Assessing the Validity of the 1-9 Non-Symbolic Numerical Range: Investigating the Confounding Influence of Subitizing on Approximate Number System Measurement
Abstract
The ability to process the quantities of sets in the world around us, referred to as ‘non-symbolic numerical processing,’ relies on two primary systems: the Approximate Number System (ANS) for processing large sets, and the Object-file system for the rapid and accurate processing of small sets (1-4 items). In numerical cognition research, sets of 1-9 are frequently used, inadvertently blending the use of these two systems. The primary objective of this study was to investigate whether the inclusion of subitizable trials within the 1-9 non-symbolic comparison task affects its validity as a pure measure of the true ratio effect of the ANS. Specifically, we sought to determine if the ANS operates independently across different numerical ranges or if subitizing introduces confounding effects. Our findings revealed that the 1-9 range is processed differently compared to other ranges, both in symbolic and non-symbolic contexts. This suggests that the 1-9 range may not reliably measure the true ratio effect of the ANS. The implications of this study recommend against using the 1-9 range for assessing ANS acuity or for cross-format numerical comparisons.