Handling word scores
Last updated
Last updated
The word-level quality score helps you assess the pronunciation accuracy of a specific word within a sentence. This score provides insight into how well that particular word was pronounced compared to the expected standard.
The scale of the quality score can be interpreted from the scoring guide.
You can use the quality_score
at the word level to display the analysis of the spoken response in your app using the following color codes:
Score | Color | Description |
---|---|---|
90 - 100 | Green | Excellent. Native or native-like |
80 - 90 | Green | Very Good and clearly intelligible. |
70 - 80 | Orange | Good. Intelligible but with one or two evident mistakes. |
60 - 70 | Red | Fair. Possibly not intelligible with several evident mistakes. |
0 - 60 | Red | Poor and must be reattempted. |
For example, in the case shown below, the quality_score
for the word "walk" is 65. Consequently, it is displayed in RED in the question based on the test-taker's response, indicating several evident mistakes in pronunciation.