<![CDATA[Abstract Standardized formative assessment instruments for early childhood science education remain scarce, a gap that can lead to observational bias. This study aimed to establish the validity and reliability of a formative assessment instrument designed for a STEAM experiment specifically, a volcanic eruption activity at the foundational education level. A Research and Development (R&D) approach was employed, focusing on the instrument's validity and reliability. Participants included children aged 5–6 years from Group B of the Lembut Early Childhood Education (PAUD) center in Kadudampit, Sukabumi, selected via total sampling. Data were collected using structured, performance-based observation sheets and analyzed using the Pearson Product-Moment formula and Cronbach's Alpha. The results demonstrated that six statement items were empirically valid and that the instrument as a whole exhibited high reliability. Instances of biased observation items were attributed to the presence of adult guidance during the activity and the tendency for children to mimic each other's responses during group sessions. Following the elimination of certain items, the resulting assessment instrument is now concise and objective. Keywords: Assessment, STEAM Experiments, Foundation Phase, Science]]>
Copyrights © 2025
