This action research explores the integration of TALPer, an AI-based learning partner, into English instruction in a rural Taiwanese elementary school. The study focused on a fifth-grade class of eight students over one academic year, aiming to enhance engagement, personalize learning, and support language acquisition in a low-resource setting. Guided by Kemmis and McTaggart’s (1988) Action Research Cycle model, the teacher planned, implemented, observed, and reflected on AI use within weekly English lessons, applying the HiTech (4 Learning) framework (Ho, 2014; cited in Wei, 2023). Data sources included classroom observations, student work samples, teaching journals, and semi-structured interviews.
Findings highlight three benefits: (1) promoting self-directed, interest-based vocabulary learning beyond textbook lists; (2) providing adaptive scaffolding for mixed-ability learners in sentence and short-presentation production; and (3) creating low-anxiety opportunities for English speaking practice. Students responded positively to the novelty and interactivity of TALPer, showing increased motivation and confidence. Challenges including time constraints for training, students’ digital readiness, technical issues such as unstable connectivity, and reduced in-person interaction were also observed. Reflective strategies emphasized hybrid lesson design to balance AI with traditional pedagogy.

Authors:
I-Chia Chou, Wenzao Ursuline University of Languages, Taiwan
Yu-Hsin Chen, Kaohsiung Municipal Biling Primary School, Taiwan

About the Presenter(s)
Dr. I-Chia Chou is currrently an associate professor at Wenzao Ursuline University of Languages. Her currently studies focus on exploring AI-enhanced English teaching with a focus on elementary education and teacher prepartation.

See this presentation on the full schedule – Wednesday Schedule

Bookmark