Experiment 3 (50% cue validity) and Experiment 4 (100% cue validity, multiple object tracking [MOT]).

(A) In Experiment 3 (50% cue validity), subjects fixated on a central point and covertly attended to 2 discs presented in the left and right visual fields for target detection. Subjects were instructed to simultaneously pay attention to both discs and were informed that the target would be equally likely to appear within the discs and that the initial cue would not predict the target location. After a noninformative red circle cue (cue validity: 50%) appeared around 1 of the 2 discs, the luminance of the 2 discs was independently and randomly modulated for 5 s (top: cued visual sequence; bottom: uncued visual sequence), during which time subjects were instructed to monitor a randomly occurring target. (B) Grand average (N = 16) time–frequency plots for cued–uncued TRF power difference in Experiment 3 (cue validity: 50%). Note the prolonged alpha-band switching (blue–red pattern), suggesting that attentional shifting is enhanced when attention is evenly distributed across the 2 spatial locations (50% cue validity). (C) In Experiment 4 (MOT experiment), a red circle cue at the beginning of each trial indicated which disc the subjects should covertly attend to for subsequent target detection. The 2 disks were then moved randomly and smoothly across the screen for 5 s, during which time the subjects were instructed to detect the appearance of a target within the cued disc. Here, the cue validity was 100%, which means that the target only appeared in the cued disk, similar to Experiment 1. (D) Experiment 4 results. Top: Grand average (N = 11) time–frequency plots for att–unatt TRF power difference. Bottom: grand average (blue lines, N = 11, mean ± SEM) time course for att–unatt TRF power within the alpha band (8–12 Hz). Red horizontal lines at the bottom indicate points showing significant power differences in the alpha band between att and unatt (p < 0.05, one-tailed, false discovery rate [FDR] corrected). Note the initial alpha inhibition followed by an alpha rebound trend, similar to Experiment 1 (see Fig 2C). The data are provided in the Supporting Information (see S3 Data).