Abstract
Ensemble perception allows us to rapidly derive summary statistical information from groups of similar objects. Although ensembles are generated quickly and efficiently, the capacity limitations of this process are still debated. The visual system can represent large numbers of items as a single average value, but current research suggests there is a limit to the number of ensembles one can simultaneously extract. We previously demonstrated that the number of ensembles the visual system can effectively represent may depend explicitly on the visual domain (Schill & Haberman, VSS, 2016). In the current study, we replicated and extended those findings. In each trial, observers viewed two ensembles presented simultaneously and were post-cued to report the average of just one of the sets. The ensembles could either be mixed, in which two different visual domains were presented (e.g., faces varying in expression and colored patches varying in hue), or unmixed, in which two sets from the same visual domain were presented. Observers then adjusted a test stimulus to match the preceding set. The results revealed an overall benefit in ensemble representation in mixed conditions relative to unmixed conditions. That is, both average color and average expression representations improved when different ensemble types were present compared to when both sets came from the same visual domain. We conclude that attending to mixed ensembles reduces competition for neural resources, as different ensemble domains rely on independently operating mechanisms.
Meeting abstract presented at VSS 2018