The 4 Writing criteria — what each one is really judging

IELTS Writing is scored on four official criteria, weighted equally at 25% each: Task Achievement (Academic Task 1) or Task Response (Task 2 and GT Task 1), Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy. The single most damaging misconception about these four is treating them as one blended impression of "good writing" — they are scored independently by the examiner, against separate band descriptors, and a response can sit at very different bands on different criteria simultaneously. A grammatically flawless essay that doesn't answer the question can score a 5 on Task Response and a 8 on Grammar in the same submission; understanding what each criterion is actually measuring is the only way to diagnose which one is capping your overall band.

Task Achievement, used for Academic Task 1, measures whether you've accurately and completely described the visual you were given, with an overview of the main trends and sufficient, accurately reported detail. A weaker response might state "the chart shows several changes in population over the period" — true, but vague enough to apply to almost any population chart. A stronger response identifies the actual shape of the data: "Overall, the population of City A grew steadily throughout the period, overtaking City B by 2015, while City B's population remained largely static after an initial decline in the first five years." The difference isn't vocabulary or grammar — both sentences could be grammatically identical in complexity — it's whether the content itself demonstrates that you understood and can report the specific data accurately.

Task Response, used for Task 2 and GT Task 1, measures something related but distinct: whether you've addressed every part of the specific prompt, taken and maintained a clear position where one is required, and developed your ideas with genuine reasoning and relevant support rather than generic assertion. A weaker Task 2 response to "Some people think the government should fund the arts; others disagree. Discuss both views and give your opinion" might spend two paragraphs restating that some people like art and some people don't, then close with "in my opinion, both sides have good points" — technically on-topic, but not actually answering with a developed position. A stronger response commits to a real, defensible opinion, develops each side with specific reasoning (funding cuts vs. economic multiplier effects of a cultural sector, for instance), and keeps that position consistent from introduction to conclusion. For GT Task 1 letters, Task Response instead means covering every bullet point in the prompt fully and matching the required tone — a technically well-written letter that omits one of three required bullet points is capped here regardless of how well the other two are handled.

Coherence and Cohesion measures how easily a reader can follow your ideas — not whether you have ideas, but whether their organisation and the connections between them are clear. This covers paragraphing (does each paragraph have one clear controlling idea, and does a new paragraph signal a genuinely new point rather than an arbitrary line break), logical sequencing (does each sentence follow naturally from the one before it), and cohesive devices (linking words, pronoun reference, and substitution used to tie sentences together). A common weakness is a paragraph where every sentence is individually clear but the connections between them are missing or unclear — "Remote work has benefits. Employees can save time. Companies save on office costs. Some people feel isolated." — versus a version where the relationships are made explicit: "Remote work offers benefits for both employees and employers alike, though not without trade-offs. Employees save the time otherwise spent commuting, while companies reduce their office overhead; however, this comes at the cost of some employees reporting greater isolation." The information is nearly identical — the coherence comes entirely from showing how the ideas relate to each other. A frequent overcorrection is stacking linking words mechanically ("Furthermore, moreover, in addition to this, it is also important to note that...") — this is penalised just as much as too few connectors, because the criterion rewards ideas that connect naturally, not the visible scaffolding of connection words layered on top of disconnected ideas.