With the advent of the Fourth Industrial Revolution, the utilization of artificial intelligence (AI) technology in the field of education is rapidly expanding. Particularly in the area of English education, innovative learning platforms such as speech...
With the advent of the Fourth Industrial Revolution, the utilization of artificial intelligence (AI) technology in the field of education is rapidly expanding. Particularly in the area of English education, innovative learning platforms such as speech recognition systems, chatbots, and personalized feedback tools are being continuously developed and applied in educational settings. Korea, characterized by an EFL (English as a Foreign Language) environment, faces structural constraints where learners have limited opportunities to experience authentic English speaking practice. As a solution to overcome these limitations in the educational environment, learning systems integrated with AI technology are gaining attention as a means to enhance learners' English communicative competence.
Based on this background, this study aims to design a model applicable to elementary school English speaking classes based on AI digital textbooks, explore ways to effectively enhance learners' English communicative abilities through this approach, and empirically verify the educational validity of the developed model.
Applying Richey and Klein’s (2007) Type II design and development research model, this study followed a three-phase process: (1) development of an initial instructional model and teaching strategies based on a literature review and expert interviews; (2) internal validation through two rounds of expert evaluation; and (3) external validation via classroom application in a real educational setting. The instructional model was grounded in the ESA (Engage-Study-Activate) approach and incorporated key functions of AI digital textbooks, including real-time feedback, individualized learning paths, automatic speech scoring, and interactive speaking tasks.
The model was implemented in six lessons with sixty-five fourth-grade students from four classes at an elementary school in Seoul. Pre- and post-tests measured learners’ communicative competence, which was evaluated in terms of content appropriateness, linguistic accuracy, fluency, and non-verbal communication. Additionally, in-depth interviews and surveys with both teachers and students were conducted to analyze their perceptions of the instructional model’s effectiveness, engagement, and areas for improvement.
The verification of the educational effectiveness of the elementary English speaking class model utilizing AI digital textbooks developed through this study confirmed that learners' English communicative competence was statistically significantly improved. Specifically, achievement was enhanced in the areas of speaking fluency, sentence construction ability, and accuracy. In-depth interviews with teachers revealed that this model was evaluated as meeting the practical needs of educational settings in terms of providing individualized education according to each learner's level, enhancing learning participation, and reducing teachers' workload.
These research findings empirically demonstrate that the educational utilization of AI digital textbooks can substantially contribute to improving learners' self-directed learning abilities and communicative competence through qualitative improvement of the teaching-learning process, rather than merely remaining as simple technological innovation. In particular, the real-time feedback and individualized learning content composition functions provided by the AI system served as key elements that systematically supported learners' speaking activities. Furthermore, by integrating elements of collaborative learning and project-based learning, authentic language use opportunities were provided to learners of various proficiency levels, thereby presenting a new direction for English speaking education.
Considering the limitations of this study and directions for future research, additional verification of the generalizability and long-term educational effects of the developed instructional model is necessary through expanded studies that include various grade levels, regional characteristics, and educational environment variables. The stable establishment and sustainable development of AI-based English education will be possible when comprehensive measures are prepared, including technological advancement of AI digital textbooks, establishment of training systems to strengthen teachers' expertise in AI utilization, and institutional support measures at the educational policy level.