Everything about CELPIP writing task 2, Responding to Survey Questions, plus 5 practice questions with model sample answers. Two options, one choice, 150 to 200 words, 26 minutes, and absolutely no greeting.
Five survey scenarios on money, heritage, schools and community. No timer. Choose your option, write your answer, then reveal the model sample and the explanation.
A parent council has raised $30,000 over two years and surveyed all parents on how to spend it.
Teaches: one idea developed deeply, a personal example with real numbers, a controlled closing contrast.
Practise this questionYour employer has set aside $1,200 per employee for staff benefits and surveyed all staff.
Teaches: reframing the question, defending the unpopular side, an example with a before and an after.
Practise this questionA structurally sound 1911 brick warehouse sits empty downtown. The council has surveyed residents.
Teaches: arguing both sides of the ledger, attacking with a fact, admitting the limit of your own case.
Practise this questionThe market runs outdoors from May to October and has grown steadily. Shoppers and vendors were surveyed.
Teaches: two reasons from two different stakeholders, evidence lists, naming the strongest objection honestly.
Practise this questionA business association has created a $10,000 education fund for graduating students and surveyed residents.
Teaches: raising the objection to your own argument, making a number tangible, never insulting the other option.
Practise this questionFormal email, informal email and survey response, all in one place with model answers.
Back to the samples hub →CELPIP writing task 2 is called Responding to Survey Questions. You are shown a short survey: a realistic situation (a parent council has raised $30,000, a town council is deciding what to do with an empty warehouse, an employer has a benefits budget) and two clearly labelled options, A and B. You choose one, and you explain why, in about 150 to 200 words, in roughly 26 minutes. It is the second of the two writing tasks, one minute shorter than the email.
The single most important fact about Task 2 is what it is not. It is not an email. There is no greeting, no "Dear Parent Council", no sign-off and no signature. Every year an enormous number of candidates arrive at Task 2 straight from writing an email in Task 1, run on autopilot, and type "Dear Organizers" at the top. That single line tells a rater you have misunderstood the task before they have read one word of your argument, and it comes directly out of your Task Fulfilment mark.
The second most important fact is that there is no correct option. Neither answer is right. Raters are not scoring your opinion, they are scoring how well you develop it, which means you should pick whichever option you can supply the best example for, not the one you imagine the marker prefers. Look at the samples below: two of the five deliberately defend the less popular side and win, because a well argued unpopular choice demonstrates far more control than a lazy agreement with the obvious answer.
What Task 2 rewards, in the plainest possible terms, is a choice stated immediately, two or three reasons that are genuinely developed rather than listed, at least one specific example, and honest treatment of the option you turned down. What it punishes is sitting on the fence. Saying "both options are good", refusing to choose, or switching sides in paragraph three costs you marks no matter how elegant the sentences are.
The 2026 CELPIP writing format for the survey task, and how it differs from Task 1.
Task 1, the email, is covered on the formal email and informal email pages.
The same four criteria as the email task, but they bite differently here.
On Task 2 this is mostly development. Is each reason actually explored, or just named and abandoned? This is where examples earn their keep.
Grammar, spelling, punctuation, sentence control. Task 2 tempts people into long tangled sentences. Resist.
Did you choose one option, stay with it, justify it, and stay off email furniture? Fence-sitting and stray greetings both land here.
Range and precision. The smallest slice, and the one candidates over-invest in. Plain and exact beats ornate and approximate.
Two of these criteria behave differently on Task 2 than on the email. Task Fulfilment on Task 1 is mostly about covering three bullets. On Task 2 it is about commitment: one option, chosen in the opening sentence, held to the end. There are no bullets to tick, so the way you demonstrate fulfilment is by never leaving the reader in doubt about where you stand.
Content and Coherence on Task 2 is almost entirely about depth rather than breadth. On the email, coverage matters. Here, one reason explored properly outscores four reasons listed in a row. The sample on the school fundraising question turns the entire answer on a single idea, breadth of benefit versus depth of benefit, and explores it for four paragraphs. That is what development looks like, and it is worth 30 percent.
The practical instruction that falls out of this: decide your option in the first three minutes and do not revisit the decision. Every minute spent internally wavering is a minute not spent developing, and wavering leaks into the prose in ways raters notice immediately.
Every sample on this page uses the same four-paragraph shape, and it is worth internalising because it removes the hardest decision (what goes where) before the clock starts.
Paragraph four is where the marks hide. Most candidates only argue why their option is good, and stop. The stronger answers do both halves of the ledger: here is why B is right, and here is why A is weaker. The warehouse sample attacks the museum option with a fact, not an adjective. The local history collection in the library basement gets about two visitors a week, by the librarian's own count. That single invented detail does more damage than a paragraph of "nobody cares about museums", which is just a feeling with a full stop after it.
The most advanced move available on Task 2, and one that almost no candidate attempts, is raising the objection to your own argument and answering it. The bursaries sample does exactly that: "The obvious objection is that $500 is too small to matter. I would answer that we are not trying to fund an education. We are trying to remove the last small obstacle standing between twenty students and a decision they have already made." Naming your own weakness paradoxically makes everything around it more credible.
One last structural tip, and it is a cheap one. Make your two reasons come from two different stakeholders. The farmers' market sample gives reason one from the shopper's point of view (winter) and reason two from the vendor's (a season you can build a business on). Deliberately switching stakeholder is the most reliable way to guarantee your second reason is not secretly your first reason wearing different words, which is the most common failure in a mid-band Task 2 answer.
Lifted from the five sample answers. Reusable, register-appropriate, and each one doing a structural job.
Look at the right-hand column again. Every phrase there is doing something generous: conceding, questioning, admitting a limit, granting the other side its due. That generosity is not politeness, it is technique. An answer that acknowledges the strength of what it rejects reads as considered rather than reflexive, and considered is what the top bands are made of. Calling the scholarship "a wonderful gift" and then explaining why it is the wrong tool is far stronger than arguing that top students do not deserve it.
The specific, repeatable ways candidates with good English land a mediocre band on the survey task.
"Dear Parent Council" or a signature at the bottom. Usually pure autopilot from Task 1, twenty-six minutes earlier. The greeting alone signals you have not understood the task, and it is the cheapest mistake on the whole writing section to eliminate.
"Both options have merit and the council should consider each carefully." This is the fastest way to lose Task Fulfilment marks on Task 2. Choose in sentence one, and never switch sides halfway through.
Raters reward development, not quantity. Two reasons explored properly beat five named and abandoned. If a reason gets one sentence, it is not a reason, it is a bullet point with ambitions.
An answer with no example stays pure opinion, and pure opinion caps your band. Equally weak: an example with no detail, like "I once went to a retreat and it was good". Give it a cost, a duration and an outcome.
It happens when both reasons come from the same point of view. Deliberately change stakeholder for reason two, shoppers then vendors, students then taxpayers, and the problem disappears.
"In my opinion I think that ..." wastes five words of a 200-word budget and delays the choice. Long introductions are why candidates run out of room before disposing of the other option, which is where the marks were.
"Nobody cares about museums" is a mood. "The local history collection gets about two visitors a week" is a fact, and it is devastating. Attack the other option with a reason, never with an adjective.
Padding about how farmers' markets build community could be written without reading the scenario, and raters can tell. Anchor every claim to the specific town, school or company the survey described.
Not the one you believe, and definitely not the one you think the marker wants. Your opinion is not scored. The development is. Whichever option you can attach a concrete story to is the correct choice for you.
"I would choose Option B and divide the fund into twenty bursaries." A rater should never have to hunt for your position, and an immediate choice buys you the whole word budget for defending it.
"I say that as someone who genuinely loves old buildings." Pre-empting the obvious accusation buys you permission to be hard-headed for the rest of the answer, and it costs one clause.
Shoppers then vendors. Students then taxpayers. It is the single most reliable technique for making sure your second reason is genuinely new rather than a paraphrase of your first.
Disposing of the option you rejected is where most of the marks live and where most candidates run out of room. Budget for it before you start, not after you notice you are at 190 words.
"It can wait. A struggling eight-year-old cannot." "Fifty-two Saturdays beats twenty-six." A deliberately short final sentence after a long one reads as confident and controlled, and it is a free high-band signal.
CELPIP Writing Task 2 is called Responding to Survey Questions. You read a short survey that presents a situation and two options, you choose one, and you justify your choice with developed reasons in about 150 to 200 words. You get roughly 26 minutes, and it is the second of the two writing tasks.
No. Task 2 is not an email and it takes no email furniture. No greeting, no Dear Parent Council, no sign-off and no signature. Adding a greeting out of habit from Task 1 signals to a rater that you have misread the task, and it costs you Task Fulfilment marks before they read a single argument.
About 26 minutes, one minute less than Task 1. Spend 3 to 4 minutes deciding your choice and your reasons, roughly 19 minutes writing, and the last 3 minutes proofreading. Deciding fast matters more here than on Task 1, because an undecided writer wastes words wobbling.
No. Refusing to choose, saying both options are good, or switching sides halfway through all fail Task Fulfilment regardless of how good the English is. Pick one option in the first sentence and defend it to the end. You may concede the other option's strengths, but the choice never wavers.
No. Neither option is right or wrong, and raters do not score your opinion. They score how well you develop it. A well argued unpopular choice scores higher than a lazy popular one, so pick whichever option you can supply the best example for, not the one you think the marker wants.
Two developed reasons beat four listed ones. Raters reward development, not quantity. A reliable trick is to make your two reasons come from two different stakeholders, for example one reason about shoppers and one about vendors, which guarantees your second reason is not just your first reason again.
Yes. An answer with no example stays pure opinion, and pure opinion caps your band. One specific example with a cost, a duration and an outcome is worth more than three additional abstract reasons. Invented examples are perfectly acceptable, as long as they are concrete.
On the same four criteria as Task 1: Content and Coherence at 30 percent, Vocabulary at 20 percent, Readability at 25 percent and Task Fulfilment at 25 percent. The result is reported on the CLB 1 to 12 scale.
Yes. Most candidates only argue why their option is good, and the marks sit in the second half: showing why the option you turned down is weaker, or at least conceding its real strength honestly before answering it. Never attack the other option with feelings. Use a reason or a fact.
Yes. All 5 survey response questions, their model sample answers and the full explanations are free, and every one follows the 2026 CELPIP format: two options, 150 to 200 words, roughly 26 minutes, scored on the CLB 1 to 12 scale.
Task 2 is only half the writing section. Task 1 has two registers, and the timed tests put both together.
CELPIP writing task 1 explained in full, with 5 formal samples.
See 5 samples →The other register of Task 1, plus the writing topics that recur most.
See 5 samples →Both tasks under the real 27 and 26 minute limits, with expert feedback.
Start a writing test →Full-length CELPIP mock tests across all four skills in one place.
View mock tests →No timer, no evaluation. Choose a side, write your 150 to 200 words, then reveal the model sample and the explanation.
