Description
This presentation is sponsored by ME Learning
Simon Moran presents a teacher's account, via their classroom data, of using AI avatars for TOEFL speaking practice: how they discovered the AI solution, what students did, what the teacher found useful, students' comments. Plus: open questions on cost, access and assessment, and an introduction to the MEQ, a universal, portable, lifelong assessment tool, and Language Lab - MEQ-informed content generation for any learner in any context.
Abstract section 3: Content/method
Over 40 students in a TOEFL preparation class practised TOEFL-type speaking activities with AI avatars, first using supplied prompts, then writing their own. The teacher also built her own prompts in a general chatbot. Students compared an animated avatar with a human-like one. The approach was then extended to other classes. Evidence consists of teacher observation, student feedback, platform usage data], not controlled testing.
Abstract section 1: Relevance
Speaking is a difficult skill to practise at scale: large classes limit each student's output and interaction time, both central to second language development (Long, 1996; Swain, 1985). Chatbots have been reviewed as promising but under-evaluated in classrooms (Huang et al., 2022), and learner interest may fade as novelty wears off (Fryer et al., 2017). Newer AI avatars add voices, faces and customisable prompts, yet teacher accounts of classroom use, particularly for test preparation, remain scarce. This presentation offers one such account.
Abstract section 4: Outcomes/results
Participants will see how the tools were used in class, examples of student-designed prompts, and student reactions to the two avatar types, with the caveat that the comparison was affected by the order of introduction. We report what worked and what did not, including engagement, availability, and technical and cost issues. Results are teacher-reported and exploratory; no test-score gains are claimed. Participants will leave with practical prompt-design tips, questions to ask their institutions, and open questions about measuring improvement.
Abstract section 5: References
Fryer, L. K., Ainley, M., Thompson, A., Gibson, A., & Sherlock, Z. (2017). Stimulating and sustaining interest in a language course: An experimental comparison of chatbot and human task partners. Computers in Human Behavior, 75, 461–468.
Huang, W., Hew, K. F., & Fryer, L. K. (2022). Chatbots for language learning—Are they really useful? A systematic review of chatbot-supported language learning. Journal of Computer Assisted Learning, 38(1), 237–257.
Long, M. H. (1996). The role of the linguistic environment in second language acquisition. In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of second language acquisition (pp. 413–468). Academic Press.
Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. Gass & C. Madden (Eds.), Input in second language acquisition (pp. 235–253). Newbury House.
Abstract section 2: Contribution/research questions
This is a practice-oriented presentation. Its purpose is to give an honest teacher's account of using AI avatars for speaking practice in a TOEFL preparation class, and to help other teachers decide whether, and how, to try them. It addresses three practical questions: how can students design effective prompts, how do animated and human-like avatars compare in use, and what problems should teachers anticipate?
| Title | Speaking Practice at Scale. Class TOEFL Prep with AI Avatars |
|---|---|
| Teaching Context | College and university education |