Detection performance is modulated at a low-theta selection rhythm

Previous studies have supported the idea that slow brain rhythms are related to behavioral performance (e.g., Busch and VanRullen, 2009). Such findings suggest that the visual environment is sampled rhythmically for processing rather than continuously. In the present study, we investigated fluctuations in attentional performance. We found evidence that accuracy on a detection task followed a slow rhythmic pattern (approximately 5 Hz). Participants viewed two drifting gratings presented on either side of a central fixation. The task was to report the appearance of a brief contrast decrement (target) within one of the gratings. Contrast decrements were pre-adjusted to equate performance across individuals. Trials were terminated by participant's response and lasted no longer than 3.8 s. Ten percent of trials contained no target (catch trials). In addition, a peripheral irrelevant probe briefly appeared surrounding one of the gratings. We varied the temporal interval between the target and the irrelevant probe: The target could appear from 750 ms before the irrelevant probe to 1000 ms after the irrelevant probe in steps of 16.7 ms. In addition, probe and target could either appear in the same or in different visual fields. Target-detection rates were analyzed as a function of the target-to-probe interval. The irrelevant probe manipulation was designed to reset performance in time and space, which indeed was achieved. When target and probe shared visual field, target detection was entirely masked. Importantly, after the probe, target detection at the probe side oscillated at 5 Hz. Interestingly, a similar oscillation occurred also for targets opposite to the probe side. The two oscillations were in anti-phase. This is consistent with a selection mechanism that samples one target location at a time in a rhythmic fashion and explores multiple locations in succession