English exam writing tests whether a student can turn a prompt into a controlled response for a reader. Whether the task is creative writing, narrative writing, descriptive writing, an essay, article, speech, letter or other directed response, the central examination problem is the same: understand the writing task, choose relevant content, organise it, control language and produce a finished piece that fulfils its purpose.
How to write better English exam answers is therefore not a matter of memorising one perfect composition. Strong examination writing begins with task interpretation and continues through planning, paragraph development, sentence control, vocabulary, grammar, editing and time management. The best language is language that does useful work: it clarifies a thought, creates an experience, develops an argument or helps a particular reader understand what matters.
This world-facing handbook continues the How English Examination Works series and complements the separate composition-planning and creative-writing guides. Its examination-writing focus is how a prompt becomes a plan, a plan becomes paragraphs, paragraphs become a complete response, and feedback becomes the next independent attempt.
The 50-second quick read
PROMPT → PURPOSE → READER → CONTENT → STRUCTURE → LANGUAGE → CHECK. Before writing, identify what must be produced and for whom. Choose content that can fulfil that job. Give each paragraph a function. Use vocabulary and syntax to carry meaning rather than display techniques. Leave enough attention for the ending and final checks. After practice, diagnose the first important breakdown instead of labelling the entire composition weak.
Part I. What an English writing examination is measuring
1. The examiner receives the page, not the plan in your head
A student may have a vivid film of a story, a sophisticated opinion or an excellent example. None of those achievements is visible until the response makes them recoverable. Examination writing is therefore an act of transmission: the intended meaning must survive the journey from thought to text.
This is why clarity is not a beginner’s concern that disappears once vocabulary becomes advanced. The more complex the idea, the more important it becomes to preserve relationships between people, events, reasons and claims.
2. The prompt is a contract
A writing prompt defines a job. It can specify genre, audience, purpose, situation, source material or a required event. A response may be fluent yet fail because it performs a different job beautifully.
Before planning content, translate the prompt into ordinary language. What must happen? What may I invent? What facts must remain fixed? What should the reader know, feel, consider or do by the end?
3. Genre changes what counts as useful detail
A narrative can dwell inside a moment. A report may need concise observation and recommendation. A persuasive speech needs a position shaped for listeners. A letter may need practical information and an appropriate relationship with its recipient.
Do not transplant surface features mechanically. A rhetorical question is not compulsory because the word speech appears. A formal greeting does not rescue a letter that ignores its purpose.
4. Planning is decision compression
A useful plan is smaller than the response. If the plan contains every sentence, it has become an early draft. If it contains only a title, it may not reduce enough uncertainty.
Plan the relationships that could otherwise break: central claim or conflict, paragraph jobs, necessary evidence or events, turning point, consequence and ending. The plan should make drafting easier without consuming the time needed to draft.
5. Paragraphs are units of work
A paragraph is not successful merely because it contains a certain number of sentences. It should perform a recognisable job in the response: establish, develop, contrast, complicate, exemplify, turn, resolve or conclude.
When a paragraph drifts, ask what changed halfway through. Sometimes the answer is to split it. Sometimes two thin paragraphs should be joined because they perform one job.
6. Vocabulary is precision under pressure
Ambitious vocabulary is valuable when it creates a distinction the response needs. Reluctant, wary and terrified do not describe the same state. A precise familiar word can be stronger than an impressive word whose connotation is wrong.
During timed writing, prefer a word you can control over one you only half remember. Practice is where uncertain vocabulary should be tested, checked and expanded.
7. Grammar carries relationships
Grammar is not a decorative layer placed on completed ideas. Tense locates events in time. Pronouns connect references. Subordination can show cause, concession or condition. Punctuation helps the reader recover sentence boundaries and relationships.
A grammar error matters most when it changes or obscures meaning. Editing should therefore ask not only whether a sentence looks correct but whether another reader can tell who did what and why.
8. Sentence variety follows varied thinking
A short sentence can isolate a decision. A longer sentence can hold a qualification or connect simultaneous observations. Variety becomes useful when syntax reflects the work of the thought.
Avoid manufacturing a pattern of short, medium and long sentences simply to tick a feature. Ask which ideas belong together and which deserve separation.
9. Development means adding consequence, not padding
A thin paragraph often needs a relationship rather than more adjectives. In narrative, an action may need a consequence. In argument, a claim may need a reason and example. In description, an image may need a viewpoint that gives it significance.
Padding repeats the same proposition in different words. Development changes what the reader can understand.
10. The ending must answer the beginning
A response creates expectations. A story establishes a conflict or question. An argument establishes a position. A proposal establishes a decision to be considered. The ending should respond to that promise rather than introduce an unrelated final flourish.
A quiet ending can be complete. Completion is about causal or logical closure, not necessarily a dramatic surprise.
11. Editing has layers
Structural editing asks whether the response fulfils the task. Semantic editing asks whether sentences mean what the writer intends. Surface editing checks grammar, punctuation and spelling.
Correcting commas while the response answers the wrong question is poor prioritisation. Check the highest-cost failure first.
12. A mark is compressed feedback
A score reports an evaluated performance under particular conditions; it does not by itself explain what to practise next. Two responses with the same score can have different weaknesses.
Reopen the response. Locate the first important breakdown: prompt interpretation, selection, organisation, development, language control, completion or checking. The next task should target that operation.
Part II. Original narrative examination laboratory
13. Prompt: write about a promise that becomes difficult to keep
The following story is original teaching material. It is not an official model answer and carries no universal examination word target. Read it first as a complete piece; the later chapters dismantle its decisions.
14. Original story: The Last Photograph
At four twenty, Ethan promised his grandmother that he would not open the envelope until she returned.
The envelope lay on the kitchen table beneath the sugar bowl. His name was written across it in the careful capital letters she used for labels. GRANDSON — ETHAN. Underneath, in smaller writing: AFTER I COME BACK.
“I am only going downstairs,” she said, searching for her keys. “Mrs Lim has a box I need. Ten minutes.”
Ethan lifted both hands. “I won’t touch it.”
His grandmother smiled, apparently satisfied by the theatrical seriousness of the gesture, and left.
For the first three minutes, Ethan ignored the envelope easily. He rinsed two cups, checked a message and opened the refrigerator without wanting anything inside it. Then his phone rang.
It was his mother. “Has Grandma shown you the photograph?”
Ethan looked at the envelope.
“What photograph?”
There was a pause. “Never mind. She wanted to tell you herself.”
“Tell me what?”
“Ethan.” His mother’s voice changed. “Wait for her.”
The call ended with the question still in the room.
He sat opposite the envelope. The paper was thick enough that he could not see through it. One corner had been folded and flattened again. He slid the sugar bowl two centimetres away, then put it back exactly where it had been.
Five minutes became eight.
From the corridor came the lift bell, footsteps, then silence. Ethan stood, expecting the key in the door. The footsteps continued past the apartment.
He laughed once at himself and sat down again.
At twelve minutes, he sent his grandmother a message: Everything okay? The message showed as delivered but not read.
Now the promise felt different. At four twenty it had been a small performance of patience. At four thirty-two it had acquired a question his mother had refused to answer and a silence his grandmother had not explained.
Ethan picked up the envelope.
It was not sealed.
That made the decision worse. He would not have to tear anything. He could lift the flap, look, close it again and leave no obvious damage. The promise had no physical lock. It existed only because he had said the words.
He put the envelope down.
His phone rang again. This time the building security desk appeared on the screen.
“Mr Ethan Tan?”
“Yes.”
“Your grandmother asked us to call. The lift stopped between floors for a few minutes. It is moving again, but she is waiting downstairs while the technician checks it. She says she is fine.”
Ethan let out a breath he had not noticed holding.
“Can I speak to her?”
“She is talking to the technician now. She said to tell you not to worry.”
After the call, the envelope looked smaller.
He understood something uncomfortable: part of his urgency had been concern for his grandmother, but part had simply been the desire to know first. The lift problem had justified his worry. It had not justified opening the envelope.
He sent another message: Security called. I’m here. No rush.
Then he carried the envelope from the table to the shelf beside the window, not because the new place was safer but because he was tired of pretending that staring at it counted as resisting it.
At four forty-one, the key finally turned.
His grandmother entered carrying a cardboard box and complaining about lifts, technicians and the fact that nobody designed buildings for people carrying cardboard boxes.
Ethan took the box from her.
“You waited?” she asked.
“Yes.”
She studied him for half a second longer than necessary. “That difficult?”
“You have no idea.”
“I probably do.”
She washed her hands, made tea and only then retrieved the envelope. Ethan decided that this delay was deliberate, but he had lost the right to complain about ten more minutes.
Inside was a photograph of a boy standing beside a bicycle. The boy had Ethan’s eyebrows and an expression of exaggerated patience.
“Your father,” his grandmother said. “The day before he left for university.”
Ethan looked more closely. His father was younger than Ethan was now.
On the back, in faded blue ink, someone had written: He said he would wait until Sunday. He lasted twenty minutes.
Ethan began to laugh.
His grandmother tapped the photograph. “That is why I wanted to be here.”
The envelope had contained no emergency and no secret that could not survive another twenty minutes. What mattered was the moment in which the photograph crossed from her hand to his, carrying a story neither the paper nor the image could have supplied alone.
Ethan placed the photograph back on the table between them.
“Did he open whatever he was waiting for?”
His grandmother lifted her tea. “Immediately.”
Part III. From model to method
15. Begin with the promise
The first sentence establishes the obligation and a clock. Later details therefore have a job: increase, reduce or reinterpret pressure around that promise.
Transfer the method, not the envelope or family. Start near the relationship the prompt requires unless a slower approach has a deliberate purpose.
16. Curiosity needs new information
The mother’s question changes the envelope from an ordinary object into withheld information. This develops the situation more effectively than repeatedly declaring Ethan curious.
Development changes what the reader knows or expects. Padding repeats the same state.
17. Make the choice possible
The envelope is unsealed. Ethan could open it easily, so the restraint remains his promise rather than a physical lock.
Do not make the wrong action impossible and then call the correct action a decision.
18. Complicate motive
Concern for his grandmother overlaps with curiosity. The security call resolves safety uncertainty without removing the desire to know.
Overlapping motives can create believable complexity without making the plot unclear.
19. Make understanding produce action
Ethan moves the envelope away after recognising that worry and curiosity are different. The action makes the turning point visible.
A turning point becomes stronger when a new understanding changes behaviour.
20. Let the ending answer the opening
The photograph reveals a family story about waiting. The grandmother’s instruction becomes understandable and the opening promise receives a payoff.
Completion is causal and emotional, not necessarily spectacular.
21. Use humour from the situation
The final immediately echoes the waiting problem and releases pressure without cancelling meaning.
Tone changes work best when they grow from established information.
22. Apply the deletion test
Remove the mother’s call and curiosity weakens. Remove the security call and worry may dominate. Remove the photograph story and the promise loses payoff.
Deletion reveals load-bearing events and exposes scenes that perform no causal job.
23. Paragraphs are units of work
A paragraph can establish, develop, contrast, complicate, turn or resolve. It does not become successful merely by reaching a preferred sentence count.
When a paragraph drifts, identify where its job changes. Split or reorganise around meaning.
24. Revise at three levels
First ask whether the story fulfils the prompt. Then ask whether each sentence means what you intend. Finally check grammar, punctuation and spelling.
Surface correction cannot rescue a story that resolves the wrong conflict.
Part IV. Directed writing
25. Original task: improve a student showcase
Visitors reported difficulty finding rooms, performances began while audiences were still moving between venues, and several exhibits had no student available to explain the work. Write to the coordinator recommending a clearer visitor route, a timetable improvement and a way to ensure exhibits can be explained. Address one disadvantage.
26. Model response
Dear Event Coordinator,
I recommend organising next year’s showcase around a simple visitor route and a timetable with short movement periods between performances. The event offers many things to see, but the reported navigation and timing difficulties suggest that visitors need a clearer way to move through them.
First, the venue could be divided into numbered zones, with the same numbers used on a one-page map and on signs outside each room. A small information point near the entrance could help anyone who still needs directions.
Second, performances could be grouped into fixed sessions with a short interval between them. This would not guarantee that every visitor reaches every venue, but it would reduce the chance of a new performance beginning while the previous audience is still moving.
For exhibits, I suggest a student rota. Each group could nominate members to cover agreed periods so that someone familiar with the project is available to answer questions. A short written explanation could remain beside the exhibit when the student is temporarily away.
The main disadvantage is reduced spontaneity. Students would need to know when they are responsible for an exhibit. This could be managed by publishing duties early and allowing groups to exchange slots before the final schedule.
A small trial could be run with one section before the full plan is adopted. Organisers could then record which navigation questions visitors ask and whether the movement intervals are sufficient rather than assuming the first redesign is perfect.
Thank you for considering these suggestions.
Kind regards,
Mira
27. Persuasion without invented proof
The response proposes mechanisms and admits a trade-off. It does not invent survey percentages, budgets or guaranteed outcomes. Persuasion is built from practical relationships rather than unsupported certainty.
28. Genre controls selection
A story could dramatise one lost visitor. This task should not become that story because the reader needs an event design. Genre determines which details deserve space.
Part V. Argument and essay writing
29. Original question: Should every student project include a public presentation?
A strong essay needs a position that can survive a reasonable objection. The following model develops mechanisms and limits without invented statistics.
30. Model argument
Student projects should sometimes include public presentation, but presentation should not be compulsory in identical form for every project. Explaining work to an audience can reveal weaknesses that remain hidden when a student works only for the teacher; however, the format should serve the project’s learning purpose rather than become a second, unrelated test.
Presenting can improve clarity because an audience needs a route through the work. A student must decide what the listener needs first, which evidence deserves emphasis and how technical ideas can be explained without losing accuracy. These decisions can expose gaps in understanding.
Presentation can also create useful questions. A listener may notice an assumption the creator no longer sees. Answering requires the student to connect evidence with reasoning, and an unanswered question can identify what should be revised.
However, compulsory presentation can distort a project when performance becomes more important than the work. A carefully researched investigation should not automatically be judged weak because its author is less fluent in an unrelated presentation format.
Some projects are also unsuitable for broad presentation. Personal reflective work may contain information not intended for a large audience, while early experimental work may benefit more from a small feedback group.
A better policy is to make communication deliberate but flexible. Projects could end with a talk, poster session, demonstration, recorded explanation or question-and-answer conference. The common requirement would be making important decisions understandable to an appropriate audience.
Public presentation is therefore valuable when it helps students clarify, defend and revise their work. It becomes less useful when imposed without regard to purpose. The real question is whether each project gives its creator a meaningful opportunity to make the work intelligible beyond the creator’s own head.
31. Build a conditional thesis
The thesis supports communication while rejecting one compulsory format. This gives the essay room for reasons, limitations and a practical alternative rather than forcing a false all-or-nothing debate.
32. Examples should do work
An example should reveal a mechanism the claim alone does not. After writing one, ask what it demonstrates. If the answer is nothing, it may be decoration.
Part VI. Language engineering
33. Topic sentences are promises
A topic sentence tells the reader what work the paragraph is about to perform. It can state a claim, distinction or shift rather than follow a fixed formula.
Check the paragraph against its opening. If half performs another job, revise the organisation.
34. Connectors cannot manufacture logic
Furthermore, however and therefore work only when the underlying relationship exists.
Remove the connector and explain the relationship plainly. If you cannot, repair the thinking first.
35. Pronouns need owners
When several people or ideas appear, this, it, they and she can become ambiguous.
Repeat a noun when necessary. Avoiding repetition is less important than preserving meaning.
36. Tense is a timeline
Narratives can shift tense deliberately for memories, general truths or speech. The problem is change without reason.
Mark the main timeline and inspect verbs that leave it.
37. Punctuation has jobs
No punctuation mark is automatically sophisticated. Use punctuation whose grammar and effect you control.
A semicolon can join closely related independent clauses; it is not a decorative replacement for any comma.
38. Specific nouns reduce adjective load
A precise noun can establish setting and implication economically.
Specificity should still fit the viewpoint and knowledge of the speaker.
39. Strong verbs clarify action
A more specific verb can sometimes carry direction or manner economically.
Do not replace every common verb. Precision matters more than novelty.
40. Figurative language needs purpose
Too many unrelated comparisons can make each sentence feel as though it belongs to a different imaginative world.
Choose imagery that grows from scene, viewpoint and tone.
41. Dialogue should alter information
Useful dialogue can reveal a problem, correct an assumption, change a relationship or force a choice.
Ask what becomes different after each exchange.
42. Description follows attention
A writer need not inventory every object. Select what the viewpoint would notice and what the task needs the reader to understand.
In argument, vividness should clarify an example rather than replace reasoning.
43. Vocabulary is precision
Ambitious vocabulary helps when it creates a needed distinction. A precise familiar word is stronger than an impressive word with the wrong connotation.
During timed writing, prefer words you can control. Test uncertain vocabulary during practice.
44. Conclusions should close the job
A conclusion can synthesise the position, sharpen a distinction or identify an implication.
A major new reason belongs earlier where it can be developed.
Part VII. Timed performance
45. Time is resource allocation
Writing requires distributing limited time across interpretation, planning, drafting and checking. There is no universal perfect percentage for every task.
Record where practice time actually goes. Different bottlenecks require different repairs.
46. Use a completion checkpoint
At a chosen practice point, ask whether enough time remains to reach the required ending or conclusion.
Reduce secondary material before sacrificing the central response.
47. Do not panic-rewrite the opening
Late in a task, an ordinary opening may be less dangerous than an unfinished conclusion.
Rewrite early material only when it creates a real task or meaning problem.
48. Use a vocabulary stopping rule
If an ambitious word will not come, use a precise alternative you know.
Record the missing word after practice and learn it then.
49. Reserve a meaning check
When conditions allow, reread for high-cost errors: missing required point, contradiction, unclear pronoun, broken sentence or omitted ending.
Train a compact check around your recurring patterns.
50. Recover from a tangled sentence
Stop adding clauses. Identify the core subject, action and relationship. Rewrite simply, then continue.
A clear sentence keeps the response moving.
Part VIII. Diagnostics and transfer
51. Ben’s thin narrative
Ben’s character changes his mind in one sentence. Adrian asks what information became available between the first decision and the second. Ben discovers there is none.
The repair is a turning-point exercise: create three observations that could reasonably change the decision, choose one and build a short scene. The target is causation, not adjective count.
52. Aisha’s overloaded argument
Aisha has five plausible reasons in one paragraph but develops none. Jo asks her to keep the strongest two and explain an example for each. The essay becomes shorter in idea count but deeper in reasoning.
The lesson is not that essays always need two reasons. Breadth has a cost under constrained time.
53. Ryan’s unfinished story
Ryan writes steadily but plans three locations, two timelines and five characters. His problem is scope, not speed. His next practice uses one location and one central decision so he can rehearse completion.
54. Clara’s polished opening
Clara rewrites her opening repeatedly while the response remains unfinished. Her next exercise uses a stopping rule: once the opening establishes the situation and fits the prompt, she continues drafting and records optional improvements for later.
55. Ethan’s grammar under pressure
Ethan’s untimed sentences are controlled, but pronoun reference and sentence boundaries break during fast drafting. His practice alternates short timed paragraphs with targeted editing of those two patterns.
The diagnosis is condition-specific. It does not justify the broad claim that Ethan does not know grammar.
56. Eight-session writing cycle
Session one: independent baseline. Session two: prompt interpretation. Session three: one paragraph or scene. Session four: compare two responses. Session five: fresh task after a gap. Session six: full response. Session seven: appropriate timed conditions. Session eight: compare the baseline with the new independent attempt and select the next repair.
This is an adaptable teaching sequence, not a promised improvement schedule.
57. Read from the receiving side
Give a reader the story without your plan and ask what changed the character’s decision. Give them the argument and ask for its position and reasons. Where their reconstruction differs from your intention, investigate the text before explaining yourself aloud.
58. Use models without surrendering ownership
Study what a model does: how it opens, develops, qualifies, turns or concludes. Then close it and perform the same underlying operation on different material.
Independent control appears when the topic, evidence and wording change but the learner can still make the necessary decisions.
Part IX. FAQs and routes
59. How many paragraphs should I write?
There is no universal paragraph count for every English examination writing task. Follow explicit requirements and organise around the jobs the response must perform.
60. Should I memorise a composition?
Study models closely, but memorising one plot is not a reliable substitute for responding to a new prompt. Preserve methods such as establishing pressure, development and consequence while allowing the content to change.
61. Do advanced words earn marks automatically?
No. Vocabulary becomes useful through appropriate, precise use. An inaccurate ambitious word can weaken meaning. Learn new language actively, but choose words that fit the response.
62. What if I cannot think of a story?
Reduce the problem. Identify one person, one valued aim, one obstacle and one decision. A manageable causal relationship is often a stronger starting point than searching for an extraordinary event.
63. What if my essay has no examples?
First ensure each claim has a reason. Then choose examples that make the mechanism visible. Do not invent research findings or statistics simply to make the paragraph appear authoritative.
64. What should I practise next?
For detailed story planning, use How to Plan a Story Before You Start Writing. For creative craft, use How to Write Your First Short Story. For vocabulary, use How to Improve Vocabulary.
65. How does reading help writing?
Reading and writing are inverse crafts. The companion reading comprehension handbook shows how readers recover meaning from evidence. Writers perform the reverse job: choosing evidence so the intended meaning can be recovered.
66. Final examination check
Prompt. Purpose. Reader. Content. Structure. Language. Check. If a response fails, locate the first broken link. If it succeeds, identify the decision worth carrying into the next unfamiliar task.
English examination writing becomes less mysterious when the learner can see the machinery. A prompt creates a responsibility. Planning chooses a route. Paragraphs perform jobs. Language carries relationships. Revision protects meaning. The finished response is where all of those decisions become visible to another reader.