Laura Edwards explains how we evaluate a test taker's performance in a speaking test. Using examples from two contrasting candidates, she demonstrates why native-level pronunciation and fluency are not enough on their own, and why test takers need to fulfil the tasks and show high-level language skills. (Video & transcript) Read more
This document summarizes the two research studies for the Dynamic Speaking Test by Associate Professor Di-feng Chueh of the Department of Foreign Languages and Literature (DFLL) at Feng Chia University and Project Lecturer Lanasari Tan from the English Language Center of Ming Chuan University. (PDF) Read more
This document outlines the development of the AI-driven "Dynamic Speaking Test", which uses natural language processing and machine learning to automate English proficiency scoring. It highlights a validation process calibrated against IELTS and CEFR standards, achieving high reliability and strong correlation with human examiners. (PDF) Read more
This session explored how the CEFR is used to design objective assessments of a student's speaking abilities. It delved into key CEFR descriptors and how to break down the marking criteria for spoken performance. Additionally, the session explained how automated marking systems are trained using AI and how these AI-powered systems evaluate student responses against CEFR criteria to produce accurate and reliable results. (Video) Read more
This webinar focused on how the CEFR is used to objectively assess English speaking ability, specifically through the Dynamic Speaking Test. Attendees learnt to evaluate student responses against CEFR descriptors and discovered how AI is leveraged to create an accurate, reliable, and user-friendly speaking test. (Video) Read more