Lost in Estimation: Untangling the Enigma between Numerical Acuity and Visual Working Memory

In daily life, humans – much like a wide variety of animal species – constantly make comparisons, such as choosing the shortest line at the store or selecting the fullest basket of strawberries. Yet, the cognitive resources that enable the ability to discriminate large numerical quantities using only visual information, and without relying on explicit counting, remain unclear. Prior studies suggest that performance on numerosity comparison tasks partly relies on domain-general visual working memory (VWM) resources, facilitating us to selectively attend to numerical information while suppressing non-numerical, potentially distracting, visual information. In this framework, the present thesis explores whether known limits in VWM capacity place any constraints on the precision of numerosity discrimination, also known as numerical acuity. To this end, forty-six university students completed a numerosity comparison task measuring numerical acuity alongside a prototypical change detection task measuring VWM capacity. The results showed that in spite of rigorous visual controls, non-numerical visual cues significantly affected the accuracy of numerosity discrimination. However, no direct inter-individual correlation was found between VWM capacity and numerical acuity, nor between VWM capacity and variation in numerical acuity across trials characterized by distinct non-numerical visual covariates. These findings suggest that while non-numerical visual covariates undoubtedly influence our ability to discriminate large numerosities, numerical acuity likely relies on a processing stage that is functionally independent from VWM. Future studies utilizing direct neural measures of VWM capacity, including its attention-based filtering efficiency, are warranted to substantiate these findings.