Interaction Effects Between Learner Characteristics and Dialogue Format in TTS Dialogue-Based Lessons

A research study investigated the impact of various dialogue formats in AI-generated, Text-to-Speech (TTS) based lessons on high school students' motivation, learning outcomes, and overall evaluation. The study specifically examined how learner characteristics, such as experiential learning style and critical thinking disposition, interact with three dialogue formats: teacher-student, student-student, and teacher-teacher. Utilizing a large language model (LLM) for dialogue generation and TTS technology, the research employed a repeated-measures design with 222 first-year high school students.

Why it matters

This research provides insights into optimizing AI-driven educational tools by understanding how different dialogue structures and learner profiles influence effectiveness. It informs the strategic development of personalized learning experiences, potentially enhancing engagement and outcomes in technology-mediated instruction.

Key insights

• The study explores the interaction effects between dialogue format and learner characteristics (experiential learning style and critical thinking disposition) in TTS dialogue-based lessons.
• Three dialogue formats were tested: teacher-student, student-student, and teacher-teacher, all generated by an LLM and using TTS technology.
• Metrics evaluated included learner motivation, learning outcomes, and overall evaluation.
• The research involved 222 first-year high school students in a repeated-measures design.
• Data analysis was conducted using linear mixed-effects models.
• The provided abstract does not detail the specific results, only that they were 'showed' but the sentence was cut off.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2608.20822