Integration of symbolic and non-symbolic numerical information in children: Task dependence and its link to math abilities

From birth, children can access the approximate number system for noisy numerical estimates. With age, they acquire an exact number system for precise numerical information representation. The relations between these two systems and their correlations with math abilities in children remain unclear. In this study, 8- to 10-year-old children (N = 119) completed two tasks to test the integration of symbolic and non-symbolic numerical information (i.e., "symbolic integration") and how this integration relates to children's formal math abilities. For the number comparison task, involving dot arrays and Arabic numerals, children indicated which of two sequentially presented stimuli was larger. These stimuli were either in the same format (dot-dot or numeral-numeral) or in a mixed format (dot-numeral or numeral-dot). For the number-letter discrimination task, participants identified numerals or letter pairs co-occurring with dot arrays that either matched or mismatched the numeral's quantity. In the number comparison task, we found that children were significantly slower when comparing mixed-format stimuli versus same-format conditions, suggesting a lack of symbolic integration (i.e., "symbolic estrangement"). In contrast, in the number-letter discrimination task, children were significantly faster in tasks where the dot arrays and numerals matched, indicating symbolic integration. While we found correlations between number processing and math skills at the condition level for both tasks, neither of the derived measures of symbolic estrangement or symbolic integration correlated with children's performance on a standardized math assessment. Thus, we conclude that numerical integration or estrangement is task dependent and that symbolic integration has limited impact on 8- to 10-year-old children's math abilities.