Abstract
The existence of an inherent link between numbers and space is a fundamental question in numerical cognition and has been highly debated in recent years. Although Spatial-Numerical Associations (SNAs) have traditionally been investigated using bimanual response settings, recent studies have moved towards unimanual paradigms to avoid potential measurement artifacts. For instance, an influential line of research adopted a modified Implicit Association Test (IAT) in which numbers and arrows were processed disjunctively in sequential trials. However, SNAs consistently emerged only when spatial and magnitude codes were explicitly contrasted, questioning the idea that numbers are inherently represented in a spatial format. To further investigate this issue, we designed a modified flanker task in which spatial and numerical stimuli were presented simultaneously. When arrows were task irrelevant (Experiment 1), no significant SNA emerged. By contrast, when participants judged arrow direction and numerical magnitude conjunctively (Experiment 2), the paradigm revealed extremely large SNAs compared to existing literature. Overall, our results provide no support for an inherent spatial-numerical link, suggesting that robust SNAs require explicit spatial codes combined with semantic number processing. Finally, the large effect size of group-level SNAs, together with remarkable mixed-model-derived psychometric properties, suggest that the modified flanker task is a promising paradigm for investigating individual differences and effect modulations. This paradigm may therefore advance research on SNAs by overcoming limitations of previous bimanual and unimanual paradigms and helping to resolve mixed findings in the literature.