Abstract
The visual system can effectively represent a complex scene by extracting summary statistical information from groups of similar objects, such as mean. However, not much attention has been given to understanding the form of mean representation, e.g., whether mean size is represented as a single size. Here, we explored this idea by examining how mean size estimation bias and precision change depending on the disparity between two comparing ensembles, by varying set size and variance. Observers were presented with a set of multiple circles followed by a probe display that contained either a single circle or a set of multiple circles. They were asked to report the mean size of the former display by adjusting the size of the single circle or the overall size of multiple circles in the probe display. We first examined how estimating mean sizes using sets with different number of items influenced performance, by varying set size disparity between the stimulus and probe displays. Next, we manipulated size variance in both the stimulus and probe displays (low or high variance) and examined how variance congruency influenced mean size estimation. Performance was measured by calculating the percentage error from the actual mean size, as well as the variance of responses, as a proxy for perceptual precision. Results showed that mean size error and response variance reduced as set size disparity decreased between the stimulus and probe displays. Moreover, they were significantly reduced when size variance was congruent between the two displays as opposed to when they were different. The fact that bias and precision of mean estimation are contingent on characteristics of the probe displays suggests that mean representation includes an ensemble of statistical properties.
Acknowledgement: This research was supported by the Brain Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF-2017M3C7A1029658)