What actually changes between band 6, 6.5 and 7 in IELTS Writing Task 2?
Across 183 IELTS Writing Task 2 essays analysed by IELTS Writing Checker, band-6.5 essays score higher than band-6 essays mainly on Task Response (+0.67 points) and Coherence and Cohesion (+0.58 points), whereas band-7 essays score higher than band-6.5 essays mostly on Lexical Resource (+0.79 points) and Grammatical Range and Accuracy (+0.71 points). In practical terms, the qualities that separate a band-6.5 essay from a band-6 essay are different from the qualities that separate a band-7 essay from a 6.5.
Imagine a learner who starts at band 6, works hard on essay structure, and reaches 6.5. Confident that the method is working, they keep polishing paragraph plans and memorising transition phrases, only to find their score staying at 6.5. Our data helps explain why that happens: the main differences between band 6 and 6.5 lie in structure and argument development, whereas the gap between 6.5 and 7 lies mostly in language accuracy and word choice.
To see how these differences play out, we analysed the essays' estimated scores across the four IELTS Writing criteria: Task Response (TR), Coherence and Cohesion (CC), Lexical Resource (LR), and Grammatical Range and Accuracy (GRA). Comparing the average criterion scores across the band-6, band-6.5, and band-7 groups shows two distinct patterns.
| Band group | Essays | Task Response (TR) | Coherence & Cohesion (CC) | Lexical Resource (LR) | Grammar (GRA) |
|---|---|---|---|---|---|
| 6 | 52 | 6.10 | 5.98 | 5.90 | 5.62 |
| 6.5 | 35 | 6.77 | 6.56 | 6.14 | 6.07 |
| 7 | 29 | 6.93 | 7.03 | 6.93 | 6.78 |
Between the band-6 and band-6.5 groups, the Task Response average is higher by 0.67 points (6.10 vs 6.77) and Coherence and Cohesion is higher by 0.58 points (5.98 vs 6.56). Over that same step, Lexical Resource is only 0.24 points higher (5.90 vs 6.14), and Grammar is 0.45 points higher (5.62 vs 6.07). Band-6.5 essays stand out from band-6 essays primarily by answering the question more fully and organising paragraphs more logically.
Between the band-6.5 and band-7 groups, the pattern shifts. Task Response barely moves, differing by just 0.16 points between band-6.5 and band-7 essays (6.77 vs 6.93), while Coherence and Cohesion is 0.47 points higher (6.56 vs 7.03). Meanwhile, Lexical Resource is 0.79 points higher (6.14 vs 6.93) and Grammatical Range and Accuracy is 0.71 points higher (6.07 vs 6.78).
Why do so many learners get stuck at band 6.5?
Grammatical Range and Accuracy is the lowest of the four criteria in roughly half of all Task 2 essays at every band from 5.5 and below right through to band 7: 54% of essays at 5.5 and below, 53% at band 6, 47% at band 6.5, and 53% at band 7. Only in essays scored at 7.5 and above does the weakest criterion spread out across Grammar (31%), Lexical Resource (30%), Task Response (25%), and Coherence and Cohesion (15%).
At band 6.5, when Grammar is not the lowest score, Vocabulary usually is. Lexical Resource is the weakest criterion in 36% of band-6.5 essays, while Task Response is the lowest in only 8% and Coherence and Cohesion is the lowest in only 9%. Together, Grammar and Vocabulary represent the lowest criterion in more than four out of five essays at the 6.5 level.
This explains why the 6.5 plateau feels so stubborn. In the band-6.5 group, the Task Response average is already 6.77 and the Coherence and Cohesion average is already 6.56. In plain words, 6.5 essays usually already answer the question and are well organised. Because Task Response barely moves after 6.5 while Grammar and Vocabulary sit near 6.0, the data suggests that the remaining gap between 6.5 and 7 is mainly language accuracy and precise word choice.
How do you move from band 6 to 6.5 in Task 2?
If your estimated score is currently sitting at band 6, the biggest score differences between band 6 and band 6.5 are in how arguments are built and connected, alongside steady attention to basic grammar (which is still the lowest criterion in 53% of band-6 essays). Across the 52 band-6 essays in our dataset, Task Response averages 6.10 and Coherence and Cohesion averages 5.98.
A common misconception at band 6 is that Task Response is about making your opinion louder. In fact, in the feedback on the 52 band-6 essays, the writer's position was described as unclear in only 1 essay, while the advice for Task Response mostly asked for deeper argument development and more varied, directly relevant examples. Strengthening Task Response means staying with each main point long enough to explain why it is true and how the cause leads to the effect, then backing it up with a specific example that directly fits the claim.
Coherence and Cohesion at band 6 often suffers from relying on mechanical, template-like connectors rather than linking the ideas themselves. In addition, memorised linking phrases can be written wrongly. In our sentence-level corrections, learners wrote "On the hand" → "On the one hand" (band 6) and "The conclusion," → "In conclusion," (band 5.5 and below). Using clear topic sentences and natural reference words—such as this policy, such measures, or these workers—helps ideas flow together without relying on stiff transitions.
What does it take to jump from band 6.5 to 7?
Where essays already answer the prompt and are logically organised, the difference between band 6.5 and band 7 lies mainly in language control. Our sentence-level corrections mark "must-fix" errors—clear mistakes in grammar, spelling, or vocabulary that hurt accuracy. Tracking the share of essays with at least one must-fix error of each type shows where band-7 essays pull ahead.
One of the sharpest differences between band-6 and band-7 essays is in spelling. Must-fix spelling errors appear in 73% of band-6 essays but only 14% of band-7 essays. Several were small slips that turn one word into another, for example "One the other hand" → "On the other hand", "there is on denying" → "there is no denying", and "excersice" → "exercise" (all band 6). Saving five minutes at the end of a practice session or test to check spelling is one of the simplest ways to cut avoidable errors.
At 6.5, the grammar corrections cluster in foundational areas: verb tense and form, prepositions, subject-verb agreement, noun plurals, and articles. As Figure 2 shows, must-fix verb tense or verb form errors fall from 75% of essays at 5.5 and below to 14% at band 7.5 and above. Preposition errors drop from 63% of band-6 essays to 41% of band-7 essays, and 17% at 7.5 and above. Subject-verb agreement errors decline from 52% of band-6 essays to 34% at band 6.5 and 31% at band 7, while errors with singular/plural and countable nouns appear in 59% of essays at 5.5 and below and 54% at band 6 before falling at higher bands.
Real fragments from our sentence-level corrections show how ordinary these errors are. In agreement and noun plurals, learners wrote "information are" → "information is" (band 6.5), "bad news spread quickly" → "bad news spreads quickly" (band 6.5), "for good or services" → "for goods or services" (band 6.5), and "every companies" → "every company" (band 6). In prepositions and verb forms, typical corrections included "attention on" → "attention to" (band 6.5), "aware to obey regulations" → "aware of the need to obey regulations" (band 6), and "It emphasized" → "This emphasizes" (band 6).
One grammar problem stands out as the most persistent error across the dataset: articles (a, an, the). Must-fix article errors appear in 60% of band-6 essays, 66% of band-6.5 essays, and still 66% of band-7 essays, falling to 34% only at band 7.5 and above. Typical corrections added a missing article: "With widespread use of internet" → "With the widespread use of the internet" (band 6), "due to lack" → "due to a lack" (band 7), and "such idea" → "such an idea" (band 7.5 and above). For learners whose first language does not use articles, checking singular countable nouns during proofreading is a habit worth building early.
On the vocabulary side, the 0.79-point difference in Lexical Resource between band-6.5 and band-7 essays coincides with fewer word-choice and collocation errors. Must-fix word choice errors fall from 67% of band-6 essays to 46% at band 6.5 and 31% at band 7, while must-fix collocation errors drop from 42% at band 6 to 21% at band 7. At band 6.5, our vocabulary feedback centres on natural collocations, precise word choice, and less repetition—such as fixing inaccurate phrases like "in some extend" → "to some extent" (band 6.5), and choosing natural word partnerships like "environmental protecting" → "environmental protection" (band 6) or "making reservations with doctors" → "booking appointments with doctors" (band 7).
While language accuracy is the final hurdle here, our analysis of IELTS Writing Task 1 common mistakes shows a different priority for 6 to 6.5 reports, where the most common must-fix problem was getting the data itself wrong or incomplete (79% of reports).
How we counted
This analysis covers 384 writing responses submitted to IELTS Writing Checker between July and October 2026 by 62 learners: 183 Task 2 essays from 41 learners (covering opinion, discussion, and four other common question types) and 201 Academic Task 1 reports from 37 learners, which we examine in our guide to IELTS Writing Task 1 common mistakes. Repeated drafts of the same essay were merged so that only the first draft was kept, texts mostly copied from published model answers were removed, and General Training letters were excluded because there were too few.
Every essay was scored on the four criteria by IELTS Writing Checker, and our sentence-level corrections mark errors that "must fix". In our published accuracy study, on 73 officially scored responses, nearly 8 in 10 of IELTS Writing Checker's scores were within 0.5 points of the official score, with an average gap of 0.49 points. All findings reported here still held after removing the five learners who submitted the most essays. As a limitation, these scores are estimates rather than official IELTS results, and because the officially scored samples used in our accuracy study give only an overall score, the criterion-level pattern could not be checked against examiners directly.
If you want to see which criterion is holding your own writing back, you can check an essay with our IELTS Writing Task 2 essay checker.
Common questions
Which criterion holds back IELTS Task 2 scores the most?
Grammatical Range and Accuracy is the lowest of the four criteria in roughly half of Task 2 essays at every band up to 7: 54% at band 5.5 and below, 53% at band 6, 47% at band 6.5, and 53% at band 7. At band 6.5, Lexical Resource is the second most common bottleneck, acting as the lowest criterion in 36% of essays.
Should I use more complex sentence structures to get from band 6.5 to 7?
The data points to fixing basic accuracy first. At band 6.5, most essays still have basic grammar slips: 66% have a must-fix article error, 51% have a preposition error, and 49% have a verb tense or verb form error. Cleaning up articles, plurals, subject-verb agreement, and prepositions comes before trying to add more complex sentences.
Do band 7 essays have to be completely free of grammar mistakes?
No, band 7 essays still contain occasional errors. In our dataset, 66% of band-7 Task 2 essays still had at least one must-fix article error, 45% had a verb tense or verb form error, and 41% had a preposition error. However, must-fix spelling errors appear in 73% of band-6 essays but only 14% of band-7 essays.
Does writing a longer essay help move from band 6 to band 7?
While 25% of essays scored at band 5.5 and below fell short of the 250-word minimum, only 8% of band-6 essays, 9% of band-6.5 essays, and 7% of band-7 essays were under 250 words. Once an essay clears 250 words, length alone does not explain the gap between band 6 and band 7; the criterion data points instead to argument depth and language accuracy.