Mean size estimation yields left-side bias: Role of attention on perceptual averaging

Rapidly estimating the mean size of a group of items of various sizes is one of the fascinating functions of our visual system. However, whether it yields left-side bias (LSB) where perceptual judgment tends to bias toward the information from the left visual field is still unclear. This issue was investigated by 4 experiments in this study. In each experiment, a set of disks were presented on the screen, followed by a central target disk. The mean size of the disks on one side (left or right) was larger than that on the other side, and which side contained disks with larger mean size was equally assigned among trials. Participants were asked to adjust the target disk to be the same as the mean size of the whole set (including both sides) of disks. Results showed a significant LSB: the size of the target disk was judged as larger when the left side contained disks of larger mean size, compared to when the situation was reversed (Experiment 1). The LSB vanished, when participants’ attention was cued to the right side by a color cue surrounding the right side before the presentation of the disks, implying that attention plays an important role on the LSB (Experiment 2). Furthermore, the magnitude of LSB increased with stimulus-onset asynchrony (SOA) when the disks on the left side were presented earlier. In contrast, the LSB first decreased and then reversed with SOA when the disks on the right side were presented earlier. This finding suggests that the LSB was enhanced by the prior entry effect on the attended side (Experiment 3). Finally, with free viewing, participants made more first saccade to the right side and had longer fixation duration there, presumably served as a compensatory role that eliminated the LSB on mean size estimation, which also supports the existence of the leftward attentional bias (Experiment 4). In summary, mean size estimation yields LSB which was induced by a leftward attentional bias that enhanced the prior entry effect on the left side.