TalkScore is an innovative tool in the dynamic world of language assessment, offering precise evaluations of spoken English proficiency. With it, recruiters can assess candidates' English skills through three audio questions: Open-ended, Listen and Repeat, and Listen and Speak.
Open-ended Questions
Measured parameters: Fluency, Grammar, Pronunciation, and Vocabulary,
For Open-ended Questions, candidates can freely answer a question asked via text. For these, TalkScore analyzes the audio recording using the question as context, providing individual scores for the four measured parameters, from 1 to 10. Other parameters will have the value "N/A" since they won't be considered for this question type.
Setup: Open-ended Question Type
Go to Campaign Settings > Questions.
Select an audio response type question, then click Edit.
On the TalkScore Question Type options, select Open-Ended.
Click the Timer button, and set the minutes for candidates to answer.
Then click Save.
Listen and Repeat Questions
Measured parameters: Pronunciation, Fluency, and Active Listening.
For Listen and Repeat questions, TalkScore needs a transcript of the expected answer and the candidate's recorded audio to generate a score. It compares the two and provides a similarity score, which serves as the Active Listening score.
💡 Remember to include a Play Media question before, with the audio your candidates will repeat.
Setup: Listen and Repeat Question Type
1. Go to Campaign Settings > Questions. Select an audio response type question, then click Edit.
2. On the TalkScore Question Type options, select Listen and Repeat.
3. On the Audio Transcript for scoring, submit a transcript of the exact audio your candidates should repeat.
4. Click the Timer button, and set the minutes for candidates to answer.
5. Then click Save.
Once completed by the candidate, you’ll see the three parameters’ scores in the Candidate Profile, all displayed on a 0 to 10 scale.
Listen and Speak Questions
Measured parameters: Pronunciation, Vocabulary, Fluency, Grammar, and Comprehension.
The process for Listen and Speak questions mirrors that of open-ended questions. TalkScore analyzes the audio recording and the question’s transcript as context to evaluate how well the candidate understands the question and stores this as the Comprehension parameter.
💡 Remember to include a Play Media question before, with the audio your candidates will repeat.
Setup: Listen and Speak Question Type
Go to the Campaign Settings > Questions. Select an audio response type question, then click Edit.
On the TalkScore Question Type options, select Listen and Speak.
On the Audio Transcript for scoring, submit a transcript of the audio question played to your candidates.
Click the Timer button, and set the minutes for candidates to answer.
Then click Save.
Once completed by the candidate, you’ll see the three parameters’ scores in the Candidate Profile, all displayed on a 0 to 10 scale.
