DEVELOPING CAUSAL REASONING THROUGH SCIENCE EXPERIMENTS: A LANDSLIDE SIMULATION IN EARLY CHILDHOOD

Abstract

Science learning in early childhood requires experiences that help children move from describing phenomena toward explaining causal relationships. This study aimed to examine the use of a landslide-simulation science experiment to develop the causal reasoning of Group B children at RA Nurus Salam Silo Jember. Previous studies have largely emphasized general science skills, process skills, or logical development, leaving causal reasoning through disaster-based experimentation less specifically addressed. The study contributes a contextual model linking concrete experimentation, causal inquiry, and early environmental awareness. The study employed Classroom Action Research (CAR) based on the planning, acting, observing, and reflecting cycle. The participants were 10 children aged 5–6 years in Group B during the 2025/2026 academic year. The intervention was conducted in two cycles, with three meetings in each cycle. Data were collected through structured observation, teacher observations, interviews, field notes, and documentation, focusing on four causal-reasoning indicators: identifying causes, explaining consequences, connecting cause and consequence, and identifying preventive actions. The findings showed a progressive improvement in causal reasoning, with the class mean increasing from 40.0% in the pre-cycle to 65.9% in Cycle I and 86.5% in Cycle II. Classical mastery increased from 20% to 53% and 90%, respectively. Children also progressed from descriptive responses toward more independent explanations of causes, consequences, relationships, and prevention.