Fostering AI Literacy in Early Childhood Education Through STEAM: Instructional Model Design, Course Effect and Teaching Strategy

In the age of Artificial Intelligence (AI), fostering AI literacy is paramount. However, promoting AI literacy in young children is rarely mentioned. This study proposes and validates the ECTRS five-stage instructional model (Explore, Create, Test, Reflect, Share) to cultivate AI literacy in young children within Early Childhood Education (ECE). The model was implemented in an 8-lesson LEGO WeDo 2.0 coding course over two weeks with children aged 4 to 6. Data for this research were collected from multiple sources, including classroom observations, artifacts, and teacher's evaluations. Empirical results show a significant improvement in children's classroom performance, thereby demonstrating the feasibility and effectiveness of the model. The children's performance differences were more pronounced in the Test, Create, and Share stages compared to the Explore and Reflect stages. Although children exhibited gains in AI knowledge and skills, improvements in their AI attitudes were modest. This study also identifies four key evaluation dimensions for young children's coding classes: construction ability, coding competence and comprehension, creativity, and attention/learning attitude. Despite the limited sample size, this study provides empirical evidence and pedagogical insights for developing AI literacy and informing instructional design in ECE.