PTE Summarize Group Discussion Practice
34 original three-speaker discussions with exam timing. Listen and take notes, prepare for 10 seconds, then speak for 2 minutes. Get a practice estimate from the optional transcript or tick the key points you covered, then compare with a model summary.
- Free, no sign-up
- Original questions
- Exam timing
Question 1 of 34
Should lectures be recorded?
Speakers: Maya, Tom, Aisha. Each speaker uses a different voice or pitch.
Audio uses your browser's voices; each speaker gets a different voice or pitch where your device allows. Recordings never leave your browser. The result is a practice estimate from an optional transcript (on-device in most recent browsers, or the browser's own in Chrome and Edge) or from key points you tick yourself, not a Pearson score.
How Summarize Group Discussion works
Three people discuss a topic from student life, such as accommodation, assessment or campus services. The audio plays once, for up to 3 minutes. You get 10 seconds to prepare and then 2 minutes to summarise the discussion in your own words.
How this task is scored (practice estimate)
After each recording you can get a practice estimate. It is labelled as an estimate and is not a Pearson score. You choose how it is made before you start.
Transcript on (automatic)
- Content (key points):every discussion has a fixed list of key points (the topic, each speaker's position and the outcome), each with a few keywords. A key point counts as covered when one of its keywords appears in the transcript.
- Fluency (rough estimate): your pace in words per minute, counted from when the recording opened to when you pressed Stop, plus hesitation and repeated words. The comfortable range is our own rule of thumb, not a Pearson figure. Long pauses (over 1.5 seconds) are counted only with the on-device transcript, which supplies word timings; with the browser transcript only pace is estimated.
- Headline percentage: the share of key points covered, lowered a little by the fluency estimate. Full coverage with the weakest fluency estimate still shows 70 percent. It is on our own scale and is never turned into a 10 to 90 score.
Transcript off (self-checked)
- You listen to your recording and tick the key points you covered. The result is labelled a self-checked estimate, rests on your own judgement, and has no fluency part because no pace is known.
What is NOT measured
- Pronunciation is never measured. A transcript cannot judge sounds, stress or accent.
- Keywords cannot check that you attributed points to the right speaker, paraphrased instead of repeating, or linked the ideas, which the content review looks at in the real test. A good idea in different words can be missed, and a keyword in the wrong place can be counted.
- Speech recognition can mishear words, and the on-device model can quietly tidy up slips, so the estimate can run generous. The on-device transcript works in most recent browsers and the browser transcript in Chrome and Edge only. If neither works, the page says so and you use the tick option.
Results are saved only in this browser. See them on the progress page.
Note-taking system
Split your notes into three columns, one per speaker, and write only keywords:
- Top line: the topic, in 3–5 words.
- Each column:the speaker's position (for / against / mixed) plus one reason or example.
- Bottom line: the outcome: what they agreed, decided or left open.
Template for a 2-minute summary
- Topic: “The speakers discussed whether…”
- Speaker 1: “[Name] argued that… because…”
- Speaker 2 (with a relationship word): “[Name] disagreed / partly agreed, pointing out that…”
- Speaker 3: “[Name] took a middle position and suggested…”
- Development: “[Name] then added that…”
- Outcome: “In the end, they agreed to… / they concluded that…”
Useful relationship verbs: agreed with, built on, challenged, objected, conceded, proposed, compromised. Using them shows how the speakers relate, which Pearson highlights as part of content.
Common mistakes
- Repeating sentences word for word. Pearson notes that retelling without paraphrasing lowers your score.
- Losing track of who said what. Attribute every point to a speaker.
- Hesitating and self-correcting. Keep a steady pace; restarts waste time and hurt fluency.
Frequently asked questions
How long is Summarize Group Discussion?
You hear a discussion between three people lasting up to 3 minutes. When the audio ends you get 10 seconds to prepare and 2 minutes to speak. You can record only once, and you can take notes while listening.
How is Summarize Group Discussion scored?
It's scored on content, pronunciation and oral fluency. For content, Pearson looks for an accurate, thorough summary that paraphrases the discussion, says which speaker contributed what, shows how speakers relate to each other, and connects ideas clearly. Content is also reviewed by a human expert.
Does this page score my summary?
It gives a practice estimate, not an official score. With the optional transcript on, the page looks for each key point of the discussion in the transcript and estimates your pace (and long pauses too, with the on-device transcript). With it off, you tick the key points you covered and the result is labelled self-checked. Keyword matching cannot check who said what or how the speakers relate, which Pearson's content review looks at, and pronunciation is never measured.
Do I need to mention the speakers by name?
Pearson's guidance says strong answers include which speaker said which piece of information. If you didn't catch a name, refer to them clearly, such as 'the first speaker' or 'the student who disagreed'.
Which skills does it count towards?
Summarize Group Discussion is an integrated task that assesses listening and speaking.
Are these real exam questions?
No. Every discussion here is original, written to match the real task's style: three students talking about an aspect of university life.