Band 7 in IELTS Writing Task 1: A Sample Answer and the Two Edges of the Band

A band 7 in IELTS Writing Task 1 is not a single score. It is an average of four separate marks, and seven different combinations of those four marks round to exactly 7.0. Two answers can both be band 7 and need completely different work.

That is the problem with every model answer that has band 7 printed beside it. You cannot see which of the four marks was the weak one, so the page cannot tell you what to change in your own writing.

This page gives the seven combinations first, then a full answer, then each criterion marked against the published descriptor wording, then the same answer walked off both edges of the band: down to 6.5, and up to 7.5.

The seven ways to score 7.0 in Task 1

Task 1 is marked on four criteria of equal weight. Task Achievement is the Task 1 name for the content criterion, and it asks whether you covered the key features, gave an overview, and reported the figures accurately. The other three are Coherence and Cohesion, which is organisation and linking; Lexical Resource, which is vocabulary; and Grammatical Range and Accuracy. Each gets a whole number. The four are averaged and the result is rounded to the nearest half band. How the writing band is calculated sets out the rounding rule in full.

The four whole numbers add to 27 or 28 if the average is going to round to 7.0. If no criterion is below 6, that leaves exactly seven combinations:

Two things in that table matter more than the rest.

First, the top row is a band 7 answer with three marks at 6. The content is near the top of the scale and the language is ordinary. Second, the bottom row is even across all four and totals 28, while the third row is almost identical and totals 27. Those two rows are both band 7 and they are not in the same position, which is the subject of the last two sections of this page.

What the marked Task 1 answers in the library actually scored

Bandly's published sample library carries the four criterion estimates on every answer, so the profiles can be counted rather than guessed. There are eight Task 1 answers in it, four Academic and four General Training. I read the four criterion bands off each record and kept the Task 1 ones.

Six of the eight reached 7.0 or 7.5. Three facts come out of those six that I have not seen on any page ranking for this query.

Grammatical Range and Accuracy was exactly 7 on all six. Not 6, not 8. On the two answers that fell below 7.0, grammar was 6 and 5. On this small set, a grammar estimate of 7 is what every band 7 answer had in common.

Lexical Resource was never estimated above 7 on any of the eight. It was 7 on two of the six, 6 on four of them, and 5 on both answers that missed band 7.

Vocabulary was the lowest or joint lowest of the four marks on every one of the eight. There is no answer in this set where vocabulary was not at the bottom.

Two cautions, and they are important. These are estimates from Bandly's own scoring engine, which has never been checked against a qualified human examiner. And Task Achievement never came back below 8 across all eight answers, which tells you something about the estimator as much as about the answers. The useful part is not the exact numbers. It is the shape: the content mark sat high, and vocabulary sat at the bottom, on every answer in the set. Lexical Resource is the criterion that pattern points at.

The question

A practice Academic Task 1 question. The figures are invented for this exercise, so please do not quote them as real data.

The line graph below shows the percentage of households with access to the internet in three countries between 2000 and 2020. Summarise the information by selecting and reporting the main features, and make comparisons where relevant. Write at least 150 words.

A band 7 answer, 177 words

The line graph shows the percentage of households with access to internet in three countries between 2000 and 2020. Overall, home internet access grew in all three countries across the two decades. Country X led throughout the period, but Country Y closed almost all of the gap by 2020, while Country Z remained far behind both of them. In 2000 a home connection was rare everywhere. Country X stood at 12 per cent, Country Y at 5 per cent and Country Z at only 2 per cent. Country X then rose quickly, reaching 65 per cent by 2010 and 91 per cent by 2020. Country Y began more slowly, with just 14 per cent in 2005. After 2010, however, its growth was much steeper than Country X. It climbed from 33 per cent to 88 per cent in the final decade, which brought it within three percentage points of the leader. Country Z also increased steadily, but from a very low base. Even in 2020 fewer than half of its households were connected, at 48 per cent.

The four marks, with the published wording behind each

Total 28, which averages to 7.0.

These four marks are mine, written against the published band descriptors so the arithmetic is visible on the page. The answer was not run through Bandly and no examiner has seen it. The wording in the right hand column is public and quoted; the number beside it is a reading of that wording, which is what any marking is until an examiner does it.

Task Achievement is a 7 and not an 8. The overview in paragraph two is clear and does real work: all three rising, X in front, Y catching up, Z far behind. Every figure is accurate and every country is covered. What holds it at 7 is selection. The answer reports 2000, 2010 and 2020 well but passes over 2015, which is where Country Y's steep phase actually begins, and it never says that Country Z is the only country still under half. An 8 selects the features that carry the most information. The Task 1 overview is where this answer earns most of its content mark.

Coherence is a clean 7. One paragraph for the general picture, one for the starting point, then one per country where the story differs. The links are relationships rather than labels: Overall, then, After 2010, however, also. Nothing is signposted mechanically.

Vocabulary is a 7 at the bottom of the band. Rose, climbed, grew, increased steadily, from a very low base and within three percentage points are precise and varied enough for the band. The occasional slips the band 7 wording allows are here too: Country Z also increased steadily says the country increased when the percentage did, and access to internet is missing an article. Both are small, and both are exactly the kind of error the band 7 bullet expects.

Grammar is a 7. There is a relative clause, a passive, a fronted adverbial with however, and several error free sentences in a row. One comparison is incomplete: much steeper than Country X compares a growth rate with a country, and needs than that of Country X. A few errors of that size is the band 7 line.

The lower edge: two marks, not one

Here is the part the labelled model answers hide. This answer totals 28. Drop any single criterion by one point and the total is 27, which still rounds to 7.0. The band does not move.

Drop two of them and the total is 26, which rounds to 6.5.

So an even 7-7-7-7 profile has one point of slack in it. That is not true everywhere inside the band. Look again at the third row of the first table, 7-7-7-6, which totals 27. That profile has no slack at all. One more point lost anywhere and it is 26 and 6.5.

Two candidates can both be band 7 and be in very different danger. The one at 28 can have a bad day in one criterion and keep the band. The one at 27 cannot.

Concretely, on the answer above, what would take it to 26? Vocabulary falling to 6 would need the precise verbs to go: rose, climbed and within three percentage points replaced by went up three times. Grammar falling to 6 would need the errors to become frequent rather than occasional, for instance tense slips across the body paragraphs. Either one alone keeps you at 7.0. Both together is 6.5.

The upper edge: one mark is enough

The arithmetic is not symmetrical, and this is the useful half of it. From 28, losing the band takes two points. Gaining the next half band takes one.

Raise any single criterion from 7 to 8 and the total is 29, which averages to 7.25 and rounds up to 7.5. One mark. That is why the question which of my four is easiest to move is worth more than general practice.

On this answer the cheapest point is Task Achievement, because the gap is selection rather than language. Add one sentence to paragraph two naming 2015 as the turning point for Country Y, and one clause observing that Country Z alone stayed below 50 per cent, and the overview is now reporting the features that carry the most information. The vocabulary and the grammar do not change at all.

That is a 7.5. The four library answers that reached 7.5 all got there the same way: Task Achievement at 9 with the language marks at 6 and 7. None of them reached 7.5 by having strong vocabulary. What band 9 in Task 1 actually requires is the other end of the same arithmetic, and a full Task 1 answer with every criterion explained walks a different chart type through the same four marks. The band 6 version of this page does the exercise one band lower.

What to do today

Take a Task 1 answer you have already written and give it four separate marks instead of one. Add them. If the total is 27 or 28, you are band 7, and the next question is which single mark is closest to moving up.

Then do the subtraction too. If your total is 27, you have no slack, and the work is defending the three marks you have rather than chasing a fourth.

If you do not know how your four marks split, paste the answer into Bandly and compare its four estimates against yours. The bands are estimates and no examiner has checked them, but the lowest of the four is the question worth asking, and the sentences the feedback highlights are text you can read and judge yourself.