Numerosity Perception and Perceptual Load: Exploring Sex Differences Through Eye-Tracking

Abstract

This study investigates sex differences in numerosity perception and visuospatial abilities in adults using eye-tracking methodology. We report the results of a controlled dual-task experiment that assessed the participants' visuospatial and numerosity estimation abilities. We did not observe sex differences in reaction times and accuracy. However, we found that females consistently underestimated numerosity. This underestimation correlated with higher perceptual load in females, as evidenced by shorter fixation durations and increased fixation rates. These findings suggest that perceptual load, rather than visual or spatial abilities, significantly influences numerosity estimation. Our study contributes novel insights into sex differences in both numerosity estimation and visuospatial abilities. These results provide a foundation for future research on numerosity perception across various populations and contexts, with implications for educational strategies and cognitive training programs.

Keywords: attention; eye movements; eye-tracking; individual differences; numerosity perception; perceptual load; sex differences; visuospatial abilities.

Figure 1

An example of different mental load conditions adopted in the experiment (perceptual load (PL) and working memory load (WML)). The central blue rectangle represents a car from the KITTI dataset, retained for naturalistic context but not counted in the numerosity task. Violet boxes represent bicycle targets, orange boxes represent pedestrian targets, and light blue boxes represent distractors. Participants were instructed to focus on violet and orange targets for identification and ignore the central blue car. To preserve the naturalistic properties of the KITTI dataset, object counts within each load condition varied slightly within predefined thresholds (e.g., 1–9 distractors in low PL—low WML; 1–9 objects, including targets, in low PL—high WML). This variability ensured ecological validity while maintaining experimental control over load categories.

Figure 2

The procedure of one session in the experiment.

Figure 3

Graphs depicting (A) average fixation duration (in ms) and (B) fixation rate (fixations per minute) across four conditions of perceptual load (PL) and working memory load (WML): Low–Low (LL: low perceptual load, low working memory load), Low–High (LH: low perceptual load, high working memory load), High–Low: (HL: high perceptual load, low working memory load), and High–High (HH: high perceptual load, high working memory load).