Analysis of Cognitive Strategies Utilized by Preschoolers in Relative Positioning Tasks by Integrating Think-Aloud Protocols, Eye-Tracking Measures, and Performance Data

This study investigates preschoolers’ cognitive strategies in relative position reasoning tasks through an integrated mixed-methods design combining think-aloud protocols, eye-tracking measures, and behavioral performance data. Eighty-two Taiwanese children aged 5 to 6 years completed map-based spatial reasoning tasks under controlled visual conditions to minimize perceptual confounds. Qualitative analysis of concurrent verbalizations identified four strategy types—Landmark-Guided Navigation, Elimination-by-Comparison, Fast Search, and Unsystematic Responding—and a cross-cutting error pattern of Perceptual-Feature Reliance. Quantitative analyses revealed that systematic strategies (Landmark-Guided Navigation, Elimination-by-Comparison) were associated with higher accuracy, longer dwell times, and greater transitions between relevant Areas of Interest, whereas efficiency-oriented approaches (Fast Search, Unsystematic Responding) favored speed at the cost of accuracy. The convergence of verbalized reasoning and oculomotor metrics validates the use of multi-modal methodologies for capturing young children’s cognitive processes. The findings challenge stage-based models of spatial cognitive development by demonstrating the coexistence of advanced relational reasoning with persistent perceptual biases, and underscore the role of spatial language as a scaffold for complex reasoning. Educationally, the results support instructional designs that foster multi-landmark coordination, systematic elimination, and balanced visual attention, contributing to the broader goals of equitable STEM preparation in early childhood.