The essay is 268 words long, printed in a slightly cramped hand, and it argues that governments should spend more on public transport than on new roads. The examiner has thirty of these to clear before lunch, and this one arrived somewhere in the middle of the stack. There is nothing remarkable about it yet, which is exactly the point: most essays are ordinary, and the marking has to be fair to ordinary work.
What follows is not a mysterious act of intuition. It is a trained reader holding four descriptors in mind at once, letting the writing tell them where it sits. The interesting part is how quickly some of that happens, and how much of the final number is settled before the reader reaches the last line.
Meet the essay landing on the examiner’s desk
The prompt asked whether the advantages of living in a big city outweigh the disadvantages. Our candidate has agreed that they do, mostly, and has written four paragraphs to say so. The word count is safely over 250. The paragraphing is visible at a glance, which already matters more than the candidate probably realises: an examiner who can see the shape of an argument before reading a word of it is starting from a calmer place.
There are two examples in the margins of the argument, one about job opportunities and one about hospitals. Neither is developed at length, but both are relevant. This is a competent, unglamorous essay, and competent essays are where the band descriptors do their most careful work.
Watch the first thirty seconds shape a first impression
Before the pen touches the page, the examiner reads the introduction and the topic sentences of each paragraph. This is not skimming to save time; it is building a hypothesis. Does the essay answer the actual question? Is there a clear position? Can the writing carry meaning without forcing a re-read?
In these opening seconds the candidate does two things right and one thing that will cost them. The position is clear, and the essay is plainly on-topic. But the introduction leans on a memorised opening — “In this modern era, this is a highly controversial issue” — that says nothing and flags a rehearsed approach. The examiner notes it, provisionally lowers the ceiling on Lexical Resource, and reads on. Candidates who rehearse their instincts through something like a structured pre-test, such as the practice materials at https://careerwiseenglish.com.au/, tend to strip out these empty openers before they become habits.
Follow the pen as it flags a weak second body paragraph
The first body paragraph holds together. The topic sentence promises an argument about employment, the example about job markets supports it, and the paragraph ends where it should. The pen makes a small approving tick and moves on.
The second body paragraph is where the essay wobbles. It opens by promising to discuss the cost of living, then drifts into a sentence about pollution, then returns to housing prices without ever finishing the pollution thought. The examiner underlines the seam where the topic changes. This is not a grammar problem or a vocabulary problem; it is a coherence problem, and it pulls at that descriptor specifically. One idea per paragraph, fully carried through, is worth more than three ideas half-stated. The pen also catches two article errors here, but those are minor beside the structural drift.
See how the final band gets decided when two paragraphs disagree
Now the examiner has a mixed picture. One body paragraph is band 7 in its organisation; the other sits closer to band 6 because the reader had to work to follow it. Task Response is solid — the question was answered, the position held — but the second paragraph’s development was thin.
The descriptors are not averaged like exam marks. Each of the four is scored on where the essay predominantly sits, and “predominantly” is doing quiet work. Two strong paragraphs and one loose one usually land the Coherence score at the lower band, because the weaker stretch is not a stray slip — it is a quarter of the essay. Grammar and vocabulary, meanwhile, are judged across the whole thing, so the rehearsed opener and the scattered article errors keep Lexical Resource and accuracy from climbing.
The four numbers go down, and this particular essay settles at a band 6.5 overall. The candidate will see one figure on a report. What they will not see is that the score was mostly decided by a single paragraph that promised one idea and delivered two.