The Association between Approximate Number Sense (ANS) and Math Achievement Depends on the Format of the ANS Test

ABSTRACT There are two lines of discussion regarding the function of the Approximate Number System (ANS). The first line focuses on the extent to which visual cues affect the estimation of numerosity. The second line investigates the extent to which ANS precision is associated with symbolic math performance. The current study combined these two lines of discussion. Data from the nonsymbolic comparison test and the math achievement of 313 first graders (mean age: 7.6 years; 51% girls) from Russia were analyzed. First, we estimated the extent to which the effect of nonnumerical visual cues on nonsymbolic numerosity estimation varied across the four formats of stimulus presentation: separated/homogenous, separated/heterogeneous, mixed/homogeneous and mixed/heterogeneous. The results revealed that in the mixed/heterogeneous format of stimulus presentation, the congruency effect was not significant with respect to accuracy and was negative with respect to reaction time (RT). The reduction in the congruency effect may indicate that in this format, the participants tended to directly estimate numerosity, ignoring visual cues. Second, we tested the association between math achievement and precision on the ANS test in each format. The results demonstrated that accuracy in the separated/homogenous format had no significant association with math performance, while accuracy in other formats had significant associations with math performance. Moreover, accuracy in the mixed/heterogeneous format had a larger association than that found in other formats. This result might indicate that ANS accuracy has a significant association with math achievement in formats that demonstrate a smaller congruency effect and greater involvement of the direct estimation of numerosity.