IELTS Writing Model Answers, and How to Use One Properly

A model answer is a finished essay written at a high level. There are thousands online, and reading them is the most popular study activity in IELTS preparation. It is also one of the least effective, because of how it is usually done: the candidate reads the model, agrees that it is good, feels reassured, and writes their next essay exactly as before.

The problem is that admiration is not analysis. A model answer only teaches you something if you can say what each sentence is doing and why it earns its place. This page takes one model apart that way, then gives you the method to do it to any other model you find.

The case against memorising

Memorised sentences are visible to a marker. The band descriptors for Lexical Resource and for Coherence and Cohesion both treat mechanical, prefabricated language as a weakness rather than a strength, and a memorised opening sits in an essay like a different font.

There is a second reason, which is more practical. A memorised structure has to fit the question you are given. If you have learned an opinion essay frame and you are given a two part question, the frame actively damages your Task Response because it makes you answer a question nobody asked. We have written separately about why memorised phrases cost marks.

Reverse engineering is the alternative. You keep the moves and throw away the words.

One model, taken apart

The question:

Some people think that the best way to reduce crime is to give longer prison sentences. Others believe there are better alternatives. Discuss both views and give your own opinion.

The model:

Public debate about crime often returns to the length of prison sentences, with one side arguing for longer terms and the other for investment in prevention. This essay will consider both positions before arguing that longer sentences alone address the symptom rather than the cause. The argument for longer sentences rests on two ideas: deterrence and incapacitation. If the cost of being caught rises, some potential offenders will decide the risk is not worth taking, and those who are imprisoned cannot offend outside prison while they are inside. For serious violent crime, where the public expects protection above all else, this reasoning carries real weight and is difficult to dismiss. The opposing view starts from who actually commits most recorded crime. A large share of offences are committed by people with limited education, unstable housing and untreated addiction, and none of those conditions is improved by a longer sentence. Programmes that treat addiction, provide training inside prison and support housing on release address the conditions that produced the offence in the first place. The same money spent on a longer sentence buys custody; spent on rehabilitation, it buys a lower chance of return. In my view, the second approach is stronger, though not universally. Deterrence works best where the offender calculates, and a great deal of crime is committed impulsively or under the influence of substances, where no calculation takes place. The sensible policy is therefore mixed: long sentences reserved for the small number of dangerous offenders, and prevention funded for the far larger number whose offending is a product of circumstance. Crime responds to conditions, not only to punishment. A system that lengthens every sentence will fill its prisons without lowering its crime rate, while one that treats the causes has a chance of reducing both.

What each paragraph is doing

Paragraph 1. Two sentences. The first names the debate in ordinary language without repeating the question's words. The second states what the essay will do and takes a position. Notice that the position is specific: longer sentences alone address the symptom rather than the cause. It is not this essay will discuss both sides, which commits to nothing.

Paragraph 2. This is the view the writer will eventually reject, and it is presented at full strength. Two named mechanisms, deterrence and incapacitation, then a concession that the reasoning carries real weight. A discuss both views essay that weakens one side deliberately loses Task Response marks, because the task was to discuss both.

Paragraph 3. The counter view, built on a factual claim about who offends, then a consequence. The final sentence does the real work: the same money spent on a longer sentence buys custody; spent on rehabilitation, it buys a lower chance of return. One idea, two outcomes, a semicolon instead of a connector.

Paragraph 4. The opinion, which the task demanded. It is qualified, and the qualification is specific: deterrence needs a calculating offender, so it fails where offending is impulsive. That is a reason, not a hedge.

Paragraph 5. A conclusion that restates the position in different words and adds no new argument.

The sentence level moves worth stealing

None of those requires rare vocabulary. Every word in the model is a word a band 6 candidate already knows. What separates the model is how the words are arranged, which is exactly the part you can copy without it being memorisation.

How to reverse engineer any model in ten minutes

1. Read it once for meaning only. Do not take notes.
2. Write one sentence per paragraph saying what that paragraph does, in verbs: presents the opposing view, gives the writer's opinion with a limit.
3. Underline every sentence that connects to the previous paragraph. Ask whether it connects by a linking word or by a reference to an idea. References score higher.
4. Find the three most specific noun phrases in the essay. These are usually where Lexical Resource marks come from.
5. Find one sentence you could not have written. Write out why: is it the structure, the tense, the word choice?
6. Close the model. Write your own paragraph on a different topic using only the moves you listed in step 2.

Step 6 is the one everyone skips, and it is the only step that changes your writing.

What to do today

Take any model answer you already have saved. Run steps 1 to 6 on it, and keep the list of verbs from step 2 somewhere you can see it during your next practice essay. A list of moves transfers to new questions. A list of sentences does not.

Then write an answer of your own to the question above and put it through Bandly to see which of the four criteria comes back lowest. The bands are machine estimates rather than marks from a qualified examiner, so treat the number as a direction to work in and the criterion breakdown as the useful part.