My essay looks band 7 but scored 6: where the gap comes from

When an essay looks band 7 against the public descriptors but comes back as a 6, the usual cause is not a harsh examiner. The descriptors were read for what the essay contains, while an examiner reads them for how well it does it — and the difference between 6 and 7 sits in qualifier words like appropriately, developed, flexibly and frequent that self-assessment slides straight past.

That is why the gap is so often exactly one band. Below: what the public descriptors can and cannot tell you, the four places a self-score inflates, one essay marked both ways, and how to check your own.

The same essay, read for what it contains and for how well it does it.

The public descriptors are a summary, not the marking sheet

The public Writing band descriptors are the four criteria written down: Task Response (Task Achievement in Task 1), Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy, each described at every band from 0 to 9. Marking is a different thing. It is the skill of placing one script against that wording, and it is what examiners are trained and certified to do. The document gives you the criteria; it cannot give you the calibration.

The reason is length. A few sentences per band have to cover every script that deserves that band, so the wording works through qualifiers — and the qualifiers are where the 6/7 boundary lives:

• Task Response. Band 6: the main parts of the prompt "are addressed (though some may be more fully covered than others)". Band 7: they "are appropriately addressed", and "a clear and developed position is presented".
• Coherence and Cohesion. Band 6: cohesive devices "are used to some good effect" but cohesion "may be faulty or mechanical". Band 7: "a range of cohesive devices including reference and substitution is used flexibly".
• Lexical Resource. Band 6: the resource "is generally adequate and appropriate for the task". Band 7: it "is sufficient to allow some flexibility and precision".
• Grammatical Range and Accuracy. Band 6: "a mix of simple and complex sentence forms is used but flexibility is limited". Band 7: "a variety of complex structures is used with some flexibility and accuracy", and "error-free sentences are frequent".

Read the band 7 lines as a checklist and almost any competent essay passes. Did I address the prompt? Use cohesive devices? Complex structures? Less common vocabulary? Yes to all — and each yes answers a band 6 question. The band 7 question is always one of degree, and a checklist cannot measure degree.

A second distortion no document can fix: you know what you meant. Rereading your own paragraph, you supply the skipped step in the argument from memory. An examiner only has the page. And because each criterion is awarded on best fit, an essay with band 7 features in places and band 6 features elsewhere is placed where most of the evidence sits, not where the best sentence sits. The full band-by-band wording is in IELTS Writing band descriptors explained.

The 4 places self-assessment inflates your score

Why does an essay look band 7 but score 6? In almost every case it is one or more of these four, each one a place where doing a thing gets mistaken for doing it well enough.

Task Response: every part mentioned is not every part answered. Two-part prompts show it most. Develop one question over two paragraphs and touch the other in half a sentence, and the main parts are unevenly covered — the band 6 line at best. If one part is not really answered at all, the best fit drops to band 5: "The main parts of the prompt are incompletely addressed." Length and fluent prose do not rescue it, because neither is what Task Response measures. The ladder below band 6 is set out in off-topic IELTS essays.

Coherence and Cohesion: linking words used is not an argument developed. A self-assessor counts connectors — Firstly, Moreover, In conclusion — and finds a well-organised essay. The descriptor asks whether there is "a clear progression throughout the response", with joins made flexibly, including by reference and substitution (this, such a shift, doing so). A connector at the head of every paragraph is what the band 6 line calls mechanical: correct, but labelling rather than linking. Linking words are keeping you at band 6 takes that line apart in detail.

Lexical Resource: difficult words are not precise words. Band 7 asks for "some ability to use less common and/or idiomatic items" and "an awareness of style and collocation". Self-assessment notices the first half — prevalent, plethora, detrimental — and misses the second. The band 6 descriptor has a sentence written for exactly this writer: "If the writer is a risk-taker, there will be a wider range of vocabulary used but higher degrees of inaccuracy or inappropriacy." One slip does not cost a band — band 7 itself allows that "inappropriacies occur". The pattern does: when the ambitious words are the ones that go wrong, the evidence stops saying precision and starts saying risk-taker.

Grammatical Range and Accuracy: complex sentences attempted are not error-free sentences. The band 7 line has two conditions — a variety of complex structures, and error-free sentences that are "frequent". Self-assessors count the first and assume the second. The wording is often misquoted, too: "the majority of sentences are error-free" is band 8, not band 7. If your errors cluster in your longest sentences, the range you are counting is exactly where the accuracy is lost. Complex sentences for band 7 sets out the structures that count, and no grammar mistakes but still band 6.5 covers the opposite case, where the accuracy is there and the range is not.

Same essay, self-scored 7 vs examiner-scored 6

One essay, marked twice, a full band apart.

Here is one Task 2 essay, written for this article, marked twice: once the way its writer marked it against the public descriptors, and once the way an examiner reads the same descriptors. The prompt is a two-part question:

More people now work from home than ever before. Why is this happening? Is it a positive or negative development?

In recent years, working from home has become increasingly prevalent owing to technological advancement. This essay will argue that, although this trend poses a significant impact on society, it is overall a positive development.

Firstly, working remotely offers a plethora of benefits for employees. Workers who no longer commute can save a considerable amount of time, which they can devote to their families or hobbies. Moreover, they are able to maintain a healthier work-life equilibrium, since they can organise their schedules in accordance with their own needs. For example, a parent who works from home can collect their children from school without having to request permission from a manager.

Secondly, companies also benefit from this arrangement. When staff work from home, employers do not need to rent large offices, which can drastically enhance their profits. In addition, firms can recruit talented individuals from other cities, thereby widening the pool of potential candidates. As a result, both small businesses and large corporations are able to operate more efficiently.

On the other hand, some people argue that remote work can lead to isolation. Employees who rarely meet their colleagues may feel lonely, and this can enhance their psychological burden. However, this problem can be solved by holding regular meetings, either online or in person, so that colleagues remain in contact with one another.

In conclusion, working from home has both advantages and disadvantages, but I believe that the advantages outweigh the disadvantages. Therefore, it is a positive development that should be encouraged.

The writer's marking is easy to reconstruct, because every judgement in it is true as far as it goes:

Criterion	What the writer saw	What the descriptor asks	Self-score	Marked
Task Response	A clear position, three body paragraphs, examples	Are both questions appropriately addressed?	7	5
Coherence and Cohesion	Linking words in every paragraph, clear structure	Is the progression clear, and are the joins flexible?	7	6
Lexical Resource	Prevalent, plethora, equilibrium, drastically	Is the less common vocabulary precise and collocated?	7	6
Grammatical Range and Accuracy	Relative clauses, a participle clause, few errors	Varied complex structures, error-free sentences frequent	7	7

Task Response — 5. Is it positive or negative? gets three developed paragraphs. Why is this happening? gets three words — "owing to technological advancement" — and nothing after. That is not a part covered less fully; it is a main part incompletely addressed, the band 5 line. The conclusion then summarises "advantages and disadvantages", the shape of a different prompt.

Coherence and Cohesion — 6. The organisation is genuinely clear, which earns the band 6 line in full. Band 7 it is not: Firstly, Moreover, For example, Secondly, In addition, As a result, On the other hand, However, In conclusion, Therefore — ten connectors in five paragraphs, most opening a sentence. There is some reference (this arrangement, this problem), but the joining is mostly announced rather than built in.

Lexical Resource — 6. The range is real; the precision is not. A trend does not pose an impact (it has one; risks are posed), equilibrium is not the idiomatic partner of work-life, and enhance means improve, so it cannot apply to a burden and sits oddly with profits. The words the writer was proudest of are the ones that went wrong — the band 6 risk-taker sentence almost word for word.

Grammatical Range and Accuracy — 7. This one the writer got right: relative clauses, since and when clauses, a participle clause with thereby, very few errors. Self-assessment is most reliable here, because grammar is what candidates are trained to check.

The arithmetic, with the four criteria weighted equally within the task:

Self-score: (7 + 7 + 7 + 7) / 4 = 7.0

Marked: (5 + 6 + 6 + 7) / 4 = 6.0

A full band, with no badly written sentence in it. The fix is not better English: a body paragraph on causes, a conclusion that answers the prompt set, half the connectors turned into references, three words swapped for plainer, correct ones. The same comparison across two complete answers is in band 6.5 vs band 7 essay.

Stop guessing: score it against the actual criteria

The fix for an inflated self-score is not to mark yourself more harshly across the board. It is to mark against the qualifier, one criterion at a time, and to demand evidence on the page for each one.

If you want to do it by hand, run four passes over one essay written under time.

1. Task Response. Underline every question in the prompt and write beside each the paragraph that answers it. A demand answered only by a clause in the introduction is not appropriately addressed.
2. Coherence and Cohesion. Circle every sentence-opening connector. If more than half your sentences start with one, rewrite two joins using this, such or a restated noun. If the paragraph reads better, the originals were labels.
3. Lexical Resource. List the five words you most want the examiner to notice and check each one's collocation in a learner's dictionary. Any that fail cost more than they earn.
4. Grammatical Range and Accuracy. Check only your three longest sentences for errors. If the errors are there, your range is outrunning your control.

Then have the essay read by someone who does not know what you meant — they will find the skipped step you keep filling in from memory. If the same criterion keeps coming back at 6, it is a habit rather than a bad day, and why am I stuck at band 6.5 in IELTS writing sorts the four usual patterns by how often each is the cap. Or paste your essay into the checker above for a criterion-by-criterion band report.

FAQ

Why does my essay look band 7 but score 6?

Because self-assessment checks whether the essay does each thing the band 7 descriptor mentions, while marking checks how well it does it. The difference between 6 and 7 is carried by qualifiers — appropriately addressed, a developed position, cohesive devices used flexibly, precise vocabulary, frequent error-free sentences — and those are the words a self-score skips.

Are the public band descriptors accurate enough to self-assess with?

They are the real criteria, so they are the right thing to mark against. What they cannot supply is calibration: a few sentences per band have to cover every kind of script at that band, so reading them accurately means asking for evidence of degree rather than presence, one criterion at a time.

Which criterion do people overrate most when marking themselves?

Task Response and Lexical Resource are the usual ones. Task Response because writers read their own intentions into the page and miss an unanswered part of the prompt; Lexical Resource because less common words are easy to notice and collocation errors are not. Grammar is usually the most accurate self-score, since it is the criterion candidates are trained to check.

Does "majority of error-free sentences" mean band 7?

No. In the current public descriptors, "the majority of sentences are error-free" is band 8 wording. Band 7 asks for a variety of complex structures used with some flexibility and accuracy, with error-free sentences that are frequent.

Can one wrong word stop me getting band 7 for vocabulary?

Not on its own — the band 7 descriptor allows that inappropriacies occur. It becomes a problem when the less common words are consistently the ones that go wrong, which is the risk-taker pattern the band 6 descriptor describes.

Are the band estimates on this site official?

No. They are estimates made against the public band descriptors. Only a certified examiner produces an official band.

Related reading

• Band 6.5 vs band 7 essay — the same question answered at both bands, scored side by side
• No grammar mistakes but still band 6.5? — what error-free writing is worth under each criterion
• Why am I stuck at band 6.5 in IELTS writing? — the four patterns that produce a repeated result
• IELTS Writing band descriptors explained — the full wording for all four criteria

All descriptor wording quoted here is from the public Writing band descriptors updated in May 2023, published by IELTS and the British Council.