Dialogue Act Patterns in GenAI-Mediated L2 Oral Practice: A Sequential Analysis of Learner-Chatbot Interactions
arXiv:2604.05702v1 Announce Type: new Abstract: While generative AI (GenAI) voice chatbots offer scalable opportunities for second language (L2) oral practice, the interactional processes related to learners' gains remain underexplored. This study investigates dialogue act (DA) patterns in interactions between Grade 9 Chinese English as a foreign language (EFL) learners and a GenAI voice chatbot over a 10-week intervention. Seventy sessions from 12 students were annotated by human coders using a pedagogy-informed coding scheme, yielding 6,957 coded DAs. DA distributions and sequential patterns were compared between high- and low-progress sessions. At the DA level, high-progress sessions showed more learner-initiated questions, whereas low-progress sessions exhibited higher rates of clarification-seeking, indicating greater comprehension difficulty. At the sequential level, high-progress sessions were characterised by more frequent prompting-based corrective feedback sequences, consist
arXiv:2604.05702v1 Announce Type: new Abstract: While generative AI (GenAI) voice chatbots offer scalable opportunities for second language (L2) oral practice, the interactional processes related to learners' gains remain underexplored. This study investigates dialogue act (DA) patterns in interactions between Grade 9 Chinese English as a foreign language (EFL) learners and a GenAI voice chatbot over a 10-week intervention. Seventy sessions from 12 students were annotated by human coders using a pedagogy-informed coding scheme, yielding 6,957 coded DAs. DA distributions and sequential patterns were compared between high- and low-progress sessions. At the DA level, high-progress sessions showed more learner-initiated questions, whereas low-progress sessions exhibited higher rates of clarification-seeking, indicating greater comprehension difficulty. At the sequential level, high-progress sessions were characterised by more frequent prompting-based corrective feedback sequences, consistently positioned after learner responses, highlighting the role of feedback type and timing in effective interaction. Overall, these findings underscore the value of a dialogic lens in GenAI chatbot design, contribute a pedagogy-informed DA coding framework, and inform the design of adaptive GenAI chatbots for L2 education.
Executive Summary
This study examines dialogue act (DA) patterns in interactions between Chinese EFL learners and a GenAI voice chatbot over a 10-week intervention. Analyzing 6,957 coded DAs from 70 sessions, the research identifies distinct interactional patterns associated with high- versus low-progress sessions. High-progress sessions featured more learner-initiated questions and prompting-based corrective feedback sequences, while low-progress sessions showed higher rates of clarification-seeking. The findings highlight the critical role of feedback timing and type in effective L2 oral practice, offering insights for designing adaptive GenAI chatbots in education.
Key Points
- ▸ The study employs a pedagogy-informed DA coding framework to analyze learner-chatbot interactions, providing a structured approach to understanding dialogue dynamics in GenAI-mediated L2 learning.
- ▸ High-progress sessions are characterized by learner-initiated questions and corrective feedback sequences positioned after learner responses, suggesting these elements are pivotal for language acquisition.
- ▸ Low-progress sessions exhibit higher rates of clarification-seeking, indicating comprehension difficulties and the need for targeted pedagogical interventions.
Merits
Methodological Rigor
The study uses a robust, human-coded annotation scheme with a substantial dataset (6,957 DAs), ensuring reliability and depth in the analysis of interactional patterns.
Pedagogical Relevance
The research integrates pedagogical insights into its coding framework, making the findings directly applicable to L2 education and GenAI chatbot design.
Novelty in Sequential Analysis
By examining sequential patterns, the study goes beyond static DA distributions to uncover dynamic interactional processes that influence learning outcomes.
Demerits
Limited Generalizability
The study focuses on Grade 9 Chinese EFL learners, which may limit the applicability of findings to other age groups, language proficiency levels, or cultural contexts.
Potential Bias in Annotation
Despite human coding, subjectivity in annotating DAs could introduce bias, particularly in interpreting the intent behind learner responses.
Short Intervention Period
A 10-week intervention may not capture long-term effects of GenAI-mediated interactions on language acquisition or retention.
Expert Commentary
This study makes a significant contribution to the evolving field of GenAI-mediated language learning by applying a dialogic lens to analyze learner-chatbot interactions. The findings underscore the importance of feedback timing and learner agency in fostering effective L2 oral practice. Notably, the sequential analysis reveals that high-progress sessions are characterized by a dynamic interplay between learner initiative and targeted feedback, a pattern that aligns with contemporary theories of sociocultural learning. However, the study's focus on a specific cohort raises questions about its broader applicability. Future research should explore the transferability of these findings across different linguistic and cultural contexts, as well as the long-term impact of GenAI-mediated interventions on language proficiency. Additionally, the ethical dimensions of using GenAI in education warrant further scrutiny, particularly in terms of ensuring equitable and unbiased interactions.
Recommendations
- ✓ Future research should replicate this study with diverse cohorts to validate and extend the findings, ensuring broader applicability across educational contexts.
- ✓ Developers of GenAI chatbots should prioritize adaptive feedback mechanisms that respond dynamically to learner needs, incorporating insights from this study on DA patterns and sequential interactions.
- ✓ Educational institutions should establish ethical guidelines for the use of GenAI in language learning, addressing issues of data privacy, transparency, and fairness in AI-generated feedback.
Sources
Original: arXiv - cs.CL