Assessing Scientific Thinking in Early Childhood: Cross-Sectional Evidence for a Six-Dimensional Hierarchical Structure in Indonesian Preschoolers

Purpose – Scientific thinking in early childhood remains understudied in non-Western contexts, with many existing models derived from Western samples and limited to two or three dimensions. This study examines a six-dimensional hierarchical framework proposing that domain-general cognitive capacities (Attention & Focus, Working Memory, Problem Solving) support domain-specific scientific competencies (Observation Skills, Prediction & Reasoning, Experimentation) in Indonesian preschoolers aged 4–6 years. Age-related patterns, gender differences, and institutional-type differences are also investigated.Design/methods/approach – Using a quantitative cross-sectional design, data were collected from 105 children (4–6 years) enrolled in secular and Islamic early childhood education institutions in South Sulawesi, Indonesia. Scientific thinking was assessed using the Scientific Thinking Assessment for Early Childhood (STAEC), a 25-item teacher-rated instrument developed through expert review and pilot testing with 30 teachers. Analyses included descriptive statistics, reliability analysis, Pearson correlations, ANOVAs, and t-tests to evaluate interdimensional relationships and group differences.Findings – Results provided initial cross-sectional evidence consistent with a six-dimensional hierarchical organization of early scientific thinking. Domain-general capacities were strongly intercorrelated (r = .796–.831) and showed higher mean scores than domain-specific competencies, suggesting a foundational role. Working memory displayed the strongest associations with advanced competencies, particularly prediction & reasoning and experimentation. A significant age-related difference emerged only for observation skills, whereas other dimensions showed non-significant developmental trends. No gender differences were observed across any dimension, and no differences emerged across secular and Islamic institution types.Research implications/limitations – The cross-sectional design limits developmental and causal inferences. Teacher ratings may introduce rater bias and do not capture moment-to-moment reasoning processes. The single-region sample constrains generalizability; future research should use longitudinal, larger, multi-region, and multi-method designs.Practical implications – Early childhood programs should strengthen foundational cognitive capacities while providing explicit, developmentally appropriate support for prediction and experimentation, and maintain equal learning expectations across genders and educational settings.Originality/value – This study offers initial empirical support for a multidimensional hierarchical model of early scientific thinking in a non-Western context, including secular and Islamic early childhood education settings.Paper type Research paper