Associations of Small, Moderate, and Large Numerosity Processing with Mathematical Ability and Attention in Primary-School Children

Source abstract

Numerosity processing across small, moderate, and large ranges may relate differently to children’s mathematical ability, but apparent differences may reflect shared developmental variation and task characteristics. We examined numerosity processing, mathematical ability, and attention in 145 primary-school children from Grades 2, 4, and 6. Large-range performance was represented by the participant-level average of Weber fractions from the 99- and corrected 134-dot reference conditions. At the zero-order level, all three numerosity indices were associated with total mathematical ability. After adjustment for age, the small-numerosity association was not statistically significant (partial r = 0.101, p = 0.230), whereas moderate (partial r = −0.248, p = 0.003) and composite large numerosity (partial r = −0.207, p = 0.013) remained associated with mathematical ability. Formal age-adjusted comparisons did not establish significant differences among their association strengths. Exploratory tests suggested grade-related heterogeneity, especially for the composite large-numerosity measure, with a stronger association in Grade 4 than in the other sampled grades. Sensitivity analyses indicated that the large-range association varied across reference conditions. These findings characterize task-dependent numerosity–mathematics associations but do not establish distinct developmental mechanisms or processing independent of continuous visual cues.