The Speaking components of Summit K12’s Connect to Literacy programs provide automated grading of student voice recordings. Students will submit audio recordings for specific activities and Summit K12 will grade those recordings using Artificial Intelligence and algorithms designed to follow the WIDA Speaking Scoring rubrics.
Students will be graded according to the 3 scoring categories:
- Discourse Level or Linguistic Complexity
- Sentence Level or Language Forms
- Word/Phrase Level or Vocabulary Usage
Discourse Level or Linguistic Complexity
The discourse level or linguistic complexity measures organization, syntax, cohesiveness of response, recording duration, and word count. Students are encouraged to develop a thorough, dense, and relevant response to increase their proficiency-level rating.
Sentence Level or Language Forms
The sentence level or language forms measure sentence patterns and grammatical structures. It also takes into consideration the usage of words and phrases that facilitate proper sentence structure. The use of unique words will also result in a higher proficiency level rating. Additionally, the use of terms that convey sophisticated, complicated ideas will result in higher proficiency level ratings.
Examples:
-
Coordinating conjunctions
-
Subordinating conjunctions
-
Correlative conjunctions
-
Relative pronouns
-
Conjunctive adverbs
Word/Phrase Level or Vocabulary Usage
Scoring is also based on the usage of precise vocabulary. Accuracy in the usage of words and phrases in the provided context will be measured. Usage of question-specific vocabulary and precise words will yield higher proficiency ratings.
Scoring also calculates the usage of high frequency words and high frequency phrases. The more that students correctly use multiple variations of these terms, the better they will score in this category.
Summit K12 AI uses a comprehensive approach to score responses based on WIDA standards. Although AI works a majority of the time, there may be infrequent outliers that would require review. We highly recommend teachers listen to student recordings to ensure the scores are accurate. To learn how to listen to a student’s response, visit our help article: How to Listen to a Student's Recording - WIDA.
If a teacher believes the AI score of a Speaking response is not accurate, they can overwrite the score by following the instructions listed in our help article: How to Overwrite AI Scoring on Recordings - WIDA.