The Five Listening Errors We See Most
None of these five is a comprehension problem. They are habits, which is why they recur across candidates at very different levels and why they can be changed in a fortnight.
What these five have in common
Across attempts on this site, the same handful of behaviours show up over and over in the way candidates work through the Listening section. They appear at every level, including among candidates whose English is clearly strong enough for the band they want.
That is the interesting part. If the recurring problems were vocabulary and grammar, the advice would be to study more English, and improvement would be measured in months. They are not. They are decisions made under time pressure, made badly, and made the same way every time because nobody has ever named them.
Each of the five below has a specific trigger, a specific cost, and a specific replacement habit. Read them as a checklist for your next practice set rather than as general advice.
One framing that helps: none of these five is about understanding the audio better. Every one is about what you do with the understanding you already have.
One: transcribing instead of anchoring
The signature is a page of neat, largely complete sentences and a candidate who reports that the clip went too fast. It did go too fast, because handwriting was consuming the attention that listening needed.
The trigger is anxiety about forgetting. Writing more feels like insurance, and it is the opposite: every complete sentence you produce by hand costs you the two sentences spoken while you produced it. Transcription in a single-pass section is not thorough, it is expensive.
There is a second cost that candidates do not anticipate. Dense notes are slow to search, and in Parts 1 to 3 you have around thirty seconds per question. A page of prose cannot be scanned in the time available, so the notes that cost you the listening also fail to pay you back at the question.
The replacement is anchoring: six or seven items per clip, drawn from four categories. Names and roles, numbers, decisions, and reversals. If something does not fall into one of those, it does not go on the page.
Two: chasing a detail that is already gone
The trigger is the moment of realising you missed something. The instinct to reach back for it is immediate, strong, and almost never worth acting on.
The cost is not one mark, it is a stretch of audio. Searching your memory and parsing incoming speech draw on the same attention, so for as long as you are reaching, you are not listening. Candidates describe arriving at the end of a clip without knowing how the conversation resolved, and this is usually why.
The replacement is a hard four-second limit and an explicit closing move: a dash on the page where the detail should be, and back to the audio. The dash is not a consolation prize. It tells you at the question that something existed at that point, which frequently eliminates two options on its own.
This one has to be rehearsed rather than merely understood. Practise letting a detail go and notice how long it takes you to re-engage, because recovery time is the variable that predicts how a real section goes.
Run one practice set per habit, with a single rule for the session, and judge yourself on whether you kept the rule rather than on the score. Policing five behaviours at once makes you listen worse than policing none.
Three: answering Parts 4 to 6 in printed order
The trigger is simply habit. The questions are numbered, so candidates answer them in sequence, as they would on almost any other test they have taken.
It is the wrong policy here because the whole screen shares one time limit and the questions are not ordered by difficulty. Ninety seconds spent on question two leaves questions five and six rushed or unanswered, and on a Part 5 screen with eight questions that can be several marks.
The cost is invisible at the time, which is why it persists. A candidate who ran out of budget remembers the hard question they wrestled with, not the easy one they never reached.
The replacement is two passes. First pass: answer everything you are certain of, quickly, and put a placeholder on anything that needs thought. Second pass: spend what is left on the placeholders. The second pass is also better informed, because settling four questions tends to reload the clip and constrain the remaining options.
Four: choosing the option whose words you heard
The trigger is recognition. Under time pressure, an option containing a phrase you remember from the audio produces a small feeling of relief, and relief is very difficult to tell apart from knowledge.
The cost is systematic rather than occasional, because of how the questions are written. Correct answers in this section are always paraphrases; they never repeat the audio's wording. Distractors, however, are frequently built from wording that was genuinely used. So the option that feels most familiar is more likely to be the trap than the key.
This is the error that catches strong listeners hardest, precisely because they heard the clip clearly enough to recognise the phrasing. Comprehension gives them a false signal that a weaker listener never receives.
The replacement is a two-second question asked before selecting: what question does this option answer? Echo distractors usually answer something real from elsewhere in the clip. Asking makes the mismatch visible in the time available.
Ask 'what question does this option answer?' before selecting any option that sounds familiar. Correct answers never repeat the audio's wording, so familiarity is a signal pointing the wrong way, and this is the error that catches the strongest listeners.
Five: leaving questions blank
The trigger is a sense that guessing is not legitimate, or an intention to come back later that the clock then defeats.
The cost is the clearest arithmetic on the test. There is no deduction for a wrong answer, so a blank returns nothing while a blind selection on four options returns one in four, and a selection with one option eliminated returns one in three. Across a section, this is regularly worth a mark or two.
The intention-to-return version is worse, because parts do not reopen. Anything unanswered when a part ends is unanswered permanently, no matter how much time you have in the part that follows. The compartments are sealed.
The replacement is an absolute rule, enforced in every practice session so that it requires no decision on the day: nothing advances with a blank on it. In Parts 1 to 3 that means choosing before the thirty seconds expires. In Parts 4 to 6 it means placing a holding answer on every question during the first pass.
Working on them
Do not try to fix all five in one sitting. Attention spent monitoring your own behaviour is attention not spent on the audio, and a candidate policing five habits at once will listen worse than one policing none.
Take one per practice session. Run a full Listening set with a single instruction to yourself, for instance that today you will not write a complete sentence, or that today every question gets an answer. Judge the session on whether you kept the rule, not on the score.
After five sessions the habits start to be automatic, and automatic is the goal. A strategy that still requires a decision under pressure is not yet a strategy; it is an intention, and intentions are what fail on test day.
Then return to scoring. When the five behaviours no longer need attention, the score you get is finally a measurement of your listening rather than of your habits, and at that point it becomes useful feedback.
Make the no-blank rule absolute in practice as well as on the test. Habits that still require a decision under pressure are intentions, and intentions are what stop working on the day.
Practise it
Five questions on the habits that cost marks independently of your English.