WiderAI › Blog › Prediction: The Thirty Seconds That Decide IELTS Listening
Prediction: The Thirty Seconds That Decide IELTS Listening
Prediction is the technique of using the short reading time before each part of IELTS Listening to work out, in advance, what kind of answer every question needs — a number, a date, a plural noun, a place, an opinion — and often to guess the likely content itself. It sounds like a small preparation habit. It is actually the dividing line between candidates who catch answers and candidates who chase them, because the recording plays once, at full speed, and does not wait.
The reasoning is simple. If you go into a question blind, you must do two jobs simultaneously when the moment comes: understand what is being said and work out whether it answers the question. At conversational speed, that is one job too many. If instead you already know that question 5 needs a price and question 6 needs a day of the week, listening stops being open-ended comprehension and becomes confirmation: you are waiting for a price, and when a price arrives you take it. Prediction converts the hardest task in the test into the easiest.
Every high scorer does some version of this, usually without naming it. This article turns it into an explicit routine you can run in the pauses the test gives you — before each part, and again in the gaps the recording announces — so that no question ever starts from zero.
What exactly is the prediction technique?
Before each part, a voice on the recording tells you to look at a set of questions, and gives you time to do it. The prediction technique is a fixed sequence for that time:
- Read the instruction and fix the word limit — it silently disqualifies answers all test long.
- Skim the scenario: who is talking, about what? A form tells you it is a booking; a map tells you directions are coming; matching names to opinions tells you a discussion is coming.
- For each gap, decide the answer's type — the grammatical shape and the category of thing.
- Where possible, guess the content — not to write it down, but to prime your ear for the area of meaning.
- Underline the anchor words around each gap: the printed words most likely to be paraphrased just before the answer arrives.
The output of those steps is a mental map: I am listening for a number, then a noun, probably a facility, then a day. With that map, you always know where you are — and knowing where you are is the real currency of IELTS Listening, because the one unrecoverable error is losing your place.
How do you predict the type of each answer?
The words around a gap constrain it tightly, and reading those constraints is a skill worth drilling on paper, without any audio at all.
- Grammar first. A gap after "a" or "the" needs a noun phrase; after "to" it may need a verb; "every ____" needs a singular, "one of the ____" a plural. If the completed sentence must be grammatical — and it must — the gap's word class is rarely in doubt.
- Printed symbols and units. "£____" needs a bare figure; "____ am" needs a clock time; "____ Road" needs a name, which will be spelt aloud.
- Category from sense. "Meet at the ____" is a place; "bring a ____" is an object; "Date of arrival: ____" is a date. Say the category to yourself in one word.
- Paraphrase the anchors. The recording will not say the printed words. If the question says "cost", expect to hear "price", "fee", "charge" or "that comes to". Spending three seconds generating one or two synonyms for each anchor is the single highest-value prediction habit, because the paraphrase of the anchor is your cue that the answer is next.
Content guessing goes one step further. "The tour does not run in ____" invites you to expect a month, probably a winter one. You are not committing to the guess — you are narrowing the space your ear must search. When the prediction is wrong, nothing is lost: the type was still right, and the type is what catches the answer.
How should you use the breaks and pauses?
The test hands you more preparation time than most candidates use. Before every part there is dedicated reading time; within and between parts the recording announces further pauses; and at the end you have transfer time on the paper test or checking time on the computer test.
Spend every announced pause on the next questions, not the previous ones. Reviewing what has already played feels responsible but buys almost nothing — the answers you have are the answers you have — while priming the next set buys marks. The one exception: if a completed section contains a gap you left blank, write a guess now (there is no penalty) and move on.
The parts get harder as the test goes on, and the later question formats — multiple choice, matching — carry heavy reading loads. Any spare seconds you can bank early by being efficient in Parts 1 and 2 should be pushed forward into reading Part 3 and Part 4 stems, because Part 4 is a single long monologue with a single reading pause and rewards preparation more than any other section.
Does prediction work for every question type?
Yes, but the object of prediction changes. For completion tasks you predict the answer's form and category, as above. For multiple choice you cannot predict the answer, so you predict the discussion: compress each option to a two-word idea and identify what actually separates the options, so you listen for the difference rather than the shared topic. For maps and plans you predict the language of location — "opposite", "just past", "in the corner" — and orient yourself on the diagram before the audio starts: find the entrance, note the compass if printed. For matching tasks you read the shorter list until you nearly know it by heart, so your eyes can stay on the longer list while the audio plays.
The common thread is that in every format, the reading time is for building expectations, and the listening time is for checking them. Candidates who read passively — just looking at the questions — get very little from the same seconds.
Prediction is also a transferable exam habit. The same discipline — know what you are listening for before it arrives — is what stops Part 2 of the speaking test collapsing into silence, which is why a timed IELTS speaking mock test with preparation minutes enforced is useful cross-training for listeners.
WiderAI gives you AI-scored IELTS mock tests with instant band scores and examiner-style feedback.
Start practising freeFrequently asked questions
How much time do you get to read the questions in IELTS Listening?
The recording itself tells you when to read and which questions to look at, typically giving around half a minute before each group of questions, plus pauses within parts and checking time at the end. The exact seconds vary, which is why a fixed prediction routine matters: it guarantees the time is spent on the same high-value steps every time.
Should I predict actual answers or just the type of answer?
Always the type; the content only when it is nearly free, as with "closed in ____" suggesting a month. The type — number, plural noun, place, name — is what lets you recognise the answer at speed, and it is never wasted. A content guess is a bonus that primes your ear, and a wrong content guess costs nothing as long as you never write it without hearing it.
Is it better to read ahead or check previous answers during pauses?
Read ahead, almost without exception. Past questions are settled and re-reading them earns nothing, while the next section's questions are where the next marks are. The only backward glance worth taking is to fill any blank with a guess, since unanswered questions score zero and wrong answers carry no penalty.
What if my prediction turns out to be wrong?
Then you discard it the moment the audio contradicts it — the prediction was scaffolding, not an answer. Even a wrong content guess usually had the right type, and the type still catches the real answer. The only dangerous habit is writing a predicted answer you never actually heard; predictions are for your ear, never for the answer sheet.