If you cannot answer these questions yet, ten more practice tests will only help you repeat the same mistakes faster.
If you cannot answer these questions yet, ten more practice tests will only help you repeat the same mistakes faster.
Most IELTS learners practise the same way: sit a test, mark it, look at the answer key, then start the next test. The score hovers around one level and refuses to climb. The reason is simple — practice tests measure the result; they do not diagnose the cause.
The twenty questions below work the other way round. Each targets a specific bottleneck — the places where most learners lose the most marks across all four skills. Before you open the next test, try answering each one in your own words, without looking at the explanation. This is active recall: the most honest check of whether you actually understand something or merely find it familiar.
The problem is not vocabulary but connected speech — the way native speakers link sounds, drop sounds and rush through function words. What do you want to do? in reality sounds close to Whaddaya wanna do?. If your only listening practice has been past papers, and you have never studied linking rules and weak forms, your ear is searching for words that do not exist in the actual sound.
IELTS Listening is essentially a test of paraphrase in real time. The question says free of charge; the recording says you won't have to pay anything. The skill to train is: read the question → predict two or three other ways the keyword could be expressed → listen for the idea, not the word.
Because all three are mentioned on purpose. The classic trap is self-correction: the speaker offers option A, then changes their mind with actually, but, having said that, on second thoughts. The correct answer almost always sits after those signal words. If you choose the moment you hear matching information, you are choosing exactly the trap that was designed for you.
Missing one question costs one mark. Panicking over one question costs three or four. The skill here is letting go at the right moment: accept the loss immediately, move your eyes to the keyword of the next question and use it as an anchor to pick the rhythm back up. This needs deliberate practice, because the brain's natural reflex is to cling to the question you just lost.
Do not listen cold. During the reading time, the layout of the questions is a map of the lecture: the headings and questions tell you in advance which points the talk will pass through, and in what order. Good listeners do not concentrate evenly for six minutes — they know when to focus hard and when the talk is merely in transition.
A working frame: 15 – 20 – 25 minutes, because the passages get harder. The more important principle: every question is worth the same single mark. A brutal Matching question is not worth more than an easy gap-fill. When you fall behind, drop whatever is eating the clock, collect all the easy marks first, and come back if time allows.
False = the passage says the opposite of the statement. Not Given = the passage does not carry enough information to judge it true or false — even when the statement sounds perfectly reasonable given what you know. The test: if you have to use knowledge from outside the passage to decide, the answer is Not Given. The exam asks about the text, not about the world.
Because you match the heading to the most striking detail instead of the paragraph's main idea. A paragraph may spend three sentences on a memorable example, but that example only serves an argument sitting in the first or last sentence. The paper always includes a bait heading that matches the detail. The question to ask yourself: What does this paragraph exist to do within the whole passage?
Only stop for words that decide the answer — words inside the sentence carrying the information the question asks about. For those, infer from context: what part of speech is it, is the sentence positive or negative in tone, what do the surrounding words suggest? Every other unfamiliar word: skip it without mercy. High-band readers do not know as many more words as you think — they are better at knowing which words they do not need to know.
Skimming = reading quickly for the main idea (used for Matching Headings, main-idea questions, author's purpose). Scanning = sweeping for specific information — proper nouns, figures, technical terms (used for gap-fill, True/False/Not Given, Matching Information). The most common mistake: reading every question type the same way — usually reading everything carefully, and running out of time.
Three filtering questions: What is the largest and smallest? What changed the most (or stayed unusually flat)? Which groups can be compared with each other? Any figure that does not fall under those three does not need mentioning. And the overview sentence is compulsory — without an overview, Task Achievement is capped at band 5 no matter how well the rest is written.
Because on topic is not the same as fully answering the prompt. A Discuss both views and give your opinion prompt contains three jobs: view 1, view 2, and a clear opinion of your own. Missing one costs you Task Response marks, however well written the essay is. Before writing, underline each part of the prompt and ask: does my outline have a place for every part?
Because Moreover, Furthermore, In addition sprinkled across sentence openings is surface cohesion. Real cohesion lives in two places: (1) the thread of ideas — each sentence developing naturally from the one before, and (2) reference — this trend, such policies, the former. Band 7+ examiners are looking for writing that still holds together if you delete every linking word. Overusing them is itself penalised as mechanical.
The principle: one correct complex sentence is worth more than three broken ones. Grammar is marked on both range and accuracy — a botched complex sentence drags down both. The safe approach: pick two or three structures you already own (relative clauses, although, second conditional…) and use them fluently, rather than trying a new structure for the first time in the exam room. Expand your range during practice, not on test day.
It is collocation. Swapping start for commence in an essay about education does not raise your score — it may well be judged unnatural. But raise awareness, pose a threat, meet the demand do, because Lexical Resource at band 7 asks for natural and accurate collocation. Learn words in chunks from the start; do not learn single words and assemble them yourself.
Do not try to generate more points — go deeper into one, through a four-layer frame: the event (what happened) → the feeling (what you felt, and why) → comparison (before and after, against other people, against expectations) → hypothetical (what if it had gone differently). One small experience taken through those four layers fills two minutes easily, and sounds far more natural than a list.
You do not need the right opinion — you need an opinion that is developed. The formula: position → reason → concrete example → concession (Having said that, I suppose some people would argue…). That concession step is precisely what lifts an answer from band 6 to band 7: it shows you can see more than one side of the issue.
These are marked as separate criteria, so do not sacrifice one entirely for the other — but in the moment, when you must choose: keep talking. Stopping to self-correct every small slip damages Fluency more than the slip damages Grammar. Safe fillers: well, let me think, that's an interesting question — in moderation. Penalised fillers: drawn-out umm… err… and meaningless repetition.
Accent is not part of the marking criteria. What is marked: word stress and sentence stress (stressing the words that carry meaning), intonation (rising and falling with the sense, not flat), and chunking (pausing between meaning groups, not inside them). A clearly Vietnamese accent, with stress and pauses in the right places, can absolutely reach Pronunciation band 7+.
Because a memorised answer has a reciting intonation, an unnaturally even pace, and collapses the moment the examiner steps off the script — they are trained to do exactly that, and memorised answers are marked down. What is worth preparing is idea frames (angles of approach for each topic group) and functional phrases (It's a bit of a double-edged sword because…) — flexible raw material, not a rigid script.
Count the ❌ marks by skill. Any skill with two or more is where your score is being held back — and also where each hour of study returns the most marks. Practice tests are only genuinely valuable when they confirm you have fixed a bottleneck, not when they substitute for fixing it.
And do not check yourself only once. Come back to this set after one week, then after one month — exactly in the spirit of spaced repetition. A question that moves from ❌ to ✅ and stays there across several checks is what real progress looks like.
The same article, told in plain words — for younger readers, or for anyone who wants the point quickly.
If you cannot answer these questions yet, ten more practice tests will only help you repeat the same mistakes faster.
Most people practise IELTS the same way: sit a test, mark it, look at the answer key, start the next one. The score sits at one level and refuses to climb. The reason is simple. A practice test measures the result. It does not tell you the cause.
These twenty questions work the other way round. Each aims at one place where people lose the most marks. Before you open the next test, answer each in your own words, without reading the explanation first. That is active recall — the most honest way to find out whether you understand something or only recognise it.
Answer each one aloud or on paper, in 30 seconds. Then compare with the explanation. Mark yourself three ways: ✅ answered clearly / ➖ vague / ❌ could not answer. Any skill with two or more ❌ — stop doing practice tests in that skill and go fix the bottleneck first.
1. Why can you understand every word on its own, then fall behind in a whole sentence at normal speed? It is not vocabulary. It is connected speech — native speakers link sounds, drop sounds and rush through the small words. What do you want to do? really comes out close to Whaddaya wanna do?. If all you have ever done is past papers, and you have never studied linking and weak forms, your ear is hunting for words that are not there in the sound.
2. How do you catch the answer when the recording never uses the words in the question? Listening is basically a paraphrase test at speed. The question says free of charge; the recording says you won't have to pay anything. Train it like this: read the question, predict two or three other ways that keyword could be said, then listen for the idea instead of the word.
3. Why do you choose the wrong option even when you clearly hear all three mentioned? Because all three are mentioned on purpose. The classic trap is the speaker changing their mind: they offer A, then say actually, but, having said that, on second thoughts. The right answer almost always sits after those words. If you choose the instant you hear something matching, you are choosing the trap that was built for you.
4. When you miss one, how do you stop yourself losing the next two as well? Missing one question costs one mark. Panicking about it costs three or four. The skill is letting go at the right moment: accept the loss straight away, move your eyes to the keyword of the next question, and use it to catch the rhythm again. This needs practising, because the brain's reflex is to cling to the question you just lost.
5. Section 4 runs six minutes with no break — how do you hold your attention? Do not listen cold. In the reading time, the questions are a map of the lecture: they tell you in advance which points the talk will pass through, and in what order. Good listeners do not concentrate evenly for six minutes. They know when to lock on and when the talk is only moving between ideas.
6. Sixty minutes, three passages — how do you split the time, and what do you drop when you are behind? A workable frame is 15 – 20 – 25 minutes, because the passages get harder. The bigger principle: every question is worth the same one mark. A brutal Matching question is not worth more than an easy gap-fill. When you fall behind, drop whatever is eating the clock, take all the easy marks first, and come back if there is time.
7. Where exactly is the line between False and Not Given? False = the passage says the opposite. Not Given = the passage does not give you enough to judge, even when the statement sounds perfectly reasonable. The test: if you have to use knowledge from outside the passage to decide, the answer is Not Given. The exam is asking about the text, not about the world.
8. Why do you still get Matching Headings wrong after understanding the paragraph? Because you match the heading to the most striking detail instead of the main idea. A paragraph may spend three sentences on a memorable example, but that example only serves an argument sitting in the first or last sentence. The paper always includes a bait heading that fits the detail. Ask yourself: what does this paragraph exist to do inside the whole passage?
9. You meet an academic word you do not know — guess or skip? Only stop for words that decide the answer — the ones inside the sentence carrying the information the question is about. For those, work it out from context: what part of speech is it, is the tone positive or negative, what do the neighbouring words suggest? Every other unknown word: skip it without mercy. High scorers do not know as many more words as you imagine. They are better at knowing which words they do not need.
10. How do skimming and scanning differ, and which question wants which? Skimming = reading fast for the main idea (Matching Headings, main-idea questions, author's purpose). Scanning = sweeping for specific things — names, numbers, technical terms (gap-fill, True/False/Not Given, Matching Information). The commonest mistake is reading every question type the same way, usually reading everything carefully, and running out of time.
11. Task 1: how do you pick the key features instead of describing every number? Three filters. What is the largest and the smallest? What changed the most, or stayed unusually flat? Which groups can be compared with each other? Any figure that fails all three does not need a mention. And the overview sentence is compulsory — with no overview, Task Achievement is capped at band 5 however well the rest is written.
12. Task 2: why is Task Response low when you wrote 250 words and stayed on topic? Because on topic is not the same as fully answering the prompt. Discuss both views and give your opinion contains three jobs: view one, view two, and a clear opinion of your own. Miss one and you lose Task Response marks however well the essay is written. Before you write, underline each part of the prompt and check that your outline has a place for every part.
13. You use plenty of linking words and Coherence still will not move — why? Because Moreover, Furthermore, In addition scattered over sentence openings is surface cohesion. Real cohesion lives in two places: the thread of ideas, each sentence growing out of the one before, and reference — this trend, such policies, the former. Band 7 examiners want writing that still holds together if you delete every linking word. Overusing them is itself marked down as mechanical.
14. How do you write complex sentences without raising your error count? One correct complex sentence is worth more than three broken ones. Grammar is marked for range and accuracy, so a botched complex sentence drags down both. The safe way: take two or three structures you already own — relative clauses, although, the second conditional — and use them fluently, instead of trying a new structure for the first time in the exam room. Widen your range in practice, not on test day.
15. Is high-band vocabulary about difficult words or natural collocation? Collocation. Swapping start for commence in an essay about education raises nothing and may be judged unnatural. But raise awareness, pose a threat, meet the demand do raise it, because band 7 Lexical Resource asks for natural, accurate collocation. Learn words in chunks from the beginning. Do not learn single words and assemble them yourself.
16. Part 2: how do you fill two whole minutes without running dry? Do not hunt for more points — go deeper into one, through four layers: the event (what happened), the feeling (what you felt and why), the comparison (before and after, against other people, against what you expected), the hypothetical (what if it had gone differently). One small experience taken through those four layers fills two minutes easily, and sounds far more natural than a list.
17. Part 3 is abstract and you have no ready opinion — how do you answer with depth? You do not need the right opinion. You need an opinion that is developed. The shape: position, reason, concrete example, concession — Having said that, I suppose some people would argue…. That last step is exactly what lifts an answer from band 6 to band 7, because it shows you can see more than one side.
18. If you cannot have both, fluency or accuracy? They are separate criteria, so do not sacrifice one entirely — but in the moment, when you must choose: keep talking. Stopping to correct every small slip hurts Fluency more than the slip hurts Grammar. Safe fillers: well, let me think, that's an interesting question, in moderation. Penalised fillers: long drawn-out umm… err… and repeating yourself for nothing.
19. What is high-band pronunciation, if it is not a British or American accent? Accent is not in the marking criteria at all. What is marked: word and sentence stress — stressing the words carrying the meaning; intonation — rising and falling with the sense rather than staying flat; and chunking — pausing between meaning groups instead of inside them. A clearly Vietnamese accent, with the stress and the pauses in the right places, can absolutely reach Pronunciation band 7 and above.
20. Why do examiners spot a memorised answer at once — and what should you prepare instead? Because a memorised answer has a reciting tune, an unnaturally even pace, and falls apart the moment the examiner steps off the script — which they are trained to do, and memorised answers are marked down. What is worth preparing is idea frames (angles for each topic group) and functional phrases (It's a bit of a double-edged sword because…) — flexible raw material, not a fixed script.
Count your ❌ by skill. Any skill with two or more is where your score is being held down — and where an hour of study pays back the most marks. Practice tests are genuinely worth doing when they confirm you have fixed a bottleneck, not when they stand in for fixing it.
And do not check yourself only once. Come back after a week, then after a month — exactly in the spirit of spaced repetition. A question that moves from ❌ to ✅ and stays there across several checks is what real progress looks like.
These articles summarize well-established research in learning science and linguistics. Key sources and further reading:
The recording plays exactly once. That single fact governs the whole paper, and every technique worth learning is a cons...
The three parts are not three difficulty levels of the same task. They test different things — and the answer length tha...
Task 2 counts twice as much as Task 1. So the most consequential decision you make is not a good phrase — it is stopping...