Band 7.5 in IELTS Writing Task 2: The Two Criterion Totals That Reach It

There is no band 7.5 in the IELTS public band descriptors. The descriptors are published as whole bands, 0 to 9, and an examiner marks each of the four criteria on that whole band scale. A 7.5 is produced afterwards, by arithmetic, and it only happens at two specific criterion totals.

Knowing which two totals changes what you practise. Most pages on this query show you an essay with 7.5 written at the top. This page shows you the sum that made it a 7.5, and the one criterion that moved it there in a set of answers you can open and read.

How a 7.5 is actually produced

Your Task 2 answer gets four whole band scores, one for each criterion:

• Task Response, whether you answered the whole question with developed ideas
• Coherence and Cohesion, whether the reader can follow the order and the links
• Lexical Resource, your vocabulary range and accuracy
• Grammatical Range and Accuracy, your sentence range and grammar accuracy

Those four are added and divided by four. The result is rounded to the nearest half band, and a result landing exactly on .25 or .75 rounds up, which is the published IELTS convention. Because four whole numbers are being averaged, the total of the four is all that matters. There are only 37 possible totals, and each one maps to exactly one reported band.

So a 7.5 in Task 2 means your four criterion scores added up to 29 or 30. Nothing else reaches it. If you want to see the same arithmetic applied to the whole Writing section, where Task 2 carries twice the weight of Task 1, we set that out in how the Writing band is calculated.

Two things follow from the table, and both are useful.

First, 28 and 29 are one point apart and half a band apart. One criterion moving from 6 to 7 is the entire difference between a 7.0 and a 7.5.

Second, a 27 is also reported as 7.0. If your total is 27, raising one criterion by one band gets you to 28, which is still 7.0. You would need two criterion points, not one. This is why two candidates can both be told 7.0, both improve one criterion by a full band, and only one of them sees the reported band move.

What the same answers show about which criterion moves

We can put numbers on this rather than guess. Bandly's sample library publishes full answers with a band estimate for each of the four criteria. Thirteen of the Task 2 answers in it were estimated at 7.0 or 7.5. These are machine estimates from Bandly's own engine, not marks from a qualified examiner, so read them as a pattern in one consistent set of judgements rather than as official scores.

Counting the criterion scores across those thirteen:

Lexical Resource is the only criterion that separates the two groups cleanly. Grammatical Range and Accuracy is 7 in all thirteen answers, so it explains none of the difference. Task Response is 8 in twelve of the thirteen. Coherence and Cohesion splits the 7.0 group in half and so predicts nothing on its own.

That is one library and one engine, and it is not a rule of the exam. But it does tell you where to look first if you are sitting on a 7.0, and it contradicts the usual advice on this query, which is to work on structure and linking. In these answers, structure was already at the level the band needed.

What band 7 Lexical Resource asks for, in the published words

The public descriptors for Lexical Resource read as follows. At band 6:

uses an adequate range of vocabulary for the task; attempts to use less common vocabulary but with some inaccuracy; makes some errors in spelling and/or word formation, but they do not impede communication

At band 7:

uses a sufficient range of vocabulary to allow some flexibility and precision; uses less common lexical items with some awareness of style and collocation; may produce occasional errors in word choice, spelling and/or word formation

Read the two together and notice what does not change. Both bands allow errors. Band 6 says errors do not impede communication. Band 7 says occasional errors may still be produced. Nobody is asking you to write a Task 2 answer with no spelling mistakes.

What changes is the positive side. Band 6 attempts less common vocabulary. Band 7 uses less common items with some awareness of style and collocation, and has enough range for precision. The move from 6 to 7 is a move from adequate to precise, not from inaccurate to accurate. Our longer breakdown of this criterion is in what Lexical Resource measures.

The same point on real sentences

Here is the evidence from one of the answers estimated at 7.0, taken from the working from home essay in the library. These were the vocabulary choices marked as the strong ones:

• balance their professional responsibilities with personal commitments
• potential for social isolation
• boundary between work and personal life often becomes blurred

Those are correct, natural and topic appropriate. They are also what any competent writer on this subject would produce. There is no precision in them that a simpler phrasing would lose.

Now the same list from the traffic congestion essay, estimated at 7.5:

• traffic congestion has become a pressing issue that worsens annually
• disproportionately affects lower-income citizens
• decentralize business districts
• financial disincentives with high-quality public transit and smart urban planning

The difference is not error count. The 7.5 answer contains benefitial for beneficial, This policy would encourages people, and If there is no alternative ways to travel. The 7.0 answer contains significent for significant and many people finds themselves. Both have slips of the same kind.

The difference is that disproportionately and disincentives do work that no ordinary word would do. They carry a precise claim about who pays and how behaviour changes. That is the flexibility and precision the band 7 wording asks for.

A rewrite, with the descriptor each change serves

Take this sentence from the food tax essay, estimated at 7.0, with Lexical Resource at 6:

many researches show that sugar addiction is very strong, and the tax has a great influence in poor people's budgets

Three faults, each a different bullet of the descriptor:

1. many researches is a word formation error. Research is uncountable in this sense. This is the "errors in spelling and/or word formation" bullet.
2. has a great influence in is the wrong preposition and a weak collocation. This is the "awareness of style and collocation" bullet.
3. very strong and great are basic adjectives doing the work of precise ones. This is the "flexibility and precision" bullet.

Rewritten:

Several studies indicate that sugar dependency is powerfully habit forming, and the tax weighs disproportionately on low income households

Nothing rare was added. Indicate, dependency, habit forming, weighs on and disproportionately are all ordinary words. What changed is that each one was chosen for the job rather than reached for first. That is the whole move from 6 to 7 on this criterion.

How to practise this in the time you have

1. Write one Task 2 answer at full length, 250 words or more in 40 minutes. A vocabulary exercise done outside exam conditions tells you nothing about what you reach for under time.
2. Mark your own adjectives and verbs. Circle every very, big, good, bad, important and a lot of. Each one is a place where the precise word was available and you did not use it.
3. Replace each with the word that carries the actual claim, not with a synonym from a list. Significant instead of big is not a gain if the real claim was doubled.
4. Check your word forms separately, at the end. Research, advice, information and equipment are uncountable. Adverbs need the ly.
5. Compare a full answer against one at the band above. Read a band 7 Task 2 answer with the band explained next to one of the band 7.5 answers and look only at the adjectives and verbs.

If you want the four criterion estimates on your own answer rather than on somebody else's, paste it into Bandly and you will get a band estimate per criterion with the sentences each judgement was based on, which is the only way to tell whether your 7.0 is a 27 or a 28.