Abstract
Purpose: To test the hypothesis that voluntary and involuntary temporal attention continuously interact to affect perceptual sensitivity. Temporal attention is the prioritization of visual information at specific points in time, and can be either voluntary (goal-directed) or involuntary (stimulus-driven). How these types of attention interact to affect perception is essentially unknown. Methods: On each trial, two Gabor targets (T1 and T2) were presented successively at the same location. We manipulated voluntary temporal attention with an auditory pre-cue that instructed participants to attend to T1, T2, or both targets (neutral condition). We presumed that involuntary attention would be elicited by T1, and we measured the timecourse of its effects on T2 performance using a variable SOA (100-800 ms, in different sessions to ensure predictable timing). Each target was independently tilted clockwise or counterclockwise (at observers' thresholds). A response cue instructed participants to report the orientation of either T1 or T2. In valid trials, the pre-cue and response cue matched (75%); in invalid trials, they mismatched (25%). We developed computational models of attentional dynamics and stimulus processing and fit them to the behavioral data. Results: Accuracy was higher for valid than invalid cues for both targets at intermediate SOAs. Overall accuracy increased across SOAs for T1 and was U-shaped for T2. Models of the dynamics of voluntary and involuntary temporal attention parsimoniously explained attentional cueing timecourses, backward masking effects for T1, and "attentional blink"-like behavior for T2, either by (a) implementing competition between targets for attentional resources across time or (b) enhancing early-stage target representations, which biased later competitive interactions. Conclusion: Attending to one stimulus improves discrimination performance for that stimulus relative to other stimuli presented slightly earlier or later. This effect depends on the time interval between stimuli, which is consistent with joint effects of voluntary and involuntary temporal attention.
Meeting abstract presented at VSS 2016