Listening Practice

Before the recording starts, you're given time to read the questions — use it to predict what KIND of answer belongs in each blank (a date, a price, a name, a place), using the words around the blank and the word-limit instruction. Decide this BEFORE you hear anything.

Once the audio starts, there's no time left to figure out what you're even listening FOR — that decision has to already be made, exactly like predicting the answer type in Reading's completion tasks, but here you only get one pass to use it.

The recording plays once, with no pause or rewind — write your best guess immediately as you hear it, even in rough shorthand, rather than waiting to compose a neat, perfect answer. You can tidy the spelling later if there's a spare moment.

Hesitating to "get it exactly right" is the single biggest cause of missing the NEXT answer too — while you're still writing the first answer neatly, the speaker has already moved on and said the second one.

Phrases like "now," "moving on to," "as for," or a clear change of topic signal that the recording has moved to a new question — use these to confirm you're on the right blank, especially if you fell slightly behind.

If you fall behind by even one answer, signpost language is your only way back on track — with no visual pause, the recording itself has to tell you when one question ends and the next begins.

A correct-sounding answer with one wrong letter or digit is marked wrong. Write numbers and spelled-out letters carefully, and double-check word pairs that sound alike, especially "-teen" vs. "-ty" endings (thirteen/thirty, fourteen/forty).

Listening is graded mechanically, exactly like Reading's completion tasks — understanding the general meaning isn't enough for credit if the spelling or number written down is wrong.

Just like in Reading, the grammar around a blank tells you what KIND of word fits — noun, verb, adjective, or adverb — before you even hear the recording. Look at the words immediately before and after the blank on the printed form.

Narrowing down the word CLASS, not just the topic, means you can instantly reject an answer that's grammatically impossible, even if it sounds similar to something you heard.

The blanks in a Listening completion task are almost always answered in the exact order the speaker presents them. Unlike Reading, you can't scan ahead or jump back — treat the form top-to-bottom as your listening roadmap, and trust that blank 3 will be answered before blank 4.

If you're unsure of an answer, staying in order is what lets you catch up — jumping around the form looking for a specific blank wastes time you don't have, since the recording never waits.

Natural speech often corrects itself mid-sentence — "the fee is thirty pounds... sorry, that's actually thirty-five." Always write the LAST version given, not the first one you hear.

This is one of the most common real Listening traps — test-writers deliberately include a spoken correction to catch students who write down the first number they hear and stop listening.

A plan/map description is usually a narrated walk-through — "as you come in through the entrance, you'll see... continuing past that, on your right..." — follow the imagined movement through the space, rather than treating it as an unordered list of facts.

Locations are described in the order you'd physically walk past them, not in the order they appear on your printed plan — following the route keeps you moving through the diagram at the same pace as the speaker.

New locations are usually described relative to one you've already placed — "the café is to the left of reception." Build the map incrementally: place the first, fixed reference point, then use it to place the next, and the next.

Trying to place every label independently means you have nothing to check a location against — anchoring each new one to an already-placed point catches mistakes immediately, since a location that doesn't fit the pattern stands out.

With two speakers, one often proposes an idea that the other then rejects or improves on — the recording is testing whether you can follow the discussion to its actual conclusion, not just note the first thing suggested.

This is the multi-speaker version of catching a self-correction (Chapter 16) — except now the correction comes from a SECOND voice agreeing, disagreeing, or proposing an alternative, which is easy to miss if you stop listening once you hear a plausible-sounding answer.

When a question asks specifically what ONE speaker thinks or plans ("What does Priya suggest?"), only that speaker's words count — the other speaker's opinion, even if stated first or more confidently, is a distractor for this specific question.

Questions in multi-speaker discussions often target one specific person deliberately — mixing up whose opinion is whose is one of the most common Section 3 errors.

In Multiple Choice, all 3–4 options are often genuinely mentioned somewhere in the recording — being discussed is not the same as being chosen. Listen for which option the speakers actually settle on by the end.

Selecting an option just because you recognise it from the recording — without tracking whether it was accepted or rejected — is a very common, entirely avoidable error.

Read a Matching task's instructions carefully before you start — some allow a letter/option to be used more than once, or state that not every option will be used at all. Don't assume a one-to-one match by default.

Assuming every option is used exactly once, when the instructions say otherwise, causes you to force a wrong match onto the last remaining blank purely because it's the only one left.

Academic lectures usually preview their own structure at the start — "I'll cover three things: firstly... secondly... finally..." — then follow that exact structure. Use this preview as a map for the whole talk, and listen for "firstly/secondly/finally" (or similar) to know which section you're in.

A long monologue with no second speaker gives you no other way to check your position — the lecturer's own stated structure is the only signpost you get, and missing the preview means listening "blind" for several minutes.

Lectures often introduce a technical term, then immediately explain or rephrase it in plainer words — or vice versa, giving the plain description first and the technical name second. Completion tasks frequently test exactly this pair.

A blank asking for a technical term is often answered a few words AFTER the term first appears, once the speaker restates it — catching the rephrasing is often easier than catching the technical word itself the first time.

Academic speech uses longer sentences with embedded descriptive clauses than everyday conversation. Don't lose track partway through — the core fact you need is often at the END of the sentence, after extra context has been given first.

Tuning out partway through a long sentence means missing exactly the part most likely to be tested — lecturers tend to build up to their main point, not lead with it.

Section 4 Short-Answer Questions often test whether you caught a specific named term, technology, or entity — answer with the EXACT word or phrase used in the lecture, not your own paraphrase of it.

Unlike some Reading questions where a paraphrase is expected, Listening short answers are graded on matching the lecture's own wording — a correct-meaning paraphrase can still be marked wrong here.