?', and that's a far harder one to answer — so a generation freed from material worry may find itself burdened instead by the weight of endless choice."
Five questions, five grounded projections, zero freezes. Q1 ageing (filial duty redefined), Q2 digitalisation (more connected, less close), Q3 mobility (city as luxury), Q4 work (informal learning lost), Q5 values (paradox of choice). Each grounded an If in a real trend, committed to the first-order, and reached a non-obvious human Beyond from one of the four sources. Q5's paradox of choice is the C·M·E mechanism from Lesson 13, projected forward — the frameworks connecting. This is what a Speculation-heavy Part 3 sounds like at the top of the band.
Two diagnostic questions before we move to the Practice Arena.
Five real Speculation answer excerpts. For each one, identify which of the four failure modes from Stage 2 the candidate fell into. Recognition first, fixing second.
Four frozen excerpts. For each, write a 3-line I·F·B — ground the If in a real trend, commit to the first-order, reach a non-obvious Beyond. Reveal each model after your attempt.
Three Speculation questions. For each, identify the trend anchor (Digitalisation / Urbanisation-Mobility / Ageing / Prosperity), then write the grounded If-statement plus the first-order consequence. The two moves that defeat the freeze, drilled together.
Three I·F answers with the first-order already given. Your job is to add the Beyond — a non-obvious human shift from one of the four sources (paradox, redefinition, displacement, new divide). Plausible path only — remember, non-obvious means "true but overlooked," not "sensational but groundless."
Pick one Speculation question — any from the Projection Builder in Stage 3, or from Exercise 2. Set a timer for 50 seconds. Deliver the full I·F·B aloud: ground the If in a real trend, commit to the first-order, reach a non-obvious Beyond. Three rounds — slow pass, natural pace, exam pressure.
Five Speculation exercises done. Here's how it landed.
Your performance across the Speculation arena suggests the I·F·B framework is taking root. Move into Stage 9 to complete the lesson — and the entire core Part 3 module.
Six sliders covering everything new in Lesson 14. Be honest about which ones still need work — that's the diagnostic that tells you which screens to revisit when you do Lionel's Speculation drill this week.
Forty-five minutes invested in the lesson that turns the most-feared Part 3 question — "what will happen?" — into a structured, gradable answer.
Fourteen lessons done — and all four Part 3 frameworks now in hand. P·E·E for Opinion, X·Y·Z for Compare-Contrast, C·M·E for Cause-Effect, I·F·B for Speculation. Lesson 15 is the one that ties them together: real Part 3 questions rarely come labelled, and the examiner often mixes types within a single thread. Lesson 15 teaches you to recognise which framework a question wants — and to switch between them on the fly.
You now have four frameworks. The final Part 3 skill is using them together: hearing an unlabelled question and instantly knowing whether it wants a P·E·E opinion, an X·Y·Z comparison, a C·M·E causal chain, or an I·F·B projection — and handling the questions that genuinely blend two types. This is where the four micro-frameworks become one fluent system.
A fast diagnostic for unlabelled questions: the verb and question-word tell you which framework to reach for. "Why" → C·M·E. "Will" → I·F·B. "Better/worse, compare" → X·Y·Z. "Do you think, should" → P·E·E.
The questions that mix two types — "why has X changed, and will it continue?" (C·M·E + I·F·B) — and how to structure an answer that does both without losing the thread.
Drills for moving between frameworks mid-conversation, so a five-question Part 3 that jumps between types never catches you flat-footed.
By the end of Lesson 15, all four frameworks work as one toolkit — the full Part 3 module complete, ready for the core-skills lessons and the mock test.
"All four Part 3 frameworks, done. P·E·E, X·Y·Z, C·M·E, I·F·B — opinions, comparisons, causes, and the future. That's the entire conceptual map of Part 3 in your hands. The freeze has a name now, and the cure is the grounded If: never refuse the future, anchor it in something real and reason forward. Do the drill for a week and 'if this trend continues' will start firing on its own. Then Lesson 15 teaches you to mix all four on the fly — and after that, Part 3 holds nothing you haven't already met. The hardest part of the whole speaking test is now behind you."
Fourteen badges. The full Part 1 and Part 2 toolkits, plus all four Part 3 frameworks — P·E·E, X·Y·Z, C·M·E, and I·F·B. The freeze has a name, a diagnostic, and a cure. Just Lesson 15 — mixing all four — finishes the Part 3 module.
The Part 3 capstone. You've built four micro-frameworks — P·E·E for Opinion, X·Y·Z for Compare-Contrast, C·M·E for Cause-Effect, I·F·B for Speculation. But in a real Part 3, the examiner doesn't announce which type each question is, and a single thread often jumps between them — an opinion question, then a "why," then a "what if." Lesson 15 teaches the last skill: reading an unlabelled question to know instantly which framework it wants, handling questions that genuinely blend two types, and switching between frameworks on the fly without losing the thread.
"Here's what nobody tells you about Part 3: the questions don't come with labels. The examiner won't say 'this is a cause-and-effect question.' They'll just ask 'why do you think that's happening?' and it's on you to hear the 'why' and reach for C·M·E. Knowing four frameworks isn't the same as knowing which one a question wants. The candidate who freezes for two seconds deciding has already lost fluency. The Band 8 candidate hears the question-word, the verb, the shape of the ask — and the right framework is just there, automatically. That recognition is a skill of its own, and it's the one thing standing between four separate tools and one real Part 3 ability. That's all of Lesson 15."
Two candidates. Same Part 3 question. Both know all four frameworks. The difference is whether they read what the question is actually asking for — or reach for the wrong framework and answer a question that wasn't asked. Same knowledge. Only one is fluent.
"Why do you think traditional skills like handicrafts are being lost in many countries?"
The candidate heard "handicrafts" and "traditions" and reached for an Opinion P·E·E — arguing they should be preserved. But the question asked why they're being lost, a Cause-Effect question wanting C·M·E. A perfectly fluent answer to a question nobody asked. The examiner notes that the candidate didn't actually address the "why."
The candidate read the question-word "why," recognised a Cause-Effect question, and reached for C·M·E — a cause (lost economic reason), a mechanism (the broken chain of transmission), and a sharp effect. The framework fits the question. The examiner hears a candidate answering exactly what was asked, with the right tool.
Same 9-stage shape as Lessons 7-14, but this one synthesises rather than introduces. Focus: the framework-selector, the recognition reflex, blended questions, and switching on the fly. This is the capstone of the entire Part 3 module.
You are here.
The 4 framework-selection failure modes — Default Framework, Keyword Misfire, Blend Blindness, Switch Lag — and why each one shows up.
The fast diagnostic for unlabelled questions — question-word and verb to framework. Plus the interactive Framework-Selector trainer.
The questions that genuinely need two frameworks — "why has X changed, and will it continue?" — and how to structure an answer that does both.
80+ phrases for signposting which framework you're using, switching cleanly between them, and stitching a blended answer together.
The Band 8 move: moving between frameworks mid-thread so a five-question Part 3 that jumps between types never catches you flat-footed.
One unbroken Part 3 thread — five questions across all four types, the candidate switching framework each time.
Five mixed-framework exercises.
Self-assessment, badge, and Lesson 16 preview (Core Skills — Pronunciation).
Here's the strange thing about this stage: by now you have all four frameworks. You can build a P·E·E, an X·Y·Z, a C·M·E, an I·F·B. So why would a candidate who owns four good tools still stumble? Because having tools and choosing the right one are two different skills. A carpenter with a full toolbox who reaches for a hammer when the job needs a saw isn't short of tools — they're short of reading the job. Part 3 is the same: the failure isn't execution, it's selection.
Under pressure, the brain reaches for whatever is most familiar or whatever a surface word triggers — not for what the question actually asks. You hear "handicrafts" and your mind jumps to "tradition is important" (an opinion) before you've registered that the question-word was "why" (a cause). The misfire happens in the first half-second, before conscious thought catches up. That's why it needs a trained reflex, not just knowledge.
An examiner grades whether you answered the question that was asked. A flawless P·E·E delivered to a "why" question doesn't read as fluent — it reads as someone who didn't understand the question, which is worse than a clumsy but on-target answer. The single fastest way to lose a band in Part 3 isn't bad English; it's confidently answering a question nobody asked.
Think back over Lessons 11-14. Each taught you to execute a framework — to build the move once you knew which one you needed. None of them taught you to choose, because each lesson handed you the question type on a plate: "this is an Opinion lesson, here's P·E·E." The real test never does that. It hands you an unlabelled question and watches whether you reach for the right tool in the half-second before you start talking.
The good news: selection is fast to learn because the cues are reliable. The question-word and the verb almost always tell you which framework fits. Once you've drilled the mapping — "why" means C·M·E, "will" means I·F·B, "compare" means X·Y·Z, "do you think" means P·E·E — the reach becomes automatic. This stage shows you the four ways selection fails, so you can catch each one.
Unlike the failure modes in Lessons 7-14, which were about executing a framework badly, these four are about choosing the wrong framework in the first place. The English might be flawless; the tool is simply mismatched to the job.
The candidate has one favourite framework — usually P·E·E, because opinions feel safest — and reaches for it no matter what the question asks. Every question becomes "I think... because... for example." The tool never changes, so most questions get answered slightly off-target.
A surface word in the question triggers the wrong framework before the candidate registers the actual question-word. The topic noun hijacks the selection. "Handicrafts" pulls them toward an opinion about tradition; "young people and old people" pulls them toward a comparison — even when the question-word was "why."
The question genuinely asks for two frameworks, but the candidate answers only the first half and stops. "Why has this changed, and will it continue?" needs C·M·E and I·F·B — but the candidate explains the cause and never projects forward, leaving half the question untouched.
The candidate can pick the right framework, but freezes when the question type changes mid-thread. Question 1 was an opinion, Question 2 is suddenly a "why," and the gear-change costs two or three seconds of dead air and a stumbling start. The tools are there; the switching is slow.
Same Vietnamese candidate, same Part 3 question, two registers. Here's "Some people prefer to live in big cities while others prefer the countryside. Why do you think these preferences differ so much between generations?" answered two ways. The first misreads it as a comparison; the second reads the actual question-word and answers it.
The words "cities" and "countryside" trigger an X·Y·Z comparison. The candidate compares the two places fluently — and never answers the "why do preferences differ between generations."
The candidate spots that the real question-word is "why" and the subject is "differ between generations" — a Cause-Effect question about generational difference, wanting C·M·E, not a comparison of places.
Before you answer, find the question-word (why, will, do-you-think, which-is-better) and the true subject of the ask. The topic nouns — cities, handicrafts, generations — are scenery; they tell you what to talk about, not which framework to use. The question-word tells you the framework. When the two seem to conflict — a comparison-looking topic with a "why" question-word — the question-word always wins.
This is a one-second habit: hear the question, locate the question-word, name the framework silently, then start. That one second of reading is what separates a fluent answer to the right question from a fluent answer to the wrong one.
Some candidates resist this stage with a fair objection: "If my English is good and my answer is interesting, does it really matter that it's a slightly different angle?" Worth addressing directly: yes, more than almost anything else. Relevance is graded explicitly, and a brilliant answer to the wrong question signals exactly the comprehension gap the test is built to catch.
"I've heard candidates give genuinely impressive answers and still score a 6, and it's almost always the same reason: they answered the question they wished they'd been asked, not the one they were. The examiner's first silent question about every answer is 'did they actually address what I asked?' If the answer is no, nothing else you do quite recovers it. So the most valuable single second in all of Part 3 is the one before you speak — the second where you actually read the question, find the 'why' or the 'will' or the 'should,' and pick up the right tool. Four frameworks are worthless if you grab the wrong one. The reading is the skill that makes the other four pay off."
Imagine the examiner asks: "Do you think people will rely on traditional medicine less in the future?" There are two cues here pulling in different directions. "Do you think" looks like an opinion (P·E·E); "in the future" looks like speculation (I·F·B). The reader resolves it in a second: the heart of the ask is "will... in the future," so it's primarily an I·F·B projection — but the "do you think" invites you to commit to a view, so you ground the projection in your own stance. One second of reading turns a confusing double-cue into a clear plan: I·F·B, lightly opinionated.
The misreader, by contrast, grabs the first cue they notice ("do you think" → opinion) and delivers a P·E·E about whether traditional medicine is good — never projecting forward, missing the "will." Same question, same one-second fork, completely different band. Reading the question is choosing the fork deliberately instead of by accident.
When you take the one second to read what the question actually asks — instead of answering the question its topic words suggest — you're respecting it. You're treating the examiner's exact wording as something worth understanding before you respond, which is the foundation every other Part 3 skill sits on. The examiner hears a candidate who listens precisely. The reading is the respect. Lessons 11-14 each gave you a "respect is in" — the abstraction, the axis, the mechanism, the projection. Lesson 15's is the one that lets the others land: the reading.
Two questions before we move to the Framework-Selector in Stage 3.
The whole selection skill comes down to one reliable rule: the question-word and the verb tell you which framework to reach for. Not the topic, not the vibe — the grammatical shape of the ask. Learn this mapping until it's reflexive, and the right framework arrives before you've finished hearing the question. This is the master diagnostic that sits above all four frameworks.
Every framework has a family of question-cues — some obvious, some disguised. Here's the complete map. The disguised cues are the ones that catch people: a question that looks like one type but is grammatically another. Read down each card and notice the "disguised as" row especially.
Selection has to be fast — you can't pause for five seconds analysing grammar before every answer. So it runs as a fixed one-second routine: three micro-steps that happen between hearing the question and starting to speak. Drill these until they collapse into a single instinct.
"The whole routine is: hear, name, open. Hear the question-word, name the framework in your head, open with that framework's first move. It takes about a second and it runs underneath the little filler you're already saying — 'That's an interesting question' buys you exactly the second you need to run hear-name-open. By the time the filler's done, the right tool is in your hand and you're already moving. Drill it on every practice question until you stop noticing you're doing it."
You don't have to run the routine in dead silence. The natural fillers you already use — "That's a good question," "Hmm, let me think," "That's interesting" — are perfect cover for the one-second routine. While your mouth says the filler, your mind runs hear-name-open. The examiner hears a thoughtful pause; you've secretly selected the right framework. This is why fluent candidates never seem to scramble: the selection is hidden inside a phrase that sounds like natural engagement.
Most questions are easy to read. The ones that trip people up are the disguised ones — where the topic or surface wording suggests one framework but the grammar demands another. Here are the four most common disguises and how to see through each.
The pattern across all four disguises: the surface suggests one thing, the operative word demands another. Disguise 1 and 3 both involve "why" — but one is a real cause question and one is a value-justification (opinion). The difference is what follows: "why does X happen" = mechanism (C·M·E); "why is X important/good" = stance (P·E·E). Train your ear to hear not just "why" but what kind of "why" it is.
Eight unlabelled Part 3 questions, including disguised ones. For each, read the question-word and pick the framework it wants. Instant feedback explains the cue. The goal isn't to get them all right first try — it's to build the half-second reflex. Work through all eight; your score tracks at the bottom.
Eight questions, four frameworks, several disguises. The more you drill this, the faster the question-word jumps out and the framework arrives on its own. By exam day, "hear, name, open" should run in under a second — no conscious analysis, just the right tool appearing in your hand.
Two questions before we move to blended questions in Stage 4.
Some Part 3 questions genuinely ask for two things at once — "why has this changed, and will it continue?" These aren't disguises; they're honest two-part questions, and the examiner expects both halves answered. The skill is hearing the "and," recognising the two frameworks, and delivering a clean two-part answer that switches between them without losing the thread. Four blend-patterns cover almost all of them.
"Why has X happened, and will it continue?" Explain the mechanism, then project it forward. The most common blend.
"Do you think X is a problem, and why is it happening?" Take a position, then explain the mechanism behind it.
"How do X and Y differ, and which do you prefer?" Weigh them on an axis, then commit to a side.
"How will X change, and is that a good thing?" Project forward, then evaluate the projection.
The most common blend in Part 3. "Why has X happened, and do you think it will continue?" The structure is natural: explain the mechanism (C·M·E), then use that same mechanism to project forward (I·F·B). The mechanism you just explained becomes the engine of your projection — the two halves connect seamlessly.
Q: "Why are urban families having fewer children, and will that continue?" "The main driver is that in a city, things that were free in a village — space, childcare, supervision — all become things you have to buy, so each child becomes a much bigger, more visible trade-off against the parents' own goals. (hinge) Whether that continues really comes down to whether that cost pressure eases — and I don't see it easing. (I·F·B) If anything, as cities get denser and more expensive, the trend deepens; and beyond simply fewer children, I suspect we'll see childhood itself become a more deliberate, heavily-invested 'project,' which may raise the bar for having any children at all."
Two more blends, both pairing a position with another framework. The key in each: lead with the half the question leads with, then pivot cleanly to the second. The signpost word ("and that's because" / "and personally") makes the switch audible.
Notice the pattern: in both blends, the second half builds on the first. The opinion in Blend 2 sets up the mechanism that explains it; the comparison axis in Blend 3 becomes the very reason for the preference. A good blended answer never feels like two stapled-together responses — the first half earns the second. That connective quality is what lifts a blend from "answered both parts" to Band 8.
The commonest blended-answer mistake after Blend Blindness is the lopsided answer: a luxurious first half and a rushed, tacked-on second. The examiner wants both halves genuinely addressed, which means roughly even time. Here's how the lopsided version fails and the balanced version works.
Pours everything into the first framework, then notices the second half and bolts on a rushed sentence. The second framework gets a token nod, not a real answer — barely better than Blend Blindness.
Gives each framework a real, compact turn. The first half is tight enough to leave room for a genuine second half. Both halves get developed; neither is rushed.
The fix for lopsidedness is counterintuitive — you have to deliberately compress the half you find easier to leave room for the other. Most candidates over-invest in the first framework because they've already started and momentum carries them. Catch yourself at the hinge: when you hear yourself finishing the first half, that's the cue to switch, not to add one more example. A blended question is a budget split two ways, not one budget with a tip added at the end.
Two questions before we move to the Master Transition Bank in Stage 5.
The previous lessons gave you vocabulary inside each framework. This bank gives you the connective tissue between them — the signposts that announce which framework you're using, the hinges that switch cleanly from one to another, and the openers that buy you the one-second selection routine. Eighty phrases that turn four separate tools into one fluent performance.
Phrases that buy the one-second selection routine. "That's an interesting one..." / "Let me think about that..." Natural fillers that hide the hear-name-open while you pick the framework.
Phrases that announce which framework you're using. "The real question here is why..." / "If I project that forward..." Telling the examiner you've read the question and picked the right tool.
Phrases that move from one framework to another in a blend. "...and whether that continues..." / "...and given that, personally..." The clean pivot that connects two halves.
Phrases that rescue a wrong start. "...although, actually, the more interesting angle is..." The graceful pivot when you realise mid-answer you grabbed the wrong framework.
Twenty phrases that buy you the one-second selection routine. While your mouth says these, your mind runs hear-name-open. They sound like natural engagement; they're secretly cover for picking the framework. Pick three favourites and make them automatic.
Twenty phrases that announce which framework you've chosen — five for each. They tell the examiner "I read the question and I'm reaching for the right tool." Using the matching signpost is how you make your selection visible. Pick a favourite for each framework.
Twenty phrases that move cleanly from one framework to another in a blended answer. The hinge is the seam between the two halves — done well, it makes a two-framework answer feel like one connected thought rather than two stapled pieces. Pick three favourites for the blends you'll meet most.
Twenty phrases for when the one-second routine misfires and you start with the wrong framework. Everyone misreads occasionally under pressure — the Band 8 skill isn't never misreading, it's recovering gracefully. These phrases let you pivot mid-answer without it sounding like a stumble. Pick three to keep in your back pocket.
Across the four categories — thinking-time openers, framework signposts, switch hinges, recovery phrases — pick three favourites per category. Twelve phrases that turn four separate frameworks into one fluent system. The openers buy you the selection second; the signposts make your choice visible; the hinges connect blended halves; the recovery phrases rescue a misfire. Twelve phrases × four categories = the connective tissue of a Band 8 Part 3.
The Band 8 move. A real Part 3 isn't one question — it's a thread of four or five, and the type often changes with each one. The examiner asks your opinion, then a "why," then a "what if," then a comparison. Each switch is a gear-change, and the candidate who freezes for two seconds at each one bleeds fluency. Switching on the fly means running the one-second routine fresh on every question, so the gear-changes become invisible.
Q1 (opinion): smooth P·E·E. Q2 (suddenly "why"): "Um... well... that's... why is it... um, I think because..." — two seconds of dead air at the gear-change, then a stumbling C·M·E that never quite recovers its footing.
Q1 (opinion): P·E·E. Q2 ("why"): "That's a good question — [one-second routine runs under the filler] — the main driver, I think, is..." The gear-change is hidden inside a four-word filler; the C·M·E starts cleanly, as if there was never a switch at all.
Here's the reassuring part: switching on the fly isn't a new skill. It's the Stage 3 one-second routine — hear, name, open — run fresh on every question instead of just the first. The reason candidates stall at switches is that they relax after the first answer and stop reading; the second question catches them still in the previous framework's gear. The fix is simply to treat every question as a fresh selection, never assuming the next one is the same type as the last.
The filler is your friend here. Between questions, a two-second "that's interesting" or "hmm, good question" is completely natural — and it's exactly long enough to run the routine. Fluent candidates aren't thinking faster; they're using the filler as cover to think at all. The dead air of Switch Lag is just the routine happening without the cover.
When a switch goes wrong, it's almost always one of three causes. Knowing them lets you catch yourself in the act and apply the fix before the stall becomes audible.
Notice that all three killers are about state, not knowledge. You know the frameworks; the switch fails because you're carrying momentum, working without cover, or depleted from over-answering. That's good news — state is controllable. A clean reset, a ready filler, and disciplined answer-length between questions, and the switches take care of themselves. The frameworks were never the problem; the transitions were.
Five mid-thread switches, each showing the move from one framework to the next with the filler-and-open built in. Audio-clickable. Listen for how the filler buys the second, then the new framework's first move starts cleanly. Notice the switch is never announced — it just happens, smoothly.
Step back and see the complete Part 3 system you've built across five lessons. Every Part 3 question runs through the same loop: read it, select the framework, deliver it, switch to the next. Here's the whole machine in one view.
Inside step 2, the selection, sits the whole library you've built:
Read, select, deliver, switch — looped across every question in the thread, with a filler hiding each selection. That's the entire Part 3 system: four frameworks held inside one selection routine, connected by transitions, repeated fluently. You don't think "which lesson was this?" any more — you hear the question-word and the right tool is just there. That's what fourteen lessons were building toward: not four separate skills, but one fluent Part 3 ability.
Two questions before we move to the worked example in Stage 7.
The same Vietnamese candidate from Lessons 10-14, now in a full, unbroken Part 3 thread. The examiner stays on one topic — work and education — but the question type changes with almost every turn. Watch the candidate read each question, select the right framework, deliver it, and switch cleanly to the next. This is everything from Lessons 11-15 working at once.
Topic thread: work, education, and skills. Five questions, four frameworks — the examiner never signals which type each one is. The candidate runs the read-select-deliver-switch loop on every turn.
Cue: "do you think... worth it" → P·E·E (opinion). Full annotated answer next.
Cue: "why" → C·M·E (cause-effect). Switch from opinion to cause.
Cue: "how does X differ" → X·Y·Z (compare). Switch from cause to comparison.
Cue: "will change... next twenty years" → I·F·B (speculation). Switch to the future.
Cue: "why... and will it last" → C·M·E → I·F·B blend. The capstone two-framework answer.
The first question. The candidate reads "do you think... worth it," recognises an opinion question, and reaches for P·E·E. Toggle annotations to see the read-select-deliver in action. The magenta tag is the selection moment; then Position, Evidence, Extension.
"Do you think a university degree is still worth it these days?"
"That's a fair question — [hears "do you think... worth it" → P·E·E]
my honest view is that it's still worth it, but for a completely different reason than it used to be.
A degree used to be worth it for the specific knowledge it gave you, but most of that knowledge is now free online; what it's actually worth now is the structure — four years of learning how to think, meet deadlines, and work with people you didn't choose.
So I'd say the degree is worth it, but we should be honest that we're now paying for the training in self-discipline more than the information itself — which makes it a much harder thing to justify for some fields than others."
Two gear-changes. Q2 switches from the opinion to a "why" (C·M·E); Q3 switches again to a "how does X differ" (X·Y·Z). Watch the filler-and-signpost open each one — the switches are clean, never announced.
Examiner: "Why do so many graduates end up in jobs unrelated to their degree?"
(filler + select) "Hmm, good question — (C·M·E signpost) the main driver, I think, is a mismatch in timing. (mechanism) Students choose a degree at eighteen based on what looks promising then, but the job market moves faster than a four-year course, so by graduation the field they trained for has often shifted, shrunk, or filled up — and meanwhile they've picked up general skills that transfer anywhere. (effect) So they drift into whatever's hiring and growing, which is rarely the narrow thing their degree named — and the degree ends up working as a general signal of capability rather than a specific qualification."
Examiner: "How does learning a skill on the job differ from learning it in a classroom?"
(filler + select) "Right — (X·Y·Z signpost: extract the axis) the real difference isn't formal versus informal, it's the order you meet the problem and the theory. (Y-values: weigh both) In a classroom you get the theory first and hunt for a problem to apply it to, so it can feel abstract and easy to forget; on the job you hit the problem first and reach for whatever theory solves it, so the learning sticks because it arrives exactly when you need it. (Zoom: the telling detail) You can see it in one detail: a classroom learner can pass a test and still freeze on a real task, whereas the on-the-job learner can do the task fluently but sometimes can't explain why it works — each is missing exactly what the other has."
Two clean switches. Q2 opened with "Hmm, good question" (filler) then "the main driver is" (C·M·E signpost) — the candidate heard "why" and was in the cause-gear instantly. Q3 opened with "Right" then "the real difference is" (X·Y·Z signpost) — heard "how does X differ," switched to comparison. Neither switch was announced; each was hidden inside a one-word filler and locked in by the framework's opening signpost.
The last two. Q4 switches to a pure I·F·B speculation. Q5 is the capstone — a blended C·M·E → I·F·B that needs two frameworks in one answer, with a clean hinge between them. The whole module, demonstrated.
Examiner: "How do you think the way we work will change over the next twenty years?"
(filler + select) "That's a big one — (I·F·B: ground the If) if the shift toward automating routine tasks keeps going, which it clearly is, (First-order) then the obvious change is that the routine, repeatable parts of most jobs shrink, and what's left is the judgement, creativity, and human-contact parts that machines can't do. (Beyond, hedged) But beyond that, I suspect the subtler shift is that 'a job' stops being a fixed thing you hold and becomes a rolling set of skills you keep refreshing — so the real security moves from the employer to your own adaptability, which is liberating and exhausting in equal measure."
Examiner: "Why has remote work grown so much, and do you think it will last?"
(select: two cues, blend) "Two parts there, so let me take both — (C·M·E half) the why is that the technology quietly removed the thing that used to require an office: real-time collaboration stopped needing physical presence, so the daily commute became a cost with no remaining purpose for a lot of knowledge work. (the hinge) And whether it lasts really depends on whether that purpose stays gone — (I·F·B half) which I think it mostly does. If anything, the tools keep improving, so I'd expect a permanent hybrid rather than a full return; and beyond the logistics, the deeper effect is that companies lose the informal, in-the-room way junior people used to learn, so the lasting challenge won't be productivity, it'll be how you grow people who were never in the room."
Five questions, four frameworks, one blend, zero dead air. Q1 read "do you think" → P·E·E; Q2 heard "why" → switched to C·M·E; Q3 heard "differ" → switched to X·Y·Z; Q4 heard "will change" → switched to I·F·B; Q5 heard "why... and will it last" → recognised a blend and delivered C·M·E → I·F·B with a clean hinge. Every selection was hidden inside a filler; every framework opened with its signpost. This is the complete Part 3 system — read, select, deliver, switch — running fluently across a whole thread. Everything Lessons 11-15 built, in one continuous performance.
Two diagnostic questions before we move to the Practice Arena.
Five candidates each answered a Part 3 question. The English is fine in every case — the problem is selection. For each, identify which of the four selection failure modes from Stage 2 occurred. Recognition first.
Four unlabelled questions, including disguised ones. For each, name the framework the question-word wants, then write a filler + the framework's first-move signpost — exactly what you'd say in the first two seconds. Reveal a model after your attempt.
Three blended questions, each needing two frameworks. For each, name the two frameworks in order, then write a short two-part answer with a hinge connecting them. Remember the proportion rule — compress the first half so both halves get real time.
Here's a four-question thread on one topic — technology and communication. For each question, write only the first line of your answer: a filler + the right framework's signpost. The skill being drilled is the clean switch, four times in a row, never assuming the next question is the same type.
The capstone drill — and the closest thing to the real Part 3 in this whole course. Below is a five-question thread that uses all four frameworks plus a blend. Set a timer and deliver the whole thread aloud, switching framework on every question. This is the complete system, performed.
Q1: "Do you think people change jobs too often these days?" (P·E·E)
Q2: "Why has changing jobs become so common?" (C·M·E)
Q3: "How does a stable career differ from a flexible one?" (X·Y·Z)
Q4: "How will the idea of a career change in the future?" (I·F·B)
Q5: "Why do young people prefer flexibility, and will that last?" (C·M·E → I·F·B blend)
Five mixed-framework exercises done — the full Part 3 system, tested. Here's how it landed.
Your performance across the mixed-framework arena shows how fluently the whole Part 3 system is coming together. Move into Stage 9 to complete the lesson — and the entire Part 3 module.
Six sliders covering everything new in Lesson 15. Be honest about which ones still need work — that's the diagnostic that tells you what to drill in Lionel's master mixed-thread exercise.
Forty-five minutes that turned four separate frameworks into one fluent Part 3 system — and completed the entire Part 3 module, the hardest part of the speaking test.
Stop and take this in. Fifteen lessons complete — and with Lesson 15, the entire Part 3 module is finished. Part 1 (Lessons 1-5), Part 2 (Lessons 6-10), and now all of Part 3 (Lessons 11-15). Every question type in the speaking test now has a framework, and you can select between them fluently. What remains isn't new question types — it's polish: the core delivery skills that lift any answer, and a full mock test to put it all together.
A different kind of lesson. The frameworks gave you what to say; the core-skills lessons sharpen how you say it. Lesson 16 starts with pronunciation — not accent elimination, but the specific features examiners actually score: word stress, sentence stress, intonation, and the connected speech that makes you sound fluent rather than word-by-word. The single highest-leverage delivery skill, and the first of four core-skills lessons.
The four pronunciation features in the band descriptors — and why a strong accent never caps your band but flat stress and intonation can.
Word stress, sentence stress, and the rising-falling patterns that carry meaning — the Vietnamese-speaker-specific patterns to watch.
The linking, weak forms, and elisions that turn word-by-word speech into the natural flow that reads as fluency.
Lessons 16-19 sharpen pronunciation, fluency, vocabulary range, and grammar — the delivery skills that lift every framework you've built.
"The whole Part 3 module — done. Four frameworks, and the skill to choose between them in a heartbeat. I want you to feel the weight of that: Part 3 is the part candidates fear most, the part that separates a 6.5 from a 7.5, and you now have a complete system for it. There's nothing in the unscripted depths of Part 3 you haven't been handed a tool for. From here it's polish — how you sound, how you flow, the range of your words, the grip of your grammar. Those lessons lift everything you've built. But the hard conceptual work, the architecture of thought? That's finished. Be proud of that. Then keep going — the summit's in sight."
Fifteen badges. All three parts of the speaking test now have complete toolkits — Part 1, Part 2, and the full four-framework Part 3 system with fluent selection between them. The hardest conceptual work of the whole course is behind you. What remains is polish and a mock test.
Welcome to the Core Skills module. The fifteen lessons behind you built what to say — every question type now has a framework. The next four sharpen how you say it, and we start with the highest-leverage one: pronunciation. Here's the first thing to unlearn — pronunciation is not about losing your accent. The examiner explicitly does not score your accent. What they score is whether English's music is there: the stress that punches the right syllables, the flow that links your words, and the melody that moves your pitch for meaning. For a Vietnamese speaker, that music is the single biggest, fastest win available.
"Let me kill the biggest myth first: you do not need to sound British or American. The band descriptors say it in black and white — a strong first-language accent is fine, as long as you're easy to understand. I've passed students with thick accents at Band 8. What holds Vietnamese speakers back isn't the accent — it's flatness. Vietnamese gives every syllable the same length and uses pitch to change the meaning of a word; English does the opposite — it stretches the important syllables, crushes the unimportant ones, links words together, and uses pitch to carry feeling and meaning across a whole sentence. When a Vietnamese speaker brings the flat, syllable-by-syllable habit into English, every word is clear but the music is gone — and without the music, the examiner has to work to follow you. This lesson gives you the music. It's three moves, and it's the closest thing to a magic trick I teach."
Here's one sentence, spoken two ways. The words are identical — what changes is the music. Click each one and listen. Then read the markup: CYAN CAPS = stressed (long & strong), grey = weak (short & reduced), ‿ = words linked together, ↗↘ = pitch rising and falling.
"I think the most important thing is to spend time with your family."
Every syllable gets the same length and weight, the pitch barely moves, and each word is pronounced separately like beads on a string. It's perfectly clear word-by-word, but the examiner has to do the work of finding the meaning — nothing in the delivery points to what matters. This is the syllable-timed habit carried straight from Vietnamese.
The content words — MOST, imPORtant, SPEND, FAMily — are stretched long and hit hard, while the little words (the, is, to, with, your) are crushed almost to nothing. Words link into each other instead of sitting apart, and the pitch falls at the end to signal "I'm finished." Now the delivery itself points at the meaning — the examiner hears what matters without effort. Same words, but the music does half the work.
The same 9-stage shape you know from the framework lessons, now aimed at delivery. The "framework" here is S·F·M — Stress, Flow, Melody — the three layers of English's music. Everything is built for the Vietnamese speaker's specific starting point: syllable-timed and tonal.
You are here.
The syllable-timed and tonal diagnosis, and the 4 pronunciation failure modes that follow from it.
Stress, Flow, Melody — the three layers of English's music. Plus the interactive Stress-Marker builder.
The specific sounds Vietnamese speakers struggle with — final consonants, clusters, troublesome pairs — and the word-stress rules that fix most errors.
Stress patterns, linking patterns, intonation patterns, and the priority sound-fixes — with audio for each.
The Band 8 move — grouping words into meaningful chunks with pauses and pitch, so long answers stay clear and natural.
A real Part 3 answer, marked up for stress, flow, and melody — and spoken, so you hear the music applied.
Five pronunciation exercises.
Self-assessment, badge, and Lesson 17 preview (Core Skills — Fluency).
Flat English isn't carelessness — it's two deep features of Vietnamese transferring directly into a language built on the opposite principles. Understand the transfer and every pronunciation error you make stops being random and starts being predictable. Two root causes explain almost all of them.
In Vietnamese, every syllable gets roughly the same length and weight — a steady, even beat, like a row of identical drops. English is the opposite: it's stress-timed, meaning a few syllables are stretched long and hit hard while everything between them is squeezed short. Carry the Vietnamese even-beat into English and you get the "machine-gun" effect — every syllable equal, none standing out, so the listener can't hear which words matter.
In Vietnamese, pitch is part of the word — change the tone and you change the meaning (ma, má, mà, mả, mã, mạ are six different words). So pitch is "spent" at the word level and held fairly fixed. English uses pitch completely differently: it moves across a whole sentence to carry meaning, emotion, and grammar — a question rises, a statement falls. A Vietnamese speaker, used to pitch belonging to words, tends to keep it flat across the sentence, and the English melody disappears.
Because the cause is structural, the fix is too. You're not "bad at pronunciation" — you're applying Vietnamese rules to English, and the two languages happen to be built on opposite principles for rhythm and pitch. Once you consciously switch rules — stretch-and-crush instead of even beat, moving pitch instead of fixed pitch — the music appears almost immediately. That's why pronunciation is the fastest core skill to improve: you're not learning a hundred sounds, you're flipping two switches.
The two transfers produce four recognisable failure modes. Two come from syllable-timing (rhythm), two from tonality and sound habits. Learn to hear each one in your own speech — recognition is the first half of the fix.
From syllable-timing. Every syllable gets equal length and weight, so speech comes out as a flat, even stream — rat-a-tat-a-tat — with nothing stretched and nothing crushed. It's clear word by word but exhausting to follow, because the listener gets no help finding the important words.
From Vietnamese sound habits. Vietnamese rarely ends syllables with strong consonants or clusters, so final sounds get dropped or softened in English — "work" becomes "wer," "asked" becomes "ass," "friends" becomes "fren." This quietly wrecks grammar too: the dropped ending is often the -ed or -s that carries tense or plural.
From tonality. Pitch stays level across the whole sentence — no rise on a question, no fall to close a statement, no lift on the important word. Because Vietnamese spends pitch on individual words, the speaker doesn't use it to shape the sentence, so everything sounds monotone and the meaning-signals disappear.
From the absence of word stress in Vietnamese (where each syllable is its own unit). English words have one strong syllable, and stressing the wrong one can make a word genuinely hard to recognise — "PHO-to-graph" vs "pho-TO-gra-phy," "com-FOR-table" stressed as "COM-for-TABLE." Even with perfect sounds, wrong word-stress trips the listener.
A short Part 1 answer, delivered two ways. The first carries all four failure modes; the second applies all three S·F·M moves. Click both, then read the markup. This is the whole lesson in miniature — what you'll be able to do by Stage 9.
Question: "Do you enjoy cooking?"
Question: "Do you enjoy cooking?"
Listen to a recording of yourself answering a question. Which of the four are you doing? Most people can hear it immediately once they know what to listen for — the even machine-gun beat, the vanished endings, the flat line, the wrong-syllable word. You don't have to fix all four at once. Fix the loudest one first; usually it's the Machine-Gun or the Flat Line, because they run through every sentence.
Many Vietnamese learners spend years trying to flatten their accent — softening vowels, chasing a "native" sound — and it barely moves their band, because accent was never what held them back. The reframe: leave the accent alone and put all that energy into rhythm and pitch, which is what's actually being scored.
"I had a student, Mai, who was convinced her accent was the problem. She'd recorded herself a hundred times trying to sound American and got nowhere. I made her do one thing: forget the accent completely, and just exaggerate the stressed syllables — almost cartoonishly, stretching them like elastic. Within a week she jumped half a band. Her accent hadn't changed at all. She'd just stopped firing every syllable at the same volume and started playing the rhythm. That's the whole secret. The examiner doesn't need you to sound English; they need to be able to ride the rhythm of your speech without effort. Give them the beat and the tune, keep your accent, and you're easy to understand — which is exactly the words in the Band 7 descriptor."
The Band 7 pronunciation descriptor centres on one phrase: the candidate "can generally be understood throughout" and "shows all the positive features of Band 6 and some of Band 8." The Band 8 version is "easy to understand throughout; L1 accent has minimal effect on intelligibility." Notice what's not there: nothing about sounding native, nothing about losing your accent. The whole scale is about effort — how hard the listener has to work. Flatness makes them work hard; the music makes it effortless. That's the entire game.
When you give your speech the beat and the tune — stretching what matters, crushing what doesn't, letting your pitch move — you're doing the listener's work for them. You're making your meaning effortless to follow instead of leaving them to decode a flat stream. That's a kind of respect, and the examiner feels it as "easy to understand." The frameworks gave the examiner your ideas; the rhythm hands those ideas over without friction. The respect is in the rhythm.
Two questions before we move to the S·F·M framework in Stage 3.
The music of English is three layers stacked on top of each other. Get all three and your delivery is Band 7-8 regardless of your accent. They build in order — Stress is the foundation, Flow connects, Melody shapes — and each one directly cures one of the failure modes you just met.
The foundation. English has two levels of stress, and both matter: word stress (one strong syllable inside each word) and sentence stress (content words punched, function words crushed). The single mental shift is from "every syllable equal" to "stretch the important, crush the rest."
"The instinct fights you here. Everything in your training says 'pronounce every sound clearly,' and now I'm telling you to swallow half of them. But that's the secret of the beat: English isn't clear-clear-clear-clear, it's CLEAR-mumble-mumble-CLEAR. The strong syllables only sound strong because the weak ones are genuinely weak. If you refuse to crush the small words, you can't punch the big ones — they all stay the same size, and you're back to the machine-gun. Be brave about crushing."
A practical test: say a sentence and tap the table only on the stressed syllables. In good English, the taps fall at roughly even time intervals no matter how many small words sit between them — that's "stress-timing." "The CAT sat on the MAT" and "The CAT has been sitting on the MAT" take almost the same time to say the stressed parts, because the extra small words get crushed to fit the beat. That's the rhythm you're building.
English doesn't leave gaps between words — it glues them together so a phrase comes out as one connected stream. Native speakers don't say "an apple," they say "a‿napple." This linking is what turns choppy, word-by-word speech into the smooth flow that reads as fluency. Three linking patterns cover most of it.
"Linking is the difference between 'I... want... to... go... home' and 'I‿wanna‿go‿home.' Vietnamese speakers tend to put a tiny stop between every word, like full stops between bricks. English smears the bricks together into a wall. You don't need to learn every rule consciously — just practise saying common phrases as if they were one long word: 'a‿lo‿tof,' 'kin‿dof,' 'wha‿da‿you‿think.' Slur them deliberately. It'll feel lazy and wrong, and that's exactly when it starts sounding right."
Flow and Stress work together: you crush the function words and link them to their neighbours, so a phrase like "a lot of" becomes a single fast "a‿lo‿təv" sitting between two stressed words. The stressed content words stay as clear islands; everything between them flows into one connected, crushed stream. That combination — clear peaks, flowing valleys — is the core texture of natural English.
The layer that defeats the Flat Line. In English, pitch moves across the whole sentence to carry meaning — and because Vietnamese spends pitch on individual words, this is the move that feels strangest and pays off most for sounding natural rather than robotic. Three pitch patterns do almost all the work.
"Vietnamese speakers often tell me moving their pitch feels 'too much,' like they're being dramatic or fake. Here's the truth: what feels like 'too much' to you sounds like 'normal and engaged' to an English ear, and what feels 'normal' to you sounds flat and bored to them. The calibration is just different. So deliberately overshoot — make your statements fall further than feels natural, your questions rise higher. You won't sound dramatic. You'll sound, for the first time, like you mean what you're saying."
Melody and Stress are linked: the pitch-lift usually lands on the most stressed content word. So in practice, the three S·F·M layers fuse into one motion — you punch a content word (Stress), the words around it flow into it (Flow), and your pitch lifts on it then resolves (Melody). One stressed word can carry all three at once. That fusion is what you'll build on the next screen.
The interactive builder. Below is a sentence with every word as a tappable tile. Your job: tap the content words (nouns, main verbs, adjectives, adverbs) to stress them — leave the function words (a, the, is, to, of, and) unstressed. When you've marked it, hear how your version sounds, then reveal the model. Three sentences to work through.
The more you mark sentences this way, the more automatic it becomes to hear which words carry meaning and deserve a punch. Content words get the beat; function words get crushed and linked. Do this with any sentence you read and your machine-gun habit dissolves.
Two questions before we move to the sounds and stress patterns in Stage 4.
S·F·M is the music; this stage is the specific notes Vietnamese speakers most often miss. There are dozens of small sound differences between Vietnamese and English, but only a handful actually affect whether you're understood. We'll focus only on the high-value ones — the sounds and stress patterns that, fixed, give the biggest jump in clarity for the least effort.
The endings Vietnamese drops — and why landing them fixes pronunciation and grammar at once.
The stacked consonants English loves and Vietnamese avoids — "asked," "strengths," "world."
The specific sounds that don't exist in Vietnamese — th, the v/w pair, the l/n endings.
A few reliable patterns that predict which syllable to stress in most words.
The two highest-value sound fixes, both from the same root: Vietnamese syllables mostly end in a vowel or a soft nasal, so English's hard final consonants and consonant clusters feel unnatural and get dropped. Deliberately landing them is the fix.
A few English sounds have no Vietnamese equivalent, so they get swapped for the nearest Vietnamese sound. These are medium-value — worth fixing, but only after endings and clusters. Here are the three that most affect being understood.
A realistic note on priority: a swapped "th" or "v" rarely stops you being understood — context usually carries it, and the descriptors forgive accent. So don't pour months into perfecting "th" while your endings are still dropping. Fix in value order: endings and clusters first (they affect grammar and clarity most), then these pairs as polish. Perfectionism on individual sounds is the classic way to spend a lot of effort for very little band.
Word stress feels random, but a handful of suffix rules predict it reliably — which cures the Wrong-Syllable failure mode for a huge number of words at once. Learn these four patterns and you'll stress most academic words correctly without memorising each one.
These suffixes appear on exactly the kind of sophisticated, multi-syllable words that earn you Lexical Resource marks — "communication," "possibility," "democracy," "psychology." So getting their stress wrong doesn't just hurt pronunciation; it makes your best vocabulary harder to recognise, undercutting the very words you reached for to sound advanced. Learning four rules protects a whole tier of your vocabulary at once.
Five multi-syllable words. Use the suffix rules to pick the stressed syllable. Instant feedback names the rule. This builds the reflex of seeing a suffix and knowing the stress before you even say the word.
The other lessons gave you phrase banks; this one gives you a pattern bank — eighty short items to say out loud, each one drilling a specific piece of the music. These aren't phrases to memorise and deploy; they're reps to build the muscle. Click each to hear a reference, then say it yourself, exaggerating the feature. Four categories, one per layer of S·F·M plus the priority sound-fixes.
Sentences and phrases with the stress marked, to drill the stretch-and-crush. Say each one tapping the table on the punched words — feel the even beat between stresses.
Common phrases that link, to drill the flow. Slur each one as if it were a single long word — "a‿lo‿təv," "kin‿dəv," "wha‿də‿yə‿think."
Statements, questions, and lists with the pitch marked, to drill the melody. Overshoot each rise and fall until it feels almost too much.
Minimal pairs and cluster words for the high-value sound fixes — final consonants, clusters, th, v/w. Say each pair slowly, exaggerating the difference.
Twenty items to drill the stretch-and-crush. The CYAN CAPS are the punches; everything else crushes. Say each one tapping the table only on the punched words — the taps should fall at roughly even intervals.
Twenty common phrases that link. The ‿ shows where words glue together. Say each as a single long word — slur it deliberately. It'll feel lazy; that's the sound of flow.
Twenty items to drill the melody. ↘ = pitch falls, ↗ = pitch rises. Overshoot every one — make the falls drop further and the rises lift higher than feels natural.
Twenty drills for the high-value sounds. Say each slowly, exaggerating the target. The minimal pairs (two words that differ by one sound) train your ear and mouth to keep them apart.
Across the four categories — stress, linking, intonation, sound-fixes — pick five from each and drill them aloud once a day. Twenty reps, two minutes. Exaggerate every feature: punch the stresses harder, slur the links lazier, overshoot the pitch, spit the final consonants. Exaggeration in practice settles to natural in speech. Two minutes a day for two weeks rebuilds the muscle. This is the one bank you train with your mouth, not your memory.
The Band 8 move. Once Stress, Flow, and Melody are working, the last layer is chunking — grouping words into meaningful units, with a tiny pause and a pitch reset between them. It's how fluent speakers keep long, complex answers clear: not one breathless run-on, not choppy word-by-word, but a series of bite-sized thought groups, each delivered as a unit. It's the rhythm of thinking itself.
Run-on, no groups: "I think the main reason is that people are very busy these days and they don't have time to cook so they eat out more and this affects their health because restaurant food is less healthy." — one long breath, no internal structure, the listener loses the thread.
Grouped into thoughts: "I think the main reason / is that people are very busy these days / so they don't have time to cook / which means they eat out more / and that affects their health / because restaurant food / is usually less healthy." — each "/" is a tiny pause and pitch reset.
A chunk — or "thought group" — is a small cluster of words that belong together as one idea: a phrase, a clause, a unit of meaning. Inside a chunk, the words link and flow with no pauses (that's your Flow). Between chunks, there's a micro-pause and the pitch resets to start the next group. Chunks are usually 3-7 words — short enough to deliver cleanly, long enough to carry a real piece of meaning. You're not pausing randomly; you're pausing at the seams between ideas.
Chunking does three jobs at once. It gives you thinking time — those micro-pauses are where you plan the next chunk. It makes you clear — the listener gets one idea at a time. And it makes you sound in control — measured rather than rushed or hesitant. It's the single biggest difference between a Band 7 and a Band 8 delivery of the same words.
Chunking isn't random pausing — the breaks fall at predictable, grammatical seams. Learn these three and your chunking will sound natural rather than arbitrary. The rule of thumb: break where a comma would go, or where a listener could take a breath without losing the meaning.
A chunk-break isn't just a pause — it's also a small pitch reset. At the end of a non-final chunk, your pitch typically rises slightly or stays up (signalling "more coming"), then resets to start the next chunk fresh. Only the final chunk falls to close. So "I think the main reason / is that people are busy / so they eat out more / which affects their health↘" has three little rises holding the listener, then one fall to release them. That combination of micro-pause plus pitch movement is what makes chunking sound like fluent thought rather than mechanical stopping.
Five real answer-openings, marked with chunk-breaks (/) and spoken. Click each to hear the pausing and pitch resets. Notice how the breaks fall at grammatical seams, and how each chunk is a clean unit of meaning.
Step back and see the complete delivery system. Four layers, stacked: chunk the answer into thought groups, and inside each chunk apply Stress, Flow, and Melody. Once it's automatic, it's not four separate things — it's one continuous musical motion.
Take the chunk "the main reason is that people are busy." In one motion you: chunk it off with a micro-pause before and after; stress the content words (MAIN REAson... PEOple... BUSy); flow the function words between them ("the," "is that," "are" all crush and link); and move the melody (a slight rise at the end signalling "more coming"). Four layers, one breath. You're not thinking about them separately any more — you're just delivering a thought with its natural music.
Chunk into thoughts, and inside each thought: stress the meaning-words, flow the rest, move the pitch. That's the entire music of English — and it works on every single answer you give, in all three parts of the test. The frameworks decide what you say; S·F·M-plus-chunking decides whether the examiner can ride it effortlessly. Get the music and your accent stops mattering — you become easy to understand, which is the whole pronunciation score.
Two questions before we move to the worked example in Stage 7.
The same Vietnamese candidate. This time we don't care what they say — the content is a solid Part 3 answer you've seen the shape of before. We care entirely about how it's delivered. We'll take one answer and mark every layer of S·F·M plus chunking onto it, then hear the difference between a flat reading and a musical one. This is the whole lesson, applied to a real answer.
Q: "Why do you think fewer young people cook at home these days?" The candidate gives a tidy C·M·E answer — we'll deliver it with full music.
First we break the answer into chunks at the grammatical seams, so it's organised before we add anything else.
Inside each chunk, we mark the content words that get stretched and punched.
We mark where the crushed function words link into their neighbours.
Finally we mark the rises (more coming) and the closing fall.
Here's the full answer with everything marked: / = chunk-break, CYAN CAPS = stress, ‿ = link, ↗ = rise, ↘ = fall. Toggle the layers to see them build up, then hear it spoken.
I think the MAIN REAson↗ / is‿that YOUNG PEOple are‿just BUSier↗ / than‿they USED to‿be↗ / so‿the TIME they‿used‿to‿spend COOKing↗ / just‿isn't THERE any‿more↗ / and‿once COOKing stops‿being‿a DAILy HABit↗ / the‿skill‿just QUIETly FADES a‿way↘
The same answer, broken down by what each layer contributes. Read down and you'll see how the four layers stack to turn a flat reading into a Band 8 delivery.
Here are the two deliveries side by side. Click both and listen for everything you've learned this lesson — the beat, the flow, the pitch, the chunking. The words are byte-for-byte identical; only the music changes. This gap is what the pronunciation score measures.
Every syllable equal, no chunk-breaks, flat pitch, endings dropped. Clear word-by-word, but the listener has to assemble the meaning themselves.
Seven thought groups, content words punched, function words crushed and linked, pitch rising through and falling to close. The delivery does the listener's work.
Took one ordinary answer and delivered it as music: chunked into seven thought groups, stressed eleven content words, crushed and linked all the glue between them, and threaded six rises and a closing fall across the whole thing. Not one word changed. The accent didn't change. But the delivery went from "followable with effort" to "effortless" — which on the pronunciation scale is the jump from Band 6 to Band 8. This is the entire lesson, and it works on every answer you'll ever give.
Two diagnostic questions before the Practice Arena.
Five short descriptions of how a candidate sounds. For each, identify which of the four failure modes from Stage 2 it is — the Machine-Gun, the Dropped Ending, the Flat Line, or the Wrong Syllable. Recognition is the first half of the fix.
Four sentences. Tap the content words you'd punch, leave the function words unstressed, then reveal each model. This is the Stage 3 builder again, but now on full answer-sentences — the kind you'll actually say.
Five more multi-syllable words, using the suffix rules from Stage 4. Pick the stressed syllable; feedback names the rule. These are exactly the sophisticated words that earn vocabulary marks, so getting their stress automatic protects your best words.
Three run-on answers with no breaks. For each, rewrite it with "/" marking where you'd chunk it — at the grammatical seams. Then reveal a model. The skill is breaking at the seams between ideas, not randomly.
The capstone drill. Below is a short answer. Your job: record yourself delivering it with all four layers — chunked, stressed, linked, pitch moving — then listen back and check each layer landed. Three rounds: marked, natural, then a fresh answer of your own.
Q: "Do you think people read enough these days?" — "Honestly, / I think people read MORE than ever, / but DIFFerently — / it's all short posts and messages now, / rather than long books, / which means we're reading constantly / but maybe thinking less DEEPly about any of it."
Five pronunciation exercises done — the music, tested. Here's how it landed.
Your performance across the pronunciation arena shows how well the music is coming together. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 16. Be honest about which ones still need work — that tells you which layer to drill in Lionel's daily music exercise.
Forty-five minutes on the fastest-moving core skill — the one where you change how you say things, not what you say, and the band moves quickly.
Sixteen lessons done — and the first core skill in hand. Pronunciation gave your speech its music; the next core skill gives it flow. Lesson 17 is Fluency: speaking at length without the hesitations, repetitions, and self-corrections that break the rhythm — and crucially, doing it without sacrificing the music you just built. Pronunciation and fluency are close cousins, because both are about effortlessness for the listener.
Fluency isn't speed — it's the ability to keep going smoothly without the stumbles that make a listener work. Lesson 17 tackles the hesitation habits, the filler-and-stall patterns, and the techniques for buying thinking time gracefully, so you can speak at length and sound in control even when you're not sure what comes next.
The repetitions, false starts, and "um"-stalls that break fluency — and where they come from.
How to think mid-answer without dead air — the fillers, the rephrasing, the chunk-pauses that double as planning time.
What to do when you lose your thread or hit a word you don't know — without it derailing the whole answer.
Two more after fluency — vocabulary range and grammar — then the full mock test brings everything together.
"Pronunciation is the core skill students underestimate the most — and the one that moves fastest once they take it seriously. You've now got the whole music: stress, flow, melody, chunking. Here's my one ask before Lesson 17: don't just know it, drill it. Two minutes a day, reading anything aloud, exaggerating the beat and the pitch, recording yourself once a week. The knowledge you have now; the habit takes two weeks of mouth-work. Do that, and your accent — which never needed fixing — will sit on top of rhythm and melody that make you genuinely easy to understand. That's the whole game, and you're holding it. Now let's make it flow."
Sixteen badges, and the first Core Skill done. You can take any answer and deliver it with stress, flow, melody, and chunking — turning flat, syllable-timed English into the music that makes you effortless to understand, accent and all. Four lessons left: fluency, vocabulary, grammar, and the mock test.
The second Core Skill. Pronunciation gave your speech its music; fluency gives it flow — the ability to keep going smoothly, at length, without the stalls and stumbles that make a listener work. Here's the first myth to drop: fluency is not about talking fast. Plenty of slow speakers are perfectly fluent, and plenty of fast ones aren't. Fluency is about continuity — keeping the stream moving and connected, so the examiner never has to wait for you to recover. For a Vietnamese learner, the biggest enemy isn't a lack of words; it's the freeze — stopping dead to hunt for the perfect word or the right grammar. This lesson teaches you to keep flowing instead.
"Students always think fluency means speaking quickly, so they rush, trip over themselves, and sound less fluent. Speed has nothing to do with it. Fluency is the absence of breakdown. It's the stream never stopping. The single most damaging thing in the whole speaking test is the silent stall — that two, three, four seconds of dead air while you hunt for a word or panic about a tense. Every one of those silences tells the examiner 'I lost control here.' The fluent speaker has a secret: they never go silent, because they've got tools to keep moving — they glide over the gap, they go around the missing word, they extend instead of stopping. The words don't have to be perfect. The stream just has to keep flowing. That's all fluency is, and it's completely learnable."
Two candidates answer the same question with roughly the same content. One keeps freezing — silent stalls, restarts, a dead-end. The other glides, goes around the words it doesn't have, and extends. Click both and listen. The difference isn't vocabulary or grammar; it's whether the stream ever stops.
"What do you usually do to relax?"
Three silent stalls, two freezes on missing words ("the place with trees" = park; "the thing you do with music"), and a dead-end ("that's all"). The ideas are fine — relaxing, music, parks — but the stream keeps stopping, so the examiner hears a candidate constantly losing and regaining control. Every freeze is a fluency penalty.
Same ideas — music, parks, trees — but the stream never stops. It glides in with "Well, to be honest"; when the exact word for "park" is slow to arrive, it maneuvers around it ("those big green spaces in the city — a park, basically"); and it extends each idea with a detail instead of dead-ending. No silence, no freeze, no full stop until it's ready.
The same 9-stage shape, aimed at flow. The framework here is G·E·M — Glide, Extend, Maneuver — the three moves that keep the stream moving. Make every answer a GEM.
You are here.
The perfectionism diagnosis, and the 4 fluency failure modes — the Freeze, the Backtrack, the Dead-End, the Word-Hunt.
Glide, Extend, Maneuver — the three moves that keep the stream moving. Plus the interactive Flow-Fixer builder.
The paraphrase patterns — how to describe a word you can't reach, so a missing word never stops you.
Glide phrases, extension connectives, maneuver phrases, and recovery lines — the connective tissue of fluent speech.
The Band 8 move — the extension engine that turns any one-line answer into a developed, flowing response without padding.
A halting answer rebuilt live into a flowing one — every Glide, Extend, and Maneuver marked and heard.
Five fluency exercises.
Self-assessment, badge, and Lesson 18 preview (Core Skills — Vocabulary Range).
Here's the surprising thing about fluency breakdowns: they rarely come from not knowing enough. They come from wanting to say it perfectly. The freeze is what happens when your standards outrun your speed — you stop the whole stream to hunt for the exact word or the flawless tense, and the silence costs you far more than the imperfect word ever would have.
You reach for the ideal word, can't find it instantly, and freeze hunting for it — three seconds of silence to avoid one "good enough" word. But the examiner can't see the perfect word you're chasing; they only hear the silence. A fluent speaker grabs the nearest workable word and keeps moving, or describes it in other words. The exact word is worth far less than the unbroken stream.
Many Vietnamese learners are taught that accuracy is everything, so a possible grammar mistake triggers a full stop and a restart. But in speaking, self-correcting every small slip breaks fluency worse than the slip itself. Examiners expect minor errors at every band — what they penalise is the constant stopping and backtracking. Letting a small error pass and flowing on is the fluent choice.
Fluency and accuracy pull against each other in real time. The more you chase the perfect word and the flawless tense, the more you stall — and stalling hurts your band more than the imperfections would. So the trade is deliberate: accept "good enough" words and let small errors slide, in exchange for a stream that never stops. This feels wrong to a careful learner. But the examiner is scoring continuity, and a flowing answer with a couple of minor slips beats a halting answer with perfect grammar every time. Done and flowing beats perfect and frozen.
Fluency breaks down in four recognisable ways. Each is a different way the stream stops — and each has a matching G·E·M move that fixes it. Learn to hear which one is happening in your own speech; recognition is the first half of the fix.
The classic silent stall. Mid-sentence, you stop dead — no sound at all — while you hunt for a word, a tense, or the next idea. Two, three, four seconds of dead air. It's the single most damaging fluency error because the silence is total and obvious, and it signals "I lost control" louder than any wrong word would.
Excessive self-correction and repetition. You start a sentence, spot a small error or a better way to say it, and go back to fix it — then again, and again. "I go — I went — I have gone there..." Each backtrack restarts the stream. A little self-correction is natural; constant backtracking shreds your fluency far worse than the original slip.
Running out after one short line. The examiner asks a question, you give a single sentence, and stop — "Yes, I like reading. It's good." — leaving an awkward silence and forcing the examiner to drag the next thing out of you. Short answers cap your fluency band hard, because the descriptor rewards speaking at length.
Freezing specifically because one word won't come. You know exactly what you mean, but the single word for it is missing — so you stop the whole answer to search for it, or you say "how to say..." and stall. The irony: you could describe the thing in five easy words, but you've fixated on finding the one hard word.
A short Part 1 answer delivered two ways. The first carries all four failure modes; the second applies the three G·E·M moves. Click both and listen for where the stream stops — and where it keeps flowing.
Question: "Do you like your hometown?"
Question: "Do you like your hometown?"
Record yourself answering a few questions and listen back for one thing: where does the stream stop? Every silent gap, every restart, every one-line answer is a place the flow broke. You'll usually find one dominant pattern — you're a freezer, or a backtracker, or a dead-ender. Name your signature failure mode, because that tells you which G·E·M move to drill hardest.
The advice that frees most Vietnamese learners is counterintuitive: stop trying to be perfect. Not because accuracy doesn't matter, but because in real-time speech, the chase for perfection is the very thing that breaks you. Aim for "good enough and flowing," and your band goes up — including, paradoxically, your grammar score, because you stop drawing attention to every slip.
"I had a student, Tuan, whose grammar was genuinely excellent — better than mine on paper. But he kept freezing, because he refused to say anything until it was perfect. He'd get a 6 for fluency and it dragged his whole score down. I gave him one rule: you are not allowed to stop, ever, even if what comes out is wrong. Just keep the stream moving. The first week it was messy — small errors everywhere. But he never went silent. And his band went up, not down, because the examiner finally heard someone who could actually hold a conversation. The errors he was so afraid of? They cost him almost nothing. The silences had been costing him everything. Flow first. Polish later, if there's room."
The Band 7 Fluency & Coherence descriptor says the candidate "speaks at length without noticeable effort or loss of coherence" and "may demonstrate language-related hesitation at times, or some repetition and/or self-correction." Read that carefully: hesitation and self-correction are expected even at Band 7 — what matters is that they don't break the overall flow. The descriptor rewards length and continuity, not perfection. So the winning strategy is exactly Lionel's: keep going, accept some mess, never let the stream stop.
When you keep the stream moving — gliding over gaps, going around missing words, extending instead of stopping — you're giving the examiner a conversation they can actually follow, without the constant stops and starts that make them wait and work. A flowing answer is generous; it carries the listener along. The frozen, perfectionist answer, however correct, leaves them stranded in your silences. Keeping the flow is a way of respecting the person listening. The respect is in the flow.
Two questions before we move to the G·E·M framework in Stage 3.
Three moves keep the stream flowing, and each one cures a failure mode you just met. Together they spell GEM — a useful reminder that a flowing answer is worth more than a perfect one. Glide over the gaps, Extend so you never dry up, Maneuver around the words you can't reach.
The foundation move. Gliding means keeping sound flowing through the small gaps where you'd normally freeze — buying yourself thinking time without dead air. The key insight: a filled pause sounds fluent; a silent one sounds broken. Same thinking time, opposite impression.
"Here's the reframe that fixes the freeze forever: you're allowed to pause — you're just not allowed to pause silently in the middle of an idea. Fill it, or move it to the seam between thoughts. 'I think the main... [silence] ...benefit' is broken. 'I think the main benefit, well, the main benefit I'd point to, is...' has exactly the same thinking time, but it's full of sound, so it's fluent. Same pause, totally different score. Learn three or four fillers so well they come out automatically, and the freeze simply stops happening — there's always a word ready to fill the gap."
A warning: gliding can tip into a different problem if you overuse fillers — "um, like, you know, um" on a loop is its own kind of disfluency. The goal isn't to fill every gap with noise; it's to have a small, varied set of natural fillers and markers ready, so that when you need a beat to think, you reach for one smoothly and move on. Variety matters: rotate through "well / the thing is / to be honest / I suppose" rather than hammering one. Used well, gliding is invisible — it just sounds like a person thinking out loud, fluently.
The move that defeats the one-line answer. Extending means every answer automatically grows beyond the first sentence — you add a reason, an example, a contrast — so you speak at length without the awkward stop that forces the examiner to drag more out of you. The secret is having a few reliable "next-move" directions always ready.
"The number one fluency killer in Part 1 is the one-line answer. 'Do you like music?' 'Yes, I like music.' Full stop. Dead air. The examiner has to fish for more, and your fluency band sinks. So I teach a simple reflex: you are never allowed to stop after one sentence. Whatever you say, the next word is 'because,' or 'for example,' or 'although' — one of four ramps that carries you into a second and third sentence. You don't plan a long answer; you just refuse to stop, and reach for a ramp. The length takes care of itself. Two ramps and any answer is comfortably long enough."
Extending isn't padding — it's developing. The difference: padding repeats the same idea in more words ("I like it, I really like it, it's something I like a lot"), which examiners see straight through. Developing adds new content — a reason, an instance, a contrast, a comparison across time. The four ramps all point at new content, which is why they work. Aim for three to four sentences on a Part 1 question and a genuinely developed response in Parts 2 and 3, every time, by reflex.
The move that defeats the word-hunt freeze. Maneuvering — the technical name is circumlocution — means describing a word you can't reach instead of stopping to search for it. Counterintuitively, examiners reward this: smoothly talking your way around a missing word shows real communicative skill, often more than just knowing the word would have.
"This is the one that feels most like cheating, and it's the one examiners love most. When you can't find a word and you smoothly describe your way around it — 'the tool you'd use to tighten a screw' instead of 'screwdriver' — you've just demonstrated exactly the skill the test is looking for: the ability to communicate any idea with the language you have, not the language you wish you had. A real conversation is full of moments where the perfect word isn't there. The fluent speaker isn't the one who knows every word; it's the one who never lets a missing word stop them. Maneuvering is that skill, and it's pure gold in the exam."
The mindset shift behind maneuvering: a missing word is not a problem, it's just a small detour. The instant you feel a word won't come, don't stop — switch immediately to describing it. The five easy words you already have always beat the one hard word you're hunting for, because the easy five keep the stream moving and the hard one stops it dead. Practise by deliberately banning yourself from using a word and forcing a description: it builds the reflex so that in the exam, the detour is automatic.
The interactive builder. Below are four answers where the stream breaks. For each, pick the G·E·M move that fixes it — Glide, Extend, or Maneuver — then reveal a flowing rewrite. Diagnosing which move a breakdown needs is the core fluency reflex.
Silence in the middle of an idea → Glide. A one-line answer that stops short → Extend. A freeze on one specific missing word → Maneuver. Once you can name what kind of break is happening, the right fix is automatic — and soon you'll apply it before the break even lands.
Two questions before we go deep on maneuvering in Stage 4.
Maneuvering deserves its own stage, because the word-hunt freeze is the single most common way fluent learners break down — and because it's the most counterintuitive to fix. The instinct is to stop and search; the skill is to keep talking and describe. This stage builds the maneuvering reflex with four concrete patterns and a drill, so a missing word never stops you again.
"The thing you use to..." — describe what the object or idea is for.
"A kind of..." — place it in a group, then add one distinguishing feature.
Grab a simpler related word and flag that it's approximate.
Give a concrete example of it instead of naming the general word.
The two workhorses of maneuvering. Function-description handles almost any object; category-plus-feature handles almost any thing or concept. Between them they cover the vast majority of missing nouns.
The two fastest maneuvers, for when even a description would slow you down. Both keep the stream moving instantly while you communicate the gist.
You don't pick consciously — with practice it's instant — but the rough logic is: for a physical object, reach for function ("the thing you use to..."); for an abstract concept or feeling, reach for category ("a kind of..."); for a strong or precise word, reach for a near-word + flag ("tired, but way beyond that"); and for a general/collective noun, reach for examples ("things like X and Y"). All four share one rule: the moment a word won't come, you switch to describing instantly — no pause, no "how to say," no freeze.
Here's the whole toolkit on one screen. The goal isn't to memorise four techniques you choose between — it's to build one reflex: word won't come → describe, don't stop. The four patterns are just the shapes that reflex takes.
The best maneuvering drill: pick a everyday object — say, a fork — and describe it without naming it, out loud, in one smooth go ("the thing you use to pick up food, with little points on the end"). Then do it with ten more objects. Then try abstract words — "freedom," "nostalgia," "convenience." The point isn't the descriptions; it's training your mouth to start describing instantly instead of freezing. Do this for five minutes and the word-hunt freeze starts to disappear, because your brain learns there's always a way around. Word won't come? Describe, don't stop.
Five missing words. For each, pick the maneuver that fits most naturally. There's often more than one workable answer, but one is the cleanest fit — and the feedback explains why. This trains the instant matching of word-type to maneuver.
The connective tissue of fluent speech. These eighty phrases are the ready-made tools that let you glide, extend, maneuver, and recover without thinking — so that in the exam, when you feel a gap coming, the right phrase is already on your tongue. Unlike Lesson 16's pattern bank, these are for deployment: memorise a handful from each category so well they come out automatically.
Fillers and markers that fill a gap with sound while you think — "well", "let me think", "the thing is". Your defence against the Freeze.
The ramps that carry you into another sentence — "because", "for example", "which means", "that said". Your defence against the Dead-End.
The frames for going around a missing word — "the thing you use to...", "a kind of...", "sort of like...". Your defence against the Word-Hunt.
The lines that rescue you when you lose your thread or stumble — "where was I", "anyway", "what I'm trying to say is". Your defence against the spiral.
Twenty fillers and markers that fill a gap with sound. Click each to hear it. The job of every one is the same: keep the stream moving while your brain catches up, so a thinking-pause never becomes a silent freeze.
Twenty ramps that carry you from one sentence into the next. Each one, said aloud, almost forces a new clause to follow — which is exactly what you want. Reach for one whenever an answer is about to stop too soon.
Twenty ready-made frames for going around a missing word. Click each to hear it. These are the exact openings that launch a description instead of a freeze — learn them and your mouth starts describing before your brain can panic about the word it can't find.
Twenty lines that rescue you mid-answer — when you lose your thread, ramble too far, or stumble badly. The fluent speaker isn't the one who never goes wrong; it's the one who recovers smoothly when they do. These phrases are the recovery.
The instinct when you stumble is to freeze in embarrassment — which makes it ten times worse. The fluent move is to name it lightly and carry on: "sorry, where was I — anyway, the main thing is..." That single phrase converts a breakdown into a tiny, natural human moment, and the examiner barely notices. Real conversations are full of these recoveries; using them well marks you as a confident speaker, not a struggling one. Pick five flow phrases from each category and over-learn them — twenty automatic phrases is the whole connective toolkit.
Gliding stops you freezing; extending stops you dead-ending. But the Band 8 version of fluency goes further: you can speak at length on anything, even a topic you know nothing about, without ever running out or padding. The secret is an "extension engine" — a small set of directions you can always push an answer in, so there's always somewhere to go next.
One idea, then empty: "I think technology is useful. It helps people. ...Yeah." The candidate had a thought, expressed it, and had nowhere else to go — so the answer collapsed into silence after two short sentences.
One idea, pushed in five directions: "Technology's useful — because it saves so much time (reason). Take my grandmother, who video-calls us daily now (example). Although, that said, it can be isolating too (contrast). When I was a kid we'd never have imagined it (time-shift). I think the next ten years will change it even more (projection)."
From any statement, you can always push in one of five directions to generate genuinely new content: Reason (why is it so?), Example (a concrete instance), Contrast (the other side), Time-shift (how it was / will be), and Consequence (so what follows from it?). You don't need to plan a long answer — you make one statement, then turn the engine: pick a direction, push, and a new sentence appears. Turn it twice and you have a developed answer; turn it three or four times and you're speaking at genuine length.
Each direction is a question you ask yourself about your last sentence — and the answer is your next sentence. Internalise the five questions and you'll never run dry, because there's always one you can answer.
Watch the directions chain on a single starter — "I enjoy cooking": Reason — "because it's the one creative thing I do all day"; Example — "like last night I tried a new curry recipe from scratch"; Contrast — "though I'll admit I hate the cleaning up afterwards"; Time-shift — "I couldn't cook at all until university, actually"; Consequence — "so now it's become this little daily ritual that genuinely de-stresses me." One starter, five turns, a forty-second answer — and every sentence is new content, not padding. You didn't plan it; you just kept turning the engine.
There's a danger in "speak at length" advice: it can tip into empty waffling — lots of words, no content — which examiners penalise just as hard as a short answer. The engine protects you from this, because every direction adds new content. But it's worth knowing the difference clearly.
"I really like it. I mean, I like it a lot. It's something I really enjoy and like very much. Yeah, it's good. I like it because it's nice and I enjoy it."
"I love it — because it's so relaxing (reason). Last weekend I spent a whole afternoon at it (example). Though it can get expensive (contrast). I only started during lockdown (time-shift). Now I genuinely couldn't live without it (consequence)."
The test for whether you're developing or padding: could you delete a sentence without losing any information? If yes, it was padding. In the developing example, every sentence carries something the others don't — a reason, an instance, a caveat, a history, an outcome. In the padding example, you could delete any five of the six sentences and lose nothing. The engine guarantees development because each of its five directions points at genuinely new information. Use the engine and you physically can't waffle.
Step back and see how it all fits. Three moves keep the stream flowing, and the extension engine is simply Extend at full power — a way to keep turning new content out for as long as you need. Together they make you genuinely difficult to stop.
You get a question. You glide in with a filler to buy a beat ("That's an interesting one — well..."). You make your opening statement, then turn the engine two or three times to develop it (reason, example, contrast). When a word won't come mid-flow, you maneuver around it without breaking stride ("the thing you use to... anyway"). And if you stumble, a recovery phrase catches you ("sorry, where was I — anyway"). The whole time, the one rule holds: the stream never stops. That's a Band 8 fluency performance, and every piece of it is now in your toolkit.
Glide so you never go silent, run the engine so you never run dry, maneuver so a missing word never freezes you, and recover so a stumble never spirals. Underneath all four is a single principle: keep the stream moving, and accept that good-enough-and-flowing beats perfect-and-frozen every time. Combine this with the music from Lesson 16 and you have delivery that's both fluent and clear — the two skills that carry every word your frameworks produce.
Two questions before the worked example in Stage 7.
We'll take a single answer that breaks down badly — every fluency failure mode in one place — and rebuild it live using G·E·M. The content barely changes; what changes is that the stream stops breaking. This is the whole lesson applied to one real answer, and you'll hear both versions.
Q: "Tell me about a hobby you enjoy." — a Part 1 / early Part 2 prompt that invites a developed answer.
The silent stalls get filled with markers and fillers, so the gaps have sound.
The freeze on a missing word becomes a smooth description that flows right on.
The one-line stop becomes a developed answer via the extension engine.
The repeated self-corrections get dropped — small errors flow by untouched.
Here's the halting answer with every breakdown marked. Click to hear it, then read the diagnosis of each failure. Four breaks in a few sentences — and notice that none of them are about not knowing enough.
"My hobby is... um... [3 sec silence] ...I like — I liked — I like to do the... the thing with... how to say... the camera... taking picture... Yeah. It's good. I like it. [stops]"
Look at what the candidate actually knew: their hobby, that it involves a camera and taking pictures, that they like it. That's a perfectly good basis for a strong answer. The problem is purely fluency: a silent freeze at the start, a triple backtrack on "like/liked", a freeze hunting for "photography", and a dead-end after one line. Every breakdown is a place the stream stopped — and not one of them was caused by missing knowledge. Now watch G·E·M close each gap.
Here's the rebuild, with every G·E·M move marked. Toggle the annotations to see where each move lands, and hear the flowing version. Same speaker, same ideas — but the stream never stops.
"Well, let me think — my main hobby is taking photos, photography I suppose you'd call it. I love it because it makes me really notice things I'd normally walk straight past. Like just last weekend, I spent a whole morning wandering around the old quarter just photographing doorways, of all things. It can be a bit of an expensive habit, mind you, but I've been into it since I got my first proper camera at university, and honestly it's become the way I switch off."
Both versions side by side. Click each and listen for the one thing that changed: in the first, the stream keeps stopping; in the second, it never does. The vocabulary and grammar are essentially the same — the fluency is two bands apart.
Four stream-stops in a few seconds: a silent freeze, a triple backtrack, a word-hunt freeze, and a dead-end. The examiner hears constant loss of control.
Glides in, maneuvers around "photography", turns the engine four times, lets the small error pass. The stream never once stops.
It didn't add vocabulary. It didn't perfect the grammar — it left an error untouched on purpose. It didn't even add much content; the core ideas (photography, makes me notice things, switch off) were all there in the broken version's raw material. What it changed was a single thing, four times over: it never let the stream stop. A filler instead of a freeze, a description instead of a word-hunt, an engine-turn instead of a dead-end, a shrug instead of a backtrack. That one change is the entire difference between Band 6 and Band 8 fluency — and it's completely within your control.
Two diagnostic questions before the Practice Arena.
Five descriptions of how a candidate's stream broke down. For each, identify which of the four fluency failure modes from Stage 2 it is — the Freeze, the Backtrack, the Dead-End, or the Word-Hunt. Recognition first.
Five words you might blank on mid-answer. For each, pick the best maneuver strategy from Stage 4 — Function, Category, Example, or Synonym — to describe it without freezing. The skill is never stopping for a word you can't reach.
Three one-line, dead-ended answers. For each, write an extended version that keeps the stream going — using the extension moves from Stage 6 (reason, example, contrast, consequence, personal angle). Don't change the opinion; just stop it dead-ending. Reveal a model after your attempt.
Four moments where a candidate is about to freeze into silence. For each, write the glide — a filler or marker that keeps the stream moving while you think — instead of going silent. Short is fine; the point is no dead air.
The capstone drill. Below is a question designed to tempt all four failures — easy to dead-end, easy to freeze on a word, easy to backtrack. Your job: deliver a 45-second answer aloud that keeps the stream moving the whole way, using all three G·E·M moves. Record it and listen back.
"Describe something you own that is important to you, and explain why." (A Part 2-style prompt — perfect for practising sustained flow.)
Five fluency exercises done — the flow, tested. Here's how it landed.
Your performance across the fluency arena shows how well the stream is staying in motion. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 17. Be honest about which ones still need work — that tells you which move to drill in Lionel's daily flow exercise.
Forty-five minutes on the skill that decides whether the examiner ever has to wait for you — the continuity that holds every other skill together.
Seventeen lessons done — two Core Skills in hand. Pronunciation gave your speech music; fluency gave it flow. The third Core Skill gives it range: the precise, varied, natural vocabulary that lifts your Lexical Resource score. And here's the neat connection — the maneuvering you just learned is what lets you reach for ambitious words without fear, because if one doesn't come, you can simply describe your way around it.
A wide vocabulary isn't about rare, showy words — it's about precision, variety, and natural collocation. Lesson 18 tackles how to sound lexically rich without memorising word-lists: upgrading vague words to precise ones, building topic word-families, using idiomatic language naturally, and avoiding the repetition that flattens your range.
Why upgrading "good," "nice," and "very" to precise words beats hunting for obscure vocabulary.
The word-partnerships and topic clusters that make vocabulary sound natural rather than translated.
How to deploy idiomatic language for the band without sounding like you swallowed a phrasebook.
One more after this — grammar — then the full mock test brings all four parts and four core skills together.
"Fluency is the skill that makes all the others usable. The best vocabulary and the cleverest frameworks are worth nothing if the stream keeps freezing — the examiner only hears the stumbles. Now you've got the tools to keep moving: glide, extend, maneuver. My one ask before Lesson 18 is the same as always — drill it aloud. One minute a day, one rule: the stream doesn't stop. You'll feel clumsy for a week, then one day you'll notice you answered a hard question without a single freeze, and you didn't even think about it. That's fluency arriving. Two Core Skills down, two to go — and then we put the whole thing to the test."
Seventeen badges, and the second Core Skill done. You can keep any answer flowing — gliding through gaps, extending past dead-ends, maneuvering around missing words — so the examiner never has to wait for you. Three lessons left: vocabulary, grammar, and the full mock test.
The third Core Skill. Pronunciation gave your speech music; fluency gave it flow; now vocabulary gives it range — the precise, varied, natural word choice that lifts your Lexical Resource score. Here's the myth to drop first: range does not mean rare, impressive, dictionary-rescued words. Examiners can smell a memorised "big word" from across the room, and using one wrongly costs you more than a simple word used well. Real range is three things — choosing the precise word instead of a vague one, having enough variety that you never repeat yourself, and using words that fit naturally together the way a native speaker's do. For a Vietnamese learner, the ceiling is almost never "not enough words" — it's the "good / nice / very / a lot" plateau, where a small set of safe words gets reused for everything.
"Students think the way to a 7 in vocabulary is to learn long, fancy words. So they memorise 'ameliorate' and 'plethora' and drop them in — and it sounds exactly like what it is: a memorised word, used slightly wrong, with no life in it. That's not range. That's decoration. Real range is when you stop saying 'the weather is very good' and start saying 'the weather's been glorious lately,' or when 'a lot of traffic' becomes 'the roads are absolutely packed.' Same idea, but the words are precise, they're varied, and they sit together naturally. None of those words is rare — 'glorious,' 'packed' — they're just right. The whole skill is learning to reach past your first, safe, vague word to the one that actually fits. That's what this lesson trains."
Two answers to the same question, with identical content. One sits on the "good / nice / very / a lot" plateau; the other reaches for the precise, varied, natural word each time. Notice that the better answer uses no rare words — just right ones.
"Tell me about a city you've visited and enjoyed."
Nothing here is incorrect — but "good," "nice," and "very" are doing all the work, repeated over and over. The examiner learns almost nothing about the city, because the words carry no specific meaning. This is the plateau: safe, clear, and capped at Band 6 because there's no range on display.
Same city, same opinion — but every vague word has been upgraded to a precise one ("stunning," "glorious," "welcoming"), the repetition is gone, and there's natural idiom ("won me over," "spoilt for choice," "in a heartbeat"). Not one rare word, yet it sounds rich. That's range: precision, variety, and natural fit.
The same 9-stage shape, aimed at range. The framework here is P·F·N — Precision, Families, Natural-fit — the three things that make vocabulary rich without making it rare.
You are here.
The safe-word diagnosis, and the 4 vocabulary failure modes — the Plateau, the Misfire, the Repeat, the Phrasebook.
Precision, Families, Natural-fit — the three layers of real range. Plus the interactive Upgrade builder.
The word-partnerships that sound natural, and how to build a topic word-family so you never repeat or run dry.
Precise upgrades for the vague words, topic word-families, natural collocations, and safe idioms — ready to use.
The Band 8 move — deploying idiomatic language so it sounds natural, not like a swallowed phrasebook.
A plateau answer rebuilt live into a high-range one — every upgrade, collocation, and idiom marked.
Five vocabulary exercises.
Self-assessment, badge, and Lesson 19 preview (Core Skills — Grammar).
Here's the thing almost every intermediate learner gets wrong about their own vocabulary: they think the problem is that they don't know enough words. Usually they know plenty — they just don't reach for them under pressure. The plateau is a habit of safety, and understanding why explains every vocabulary failure you'll meet on the next screen.
"Good," "nice," "very," "a lot," "interesting," "important" — these words are safe because they fit almost anywhere and you can't really misuse them. Under the pressure of a live test, the brain reaches for the safest available option, every time. The result is technically correct, perfectly clear, and completely flat. You're not failing to know better words — you're defaulting to the safe ones because reaching for a precise word feels risky.
Vietnamese learners often think in Vietnamese and translate, and translation tends to land on the most general English equivalent. A rich Vietnamese word for a specific kind of beauty or busyness comes out as plain "nice" or "busy," because that's the safe one-to-one swap. The precision lived in the original; it gets sanded off in translation. So the words you produce are blunter than the thoughts you had.
Because the problem is reaching, not knowing, the fix is mostly about retraining the reach — building the habit of pausing on a vague word and upgrading it, and pre-loading a few precise words per topic so the better word is already at hand. You don't need to memorise thousands of new words. You need to stop defaulting to the safe ten you overuse, and make the precise word the one your brain grabs first. That's a habit, and habits are trainable in weeks.
The plateau produces four recognisable failure modes. The first two come from staying safe, the second two from over-reaching the wrong way. Learn to hear each in your own answers — recognition is the first half of the fix.
The core mode. A small set of safe, vague words — good, nice, very, a lot, thing, stuff, interesting — does all the work, across every topic. Nothing is wrong, but nothing is precise, so the examiner can't see any range. This is the Band 6 ceiling: clear but generic.
Using the same word again and again within one answer because no alternative comes to mind. "The city is beautiful, the beaches are beautiful, and the buildings are beautiful." Repetition is one of the most visible range-killers, because the examiner literally hears the same word stack up.
Reaching for an impressive word and using it slightly wrong — wrong meaning, wrong context, or wrong word-partner. "I was very enormous about the news" (meant "excited"). A misfire costs more than the safe word would have, because it shows the word was memorised, not owned.
Dropping in memorised idioms or "high-level" phrases that don't fit the moment, so they sound bolted on. "It is raining cats and dogs and I am over the moon to be here today, every cloud has a silver lining." Cramming idioms in unnaturally is worse than using none — it flags the whole performance as rehearsed.
A short answer delivered two ways. The first carries the safe-word habit; the second applies all three P·F·N moves. Read both — this is the whole lesson in miniature, and what you'll be able to do by Stage 9.
Q: "What kind of films do you like?"
Q: "What kind of films do you like?"
Record yourself answering a question, then count two things: how many times you used "good," "nice," "very," "a lot," or "interesting"; and whether any single content word repeated three or more times. Those two counts are your Plateau and your Repeat, made visible. Most learners are shocked the first time — the safe words are invisible while you speak and obvious on playback. Once you can see them, you can start upgrading them.
The instinct, once you know the plateau is holding you back, is to reach for bigger, more impressive words. That's the trap — it leads straight to the Misfire. The reframe: don't reach higher, reach truer. Aim for the word that most precisely captures what you actually mean, and the richness takes care of itself.
"A student once told me proudly that she'd learned fifty 'band 8 words.' I asked her to describe her morning coffee. She said it was 'an exquisite and sublime beverage.' I laughed — and then I told her the truth: that sentence is worse than 'my coffee was lovely.' Because 'lovely' is true and natural, and 'exquisite sublime beverage' is a costume. Precision isn't about the size of the word — it's about the fit. 'The coffee was rich and proper strong, just how I like it' — those are tiny words, and it's miles better, because every one is exactly right. Stop trying to sound clever. Try to sound exact. Exact is what range actually is, and exact is always within your reach."
The Band 7 descriptor for Lexical Resource talks about using vocabulary "with some flexibility" to discuss a variety of topics, "some less common and idiomatic" items, and "an awareness of style and collocation." Band 8 adds "skilfully" and "rarely" any inappropriacy. Notice the key words: flexibility, collocation, appropriate. Nothing about rare or advanced. The scale rewards using the right word in the right place with variety — which is precisely what reaching truer, not higher, produces.
When you reach for the exact word instead of the safe one, you're handing the examiner a clearer, sharper picture of what you mean — "stunning" tells them more than "nice," "packed" more than "busy." That precision is a kind of respect: you're doing the work of being understood exactly, rather than leaving them with a vague approximation. The frameworks gave the examiner your ideas; precision gives them the ideas in full colour. The respect is in the precision.
Two questions before we move to the P·F·N framework in Stage 3.
Real vocabulary range is three things stacked together. Get all three and your Lexical Resource sits at Band 7-8 without a single rare word. They build in order — Precision is the foundation, Families give you variety, Natural-fit makes it sound real — and each one cures one of the failure modes you just met.
The foundation. Every time you're about to say a safe, vague word, there's a more precise one that says exactly what you mean. The skill is the pause-and-upgrade: catch the vague word on its way out and swap it for the exact one. It's a reflex you build, not a vocabulary you cram.
"The single fastest upgrade in the whole language is killing 'very.' Almost every 'very + adjective' has a single word that means the same thing with more punch. Very happy — delighted. Very angry — furious. Very hungry — starving. Very clean — spotless. The moment you stop saying 'very' and start reaching for the strong single word, you sound a band higher, instantly, with words you already know. Try it for one day: ban 'very' from your speech and force the upgrade every time. It rewires the reach faster than anything else I teach."
A crucial guardrail: only upgrade to words you've actually heard in real sentences. The upgrade must be a word you own, not one you found in a thesaurus thirty seconds ago — that's how the Misfire happens. Precision means reaching for the exact word from the words you genuinely know, which for most learners is a far bigger set than the ten safe ones they actually use. You're promoting words from your passive vocabulary into active use, not importing strangers.
The variety layer. The Repeat happens because, under pressure, only one word for an idea comes to mind. The fix is to pre-load a small family of related words for each common topic, so when you need a second or third way to say something, it's already there. You think in clusters, not single words.
"Don't learn words in a list — learn them in families, around a topic, because that's how you'll need to reach for them. When the examiner says 'describe your hometown,' you don't want to be rummaging through every adjective you've ever met; you want your 'places' family to light up — vibrant, sprawling, bustling, charming — all there at once. It's the difference between a tidy toolbox where everything's grouped, and a drawer where everything's jumbled. Same tools. But under pressure, the grouped one is the only one you can actually use."
Families and Precision work together: Precision gets you from "nice" to "vibrant" once; Families make sure that when you need a second word in the same answer, "bustling" and "charming" are right there instead of "nice" again. Precision lifts a single word; Families keep the whole answer varied. Together they kill both the Plateau and the Repeat.
The layer that defeats the Misfire and the Phrasebook. A word isn't just a meaning — it has partners it naturally goes with (collocations) and a register it belongs to. Natural-fit means using words in their real partnerships, so your vocabulary sounds lived-in rather than translated or memorised.
"Collocation is the secret examiners notice but learners ignore. Anyone can learn 'enormous.' But knowing it's 'an enormous amount' and 'a huge impact' and 'a major issue' — that's range, because it shows you don't just know the word, you know how it lives. The way to build it isn't to study collocation lists; it's to learn words inside phrases. Never learn 'decision' alone — learn 'make a tough decision.' Never learn 'rain' alone — learn 'heavy rain' and 'it's pouring.' Store the partnership, and the natural fit comes free."
All three layers fuse in a single phrase. "The beaches were absolutely stunning" — that's Precision (stunning, not nice), drawing on a Families cluster (your "places" words), in a Natural-fit partnership ("absolutely stunning" is a real collocation; "very stunning" is not). One short phrase, all three moves. That fusion is what you'll build on the next screen.
The interactive builder. Each card shows a flat sentence with a vague word, and three possible upgrades. Pick the one that's precise AND a natural fit — beware the Misfire option (a big word that doesn't fit) and the no-change option (still vague). Five to work through; feedback explains each.
Notice the pattern: the best answer was never the biggest word, and never the unchanged vague one — it was the precise word in a natural partnership. Reach truer, not higher. Do this with your own flat sentences and the plateau dissolves.
Two questions before we move to collocation and word-families in Stage 4.
P·F·N is the principle; this stage is the practical machinery. Two tools build real range fast: collocations (the word-partnerships that sound native) and word-families (topic clusters that give you variety on demand). Master these two and you stop translating word-by-word and start speaking in natural chunks.
The most common and most useful — make a decision, take a risk, raise an issue, meet a deadline.
The natural pairings — heavy rain, strong accent, a close friend, a major issue, a vague idea.
The natural boosters — absolutely stunning, deeply moving, utterly exhausted, highly unlikely.
The pre-loaded clusters for IELTS's recurring themes — places, people, work, study, technology.
A collocation is a pair of words that natives habitually use together. They're not governed by logic — they're just convention — which is exactly why they signal range: getting them right shows you've absorbed real English, not translated it. Here are the patterns that matter most.
A word-family is your pre-loaded variety for a recurring topic. Because IELTS recycles the same dozen themes, building one rich cluster per theme covers almost any question. Here's what a strong family looks like, and how to grow your own.
Take one common topic. Write the vague word you'd normally use ("nice place"). Then collect five to seven precise alternatives across word-types — not from a thesaurus, but from things you've actually heard in films, songs, podcasts, conversations. Group them on one page under the topic heading. Review the page out loud, in example sentences, not as a bare list. Do one topic a week and in a couple of months you'll have pre-loaded variety for every theme IELTS can throw at you — and because you gathered them from real input, they'll be words you can actually deploy.
Everything in this stage rests on one principle that quietly separates Band 6 from Band 8 vocabulary: fluent speakers don't store and retrieve single words — they store and retrieve chunks. Phrases, partnerships, families. It's faster, more natural, and far more accurate.
You store "decision," "rain," "stunning" as isolated words. At speaking time you assemble them from scratch — and guess the partners. You get "do a decision," "strong rain," "very stunning." Each guess is a chance to misfire, and the assembly is slow because you're building every phrase fresh.
You store "make a tough decision," "heavy rain," "absolutely stunning" as whole units. At speaking time you retrieve the chunk intact — partners already correct, no assembly, no guessing. It's faster and it's right. This is what native fluency actually is: a huge store of pre-assembled chunks, not a dictionary plus grammar rules.
Learn words in their natural company — verb with noun, adjective with noun, word within family. Store "meet a deadline," not "deadline." Store the "places" cluster, not "vibrant" alone. When you store chunks, the precision, the variety, and the natural fit all come pre-attached, and you retrieve correct, native-sounding English under pressure instead of assembling risky guesses. Range isn't a bigger dictionary — it's a bigger store of chunks.
Five pairs. Pick the partnership a native speaker actually uses. Instant feedback explains the convention. This trains the collocation instinct that examiners hear as range.
Your pre-loaded range, organised the way you'll reach for it. Eighty items across four categories: direct upgrades for the vague words you overuse, three full topic word-families, the highest-value collocations, and a short set of genuinely safe idioms. Everything here is common — no rare words — and everything is a chunk you can lift straight into an answer.
The precise replacements for your top offenders — good, nice, very, a lot, big, interesting. The single highest-impact set: learn these and the Plateau breaks.
Three full clusters — places, people, work — so you have instant variety on the most common prompts and never repeat within an answer.
The verb+noun and adjective+noun partnerships that work across every topic — make a decision, meet a deadline, heavy traffic.
A small set of natural, low-risk idioms with the situation each one fits — the ones you can deploy without sounding like a phrasebook.
Twenty direct upgrades for the words you overuse most. Each shows the vague word and the precise options. The arrow means "reach for one of these instead" — pick the one that fits what you actually mean.
Three full word-families for the most common prompts. Each spans word-types so you can rephrase a whole idea, not just swap one word. Build your other seven topics on this model.
Twenty high-value collocations that appear across almost every topic. Learn each as a whole chunk — the verb with its noun, the adjective with its noun. These are the quiet signals of range examiners notice most.
Twenty natural, low-risk idioms — each with the situation it fits. These are common in real speech and hard to misfire. Learn a handful with their context, and deploy one only when the moment genuinely calls for it.
Don't memorise this bank — graze it. Each day, pick three items total across the four categories: maybe one vague-word upgrade, one collocation, one idiom. Say each in a sentence about your own life, out loud, three times. That's thirty seconds. Over a few weeks you'll have genuinely absorbed a few dozen chunks — owned, not crammed — and that's exactly enough to lift your Lexical Resource a full band. A handful you own beats eighty you half-remember.
The Band 8 move. The descriptor explicitly rewards "idiomatic" language — but it's the single most dangerous thing a learner can reach for, because forced idiom (the Phrasebook) does more damage than no idiom at all. The skill isn't knowing idioms; it's deploying them so naturally that they sound like they arrived on their own. Idiom is seasoning: a pinch lifts the dish, a fistful ruins it.
"To be honest, every cloud has a silver lining, so I was over the moon and on cloud nine. It was raining cats and dogs but I killed two birds with one stone and the early bird catches the worm, so it was a piece of cake at the end of the day."
"Honestly, it was a tough first month — a really steep learning curve — but it turned out to be a blessing in disguise, because it pushed me right out of my comfort zone and I came out far more confident."
An idiom dropped where it doesn't fit does three bad things at once: it sounds unnatural (failing the "appropriacy" the descriptor demands), it reveals the language as memorised rather than owned, and it distracts from whatever you were actually saying. One idiom too many can flip an answer from "impressively natural" to "obviously rehearsed" in a single phrase. That's why the rule is the opposite of what learners expect: when in doubt about an idiom, leave it out. The natural answer with no idiom beats the stuffed one every time.
How do you know if an idiom is safe to use? Run it through three quick tests. If it passes all three, deploy it; if it fails any, leave it out and say the idea plainly. With practice this becomes instant.
Notice that the natural example on the last screen used "a steep learning curve," "a blessing in disguise," and "out of my comfort zone" — none of them flashy, all of them current and common in real speech. The flashiest idioms ("raining cats and dogs") are the riskiest, because they're the ones textbooks over-teach and natives under-use. The quiet, slightly mundane idioms are both safer and more impressive, precisely because they sound like something a real person would actually say.
Beyond the three tests, there's a question of placement. A natural idiom tends to arrive at the moment of feeling or judgement in an answer — the emotional beat — not scattered through the factual parts. Place it there and it feels like a genuine reaction; place it elsewhere and it feels bolted on.
Step back and see the complete vocabulary system. Three layers, plus idiom as the finishing touch: upgrade the vague words (Precision), vary them from a topic family (Families), keep every partnership natural (Natural-fit), and add at most one earned idiom at the emotional beat. Once it's a habit, it's not separate steps — it's just how you talk.
Take "It was a nice trip and I saw a lot of nice places." Run the full system: Precision upgrades "nice trip" to "a fantastic trip"; Families varies "nice places" into "stunning beaches and charming little towns"; Natural-fit keeps the partnerships real ("absolutely stunning," not "very stunning"); and one earned idiom lands at the emotional beat: "honestly, I'd go back in a heartbeat." Result: "It was a fantastic trip — stunning beaches, charming little towns — and honestly, I'd go back in a heartbeat." Same trip, same length, Band 6 to Band 8, no rare words.
Upgrade, vary, keep it natural, season with one idiom. That's the entire P·F·N system, and it works on every answer in all three parts of the test. None of it requires rare words — it requires reaching for the precise common word, carrying topic families for variety, respecting natural partnerships, and trusting one earned idiom over five forced ones. Range is precision and variety and natural fit — not a fatter dictionary.
Two questions before we move to the worked example in Stage 7.
We'll take a real plateau answer — all "good," "nice," "very," and repetition — and rebuild it live with the full P·F·N system. Same ideas, same length, same opinion. We change nothing but the words, and watch it climb from Band 6 to Band 8 without a single rare word. This is the whole lesson, applied to a real answer.
Q: "Describe a restaurant you like." A (plateau): "There's a very nice restaurant near my house. The food is really good and there are a lot of dishes. The staff are nice and the place is very nice inside. I go there a lot because it's good. It's a really good restaurant."
Every "good," "nice," and "very" gets upgraded to the precise word that says what's actually meant.
"Nice" appears four times; we vary it from the food/places families so no word repeats.
We swap in genuine collocations ("absolutely delicious," "spoilt for choice") so every pairing sounds native.
A single earned idiom lands at the emotional peak, where the answer turns to feeling.
Here's the same answer rebuilt. Toggle the annotations to see which move each upgrade came from — precision, family/variety, natural-fit, idiom. Then read it clean and hear how natural it sounds.
"There's this absolutely fantasticprecision: very nice → fantastic little restaurant near my place. The food is genuinely deliciousprecision: good → delicious, and you're honestly spoilt for choicenatural-fit: a lot of dishes → spoilt for choice — the menu goes on forever. The staff are so warm and welcomingfamily: nice → warm/welcoming, and inside it's really cosy and charmingfamily: nice → cosy/charming. I'm there all the timeprecision: a lot → all the time — honestly, it's become a bit of a second homeidiom, at the emotional beat."
The same rebuild, broken down by which P·F·N move fixed what. Read down and you'll see how the four moves stack to turn a flat plateau answer into a Band 8 one — using only words you already know.
The two versions side by side. Click both and listen for everything you've learned this lesson — the precision, the variety, the natural partnerships, the single earned idiom. Same restaurant, same opinion, same length; only the words changed. This gap is what the Lexical Resource score measures.
"nice" ×4, "good" ×4, "very," "a lot." Clear and correct, but the words carry no specific meaning and the same ones recycle. The examiner learns almost nothing about the restaurant.
Every vague word upgraded, no word repeated, every partnership native, one earned idiom at the beat. Rich and specific — yet not one rare word in it.
Took one ordinary plateau answer and rebuilt it with the full P·F·N system: five precision upgrades, two family swaps to kill the repetition, a natural collocation, and one earned idiom at the emotional beat. Not one word changed in meaning. Not one rare word appeared. But the answer went from "clear but flat" to "vivid and natural" — which on the Lexical Resource scale is the jump from Band 6 to Band 8. This is the entire lesson, and it works on every answer you'll ever give.
Two diagnostic questions before the Practice Arena.
Five answer snippets. For each, identify which of the four failure modes from Stage 2 it is — the Plateau, the Repeat, the Misfire, or the Phrasebook. Recognition first.
Five vague sentences, three upgrades each. Pick the one that's precise AND a natural fit — avoid the still-vague option and the Misfire. This is the Stage 3 builder again, on fresh sentences.
Five sentences with the wrong word-partner. Pick the natural collocation a native would actually use. Feedback names the convention. This trains the partnership instinct that examiners hear as range.
Three answers that overuse one word. For each, rewrite it varying that word from a topic family, so it never repeats — and upgrade any other vague words while you're there. Reveal a model after your attempt.
The capstone drill. Below is a question. Your job: deliver a 45-second answer aloud that consciously applies the full P·F·N system — precise upgrades, varied words from a family, natural collocations, and one earned idiom. Record it and listen back for the plateau words.
"Describe a place in your country that you'd recommend to a visitor." (A Part 2-style prompt — rich ground for the places word-family.)
Five vocabulary exercises done — the range, tested. Here's how it landed.
Your performance across the vocabulary arena shows how well the range is coming together. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 18. Be honest about which still need work — that tells you which move to drill in Lionel's daily upgrade exercise.
Forty-five minutes on the skill that's almost never about knowing more words — only about reaching for the ones you already have.
Eighteen lessons done — three Core Skills in hand. Pronunciation gave your speech music, fluency gave it flow, vocabulary gave it range. The fourth and final Core Skill gives it structure: the grammatical range and accuracy that complete the four marking criteria. Then only the full mock test remains.
Grammatical Range and Accuracy is the last of the four scores — and the one learners most fear and most over-think. Lesson 19 cuts it down to what actually moves the band: a small set of high-value structures used naturally, the accuracy habits that stop small errors stacking up, and the same reach-don't-cram principle you've just learned for vocabulary, applied to sentence structure.
The handful of complex structures — conditionals, relative clauses, the perfect aspect — that signal range when used naturally.
The handful of small errors that stack up fastest — articles, tense, agreement — and the habits that keep them in check.
Reaching for a complex structure naturally, the same way you reach for a precise word — not forcing it in.
After grammar, all four criteria are covered — and Lesson 20 puts everything together in a full mock test.
"Three of the four marking criteria are now yours — pronunciation, fluency, vocabulary. That's three-quarters of your score, built on principle rather than luck. And notice the thread running through all of them: reach, don't cram. Reach for the music, reach for the flow, reach for the precise word — never force, never decorate. Grammar is the same story, and it's the last piece. My ask before Lesson 19 is the one you know by now: drill the daily upgrade — ban the plateau words, graze three chunks. The reach is a habit, and habits are built by reps. One core skill left, then we put the whole thing to the test. You're at ninety percent. The summit's right there."
Eighteen badges, and the third Core Skill done. You can take any flat answer and give it range — upgrading the vague words, varying from a family, keeping every partnership natural, and seasoning with one earned idiom — all without a single rare word. Two lessons left: grammar, and the full mock test.
The fourth and final Core Skill — and the one learners fear most. Pronunciation gave your speech music, fluency gave it flow, vocabulary gave it range; grammar gives it structure. The marking criterion is Grammatical Range and Accuracy — two halves — and the fear comes from thinking you must perform difficult structures flawlessly under pressure. You don't. Here's the truth that changes everything: a Band 7 isn't built from rare, complex grammar performed perfectly. It's built from correct simple sentences, with a few well-chosen complex structures mixed in, and the common errors kept under control. Reach for grammar the way you learned to reach for a precise word — naturally, never crammed.
"Grammar is where I see the most self-sabotage. A student who could get a clean 7 ties themselves in knots trying to force a third conditional mid-sentence, the structure collapses, and they take three errors trying to sound clever. The examiner would far rather hear a correct simple sentence than a broken complex one. Range matters — but it's range used naturally, not range forced in. The secret is to build on a foundation of correct simple sentences you can produce without thinking, then upgrade a few of them to complex structures you genuinely control, and meanwhile stop the small, repeated errors — the articles, the tenses, the agreement — that quietly stack up. Build, upgrade, fix. That's the whole of grammar, and none of it requires you to be a grammarian."
Two answers to the same question. The first tries to sound advanced and collapses into errors; the second builds on correct simple sentences, upgrades a couple naturally, and stays clean throughout. Notice the second isn't "simpler" — it's solid, with range that holds its weight.
"How has technology changed the way people work?"
This candidate reached for a complex conditional and a relative clause, and the whole sentence collapsed — a mangled conditional, wrong tenses, a missing plural, "cannot to," subject-verb disagreement. Six errors in one breath, all from over-reaching. The ambition is admirable; the execution costs band in both halves of the score.
Same ideas, built solidly: a couple of correct simple sentences carry the meaning, then two complex structures are upgraded in cleanly — a relative clause ("which simply wasn't possible") and a third conditional ("if you'd suggested it... would have laughed"), both controlled and natural. Real range, zero errors. That's Build, Upgrade, Fix in action.
The same 9-stage shape, aimed at structure — and the last teaching lesson before the mock test. The framework is B·U·F — Build, Upgrade, Fix — the three moves that give you grammatical range and accuracy without the over-reaching that breaks both.
You are here.
The perfectionism diagnosis, and the 4 grammar failure modes — the Collapse, the Stacking Error, the Safe Monotone, the Tense Slip.
Build, Upgrade, Fix — the three moves to range and accuracy. Plus the interactive Sentence-Upgrade builder.
The handful of high-value complex structures — relative clauses, conditionals, the perfect aspect, and more — used naturally.
Ready-made sentence frames for each high-value structure, plus the top accuracy fixes — all natural and deployable.
The Band 8 move — the handful of errors that stack up fastest (articles, tense, agreement) and the habits that keep them in check.
A fragile, error-filled answer rebuilt live into a solid one — every Build, Upgrade, and Fix marked.
Five grammar exercises.
Self-assessment, badge, and Lesson 20 preview (the Full Mock Test).
Here's what surprises most learners about their own grammar: the majority of their errors don't come from not knowing the rules. They come from over-reaching — attempting a structure more complex than they can control in real time, so it collapses and takes two or three errors down with it. The fear of "simple = low band" pushes them past what they can hold together. Understanding this explains every failure mode on the next screen.
Learners hear "you need complex structures for Band 7" and conclude that every sentence must be elaborate. So they force conditionals, passives, and multi-clause sentences they can't yet control — and the structures break mid-delivery. The irony: chasing range this way destroys accuracy, and you lose marks in both halves at once. The examiner sees the ambition, but scores the wreckage.
Vietnamese grammar works very differently from English — no verb conjugation for tense, no articles, no plural -s, a different word order for questions and clauses. When a learner translates a complex Vietnamese thought directly, all those English machinery points (tense, articles, agreement, word order) have to be added at once, under pressure. That's where the stacked errors come from — too much being assembled live.
Because the problem is over-reaching, not ignorance, the fix is mostly about control — building on simple sentences you can produce flawlessly, then upgrading only to the complex structures you've genuinely mastered, one at a time. You don't need to learn more grammar; you need to deploy the grammar you have within your control, and stop forcing the grammar you don't. That's a strategy you can apply today, not a syllabus you have to complete.
Grammar produces four recognisable failure modes. The first two come from over-reaching, the third from under-reaching, the fourth from the single most common error type. Learn to hear each in your own speech — recognition is the first half of the fix.
Reaching for a complex structure you can't control, and having it break down mid-sentence. You start a conditional or a multi-clause sentence, lose the thread, and the grammar falls apart — often with a restart. One ambitious structure, several errors, plus a fluency hit.
One over-reach that pulls several errors into a single sentence — a wrong tense, a missing article, a disagreeing verb, all at once. The sentence is trying to do too much, so the small machinery points all slip together. Stacked errors read as much weaker than the same errors spread out.
The opposite error: playing it so safe that every sentence is short, simple, and identical in shape — "I like it. It is good. I do it every day." Accurate, but with no range at all, which caps the Grammatical Range half of the score. Correct, but flat and repetitive.
The most common error of all: drifting between tenses, or using the wrong one, especially when telling a story or describing a change. Present where past is needed, dropped past-tense endings, or inconsistent time reference within one answer. Small, but frequent, and it accumulates fast.
A short answer delivered two ways. The first over-reaches and stacks errors; the second builds solidly and upgrades with control. Read both — this is the whole lesson in miniature, and what you'll be able to do by Stage 9.
Q: "Do you think people read less now than before?"
Q: "Do you think people read less now than before?"
Record yourself answering a question, then check two things: do any of your sentences "collapse" — start one way and restart or break — and do you stay in one consistent tense throughout? Those two checks catch the Collapse and the Tense Slip, your two most likely band-droppers. Most learners are surprised how often a sentence breaks mid-structure on playback; it's invisible while speaking and obvious on the recording.
The instinct, once you know range matters, is to cram in more difficult structures. That's the trap — it leads straight to the Collapse. The reframe: don't reach harder, reach cleaner. Build on what you control, and upgrade only when you can land the structure without breaking it. A clean simple sentence beats a broken complex one, every time.
"A student once gave me an answer so tangled with half-built conditionals and passives that I genuinely couldn't follow what he meant. I asked him to say it again, simply. He said: 'People work from home now. Ten years ago, that was rare.' Two clean sentences — clear, correct, and honestly a better answer. Then I showed him how to upgrade just the second one: 'Ten years ago, that would have been almost unthinkable.' One controlled structure, landed cleanly. That's the move — build it right, then upgrade what you can hold. Grammar isn't a competition to use the hardest structure. It's a quiet demonstration that you can build correct sentences and stretch a few of them without breaking. Reach cleaner, not harder."
The Band 7 descriptor for Grammatical Range and Accuracy asks for "a range of complex structures" with "some flexibility," and "frequent error-free sentences." Band 8 wants "a wide range" used flexibly with only "occasional" errors. Notice what's there and what isn't: it asks for error-free sentences (accuracy matters as much as range) and for range used with flexibility (naturally, not forced). Nothing rewards difficulty for its own sake. A few controlled complex structures among correct simple ones is exactly the recipe — which is what building solidly and upgrading cleanly produces.
When you build a correct, well-formed sentence — and stretch it only as far as you can hold it together — you're giving the examiner something solid to follow, something that carries your meaning cleanly from start to finish. That structural soundness is a kind of respect: you're handing them a sentence that holds its weight, not a half-built one they have to prop up. The frameworks gave them your ideas; precision gave them the exact words; sound structure gives them all of it on a foundation that doesn't wobble. The respect is in the structure.
Two questions before we move to the B·U·F framework in Stage 3.
Strong grammar in the test is three moves, in order. Build a correct simple base, Upgrade a few sentences to controlled complex structures, and Fix the errors that recur most. Get all three and your Grammatical Range and Accuracy sits at Band 7-8 — without forcing a single structure you can't control. Each move cures one of the failure modes you just met.
The foundation. Before any complex structure, you need a base of correct simple sentences you can produce without thinking — clear subject, correct verb, right tense. This isn't the low-band option; it's the platform that makes range safe. A correct simple sentence is never wrong, and it's where every good answer starts.
"The full stop is the most underrated tool in grammar. Learners are terrified of short sentences, so they join everything with 'and' and 'which' and 'because' until the whole thing topples. Stop. Literally — use a full stop. 'People are busier now. They read less than before. It's a real shame.' Three short, correct sentences. Clear, accurate, and a perfect base to upgrade. The full stop is what keeps you out of the Collapse. Build on full stops, and you'll never wreck a sentence again."
A crucial reframe: a correct simple sentence is never penalised. The Grammar band doesn't dock you for simplicity — it caps you only if there's no range at all. So the base costs you nothing; it's pure safety. What lifts the band is adding a few upgrades on top of the base, which is the next move. Build first means you always have a correct sentence on the table before you risk a complex one.
The range layer. Once you have a correct base, you upgrade some of those simple sentences into complex structures — but only ones you can land cleanly. The skill is choosing which sentences to upgrade and which structures you genuinely control, so you add range without inviting the Collapse. A few clean complex structures is all Band 7 needs.
"Range isn't quantity, it's evidence. The examiner needs to see that you can handle a complex structure — they don't need every sentence to be one. So give them the evidence cleanly: one well-formed relative clause, one controlled conditional, in a two-minute answer, and you've shown range. The rest can be correct simple sentences, and should be. I'd rather hear nine correct simple sentences and one beautiful conditional than ten ambitious ones that all wobble. Evidence, landed cleanly — that's the upgrade."
Build and Upgrade work as a pair: Build gives you correct simple sentences; Upgrade lifts a couple of them into complex ones you control. Because you started from a correct base, even if an upgrade doesn't quite come, you still have a correct sentence underneath. That safety net is why the order matters — Build first means Upgrade is never a gamble with nothing to fall back on.
The accuracy layer that runs underneath the other two. A small number of error types account for most of the accuracy marks lost — and they're the same ones, again and again. Fix means knowing your top few and targeting them one at a time, until each stops happening. You don't fix all of English; you fix the handful that actually cost you.
"You cannot fix every error at once — your brain can only watch one thing while it speaks. So don't try. Pick your single most frequent error — for most of my Vietnamese students it's tense consistency or articles — and hunt only that one for two weeks. Record yourself, count it, drive it down. When it's mostly gone, move to the next. Fix them one at a time and they actually stay fixed. Try to fix all of them at once and none of them improve. One target at a time is the whole secret of accuracy."
The three moves fuse into a single habit. Build a correct base on full stops; Upgrade one or two sentences with a structure you control; and keep your one Fix-target in the back of your mind the whole time. "People are busier now. They read less than they used to — which is a shame, because reading really enriches you." Correct base, one clean upgrade ("which... because..."), consistent tense throughout. That's B·U·F in one breath, and it's what you'll build on the next screen.
The interactive builder. Each card shows two correct simple sentences, and three ways to combine them into one complex structure. Pick the one that's a clean, controlled upgrade — beware the Collapse option (a broken structure) and the no-change option (still two flat sentences). Five to work through; feedback explains each.
Notice the pattern: the best answer was never the broken ambitious one, and never the two unchanged flat sentences — it was the clean, controlled complex structure. Build correct, upgrade what you can hold. Do this with your own simple sentences and your range climbs without the Collapse.
Two questions before we move to the structures that score in Stage 4.
B·U·F is the principle; this stage is the toolkit. You don't need every complex structure in English — you need four or five high-value ones you can produce cleanly and naturally. Master these, and you have all the range the descriptor asks for. Each one is an upgrade you can apply to a correct simple base.
The easiest, most flexible upgrade — join two ideas with which, that, who, where.
The "if" structures — first, second, and third — that show real grammatical control.
Present perfect and past perfect — for experience, change, and time spans.
Cause-effect linkers and the occasional passive — for flow and formality.
The two most valuable structures, because they're flexible, common, and clear evidence of control. Learn the clean pattern for each, and a couple of ready starters you can drop onto almost any answer.
Two more upgrades that add range cleanly. The present perfect is essential for talking about experience and change; linkers weave your sentences together so they flow as connected reasoning rather than a list. Both are lower-risk than conditionals and easy to land.
Everything in this stage rests on one principle that separates Band 6 grammar from Band 8: you don't need many structures, you need a few that you control so well they come out naturally. Depth of control beats breadth of coverage, every time.
You try to use ten different structures, none of them mastered. Each one is a gamble — sometimes it lands, often it collapses. The examiner hears ambition and instability: lots of attempts, frequent errors, no structure you can clearly produce on demand. That reads as Band 6 — range attempted, accuracy lost.
You have four structures you can produce cleanly every time — relative clause, second conditional, present perfect, a linker. You deploy them naturally, where they fit, with no errors. The examiner hears clear evidence of range, landed accurately and repeatedly. That reads as Band 7-8 — range demonstrated, accuracy intact.
Pick four or five structures and drill them until they're automatic — until they come out clean without you thinking about the mechanics. A relative clause, the second and third conditionals, the present perfect, and a couple of linkers is a complete, Band-8-capable toolkit. Master those few deeply and deploy them naturally on top of a correct simple base, and you have all the range the test will ever ask for. Grammar range isn't how many structures you attempt — it's how many you control.
Five pairs. Pick the version where the complex structure is formed cleanly — the controlled upgrade, not the Collapse. Instant feedback names the structure and the error. This trains the control instinct.
Your grammar toolkit, organised the way B·U·F works. Eighty items across four categories: ready sentence-frames for each high-value structure, the natural linkers that connect your reasoning, the top accuracy fixes for the errors that stack up, and a set of fluent self-correction phrases for when something slips. Every frame is a chunk you can drop straight onto a simple base.
Ready sentence-starters for each high-value structure — relative clauses, conditionals, the perfect aspect. Drop them onto a simple base to upgrade instantly.
The cause, contrast, and result connectors that turn two facts into a piece of reasoning — and double as Part 3 coherence tools.
The top error-types — tense, articles, agreement, prepositions — with the quick rule of thumb that catches most of each.
The fluent little phrases that let you fix a slip mid-sentence without breaking flow — turning an error into a sign of control.
Twenty sentence-frames for the high-value structures. Each is a starter you complete with your own content — the structure is pre-built, so you can't collapse it. Tap any to hear it.
Twenty linkers that join two ideas into one piece of reasoning. They add grammatical range and Part 3 coherence at the same time. Grouped by what they do — cause, result, contrast, adding.
Twenty quick fixes for the high-frequency errors, each with a rule of thumb that catches most cases. You won't get every one perfect — but catching the common patterns clears the bulk of your accuracy losses. Grouped by error type.
Twenty phrases that let you correct an error mid-sentence smoothly. A clean self-correction isn't a weakness — it's evidence of control, because it shows you know the right form. The key is to fix and move on, not to stall. Grouped by what they do.
Don't memorise this toolkit — graze it. Each day, pick three items total: one upgrade frame, one linker, and one accuracy fix (or a self-correction phrase). Use each in a sentence about your own life, aloud, a few times. Over a few weeks you'll have internalised a working set of frames and fixes — reflex, not recall — and that's exactly enough to lift your Grammatical Range and Accuracy a full band. A handful you produce automatically beats eighty you have to recall.
The Band 8 move. The gap between Band 7 and Band 8 in grammar isn't more complex structures — it's accuracy that holds up across the whole test. Band 7 allows "frequent error-free sentences"; Band 8 wants the majority error-free, with only "occasional" slips. The candidates who get stuck at 7 almost always have the range — they lose the band to a steady trickle of small, repeated errors. Closing that trickle is the move, and it's the most controllable thing in the whole test.
"I've been to many country and I really enjoy travel. Last year I go to Japan and it was amazing. The people there is so polite, and I learn a lot of thing about they culture."
"I've been to quite a few countries, and I genuinely love travelling. Last year I went to Japan, and it was amazing. The people there are so polite, and I learnt a lot about their culture."
Individually, none of these errors is serious — "many country" is perfectly understandable. But the Grammar band measures the proportion of your sentences that are error-free, so a steady trickle of small slips drags the whole score down even when every idea is clear and every structure is ambitious. The good news is that this is the most fixable thing in the entire test: the errors are few in type, predictable, and yours specifically. Find your three most frequent, drive them down one at a time, and you convert a Band 7 into a Band 8 without learning anything new.
Accuracy improves through one disciplined method: find your most frequent error, hunt only that one until it's rare, then retire it and move to the next. Trying to fix everything at once fixes nothing. This is how you close the trickle, one error type at a time.
Speaking already takes most of your attention — you're choosing words, building structures, staying fluent, all at once. There's almost no spare capacity left to monitor grammar, and what little there is can watch exactly one thing. Try to watch tense and articles and agreement simultaneously and you'll watch none of them, because your attention can't split that finely under the pressure of live speech. One target at a time isn't a gentle suggestion; it's the only way the monitoring actually works.
Here's a counter-intuitive truth: catching and fixing your own error mid-sentence doesn't hurt your score — it can help it. A clean self-correction shows the examiner you know the right form, which is exactly what they're assessing. The skill is correcting smoothly and moving on, not stalling or over-apologising.
First, fix it fast and move on — a quick "sorry, I mean..." and straight back to your point. Don't stall, don't repeat the whole sentence, don't apologise three times; that turns a small win into a fluency problem. Second, only correct real errors — if you're not sure something's wrong, leave it and keep going, because chasing imaginary errors shreds your fluency for nothing. A clean, confident self-correction is a quiet flex; a flustered, repeated one is a stumble. Aim for the flex.
Step back and see the complete system. Three moves, with accuracy as the finishing polish: build a correct simple base, upgrade a couple of sentences with controlled structures, fix your trickle of recurring errors, and self-correct cleanly if something slips. Once it's a habit, it's not separate steps — it's just how you build an answer.
Take "Technology change how we work. People can work from home now which is good." Run the full system: Fix the agreement ("Technology has changed"); Build the base cleanly ("A lot of people can work from home now"); Upgrade with a controlled "which" comment ("which simply wasn't possible before"); and if "wasn't" came out as "weren't," a quick self-correct ("weren't — wasn't possible"). Result: "Technology has changed how we work enormously. A lot of people can work from home now, which simply wasn't possible before." Same idea, Band 6 to Band 8 — correct base, one clean upgrade, no trickle.
Build correct, upgrade a few, fix the trickle, self-correct cleanly. That's the entire B·U·F system, and it works on every answer in all three parts of the test. None of it requires rare or difficult grammar — it requires a solid base, a few controlled structures, and accuracy that holds up. Grammar is range and accuracy together — built solidly, not performed riskily.
Two questions before we move to the worked example in Stage 7.
We'll take a real fragile answer — over-reaching, error-stacked, collapsing — and rebuild it live with the full B·U·F system. Same ideas, same intent. We build a correct base, upgrade a couple of sentences with control, fix the trickle of errors, and watch it climb from Band 6 to Band 8 without forcing a single difficult structure.
Q: "Has the way people communicate changed in recent years?" A (fragile): "Yes it change a lot because before people is writing letter but now if you would want to talk someone you just send message and it have many app for this which everyone using it and make communication more faster than before."
Break the runaway sentence into short, correct ones on full stops, each carrying one clear idea.
Add one clean "which" comment and one controlled conditional — evidence of range, landed cleanly.
Repair the stacked errors: agreement, tense, plurals, "more faster," "talk someone."
Correct base, two clean upgrades, zero errors. Band 6 to Band 8, no difficult grammar.
Here's the same answer rebuilt. Toggle the annotations to see which move each part came from — build, upgrade, fix. Then read it clean and hear how solid it sounds.
"Yes, it's changed enormously.fix: "it change" → "it's changed" People used to write letters,build: short correct base, on a full stop whereas now we just send a quick message.upgrade: "whereas" contrast linker There are so many apps for it,fix: "it have many app" → "there are many apps" which has made communicating far faster than before.upgrade: clean "which" comment + "far faster" If you'd wanted to reach someone instantly twenty years ago, it would have been almost impossible.upgrade: controlled third conditional"
The same rebuild, broken down by which B·U·F move fixed what. Read down and you'll see how the moves stack to turn a fragile, collapsing answer into a solid Band 8 one — without any difficult grammar.
The two versions side by side. Click both and listen for everything you've learned this lesson — the correct base, the controlled upgrades, the closed trickle. Same ideas, same intent; only the construction changed. This gap is what the Grammatical Range and Accuracy score measures.
One runaway sentence trying to do everything at once, a broken conditional, and seven small errors trickling through. The ideas are clear, but the construction wobbles and collapses.
A correct simple base on full stops, three controlled upgrades woven in cleanly, and every small error fixed. Clear range, accuracy that holds from start to finish.
Took one fragile, collapsing answer and rebuilt it with the full B·U·F system: broke it onto a correct base, upgraded three sentences with controlled structures, and fixed every small error in the trickle. Not one idea changed. Not one difficult structure appeared. But the answer went from "ambitious but wobbling" to "solid with clear range" — which on the Grammar scale is the jump from Band 6 to Band 8. This is the entire lesson, and it works on every answer you'll ever give.
Two diagnostic questions before the Practice Arena.
Five answers. For each, name the failure mode from Stage 2 — the Collapse, the Stacking Error, the Safe Monotone, or the Tense Slip.
Five sentence pairs, three ways to combine each. Pick the clean, controlled complex structure — avoid the Collapse and the no-upgrade option.
Five sentences, each with one high-frequency error. Pick the corrected version.
Three fragile answers. Rewrite each with the full B·U·F system — build a correct base, upgrade one or two sentences cleanly, fix every error. Reveal a model after your attempt.
The capstone drill. Deliver a 45-second answer aloud applying the full B·U·F system — a correct base, one or two clean upgrades, no error trickle. Record it and listen back.
"How do you think the way people travel will change in the future?" (A Part 3 prompt — rich ground for conditionals.)
Five grammar exercises done — the structure, tested.
Your performance across the grammar arena shows how well the structure is coming together. Move into Stage 9 to complete the lesson.
Six sliders covering everything new in Lesson 19. Be honest about which still need work.
Forty-five minutes on the final core skill — the one that completes all four marking criteria.
Nineteen lessons done — all four core skills in hand, all three parts mastered. Everything you've built now comes together. Lesson 20 isn't a teaching lesson; it's the real thing — a complete simulated IELTS Speaking test from start to finish.
A complete end-to-end simulation: Part 1 (interview), Part 2 (the long-turn cue card with one minute's preparation), and Part 3 (the discussion) — delivered in sequence, under timing, just like the real exam. Then you self-assess against all four marking criteria using the full rubric.
Part 1 interview, Part 2 cue card with prep, Part 3 discussion — the full 11-14 minute exam.
Real timings throughout, including the Part 2 preparation minute and long turn.
Score yourself on fluency, vocabulary, grammar, and pronunciation with the full band rubric.
Complete it and you've finished the entire course — every part, every core skill.
"All four criteria are yours now — pronunciation, fluency, vocabulary, grammar. Every one built on the same idea: reach, don't cram. There's one thing left, and it's the most important: putting it all together under real conditions. A mock test is where you find out what's become automatic and what still needs drilling. Don't aim for perfection — aim to use what you've built, all at once, the way you will on the day. Nineteen lessons down. One to go. Let's see what you can do."
Nineteen badges, and the final Core Skill done. You can build any answer solidly — a correct base, a few controlled upgrades, the trickle of errors closed. Pronunciation, fluency, vocabulary, grammar: the complete set. One lesson left — the full mock test.
Every lesson with your child counts.
Lanbridge Bank Corporation · Training Manager
Readiness sphere, capability map, ordered development tracks, interactive lessons and workplace simulations—inside this dashboard.
5-minute AI assessment · Get a personalized Corporate Readiness Report. No signup required.
Get an instant Corporate Readiness Report in 5 minutes.
No email required. No signup. Just answers.
"Following the quarterly review, the board expressed concerns about the team's ______ to deliver on key performance targets. A follow-up meeting has been scheduled to discuss remediation strategies."
"Describe a time when you had to explain a complex idea to a colleague. What approach did you take?"
Or skip recording and rate your team's confidence:
Your answer personalizes the Corporate Readiness Report.
Based on 5-minute micro-assessment ·
How your team compares to other companies in your sector.
Personalized for your team's biggest gaps.
A boardroom-ready scenario tool. Adjust inputs to see strategic projections, peer benchmarks, and the cost of delay — all backed by industry data.
Your team could qualify for 23% more international projects with Band 7.0+ English.
Comparable to teams at leading Vietnamese corporates.
Every quarter your team stays at Band 5.5 = estimated lost contract eligibility, based on industry win-rate differentials.
Get a board-ready PDF with your company name, full scenario projections, and a 90-day implementation roadmap.
We'll generate a PDF with your inputs, peer benchmarks, and a 90-day rollout plan. Reviewed by Lionel before sending.