Note-Taking That Survives a Single Listen
You are given paper and a pen, nobody ever reads what you write on it, and most candidates still use it wrongly. Good notes are short, ugly and fast.
Your notes are a tool, not a submission
The test centre gives you paper and something to write with. What goes on that paper is never collected, never marked, and never seen by anyone. This sounds obvious and yet it is the fact most candidates fail to act on, because habits formed in classrooms produce notes written to be readable by a teacher.
Notes that would embarrass you are usually the right notes. Arrows, initials, single digits, a circle around something that changed. If a symbol gets information from the clip to the question, it is correct. If it is neat, it probably cost too much.
This also means your first language is allowed on that paper, and for many candidates it is faster. Writing a concept in the language you think in takes less working memory than translating it into English while the next sentence is playing. Nobody is assessing the paper, so use whatever gets the meaning down quickest.
Write anchors, not sentences
An anchor is a single item that lets you reconstruct a stretch of audio later. Six or seven anchors per clip is a realistic target. Thirty lines of half-finished sentences is not notes, it is a failed transcription attempt with a comprehension loss attached.
Four kinds of thing are worth anchoring. Names and roles, because questions routinely ask which speaker did or wanted something. Numbers, because they leave memory first and arrive too fast to reconstruct. Decisions, because most conversations in this section exist to reach one. Reversals, because a plan that changes is where the test hides its distractors.
Everything else is padding. Descriptions, pleasantries, a speaker agreeing with another speaker, the second and third example of a point already made: none of it produces questions and all of it consumes the hand and the attention at the same time.
There is a quick test for whether something is an anchor. Ask whether a question could be built on it. A figure, a name, a decision and a change of plan all support questions. The adjective a speaker chose about the weather does not, and neither does the fact that one of them agreed politely.
A shorthand you build before test day
Shorthand invented during the test is not shorthand, it is one more thing to think about. Decide on your symbols in advance and use the same ones in every practice session until they are automatic.
A workable starter set: an arrow for 'leads to' or 'then', a crossed-out arrow for a plan that was abandoned, a question mark for something you suspect but did not catch cleanly, a dollar sign for any amount, a clock shape for any time, and a plain dash for a gap you know exists. Initials for speakers, always the same position on the page.
The test of a symbol is whether you can write it without deciding to. If you have to think about which mark means a reversal, the mark is costing more than it saves. Cut your set down until every symbol is reflexive, even if that leaves you with only five.
Keep each symbol to a single stroke wherever you can. A tick, a cross, a slash and an arrow are one movement of the hand each, whereas a three-letter abbreviation is three decisions and three movements. Across a clip with ten anchors, that difference is measured in seconds of lost audio.
Write numbers as digits, always, and write them the instant you hear them, before you have worked out what they refer to. You can attach a label to a number afterwards. You cannot recover a number you did not write.
Notes do different jobs in Parts 1 to 3 and Parts 4 to 6
In Parts 1 to 3 the questions are spoken to you one at a time, with about thirty seconds each, and they follow shortly after the audio they relate to. You are answering from memory that is still warm. Notes here are a safety net for the things memory is bad at, which is essentially numbers, names and the order of events.
In Parts 4 to 6 the whole set of questions appears on screen after the recording has finished, sharing one time limit. You are now answering minutes after hearing the content, with five or six questions competing for the same recall. Here the notes are not a safety net, they are the primary source.
That difference should change how much you write. It is reasonable to write noticeably more in Part 4, Part 5 and Part 6 than in Part 1, because in those parts the paper is doing work that your short-term memory simply will not do for you. Even then, more means eight or ten anchors rather than twenty lines of prose.
Rule your page into three columns before Part 5 begins, during the introductory statement. Doing it while the audio runs costs you the opening exchange, which is precisely where the three speakers state their initial positions.
The three-column layout for Part 5
Part 5 has three speakers, and you are told so before the clip begins. On the real test this part is presented as a video and the speakers are introduced in a fixed arrangement on screen, left, centre and right, which gives you a labelling system for making no decisions about.
Draw two vertical lines before the clip starts, making three columns. As each speaker is introduced, put their name or initial at the top of the column matching their position on screen. From then on, everything a speaker says goes in that speaker's column, and you never have to record who said it.
This matters because Part 5 questions are frequently about attribution and agreement. Which speaker raised the cost objection, which two of the three ended up on the same side, what the group collectively concluded. With three columns those answers are visible at a glance. With one running list of remarks, you have to remember attribution as well as content, and attribution is exactly what memory loses first when three voices are involved.
Common note-taking failures and what they look like
Transcription. The page fills with real sentences and the candidate reports that the audio went by too fast. It did, because half the listening was being spent on handwriting. If your notes contain a complete grammatical sentence, that sentence cost you the two sentences spoken while you wrote it.
Neatness. Going back to tidy a word, straighten a line or complete an abbreviation. The clip does not pause for cosmetics, and nothing on the page is ever read by anyone else.
Waiting for something important. Some candidates hold the pen still, listening for the part that matters, and then write nothing because it never announced itself. Detail arrives without warning. The fix is to commit to the anchor categories in advance so that hearing a number triggers the hand without a decision in between.
Notes with no structure. A single column of remarks in the order they occurred, with no speaker attribution and no separation between the clip's three sections in Part 1. When the question asks who suggested something, an undifferentiated list cannot answer it.
Use a dash for a gap and a question mark for a guess, and keep the two distinct. A dash means you know something was said and you did not get it; a question mark means you got something and are not sure it is right. They lead to different guessing strategies at the question.
Practise it
Five questions on what belongs on the paper and what does not.