생성형 AI 활용 경험에 따른 유아교사의 그림책 읽기 질문과 AI 활용 인식 분석

초록

This study explored the distribution of question types used by early childhood teachers during picture book reading activities and their perceptions of generative AI use. Question types included literal, inferential, applied, and critical questions, and perceptions of AI use focused on usefulness, challenges, limitations, and needs. Participants consisted of eight in-service teachers in charge of 4-year-old classrooms, who were split into two groups of four: a generative AI use group and a non-use group. From August 28 to November 30, 2025, a total of 80 picture book reading sessions (ten per teacher) were conducted using the same ten picture books. All sessions were transcribed. The activities were examined across three stages introduction (cover exploration), reading, and post-reading, and the frequency of open-ended questions was analyzed. Two rounds of semi-structured interviews were also conducted with the four teachers in the generative AI use group. The results showed that the generative AI use group used inferential questions more frequently across all stages and applied critical questions more, particularly in the post-reading stage, whereas the non-use group tended to focus more on literal questions. Interviews also showed both the usefulness and limitations of generative AI in supporting teachers’ preparation.

키워드