What a band score comparison can and cannot show
IELTS Speaking is assessed across the whole live test using four equally important criteria: Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, and Pronunciation. Official band descriptors describe patterns across performance, not a single perfect sentence. An examiner hears hesitation, pace, self-correction, rhythm, stress, intonation and intelligibility—evidence that disappears from a clean transcript.
For that reason, the examples below are teaching snapshots. Labels such as “around Band 6 features” mean the response illustrates several characteristics associated with that level; they are not official awarded scores. Your final band can also contain different criterion scores. A speaker might organise ideas well but have variable pronunciation, or use wide vocabulary while making frequent grammar errors.
Ask what the listener can follow, how precisely meaning is expressed, whether structures remain controlled, and how delivery makes the message understandable.
Read the published Speaking assessment criteria guide alongside the official descriptors before judging your recordings.
How performance develops across the four criteria
At developing levels, repetition and searching may interrupt the idea. Stronger performance sustains longer turns, sequences ideas clearly and uses linking naturally rather than mechanically.
Development means more than rare words. Vocabulary becomes flexible across topics, collocations become more natural, meanings become precise and paraphrase helps when a word is missing.
Speakers move from mainly basic forms toward a flexible mix of structures. Errors become less frequent and less disruptive, especially when expressing conditions, comparisons and complex time relationships.
More phonological features support meaning consistently. Thought groups, word stress, sentence focus, rhythm, connected speech and intonation reduce the effort required from the listener.
A higher band does not require a native accent. Official descriptors focus on intelligibility and the effect of accent, not imitation of one variety of English.
Part 1 comparison: the same everyday question
Question: Do you enjoy cooking?
“Yes, I enjoy cooking because it is good and, um, I can make many foods. Usually I cook rice and chicken, and sometimes pasta. I cook at weekend because in weekday I am very busy. Cooking is interesting for me and also it save money.”
“Yes, especially at weekends, when I have enough time to experiment a little. During the week I normally make quick meals, but on Saturday I might try a new curry or bake bread. I find the process relaxing, and it is satisfying to share something I have made myself.”
“I do, although my enthusiasm depends on how much time I have. On a quiet weekend I enjoy making something slowly—bread, for instance—because the process is almost as rewarding as the meal. After work, however, cooking becomes a practical task, so I rely on a few dishes I can prepare without thinking.”
Part 2 comparison: sustaining a long turn
Task: Describe a useful object you own. Explain what it is, when you got it, how you use it and why it is useful.
Developing response
“I want to talk about my laptop. I bought it three years ago for my study. It is black and not very heavy. I use every day for assignments and watching videos. It is useful because I can find information and contact my friends. Also I use it for many things. Before I had old computer and it was very slow, so this laptop is better. Yes, it is very important in my life.”
The speaker covers the prompts but relies on listing and repetition. “Useful,” “many things” and “important” are not made precise. Time reference and articles contain errors, although communication is still possible.
More controlled response
“The object I use most is a lightweight laptop I bought when I began an evening design course three years ago. My old computer could not run the design software reliably, so I saved for several months and chose a model that was powerful enough but still easy to carry.
I initially used it for assignments, particularly projects that involved editing large image files. Since finishing the course, its role has changed. I now take it to client meetings, sketch ideas on screen and show several versions without printing them.
What makes it useful is not simply that it is fast. It allows me to move smoothly between learning and work. It also reminds me that the course led to something practical, because a tool I bought as a student is now central to my job.”
The response establishes a reason for buying the object, moves from past study to present work and reflects on its wider meaning. Vocabulary and grammar serve the timeline. Coherence comes from the changing role, not a list of connectors.
A stronger long turn is not automatically faster or longer. It is easier to follow and appropriately extended. The published Part 2 sample-answer guide shows how keyword notes can support this development.
Part 3 comparison: moving from assertion to analysis
Question: Should cities discourage people from driving?
“Yes, because cars make pollution and traffic. There are too many cars in cities and it is very bad. Government should make people use bus and train, because public transport is better. For example, they can make more buses and people will not use cars. But some people need car, so it is difficult.”
“Yes, particularly in city centres where road space is limited. Congestion affects buses as well as private drivers, and poor air quality harms residents who may not own a car. However, restrictions need to be introduced alongside reliable public transport. If commuters have no practical alternative, higher charges will feel like a punishment rather than an environmental policy.”
“Cities should probably discourage unnecessary car use rather than treat every journey in the same way. A commuter travelling alone along a well-served rail route has realistic alternatives, whereas a disabled passenger or a night-shift worker may not. Policies such as congestion charges can therefore be effective, but only when exemptions are fair and the revenue improves transport. Otherwise, the policy risks reducing mobility for lower-income residents while wealthier drivers simply absorb the cost.”
What the transcript hides
Imagine the high-level answer delivered word by word with long searches, misplaced stress and unclear consonants. The text remains sophisticated, but the live performance would not show the same control. Now imagine the middle answer delivered smoothly with clear thought groups and expressive contrast. Its pronunciation evidence may be stronger than the written comparison suggests.
Continuity: Are pauses mainly used to organise content, or to search for basic language?
Chunking: Are related words grouped so the listener can hear each idea?
Focus: Does stress reveal the important contrast or new information?
Intelligibility: Can the listener understand without repeated effort?
Keep audio whenever you compare bands. Use the published pronunciation guide to mark thought groups, sentence stress and intonation on your own transcript.
Compare one criterion at a time
Fluency and coherence
Do not count seconds without pauses. Listen for whether hesitation breaks the relationship between ideas. A speaker can pause before choosing content and remain coherent. Repeated restarting, functionless repetition and searching inside simple phrases are more limiting. Map the answer: could you label the role of each section?
Lexical resource
Highlight exact meanings, natural collocations and successful paraphrases. Cross out forced idioms or awkward synonyms. “Affordable public transport” may demonstrate more control than an unusual word used incorrectly. Ask whether the speaker can discuss unfamiliar subjects without falling back on vague vocabulary.
Grammatical range and accuracy
Look for the structures required by meaning: past and present time, conditions, comparison, causes and qualification. Count patterns, not isolated slips. Frequent agreement errors may matter more than one accidental mistake in an ambitious sentence. Range without control does not automatically raise performance.
Pronunciation
Listen beyond individual sounds. Mark rhythm, connected speech, word stress, sentence focus and intonation. Note when pronunciation obscures a word or relationship. Accent is not the problem unless it affects intelligibility.
Common mistakes when chasing a higher band
- Memorising a Band 8 transcript: the language may not fit a new question and delivery can sound inflexible.
- Replacing every simple word: precision matters more than rarity.
- Forcing long sentences: uncontrolled complexity can damage clarity.
- Speaking extremely fast: rate without chunking increases listener effort.
- Counting linking words: connectors cannot repair illogical development.
- Removing every pause: content-related thinking pauses are natural.
- Judging one answer: criteria describe performance across the test.
- Ignoring pronunciation: a transcript contains no evidence of live intelligibility.
- Assigning an exact score yourself: use descriptors to choose practice priorities, not promise a result.
Turn a comparison into useful practice
- Blind ranking: compare two recordings before seeing labels and cite evidence under all four criteria.
- One-criterion replay: listen four times, focusing on a different criterion each time.
- Meaning upgrade: replace three vague phrases with precise everyday language.
- Coherence repair: remove repetition and add the missing relationship between two ideas.
- Grammar pattern: find one recurring error and practise five accurate spoken examples.
- Pronunciation map: mark thought groups and focus words, then record again.
- Paraphrase pause: deliberately forget a word and explain it without stopping.
- Complexity test: split an uncontrolled sentence, then reconnect it only if the relationship stays clear.
- Topic transfer: use one improvement on a completely different question.
- Weekly evidence: save a first and revised recording with one measurable change.
A fair self-review
Can I sustain answers without excessive searching?
Does each sentence add or connect meaning?
Can I paraphrase a missing word?
Are vocabulary choices natural and precise?
Do I use structures flexibly for the idea?
Which errors repeat across topics?
Are thought groups easy to hear?
Does stress show new or contrasting information?
Can listeners understand without much effort?
What single change should I practise next?
References
These official IELTS-owner sources were consulted for current descriptors, assessment criteria and preparation guidance.
Improve the evidence, not the label
Choose one criterion, make one observable change and record again. A clearer, more flexible performance is the useful goal; the band estimate follows from complete live evidence.
Record your next four-criteria comparison