Location- and object-based attention enhance number estimation

Antonella Pomè,Diego Thompson,David Charles Burr,Justin Halberda

doi:10.3758/s13414-020-02178-w

Abstract

Humans and non-humans can extract an estimate of the number of items in a collection very rapidly, raising the question of whether attention is necessary for this process. Visual attention operates in various modes, showing selectivity both to spatial location and to objects. Here, we tested whether each form of attention can enhance number estimation, by measuring whether presenting a visual cue to increase attentional engagement will lead to a more accurate and precise representation of number, both when attention is directed to location and when it is directed to objects. Results revealed that enumeration of a collection of dots in the location previously cued led to faster, more precise, and more accurate judgments than enumeration in un-cued locations, and a similar benefit was seen when the cue and collection appeared on the same object. This work shows that like many other perceptual tasks, numerical estimation may be enhanced by the spread of active attention inside a pre-cued object.

Highlights

Perceiving the number of objects is a fundamental skill for many animal species (Agrillo & Bisazza, 2014; Dehaene, 2011), as well as for humans, including infants (Izard, Sann, Spelke, & Streri, 2009)
We model likelihood with a Gaussian distribution centered on the stimulus, the width of which is estimated by the variability of the estimates, averaged over participants
Since the previous literature on location- and objectbased attention has focused on the differences in reaction times (RTs) between the condition tested, we analyzed RTs to respond on the number line

Summary

Introduction

Perceiving the number of objects is a fundamental skill for many animal species (Agrillo & Bisazza, 2014; Dehaene, 2011), as well as for humans, including infants (Izard, Sann, Spelke, & Streri, 2009). In a seminal study, Egly et al (1994) used a double-rectangle display to demonstrate that attention can be delineated by the boundaries of objects, effectively spreading through attended objects. This has been shown by cueing one end of the two rectangles and displaying a target either at the cued location (valid trials) or at one of three remaining locations (invalid trials). This explanation is based on the idea that all the two locations (i.e., same object different location, and different object) are equidistant from the cued location – responses on the attended object benefit from facilitated processing

Methods

Results

Conclusion