Abstract
Endogenous attention, exogenous attention, and visual masking are processes that control the visibility of a stimulus and determine whether it will register into consciousness and have access to short-term memory. Do these processes operate independently or interact as they control these common perceptual and cognitive outcomes? While earlier studies provided evidence for interactions between attention and masking, recent studies on common-onset masking reported that the interactions were artifacts of ceiling/floor effects. Our review of previous studies reporting metacontrast-attention interactions revealed similar artifacts. Therefore, we investigated in separate experiments whether exogenous attention and endogenous attention interact with metacontrast by using an experimental paradigm in which ceiling/floor effects are avoided. Observers fixated at the center of the display and reported the orientation of a target bar. An array of six or four oriented target bars followed by surrounding mask rings (with varying stimulus-onset-asynchronies) were presented around a virtual circle. The focus of attention was controlled by spatial cues with varying cue-target onset times. The endogenous cue (100% validity) consisted of an arrow at the center of the display that pointed to a randomly selected target bar. The exogenous cue (uninformative cue) was a small square flashed in the vicinity of a randomly selected bar. We also measured performance without masks (baseline). Critically, the target-mask luminances were adjusted individually to avoid saturation artifacts. Response errors were defined as the difference between the actual and reported orientations. We analyzed two different aspects of performance: The mean absolute response-errors, and the statistical distribution of signed response-errors. Our results show that both endogenous and exogenous attention modulate responses without interacting with masking, suggesting that they operate independently. Statistical modeling of the distribution of signed response-errors suggests weak interactions in modulating the probability of "guessing" behavior for some observers in both types of attention.
Meeting abstract presented at VSS 2016