Summarize Written Text and Summarize Spoken Text look alike, but their word ranges are different, and mixing them up costs points directly. This lesson treats them side by side.
The two tasks side by side
| Detail | Summarize Written Text | Summarize Spoken Text |
|---|---|---|
| Input | Text up to 200 words | Audio 45–75 seconds |
| Items per test | 1–2 | 1–2 |
| Time | 10 minutes | 8 minutes |
| Summary length | 25–50 words | 20–30 words |
| Skills | Reading and writing | Listening and writing |
| Scored traits | Content, form, grammar, vocabulary | Content, form, grammar, vocabulary, spelling |
| Source | Pearson: Speaking & Writing format | Pearson: Listening format |
Figures come from Pearson's PTE Core test format pages linked in the table; items per test come from the PTE Core Score Guide. The most common mix-up is the word range: written is 25–50, spoken is 20–30. Once you get used to one, it is easy to overshoot the other. The second difference is easy to miss: the spoken summary has a spelling trait, and its top band allows only one error.
Summarize Written Text: what Form really requires
The Score Guide's Form bands for the written summary only set length and format. There is no requirement to write a single sentence:
- 2 points: Contains 25–50 words
- 1 point: Contains 5–24 words or 51–60 words
- 0 points: Contains less than 5 words or more than 60 words. Written in capital letters, contains no punctuation or consists only of bullet points
The task instruction also just says to write a short 25–50 word summary. One sentence or two, Form is not affected. The rule that it must be one sentence joined with and / but / so comes from a different PTE test, not PTE Core.
The other traits are Content, Grammar and Vocabulary, each worth 0–2. Squeezing everything into one long sentence often creates grammar errors. Choose one or two sentences based on a single question: which is less likely to go wrong for you?
Worked example: cutting preschool funding.
Cutting funding from Ontario's community preschool programs will negatively affect early childhood educators and their families, so the city should reconsider this budget decision before it takes effect.
28 words, inside 25–50, so Form scores 2. It covers both the effect and the recommendation, so the content holds up. Writing it as two sentences ("...their families. Therefore, the city should...") would also score 2 for Form, as long as both sentences are correct.
The method: read the whole passage and find the main idea, then find the 2–3 sentences that support it. Cut examples and descriptive detail, and compress into 25–50 words. The key step is paraphrasing rather than copying. Pearson's Test Tips say: "Try to paraphrase (restate the information in different words)." Copying long sentences from the passage makes it easy to go over the limit and does nothing for your vocabulary score.
Worked example: children and the outdoors. The passage says children used to spend a lot of time playing outside; now many parents just tell them to hurry up and not be late for class; children are no longer close to or comfortable with nature; they need more time to play and explore outdoors.
Although children used to spend much time playing outside, many parents now rush them to class, so they have grown distant from nature and need more time to explore outdoors.
30 words. It restates the passage in new words ("rush them to class" for "hurry up, do not be late for class"; "grown distant from nature" for "are not close to nature"), which is exactly the paraphrasing the tips ask for.
SWT worked example
This example runs the method from start to finish: read the passage, decide what to keep, write the summary, then compare it with three versions that lose marks. The templates lesson Summary frames for SWT and SST sets out the same rules side by side.
The passage
An original practice passage (188 words).
Many cities now plant trees along their streets, and the reasons go well beyond appearance. The most immediate benefit is shade. Paved roads and rooftops absorb heat during the day and release it slowly at night, so neighbourhoods without trees can stay uncomfortably warm long after sunset. A row of mature trees shades the pavement, cools the air through the water its leaves release, and makes walking in summer far more pleasant. Trees also help with rain. Their roots loosen the soil and their branches slow falling water, which means less runoff rushing into drains during a storm. Some residents worry that roots crack pavements and that falling leaves block gutters, and these costs are real. However, planners argue that careful choice of species and regular maintenance can keep such problems under control. There is a social side as well: many people find a green street more inviting to walk along, and busier pavements can make local shops livelier and neighbours more familiar with one another. For these reasons, street trees are increasingly treated not as decoration but as basic infrastructure that a city should plan and fund.
What to keep and what to drop
The main idea sits in the last sentence: cities treat street trees as infrastructure, not decoration. Everything you keep should support that idea.
| Keep | Drop |
|---|---|
| Main idea: street trees are planned and funded as infrastructure, not decoration | How paved roads store heat and release it at night |
| Trees cool hot streets | How leaves cool the air by releasing water |
| Trees reduce storm runoff | How roots loosen the soil and branches slow the rain |
| Trees encourage walking; the costs (cracked pavements, blocked gutters) can be managed | Livelier shops and friendlier neighbours as examples of the social side |
Model summary
Cities increasingly plant street trees as essential infrastructure rather than decoration, because they cool hot neighbourhoods, reduce storm runoff and encourage people to walk, while problems such as damaged pavements and blocked gutters can be controlled through sensible species selection and upkeep.
42 words, inside 25–50, so Form scores 2. The main idea comes first, the three benefits and the managed costs follow, and no run of five words is copied from the passage. "Because" and "while" carry the cause and the contrast.
Three common mistakes
1. Too long.
Many cities now plant trees along their streets for reasons beyond appearance, because paved roads and rooftops absorb heat during the day and trees shade the pavement and cool the air, their roots loosen the soil so less runoff reaches drains, and greener streets attract walkers, although residents worry that roots crack pavements and falling leaves block gutters.
58 words. This is outside 25–50; at 51–60 words Form drops from 2 to 1. It happens when you keep the drop column as well as the keep column, and the long copied runs ("paved roads and rooftops absorb heat during the day") do nothing for Vocabulary.
2. Broken sentences.
Cities increasingly treat street trees as essential infrastructure rather than decoration, they cool hot neighbourhoods, reduce storm runoff and encourage people to walk, and their drawbacks can be managed with sensible planning.
32 words, so Form is fine, but two complete sentences are joined with only a comma after "decoration". That is a grammar error and costs Grammar. Writing more than one sentence is not the problem: in PTE Core, two correct sentences are fine. Put a full stop after "decoration" and start the next sentence with "They cool".
3. Missing the main idea.
Trees shade the pavement and cool the air on summer evenings, their roots help rainwater soak into the ground, and their leaves can block gutters unless someone clears them regularly.
30 words, correct grammar, right length, but it only lists details. It never says why cities plant street trees or that they now treat them as infrastructure, so Content loses marks. Always write the main idea first, then add support.
Summarize Spoken Text: plan a word budget
A popular approach is a fixed three-part frame: "The lecture discusses… / It mentions that… / In conclusion, …". This course does not use it, for three reasons you can check yourself:
- It is expensive. Full marks allow only 20–30 words, and the frame alone uses 8 of them (The lecture discusses = 3, It mentions that = 3, In conclusion = 2). That is 20–40% of your response on words that carry nothing from the recording, taken straight out of your content budget.
- It pushes you to invent. To fill "In conclusion", people add a conclusion the speaker never made. It adds no marks and takes up words.
- The opening is often wrong. Recordings can be lectures, interviews, announcements or broadcasts. A fixed "The lecture discusses" is wrong whenever the recording is not a lecture.
Instead, split your 20–30 words like this:
| Part | Budget | Needed? | Notes |
|---|---|---|---|
| Main point (what the recording is mainly about) | 8–14 words | Required | Without it, content drops a band. It must be something the speaker actually said. |
| Supporting point 1 | 6–10 words | Strongly recommended | The main reason, detail or example behind the main point |
| Supporting point 2 | 6–10 words | Only if words remain | After the first two, count; add this only if 6 or more words are left |
| Frame words (The lecture / It mentions / In conclusion) | 0 words | Not needed | See above |
Write the main point, count it, and see what remains. If fewer than 6 words are left, write only one supporting point. A complete, correct sentence beats squeezing in a second point, and grammar and spelling are scored separately.
Keep linking words that carry meaning: because, although, so, while, rather than. They cost one or two words and show the speaker's cause and effect or contrast. Drop words that only fill space: if you can delete a word and the reader still knows what the speaker said, it should not use up your budget.
Worked example: job interviews. The recording says that in an interview, a manager wants your thoughts, emotions and story, not just answers to questions, and that being able to communicate them matters. The main point takes about 17 words, leaving room for one supporting point:
In a job interview, managers look for your thoughts, emotions and personal stories rather than simple answers, so the ability to communicate these experiences clearly matters most.
27 words. Only "rather than" and "so" are not content words, and both carry the speaker's logic.
Worked example: a park cleanup announcement. The speaker announces cleanup teams: one picks up rubbish, one checks for plants that do not belong, and one repaints the tables. This is an announcement, not a lecture.
The speaker announced the cleaning teams: one will collect rubbish, another will check for plants that do not belong, and a third will repaint the tables.
26 words, all carrying information. Two lines to hold: do not add conclusions the speaker did not make (the content band asks for "a good summary of the text", not your opinion), and match the opening to the recording type. The simplest option is to skip the recording type and start with the content.
Keep counting. At 31–40 words, Form drops from 2 to 1. And remember spelling: the Score Guide gives 2 points for one spelling or typing error, 1 for two or three, 0 for more than three. If you are unsure how to spell a word you heard, use a synonym you can spell.
Taking notes while you listen
Pearson's Test Tips suggest shorthand only for the spoken summary: "As you listen, you will need to take notes, but do not try to write everything you hear. Note only main and supporting ideas. Use abbreviations, symbols, and arrows to show relations between them."
That gives three instructions: do not write everything; note only the main and supporting ideas (the first rows of your budget); and record how ideas relate, not just the words. The relations are what let you write because, although and so later.
| What to note | How | Example (park cleanup) |
|---|---|---|
| Main point | In full; do not abbreviate this | teams / park cleaning |
| Each supporting point | First letters or first 2–3 letters | Kev → rubbish; Mar → plants; spk → tables |
| Relations | → result, ← cause, vs contrast, ≠ negative | Write ← when you hear because, vs when you hear but |
This works here because a whole recording holds more than short-term memory can. It is not useful for Write from Dictation, where a sentence is only about 9 words long.
SST worked example
There is no audio here: the transcript stands in for the audio. In the test you hear a 45–75 second recording once and cannot read it.
The transcript
An original practice talk (165 words).
Today I want to talk about why sleep matters for learning. Many students treat sleep as lost study time, so before an exam they stay up late and read their notes again and again. The trouble is that memory does not work like a notebook that simply fills up during the day. When you learn something new, the memory is fragile at first. During sleep, and especially during deep sleep, the brain appears to replay what you learned and move it into more stable, long term storage. So sleep is not a pause in learning; it is part of learning. This has a practical consequence. If you cut sleep to gain extra hours of revision, you may end up remembering less of what you studied, not more. A better plan is to spread your study over several days and protect a full night of sleep, particularly the night after you learn something important. In short, rest is when the work of remembering actually gets done.
Notes
Notes in the format from the section above: the main point in full, short forms for the rest, and symbols for how ideas relate.
| What to note | Note |
|---|---|
| Main point | sleep is part of learning, not lost study time |
| Supporting point | new mem fragile; deep slp → replay → stable storage |
| Supporting point | cut slp for revision → remember less |
| Advice | spread study vs cram; protect night after learning |
Model summary
Sleep is an essential part of learning, not wasted revision time, because the sleeping brain replays newly learned material and stores it more securely.
24 words, inside 20–30, so Form scores 2.
Why it works
- Main point first. "Sleep is an essential part of learning, not wasted revision time" is the speaker's central claim, taken from the first line of the notes.
- One supporting point. The replay and storage idea explains why. The advice about cramming is a second supporting point; adding it ("so cutting sleep to revise can reduce recall", 8 words) would push the total to 32, past 30, so it stays out.
- No frame words. There is no "The lecture discusses" or "In conclusion", so every word carries information.
- No invented conclusion. It does not add advice the speaker never gave, such as a number of hours to sleep.
- One link word carries the logic. "Because" shows the cause the speaker gave.
One method for both tasks, and hard form rules
Whether you read or listen, the steps are the same:
- Find the topic sentence, usually at the start or the end.
- Find 1–2 supporting points: examples, figures, reasons. Skip the extra detail.
- Remove repetition; keep repeated ideas once.
- Rebuild in your own words, then trim to the range: 25–50 for written, 20–30 for spoken.
The only difference is where you trim. Number of sentences is not a difference; neither task's Form band counts sentences. If you aim for 30–40 words on the written summary to stay safely inside 25–50, do not carry that habit over: 31–40 on the spoken summary drops Form to 1.
These form rules apply to both:
- No all-capitals. Writing in capitals scores zero for Form.
- Use punctuation. Missing basic punctuation, such as a full stop at the end, costs marks.
- No bullet points only. The zero band says "consists only of bullet points". For the spoken summary it also mentions very short sentences, so write full sentences.
These errors are rare, but when they happen Form goes to zero, which is never worth it. The final check covers three things: format (word count, punctuation, capitals), spelling (swap uncertain words for ones you can spell) and grammar (agreement, singular and plural, linking words).
Score your own summaries
As with Write Email, these checklists turn the official bands into something you can apply yourself. The written summary has four traits (8 points in total); the spoken summary has five (10 points, adding spelling).
Summarize Written Text (out of 8)
| Trait | Max | Official top band | How to judge it yourself |
|---|---|---|---|
| Content | 2 | Provides a good summary of the text. All relevant aspects mentioned | List the main idea and main aspects of the passage (usually 2–3) and find each in your answer. All there = 2; one or two missing = 1; main idea wrong or missing = 0. |
| Form | 2 | Contains 25–50 words | Count words; 25–50 = 2. Also confirm: not all capitals, punctuated, not bullet points. |
| Grammar | 2 | Has correct grammatical structure | Check agreement, number, tense and linking words. No errors = 2; errors that do not block meaning = 1. |
| Vocabulary | 2 | Has appropriate choice of words | Put your answer beside the passage and mark any run of 5 or more identical words. None, and accurate word choice = 2; copied runs or imprecise words that still make sense = 1. |
Summarize Spoken Text (out of 10)
| Trait | Max | Official top band | How to judge it yourself |
|---|---|---|---|
| Content | 2 | Provides a good summary of the text. All relevant aspects mentioned | Is the main point there? How many supporting points? Also check: did you add anything the speaker did not say? If so, deduct. |
| Form | 2 | The response contains 20–30 words | Count words; 20–30 = 2. This is the trait people lose most, so always count. |
| Grammar | 2 | Correct grammatical structures | As above. Errors here usually come from forcing two points into one long sentence. |
| Vocabulary | 2 | Appropriate choice of words | Do your words fit the topic? Swapping a word you cannot spell for a synonym is normal and protects your spelling mark. |
| Spelling | 2 | One spelling/typing error | Count wrong words, not letters. 1 or fewer = 2. |
The right-hand columns are our self-check guides based on the official descriptions, not official wording.
Four steps: count words first; score each trait; look at which row is not full marks rather than at the total; fix only that row (missing aspect: go back to the source; copied phrases: paraphrase; grammar: split the long sentence; spelling: swap the word).
Written summaries most often lose vocabulary, because the passage is right there and copying is easy. Spoken summaries most often lose form and spelling, because 20–30 words is a narrow range and you are spelling words you just heard. The fix for spelling is the same in both: if you are not sure, use a simpler word.
Self-check: score the worked examples in this lesson first. All three written model summaries should score 8/8 and all three spoken model summaries 10/10; if yours do not, adjust how you read the checklist. Then aim for 8/8 and 10/10 on your own summaries, three in a row for each task. If you lose the same trait every time, that is your real weak spot. Repeated content losses mean practicing how to find the main idea and its aspects, not learning more phrases.