Abstract
The mechanism of attention is one of the central topics in psychology, neurophysiology, and functional imaging. However, neurophysiology experiments often require an animal to maintain attention to one location (sustained attention1), while human behavioral studies often direct subjects to attend to different locations in every trial (transient attention2). It has been difficult to integrate the two levels of investigation. To facilitate integration, we conducted a human behavioral study3 based on a paradigm in a neurophysiology experiment1. Four synchronized RSVP streams of Gabors were presented in each trial. Each stream consisted of 10 distractors, all oriented at 45° clockwise, and one potential target, either +11° or -11° from 45°, that occurred between the 3rd and the 9th temporal positions simultaneously. Subjects were instructed to maintain fixation (monitored by an eye-tracker) and attend to one location in each block of 70 trials. A central report cue, presented 150 ms after target presentation and consistent with the block cue 62.5% of the time, directed subjects to report the orientation of the target in the cued location. In the invalid condition, the report cue pointed to one of the other three locations with equal probability. Full psychometric functions were measured at 4-5 external noise levels. Block cuing improved performance in high external noise conditions for three out of four subjects and increased asymptotic performance across all noise levels in one of them. The other subject showed only increased asymptotic performance. The data were fit with the elaborated perceptual template model4 (r2=0.9529). On average, sustained attention excluded 17% of external noise, and narrowed the perceptual template by 16%, but did not enhance the stimulus. The results are consist with behavioral studies on transient attention2,5. (1.Cohen & Maunsell, 2009; 2. Dosher & Lu, 2000; 3. Zhao et.al., 2013; 4. Jeon et.al., 2009; 5. Hetley et.al., 2014)
Meeting abstract presented at VSS 2015