Is your IELTS Task 1 answer a data dump, or does it actually compare?
Is your IELTS Task 1 answer a data dump? Two answers to one chart, a sentence-by-sentence check, and three fixes that turn listing into comparing.
Mark your own IELTS essay against the band descriptors one criterion at a time: a four-pass procedure, what to count, and where self-marking fails.
To self-assess IELTS writing against the band descriptors, mark one criterion at a time, collect evidence from the page before you choose a band, and only combine the four marks at the very end. The order matters more than the descriptors themselves: once you have an overall impression of the essay, every later judgement gets pulled towards it.
What follows is the step most explanations of the descriptors leave out — turning the wording into a mark for your own essay — written as a procedure. It works for Task 2 and, with Task Achievement in place of Task Response, for Task 1. Keep the public descriptors for the right task open while you do it; IELTS Writing band descriptors explained sets out the wording band by band if you want it in one place.

Read one criterion's wording at a time, with the others covered.
The usual way people mark their own essay is to read it through, decide it feels like a 6.5 or a 7, and then open the descriptors to check. That order quietly decides the result. With a number already in mind, you read the band 7 paragraph looking for reasons to agree with it, and the descriptors are written in words of degree — "appropriately", "frequent", "some flexibility" — that will stretch to fit almost any impression.
The impression also leaks across criteria: confident vocabulary makes a thin argument look developed, and smooth sentences hide a paragraph that answers a slightly different question. An examiner is trained to keep the four criteria apart; reading your own essay, you have to force that separation with the procedure itself.
So the rules for the whole process are:
Use an essay written under time, and if you can, leave a day between writing and marking. Both make the passes below more honest.
Task Response comes first because it is the criterion where a single gap matters most, and you want to find that gap before anything else in the essay has impressed you.
Take a prompt such as:
In many countries, more people now work from home for part of the week. What are the advantages and disadvantages of this for employees and for employers?
Split it into the nouns the question is actually about. Here there are two pairs: advantages and disadvantages, crossed with employees and employers. That gives four boxes — advantages for employees, disadvantages for employees, advantages for employers, disadvantages for employers. For an opinion question, one of the boxes is always "your position".
Now go through the essay and, for each box, copy the one sentence that answers it most directly. Not the paragraph; the sentence.
If any box is empty, stop. Do not move on to how well the other boxes are handled. An empty box is the band 5 wording: "The main parts of the prompt are incompletely addressed." Nothing else in Task Response can lift the mark above that, and you need to know it before the rest of the essay persuades you otherwise.
If every box has a sentence, look at how much sits behind each one. A box answered by a clause tucked into a paragraph about something else is the band 6 wording: the main parts "are addressed (though some may be more fully covered than others)". For band 7, the main parts "are appropriately addressed" and "Main ideas are extended and supported". The test is simple: under each copied sentence, is there a sentence explaining why it is true, not just an example of it? If the idea is stated and then illustrated but never explained, it has not been extended.
Finally, for opinion questions, write your position in one sentence and check that each body paragraph argues towards it. Band 7 asks for "a clear and developed position". A view stated in the introduction and repeated in the conclusion, with the body arguing around it, has been presented but not developed.
Most self-assessment of cohesion is a count of linking words, and that measures the wrong thing. The second pass removes them instead.
Take a copy of the essay and strike through every connector that opens a sentence: Firstly, Moreover, In addition, However, On the other hand, In conclusion. Then read what is left, paragraph by paragraph.
If the order still makes sense, the progression is real. Each sentence follows from the last because of what it says — it picks up a word, narrows a claim, gives the reason for the one before. If the paragraph falls apart into a list of separate statements that could be shuffled without loss, the connectors were doing all the work. That is the band 6 description: cohesion that "may be faulty or mechanical due to misuse, overuse or omission".
Then look at what survived the strike-through. The band 7 paragraph names the devices it wants: "A range of cohesive devices including reference and substitution is used flexibly". Underline every reference word and phrase — this, these, such a policy, the same problem, doing so. Sentences held together by underlines rather than by struck-out openers are what the band 7 wording describes. The difference is explained in more detail in linking words are keeping you at band 6.

Tick the error-free sentences, then sort the words into two lists.
The last two passes are about language, and both work best as counting exercises, because counting stops you from reading for meaning.
Grammatical Range and Accuracy. Number every sentence. Put a tick next to each one with no error in grammar or punctuation, and a separate mark next to each one built on a complex structure — a relative clause, a conditional, a clause of reason or concession. You now have two numbers: error-free sentences out of all sentences, and error-free sentences out of the complex ones.
The descriptors do not give a percentage, so do not invent one. Band 7 says "error-free sentences are frequent"; band 8 says "The majority of sentences are error-free." What the count gives you is a comparison. If the complex sentences carry most of the errors, you are reading the band 6 line almost word for word: "Examples of more complex structures are not marked by the same level of accuracy as in simple structures." If the essay is accurate but nearly every sentence is simple, the problem is range, not accuracy, and the band 7 wording asks for both.
Lexical Resource. Make two separate lists, and keep them separate.
The first is repetition: every content word you used three or more times where a different word or a reference would have worked. The second is misuse: every word that is slightly wrong — wrong collocation (make a research), wrong form, wrong shade of meaning, or a register that does not fit an essay.
They are separate because they point to different parts of the descriptor. A long repetition list and a short misuse list is a range problem, close to band 6's "rather restricted range". A short repetition list and a long misuse list is the opposite pattern, and the band 6 descriptor has a sentence for it too: "If the writer is a risk-taker, there will be a wider range of vocabulary used but higher degrees of inaccuracy or inappropriacy." Band 7 sits where the resource allows "some flexibility and precision" and "An awareness of style and collocation is evident". Whether you need more words or fewer mistakes depends entirely on which list is longer, which is why one combined count hides the answer.
Now write the four marks in a row, average them, and note the lowest. That lowest mark is the one to work on; which IELTS writing criterion should you fix first? explains why the order of fixes matters.
The procedure removes most of the drift that a single read-through produces. It cannot remove all of it, because the person collecting the evidence is the person who wrote the essay. Two limits survive.
You cannot see your own habitual errors. The ticks in Step 3 are only as accurate as your eye, and an error you have made for years looks correct to you. A grammar pass on your own essay reliably finds slips and reliably misses habits.
You can read your intentions into the page. In Step 1, you will find a sentence for a box partly because you know what you meant it to say. A sentence that nearly answers the question gets counted as answering it. Why this pushes self-assessed bands up, and where it does so most, is covered in why your self-assessed IELTS writing band is higher than your real one.
Both limits have the same fix: a second reading that does not share your memory of the essay.
If you have just run the four passes, paste the same essay below. You get a band for each of the four criteria and the sentences each judgement is tied to, so you can set it directly next to your own four marks and see where they split.
Every mark in this process, yours or an automated one, is an estimate made against the public descriptors. Only a certified examiner produces an official band. What the estimate can do is tell you which criterion to spend your next essay on.
How do I self-assess my IELTS writing against the band descriptors?
Mark one criterion at a time — Task Response, then Coherence and Cohesion, then Grammatical Range and Accuracy, then Lexical Resource — and write down evidence from the essay before choosing a band for each. Only average the four marks at the end. Forming an overall impression first pulls every later judgement towards it.
Which band paragraph should I start reading from?
Start one band below where you expect to land, and move up only when the evidence on the page clears the next paragraph. Starting at the band you hope for and checking downwards tends to push marks up.
Can I turn the descriptors into a checklist and add up the points?
No. Each band is a paragraph, and the examiner chooses the paragraph that best fits the essay as a whole for that criterion. Meeting one clause of the band 7 wording does not make a criterion band 7.
How many error-free sentences do I need for band 7?
The descriptors do not give a number. Band 7 says error-free sentences are "frequent" and band 8 says "the majority" are error-free. Counting your error-free sentences is still useful for comparing essays and for seeing whether your complex sentences carry most of the errors.
Why does my own mark come out higher than an outside one?
Because you cannot see your own habitual errors, and you read what you meant into the page. Marking the essay again a few days later, swapping with a partner, or comparing with an outside reading criterion by criterion are the ways to check for it.
Is a self-assessed band official?
No. It is an estimate made against the public band descriptors, and so is any automated band. Only a certified examiner produces an official IELTS band.
Descriptor wording quoted here is from the public Writing band descriptors published by IELTS and the British Council.
Bandly grades your essay or letter against all four criteria and shows the gap to your target band, sentence by sentence.
Is your IELTS Task 1 answer a data dump? Two answers to one chart, a sentence-by-sentence check, and three fixes that turn listing into comparing.
Fluency reaches two of the four IELTS writing criteria. Here is where native English speakers lose Task Response and Coherence, and what that means for you.