Early Childhood Numeracy Practice Model Through Deep Learning

Early numeracy refers to the ability to think critically using mathematical concepts, procedures, facts, and tools to solve everyday problems in various relevant contexts for early childhood. This study aims to develop early numeracy practices for young children through deep learning approaches. The subjects of this research are 60 Kindergarten students, divided into two developmental stages: 4–5 years old and 5–6 years old. The study employed a development research method using the ADDIE model, which consists of Analysis, Design, Development, Implementation, and Evaluation. Data was collected through observations, interviews, documentation, and numeracy performance assessments tailored to the children’s cognitive stages. The data were analyzed using both qualitative techniques to evaluate the effectiveness and developmental appropriateness of the numeracy practices. The result of this study is a validated early numeracy product that aligns with children's levels of cognitive development: knowing, applying, and reasoning. This product is adapted to each age group to ensure developmental suitability. The novelty of this research lies in the implementation of early numeracy through deep learning, which integrates meaningful, mindful, and joyful learning strategies to enhance young children’s mathematical understanding. Future research is recommended to explore the scalability of this model in various cultural and socio-economic contexts, as well as to examine its long-term impact on children's mathematical thinking beyond the kindergarten years.