Decoding the Band Descriptors
The 4 marking criteria: what every band level actually looks like.
Most students plateau at Band 6 not because their English is weak, but because they don't know what examiners are marking. IELTS Writing has 4 separate criteria, each worth 25%. Ignore one, and your score caps. This lesson shows you what every band level actually looks like, so you know what target you're shooting at.
How Writing is scored
Your essay receives 4 separate scores, then averaged. The 4 criteria are equally weighted:
- Task Response (TR). Did you fully answer the question?
- Coherence and Cohesion (CC). Is your essay organised? Does it flow?
- Lexical Resource (LR). Is your vocabulary varied, accurate, and natural?
- Grammatical Range and Accuracy (GRA). Do you use complex structures correctly?
Your final Writing band is the average of the four. A student who scores 8 / 8 / 8 / 5 averages 7.25 → Band 7.5. One weak criterion drags everything down. To hit Band 7+, all four must be at 7 or above.
Task 1 and Task 2 do NOT count equally
The Writing test has two tasks, but they are not weighted the same in your final band.
- Task 1 (chart / graph / map / process). Weighted 1/3 of your Writing score.
- Task 2 (essay). Weighted 2/3 of your Writing score.
In plain numbers: a student who scores T1 = 6.0 and T2 = 7.0 ends up with Writing = (6.0 + 2 × 7.0) / 3 = 6.7 → reported as Band 6.5. Move T2 from 7.0 to 7.5 and the band jumps to 7.0. Move T1 by the same amount and it barely budges. Every minute spent improving T2 is worth roughly twice every minute spent improving T1.
The band levels at a glance
Here is what each band level means across the four criteria. Memorise the row for your current band AND the row for your target. Those are the two reference points you need cold.
TR: Partially addresses the prompt. Position unclear or shifting.
CC: Some paragraphing. Linkers repeat or misused.
LR: Limited vocabulary. Frequent repetition + word-choice errors.
GRA: Mostly simple sentences. Frequent grammar errors.
TR: Addresses prompt but underdeveloped. Some irrelevant detail.
CC: Information arranged coherently. Linkers used but not always precise.
LR: Adequate vocabulary. Some inaccuracies and repetition.
GRA: Mix of simple and complex. Some grammar errors persist.
TR: Addresses ALL parts of the prompt. Clear position throughout.
CC: Logical paragraphing. Range of cohesive devices used precisely.
LR: Sufficient range. Less common items used. Occasional errors.
GRA: Range of complex structures. Frequent error-free sentences.
TR: Sufficiently addresses all parts. Position fully developed.
CC: Logical sequencing throughout. Skilful paragraphing.
LR: Wide range used fluently. Skilful with less common items. Rare errors.
GRA: Wide range used flexibly. Majority of sentences error-free.
Task Response, answering the actual question
TR measures whether you addressed the prompt. Not whether you wrote well. Whether you wrote about the right thing. It is the criterion that catches out the most students.
Prompt: Some people think children should learn academic subjects, while others think practical skills are more useful. Discuss both views and give your own opinion.
Student writes: "Education is very important. Children need to learn many things. Academic subjects help them get good jobs. They learn maths and science. They can become doctors, engineers, and scientists. This is good for the country..."
Discusses only one view. Doesn't engage with practical skills. No clear opinion. Caps at Band 5-6 for TR even if grammar is perfect.
Same prompt.
Student writes: "Both views feature prominently in modern education debates. Those who favour academic subjects argue that they develop critical thinking and lay a foundation for further study, while supporters of practical skills emphasise their direct usefulness in adult life. While I see merit in both, I believe a balance, with practical skills given more weight in later years, best prepares students for the world they will enter..."
Addresses both views in equal depth. Takes a clear position. Stays on the prompt throughout. Band 8.
Coherence and Cohesion, organising your essay
CC measures structure, paragraphing, and flow between sentences and paragraphs. Two parts:
- Coherence. Your ideas progress logically.
- Cohesion. Your sentences connect with the right linking words and references.
"Pollution is bad. Cars cause pollution. People drive a lot. Governments should make rules. Public transport is also important. Many cities have buses. People can take the bus instead of cars..."
One sentence per idea. No clear paragraphing logic. No linkers. Choppy, mechanical. Band 5-6.
"Air pollution in urban areas stems primarily from private vehicle use. Consequently, governments need to implement policies that discourage car use, such as congestion charges or low-emission zones. At the same time, investment in public transport remains essential, since people are unlikely to abandon driving without a reliable alternative..."
Sentences connect with precise linkers. Each idea develops the previous one. The paragraph has one controlling idea. Band 8.
Lexical Resource, your vocabulary
LR measures three things: range (variety of words you use), accuracy (correct word choice and collocation), and naturalness (do native speakers actually phrase it like that?). It is where most B2 students lose marks.
"There are many problems with pollution. It is a big problem. We need to do something about this problem. Pollution is very bad for our health. Cars make a lot of pollution. We should reduce pollution. Pollution affects everyone..."
"problem" 3x. "pollution" 5x. "very bad" / "a lot", vague. No precise word choice. Band 6.
"Air pollution poses serious threats to public health. The issue stems primarily from vehicle emissions, which release toxic particulates into the atmosphere. Long-term exposure to these contaminants has been linked to respiratory illness and cardiovascular disease..."
Precise terminology (particulates, contaminants, respiratory). Each repeat is a deliberate paraphrase. Band 8.
Grammatical Range and Accuracy, your sentences
GRA measures sentence variety and grammatical correctness. Two parts:
- Range. Do you use complex structures? Conditionals, relative clauses, subordination, passive voice, participle clauses?
- Accuracy. Are they correct? How often do you make errors per 100 words?
"Cars cause pollution. Pollution is bad. People should drive less. Governments can help. Many countries have done this. The result was good. We need more of this."
All simple sentences. No subordination. No relative clauses. Even with zero errors, this caps at Band 6 for GRA.
"While vehicle emissions remain a primary contributor to urban air pollution, the problem cannot be solved through driver behaviour alone. If governments invested heavily in cycling infrastructure, which has proven successful in cities such as Copenhagen, car dependency could decline significantly."
Subordinate clause, conditional, relative clause. Three complex structures in two sentences. Errors are rare. Band 8.
The four traps that drop your score
- Off-topic introduction. You paraphrase the prompt incorrectly and the rest of the essay drifts. → TR drops.
- No paragraphs. One giant block of text. → CC drops to 4-5.
- Memorised phrases. Examiners are trained to spot them. → LR penalised.
- Only simple sentences. Even if every one is perfect. → GRA caps at 6.
Exam mechanics you must know
A handful of mechanical rules govern the test itself. These are not band-level skills; they are pass/fail facts about how the exam runs, and most students don't know them.
- 250-word minimum on Task 2. Going under is mechanically penalised. Your TR drops. Aim for 260-280 to leave a safety margin.
- 150-word minimum on Task 1. Same rule. Under and you lose marks on TR.
- 60 minutes total, both tasks combined. Task 1 budgets 20 minutes; Task 2 budgets 40. There is no separate timer. You manage the split yourself.
- Task 2 is weighted DOUBLE Task 1. Your final Writing band is roughly (Task 1 × 1 + Task 2 × 2) ÷ 3. So 30 minutes invested in Task 2 quality earns roughly twice the marks of the same effort on Task 1.
- Every essay is double-marked. Two examiners score independently using the published rubric; if they disagree by more than half a band, a third reviewer settles it. There is no "lucky examiner".
- The rubric is public. The official Band Descriptors PDF is on the IELTS website. Read it once. Your marker reads it before every essay, and so should you.
Find your weakest criterion
Before you build any other skill, know which of the four criteria is dragging your score down. Here is a fast self-diagnostic. Take any essay you have written in the last month, graded or ungraded, and answer these four questions honestly:
- TR check. Did I fully answer every question mark in the prompt, including any sub-questions? → If no, TR is weakest.
- CC check. Does each paragraph have one clear controlling idea? Do my linkers vary, or do I repeat "also" / "moreover" three times? → If no / yes-repeats, CC is weakest.
- LR check. Did I repeat any content word more than twice in one paragraph? Did I use any precise collocations (not just "good", "important", "many")? → If yes-repeats / no-collocations, LR is weakest.
- GRA check. Did I write any complex sentences with subordinate clauses, conditionals, or relative clauses? Did I make any grammar error I can name now? → If no / yes-errors, GRA is weakest.
Summary
- 4 criteria, each 25%: TR, CC, LR, GRA.
- Your final score is the average. One weak criterion drags everything.
- Task 2 is weighted 2/3, Task 1 is weighted 1/3. Every minute on T2 is worth roughly twice every minute on T1.
- Most students lose marks on TR (didn't fully answer) and LR (vocabulary repetition), not grammar.
- Paragraph properly. Take grammar risks. Use precise words. Answer the actual question.
- Task 2 weights double Task 1. Always give it the time it needs.
In the next lesson we cover the 5 prompt types, recognising exactly what each question demands. Mis-reading the prompt is the #1 source of TR failure.