skip to main content skip to footer

Exploring the Provision of Artificial Intelligence–Based Feedback for TOEFL Junior Writing Practice Tasks TOEFL EFL AWE AD YL AI

Author(s):
Wolf, Mikyung Kim; Suhan, Michael; Alderman, Jean
Publication Year:
2024
Report Number:
RM-24-05
Source:
ETS Research Memorandum
Document Type:
Report
Page Count:
34
Subject/Key Words:
TOEFL Junior, Artificial Intelligence, Writing Skills Test, Assessment Development (AD), English Language Skills, Young Learners (YL), Automated Writing Evaluation (AWE), Automated Feedback, English as a Foreign Language (EFL), South Korea, Turkey, Prototype, Cut Points, Usability Studies, Children's Perceptions, Instructors' Perceptions

Abstract

This research memorandum reports on an exploratory study of developing an artificial intelligence (AI)–based feedback prototype tool and examining its usability for potential TOEFL Junior Writing test users. The recently launched TOEFL Junior Writing test employs ETS’s automated writing evaluation (AWE) tool to score test takers’ responses. Utilizing the ETS’s AWE engine and an existing database, we designed an approach to feedback that was closely linked to the key writing dimensions and proficiency descriptors as outlined in the TOEFL Junior Writing rubrics. A small-scale usability study was conducted with 14 students and seven teachers from South Korea and Turkiye. The participants perceived the usefulness of the feedback tool positively, but they also provided practical suggestions to further improve the tool. The promises and limitations of the tool are discussed.

Read More