5.1 Setup and Speaking to the ASR
Key Takeaways
- Versant is scored by an automatic speech recognizer, so a clean audio signal from a boom-mic headset in a quiet room directly protects your score
- Position the microphone to the side of your mouth, about two finger-widths away, to avoid plosive pops on p, b, and t sounds
- Speak at a normal volume and steady pace; over-articulating or whispering both lower Fluency and Pronunciation scores
- Start speaking within about 6-8 seconds on Read Aloud, Repeat, Short Answer, and Sentence Builds (no beep); wait for the beep only on Story Retellings and Open Questions
- You cannot go back or re-record once an item starts, so use each part's sample item to confirm the task first
Your score is decided before you say a word
The Versant English Test is scored by an automatic speech recognizer (ASR) — a machine that measures the sound signal your microphone captures. It does not "listen" the way a human examiner does. It cannot lean in, ask you to repeat yourself, or give you the benefit of the doubt. If your headset drops a consonant or a fan smears your vowels, the engine scores the corrupted signal, not the English you actually produced. A capable B2 speaker on a laptop's built-in mic in a noisy kitchen can post an A2-looking result. Setup is not housekeeping you do before the "real" test — for an ASR-scored exam, setup is part of the test.
So the first job on test day has nothing to do with grammar or vocabulary. It is to hand the engine a clean, loud, undistorted copy of your voice.
Equipment checklist
Work through this list before you enter your Test Identification Number (TIN), because you cannot go back and re-record once the test starts:
- Use a headset with a boom microphone, not your device's built-in mic. The boom keeps the mic a fixed distance from your mouth even when you move.
- Position the mic to the side of your mouth, roughly two finger-widths away — near the corner of your lips, not directly in front of them. Straight-on placement makes "p", "b", and "t" sounds burst into the mic (plosive "pops") that the ASR reads as noise.
- Prefer a wired headset over Bluetooth. Wireless can compress or drop audio; a cable does not.
- Confirm your delivery mode. Pearson delivers the test on a computer or a smartphone, depending on what your administrator chose. On a computer you may use an up-to-date browser or Pearson's Computer Delivered Testing (CDT) software on Windows.
- Plug in or fully charge the device so it does not sleep or throttle audio mid-test.
- Test your internet. The test is taken online and autoscored; a stable connection prevents the audio upload from stalling.
The room
Pearson is explicit that background noise degrades scoring — the official guidance warns that even loud fans and air conditioning can cause interference. Treat the room as part of your equipment:
- Choose a quiet room where no one will interrupt you for the whole test — about 17 minutes for the speaking Versant English Test, and about 50 minutes if you have been assigned the 4-skill Placement Test.
- Close windows and doors, silence phone and computer notifications, and tell anyone nearby not to talk to you until you are done.
- Avoid rooms with a lot of echo. Soft furnishings — curtains, a rug, a sofa — absorb reflections; a bare tiled bathroom is the worst possible choice.
Use the system check and the sample items
Each part of the test opens with an instruction screen and a sample item before the scored questions begin. This is a gift: it lets you confirm the task type and hear the examiner's voice at the volume the engine expects. During the initial audio and microphone check, speak at the exact volume you plan to use for the test — not louder, not softer — so the level the system sets matches your real answers.
How to speak so the machine hears you
The engine is trained on natural human speech, so your job is to sound like a clear, ordinary speaker — not a robot, and not a whisper.
| Do | Don't | Why it matters |
|---|---|---|
| Speak at a normal volume | Shout or whisper | Both push the signal outside the range the ASR expects |
| Speak at a normal, steady speed | Rush, or crawl word-by-word | Over-slow, over-careful speech reads as disfluent and unnatural |
| Pronounce final consonants clearly | Let word endings trail off | Dropped endings cost you phonemes the engine is counting |
| Keep a natural rhythm | Over-articulate every syllable | Robotic delivery lowers Fluency and Pronunciation scores |
The paradox worth remembering: trying too hard to "enunciate" usually lowers your score, because it destroys the natural rhythm and timing the engine rewards. Speak the way a confident colleague would in a normal conversation.
Why does Pearson recommend a headset with a boom microphone rather than a laptop's built-in microphone for the Versant test?
Timing: when to start, and the beep
The most common avoidable mistake on setup day is misunderstanding when to begin speaking. It differs by task:
- On Reading (Read Aloud), Repeat, Short Answer Questions, and Sentence Builds, there is no beep. The prompt or timer starts, and you must begin speaking within about 6 to 8 seconds or the test assumes you are not answering and moves on. Do not wait for a signal that never comes.
- On Story Retellings and Open Questions, you do wait for a beep. You hear the story or question, then a beep marks the start of your response window (30 seconds for a retelling, 40 for an open question), and a second beep marks the end.
So the rule is not simply "wait for the beep." It is: start promptly on the short tasks; wait for the beep only on the two long-form speaking tasks. Once you begin, you cannot go back or re-record, so commit to your answer and keep going.
On the Repeat task, you hear the sentence and then wait. What should you do?
During the initial microphone check and while answering, how should you speak for the best automated score?