Abstract
Natural scenes consist of objects of varying sizes and shapes. Recent evidence shows that real-world object size influences topography of object representations in occipitotemporal cortex (Konkle & Oliva, 2012). The degree to which real-world object size influences attentional allocation remains unexplored. Here, we ask whether real-world object size influences attentional allocation. A spatial cueing paradigm was used, with cues highlighting one end of a single line-drawing of a real-world object, presented in the center of the screen. Participants performed a T/L discrimination task on targets that appeared either at the cued (valid) or uncued (invalid) location within the object. While objects subtended the same physical size, they varied in real-world size (e.g., domino, billiards table). Prior to experiments, objects were classified into two groups (small and large) according to a pre-determined metric. Across four experiments, the validity of the spatial cue (50% vs. 75% valid) and the temporal profile of attention (250ms vs. 500ms SOA) were manipulated. Upon completion of the experiment, participants rated the real-world size of each object in a post-experiment survey. In addition to validity effects, we observed direct evidence of real-world object size modulating attentional selection. In all experiments, targets were identified faster in small objects as compared to large objects. Importantly, no interaction between size and validity was observed, providing evidence for additivity of object size and validity. The direct relationship between inferred object size and target identification was validated by a strong correlation between individual's ratings of real-world object size and reaction times for individual objects (longer target identification RTs for larger objects). These findings offer a novel suggestion that knowledge of real-world size modulates attentional shifts, with overall more sluggish processing of spatial locations within larger objects and more efficient processing in smaller objects (i.e., scaling of attention).
Meeting abstract presented at VSS 2017