Enhancing early childhood science literacy through deep learning–based instruction using simple coding games

Abstrak

This study addresses the need to strengthen science literacy in early childhood education in response to rapid technological and scientific development in the digital era. Science literacy in early years is essential for developing children's ability to think critically, solve problems, and understand basic scientific concepts through meaningful learning experiences. This study examined the effect of deep learning–based instruction using simple coding games on science literacy. Using a quantitative quasi-experimental nonequivalent control group design, 63 children aged 5–6 years from TK Dewi Kunti I and II were selected via cluster random sampling, with intact classes assigned to the experimental and control groups. The experimental group received coding-based deep-learning instruction, while the control group received conventional instruction. Data were analyzed using descriptive and inferential statistics. The results indicate that children who participated in deep-learning–based coding activities demonstrated better development of science literacy than those in conventional learning settings. The findings suggest that integrating simple coding games into a deep learning framework can be an effective and engaging strategy to support science literacy in early childhood education. This study recommends further research with broader samples and longer intervention periods to strengthen implementation in diverse educational contexts.