Speaker
Description
This research investigates a hybrid AI-human assessment model using Scribo and Progos to evaluate English proficiency at the tertiary level in Japan. Findings from a mixed-methods study (N = 41) reveal significant improvements in speaking and writing proficiency, with students advancing to the B2 level of the CEFR. While AI enhances learner autonomy and provides immediate feedback, the results highlight the need for human evaluation to assess cultural nuance and complex discourse, thereby ensuring a balanced and practical language assessment approach.
Summary
This presentation explores the integration of AI-powered tools, Scribo and Progos, in English language assessment at the tertiary level in Japan. The study demonstrates how AI-driven feedback enhances learner autonomy and proficiency while highlighting the necessity of human evaluation for higher-order skills. Participants will gain insights into implementing a hybrid AI-human assessment model to optimize learning outcomes, ensuring a balance between technological efficiency and pedagogical expertise in language education.
| Teaching Context | College and university education |
|---|
