IELTS Writing — What Is Actually Being Assessed
A structured analysis of the four equal criteria used by IELTS examiners, how they translate into band distinctions, and how automated evaluation aligns with official descriptors.
On this page
Writing Overview
Both Academic and General Training Writing are marked on four equal criteria. Task 1 uses Task Achievement (Academic) or Task Fulfilment (General); Task 2 uses Task Response. The other three criteria are shared across tasks and modules.
Task Response
How fully and directly the question is answered; development of ideas and position.
Coherence & Cohesion
Logical organisation, paragraphing, and purposeful use of cohesive devices.
Lexical Resource
Range, precision, and appropriacy of vocabulary; collocation and spelling.
Grammatical Range & Accuracy
Variety and correctness of sentence structures; control under length and complexity.
Four Criteria Breakdown
Task Response
What it measures: Extent to which the prompt is addressed, ideas are developed, and a clear position is presented and supported.
Examiner insight: Examiners look for relevance, specificity, and progression. Generic or tangential content is penalised even if well expressed.
Band 6 vs 7 gap: At 6, the position may be less clear or less fully developed; at 7, ideas are extended, the position is clear throughout, and main ideas are adequately developed with support.
Automated evaluation maps prompts and responses to official descriptors to assess task fulfilment and idea development.
Coherence & Cohesion
What it measures: Logical flow, paragraph structure, and effective use of linking words and referencing.
Examiner insight: Cohesion must serve meaning. Overuse of formulaic linkers (“Firstly,” “Moreover,” etc.) without clear logical relationships can reduce the score.
Band 6 vs 7 gap: At 6, organisation may be adequate but progression can be mechanical; at 7, there is clear progression and logical use of cohesive devices without overuse.
AI analysis checks paragraph logic, progression, and appropriate use of cohesive devices against descriptor bands.
Lexical Resource
What it measures: Range and appropriateness of vocabulary, collocation, word choice, and spelling.
Examiner insight: Repetition, imprecise wording, and unnatural collocations limit the score. Flexibility and control matter more than rare or showy words.
Band 6 vs 7 gap: At 6, vocabulary is generally adequate but may be less flexible or precise; at 7, there is sufficient range and flexibility with fewer inappropriate or repetitive choices.
Automated evaluation assesses vocabulary diversity, collocation quality, and common error patterns against lexical descriptors.
Grammatical Range & Accuracy
What it measures: Variety of sentence structures and the accuracy with which they are produced.
Examiner insight: Accuracy under complexity matters. A mix of simple and complex structures with fewer errors scores higher than ambitious structures with frequent mistakes.
Band 6 vs 7 gap: At 6, a mix of simple and complex forms exists but errors may be noticeable; at 7, a variety of structures is used with good control and few errors.
AI systems parse grammar, detect error types, and map outputs to the grammatical range and accuracy descriptors.
Common Candidate Weakness Patterns
Systematic diagnostic patterns observed across mid-band candidates:
- Answering a different question or addressing only part of the prompt (Task Response)
- Weak or absent thesis; position implied rather than stated clearly
- Paragraphs with unclear boundaries or single-sentence paragraphs
- Formulaic linking without logical relationships between ideas
- Repetition of key terms instead of paraphrasing or varying vocabulary
- Inappropriate register or collocation (e.g. informal phrases in formal essays)
- Recurrent grammatical errors in articles, tense, agreement, or clause structure
- Errors increasing in longer or more complex sentences
- Spelling and punctuation errors that impede readability
- Word count below minimum (150 / 250), reducing scores across criteria
AI Evaluation Alignment
Automated writing evaluation systems are designed to map outputs to the same four criteria used by human examiners. The underlying logic mirrors the public band descriptors: task fulfilment, coherence and cohesion, lexical resource, and grammatical range and accuracy. Each dimension is operationalised through measurable signals (e.g. prompt coverage, paragraph structure, vocabulary diversity, error density) and aligned to descriptor language. This alignment does not replace examiner judgement but provides a structured diagnostic against the same reference framework.
See the Band Scores & Descriptors page for the official public descriptors. Band descriptor PDFs: English, Chinese.
Evaluate your writing
Evaluate My Writing with AIELTS →IELTS is a registered trademark. This content is for informational purposes only and is not affiliated with the test owners.