Abstract
The phenomenon of subitizing (the rapid detection and individuation of a small number of items in parallel, in contrast to the serial set-size slope for counting larger numbers) has been proposed to reflect a parallel visual perception process with limited capacity. A growing body of studies has related subitizing to visual attention, demonstrating shared capacity between subitizing and various visual attention tasks, such as object tracking (Chesney & Haladjian, 2011); feature versus conjunction target discrimination (Vetter et al. 2008); and the attentional blink (Burr et al. 2010; Olivers & Watson 2011). Indeed, recent work demonstrates that individual differences in subitizing capacity are associated with a common ‘perceptual capacity’ factor underlying performance across diverse visual attention tasks (Eayrs & Lavie, 2018). Here, we tested the perceptual capacity hypothesis of subitizing further by investigating the pupil dilation pattern in response to subitizing versus counting in visual enumeration. Pupil dilation is a well-established physiological marker of attention demands (e.g. Alnaes et al. 2014). The parallel perceptual processing with limited capacity model of subitizing predicts a bifurcation of the pupil dilation by set-size function at the subitizing capacity limit, with a parallel slope up to the limit and a serial slope beyond. To test this prediction we recorded pupil diameter while participants reported the number of target shapes from brief, luminance-matched displays containing 1–7 target shapes among 8–14 non-target shapes. The overall results showed no linear effect of set-size up to 4 targets and a linear increase thereafter. Importantly, the bifurcation point of an individual’s RT by set-size function was correlated with their capacity estimate based on the bifurcation of the pupil diameter function. These results support the perceptual capacity hypothesis of subitizing and demonstrate that individual differences in subitizing capacity can be predicted from pupil dilation patterns.
Acknowledgement: The research was funded by a grant from Toyota Motor Europe