Oral practice — teacher's guide

The student gets one sentence on the screen, can listen to it being read aloud, reads it aloud into the microphone themselves, and receives an assessment back after a few seconds. One sentence at a time, no conversation — short and repeatable.

How it works

The student opens Pronunciation and selects a level (A0–B2) and source under Practice. The source is either the pre-made level list or one of the student's own lists — the same lists found in Word lists. The sentences are pulled from the example sentences in the dictionary at the chosen level.

1. The sentence is displayed under Say this sentence. Listen to the sentence plays it aloud, so the student hears the target first.
2. The microphone button starts the recording (Tap to record), and the same button stops it (Tap to stop). The recording automatically stops after 30 seconds regardless.
3. Listen to recording plays back the student's own recording. Get feedback sends it for assessment.
4. Afterwards, the student can repeat the same sentence with Try again, or move on with Next sentence. Skip switches the sentence without attempting it.

What the feedback says

Under How it went, there is an overall score from 0 to 100 and a short status text (Great job!, Almost there, Try again, or Unclear recording), plus three subscores: Correct words, Pronunciation, and Fluency.

• What Språki heard is a transcription of what the student actually said — not the target sentence. This is often the most useful part: here the student sees in black and white which words were dropped or turned into something else.
• Try this next time is one specific piece of advice for the next attempt.
• Word by word shows each word as a colored tile. If the student taps a tile, a short note appears. The labels are Good, Almost, Missing, Unclear, Different word, and Extra word.

The assessment is performed by a language model that listens to the recording and makes an informed evaluation of what it hears. This means it is a qualified assessment, not a phonetic measurement: the score may vary slightly between two fairly similar attempts, and background noise or a weak microphone will lower the score. Use the number as a general indicator over time, not as a precise measurement of a single attempt.

Feedback is written in the student's native language at A0–A2. From B1 onwards, it is mainly in the language the student is learning.

Who it is for

Pronunciation works best from A1 to B2, where the student can read a sentence without sounding out every letter. A0 works as well — short sentences are then pulled from the A0 and A1 material — but absolute beginners will often need you to read along with them the first time.

Two things are worth clarifying to the students: this measures reading a given sentence aloud, not free or spontaneous speech, and the score is practice support, not a test. It is not a certified pronunciation test and should not be used as a basis for grades.

Suggestions for working with oral practice

1. Assign it as homework. Adult learners speak more freely alone than in front of the class, and at home there is neither an audience nor classroom noise.
2. If you have headsets with microphones, the entire class can practice simultaneously without overhearing each other. Without headsets, the hallway, a breakout room, or pairs at opposite ends of the room work well.
3. Link the session to the vocabulary you are working on: create a list in Word lists, share it with the group, and ask the students to select it under Practice.
4. Ask for two attempts on each sentence. Between attempts: listen to Listen to the sentence, then Listen to recording, and identify the difference. The progress between attempt 1 and 2 is the main point, not the number.
5. Focus on one sound at a time. Set up a list of words containing the target sound, and have the students complete five sentences using that list.
6. Use it as a five-minute warm-up or wrap-up with a fixed goal, such as five sentences. Short and frequent sessions yield better results than one long session.

Good to know

• Audio recordings are not stored. The recording is sent for assessment and discarded afterwards. What remains stored on the student's account is the transcription, the four scores, and the written feedback — and all of this is tied to the account and deleted with it.
• Time counts toward learning time. Pronunciation is categorized under "Speaking" in Progress, along with the number of attempts and the average score.
• The microphone must be allowed in the browser. The first time, the student must click "Allow" when prompted by the browser. If issues occur, see Microphone is not working.
• Pronunciation and Conversation address two different things. Conversation is a role-play conversation using voice and free responses — great for fluency and building confidence to speak. Pronunciation is short, precise practice on a single sentence with clear feedback on every word. Many students benefit from both: Pronunciation for precision, Conversation for extensive practice.

See also

• How do I practice speaking? — the student version, which you can forward
• Microphone is not working
• For teachers