Attentional modulation of hierarchical ensemble coding for the identities of moving faces.

Observers can accurately compute summary statistics from a set of similar objects, which is referred to as ensemble coding. Few studies have investigated whether ensemble coding can be modulated by attention (i.e., whether attention can be selectively deployed to exclude irrelevant items from the ensemble coding of relevant items), and no studies have tested whether this coding represents the average of a relevant subset, the average of an irrelevant subset, or the average of the entire set. The present study employed an adapted paradigm of the multiple object tracking (MOT) task in which participants viewed multiple moving target faces and were asked to indicate whether a subsequent test face was one of the preceding targets. Experiment 1 showed that ensemble coding occurred for moving faces, and Experiments 2, 3, and 4 showed that participants were more likely to judge the individual and the average of tracked target faces than the average of all (target and nontarget) faces in the display and average nontarget faces. These findings suggest that ensemble coding for facial identities is hierarchical and can, at least partially, be modulated by attention. (PsycINFO Database Record