Differential developmental trajectories of non-symbolic numerical comparison skill across congruency conditions: A two-year longitudinal study

Abstract

Non-symbolic numerical comparison is a fundamental and persistently developing mathematical skill in early childhood. This two-year longitudinal study examined two explanations of how nonsymbolic numerical precision improves in early childhood: the filtering hypothesis (that children learn to suppress irrelevant continuous cues such as dot area) and the sharpening hypothesis (that maturation of the approximate number system refines numerical representations). We tested these hypotheses by comparing developmental trajectories under congruent (CON) and incongruent (INC) conditions. Additionally, this study examined executive function predictors of the trajectories and how they related to mathematical outcomes. Ninety preschoolers (T1 M_age = 5 years and 1 month old, SD = 11 months old) completed non-symbolic numerical comparison tasks (congruent/incongruent trials), working memory, inhibitory control, and cognitive flexibility measures, along with mathematical assessments. Results revealed significant growth only for INC trials, supporting the filtering hypothesis. Two distinct trajectories emerged for both congruent and incongruent conditions: CON low-steady, CON high-steady, INC high-increasing, and INC low-increasing groups. Higher levels of cognitive flexibility were associated with decreased likelihood of CON low-steady group membership. Notably, CON high-steady and INC high-increasing groups demonstrated higher mathematical achievement, highlighting the educational implications of non-symbolic numerical development.