Loading...
Teaching with Conversational AI: Exploring Human-AI Collaboration and the Beliefs and Experiences of Language Teachers
Citations
Altmetric:
Genre
Thesis/Dissertation
Date
2025-08
Advisor
Committee member
Group
Department
Teaching & Learning
Permanent link to this record
Collections
Research Projects
Organizational Units
Journal Issue
DOI
Abstract
While generative artificial intelligence (GenAI) tools such as ChatGPT have become increasingly accessible in education, their classroom use, particularly for speaking instruction in second language education, remains underexplored. Most existing research has emphasized learner autonomy and out-of-class applications, overlooking the teacher’s evolving role in facilitating AI-supported instruction. Yet, as language learning is inherently social and interactional (Long, 1996; Vygotsky, 1978), understanding teacher-AI collaboration is critical. Further, as Baker (2016) argues, the intelligence in AI-supported education should ultimately reside in the teacher, not in the system. This highlights the importance of investigating how language teachers actively shape their roles when working with conversational AI tools in the classroom.This multi-paper dissertation addresses these gaps through three interrelated studies. Grounded in sociocultural theory and interaction theory, the research explores how language teachers engage with conversational AI tools in speaking-focused classrooms. Study 1 presents a systematic review of empirical studies, identifying gaps in understanding how conversational AIs and human teachers collaborate. Study 2 uses social network and content analysis to investigate teacher-AI interaction in classrooms using Google Assistant. Study 3 draws on multiple-case study methods to examine how three teachers with diverse backgrounds implemented ChatGPT for speaking instruction in a community-based language program.
Findings across studies show that GenAI reshapes, rather than replaces, teachers’ instructional roles. Teachers employed varied facilitation strategies and adapted to affective and contextual constraints while maintaining pedagogical authority. These practices were shaped by teachers’ prior experiences, technological familiarity, and participation in structured, reflective professional development. Together, the findings contribute to theory and practice by illustrating conversational AI as a dynamic, co-mediating agent in language classrooms, where teacher agency remains central to meaningful integration. Future research should continue to explore and expand understanding of how teachers engage with emerging AI systems, how orchestration unfolds in multi-agent learning environments, and how AI-mediated instruction can be designed to support human-centered design and context-responsive pedagogy across diverse learning contexts.
Description
Citation
Citation to related work
Has part
ADA compliance
For Americans with Disabilities Act (ADA) accommodation, including help with reading this content, please contact scholarshare@temple.edu