Fig 1.
Task design and behavioral results.
(a) Trial schematic. Two colored, tilted bars (one on the left and the other on the right, one tilted to the left and the other to the right, with location and tilt being orthogonally manipulated) were displayed at encoding. Participants were probed to report the orientation of one of the two encoded bars at the end of each trial. In half of the trials (informative) a retro-cue matching the color of one of the two bars was shown; in noninformative trials (50%), a cue with a different color appeared instead. In informative trials, if the delay after cue offset was short (1 s), participants were probed about the cued item. Alternatively, if the delay was long (3 s), they had to report the other (uncued) item. In noninformative trials the probed item was unpredictable. Participants received feedback about their response accuracy on that trial as a percentage. Reaction time (s; panel b) and report error (°; panel c) are plotted in informative and noninformative and short and long trials. Thin grey lines represent individual participants (N = 30), error bars represent the standard error of the mean (SEM), light grey represents noninformative trials, and dark grey depicts informative trials. (d) Average response density as a function of the reported orientation and in relation to the orientation of the reported bar. Shaded areas represent SEM (N = 30). The data in this figure can be found in OSF under data/behav [52]. The numerical RT and report error values displayed in (b) and (c) can be found in S1 and S2 Data, respectively.
Fig 2.
Lateralized frequency-specific EEG activity locked to cue onset.
(a) Contrast between EEG time–frequency activity contralateral vs. ipsilateral to the cued bar location in occipital sensors (PO7/PO8) divided by summed contralateral and ipsilateral activity (expressed as a percentage) in informative trials vs. noninformative trials. (b) Contrast between EEG time–frequency activity contralateral vs. ipsilateral to the cued prospective action in central sensors (C3/C4) divided by summed contralateral and ipsilateral activity (expressed as a percentage) in informative trials vs. noninformative trials. (c, d) Cross-participant average alpha (8−12 Hz; c) and mu/beta (8−30 Hz; d) activity difference between contralateral and ipsilateral sensors to the cued location and action, respectively, in informative trials. Topographies represent the average frequency-specific activity in contra-vs-ipsi contrasts in informative trials across all sensor pairs during the time-windows which correspond to the alpha clusters in panel c or mu/beta clusters in panel d, respectively. Black outline in time–frequency spectra (a, b) indicates statistically significant clusters. Shaded areas represent the SEM and cluster-permutation corrected significant time points are indicated with horizontal lines in c and d (N = 30). The first part of the time–frequency spectra in panels a and b and of the time course in c and d (−0.2–1.2 s) corresponds to the average of short and long trials, and the second part (1.2–3.2 s) corresponds to long trials only. The vertical dotted lines represent (from left to right) the onset (0 s) and offset (0.2 s) of the retro-cue and the time of probe appearance in early trials (1.2 s). The data in this figure can be found in OSF under data/eeg/trf [52].
Fig 3.
The lateralized alpha and mu/beta modulation time courses do not co-evolve in lockstep.
(a) Average alpha (8−12 Hz) activity difference between contralateral and ipsilateral sensors to the cued location (burgundy) and average mu/beta (8−30 Hz) activity difference between contralateral and ipsilateral sensors to the cued action (green) scaled to range between −1 (minimum) and 1 (maximum) per participant (N = 30). Horizontal grey lines depict time points that were significantly different between the normalized alpha and mu/beta modulation time courses in a participant-wise cluster-based permutation test. Shaded areas represent the SEM. The vertical dotted lines represent (from left to right) the offset (0.2 s) of the retro-cue and the time of probe appearance in early trials (1.2 s). (b) Schematic representation of the average visual and motor shift time calculation procedure. For each participant, we calculated the minimum and maximum points of alpha and mu/beta modulation and estimated the average zero-crossing time (shift time; see “Materials and methods”). (c) Violin plots depicting the distribution of visual (top) and motor (bottom) average shift times across participants. Colored lines represent individual participant mean values, the white line inside the grey box represents the mean, and the edges of the grey box represent the first (left) and third (right) quartiles. Statistical significance is depicted with asterisks (***p < .001). (d) Violin plot depicting the difference between average motor and visual shift time for each participant (colored dots). The data in this figure can be found in OSF under data/eeg/trf [52]. The numerical values of the visual and motor shift times and their difference can be found in S3 Data.
Fig 4.
Lack of correlation between the shift times of the lateralized alpha and mu/beta modulation time courses.
(a) Top and right: the histograms represent the distribution of average mu/beta (green) and alpha (burgundy) shift times, respectively, on all the iterations of the bootstrapping procedure (50 bins per histogram). Centre: scatter plot of the average alpha and mu/beta shift times in each of the 100 bootstrapping iterations across all 30 participants as plotted on the mu/beta (x-axis) and alpha (y-axis) time space. Both axes are centered around the time of early probe onset. Vertical and horizontal dotted lines represent the time of early probe appearance (1.2 s after retro-cue onset). The burgundy circle represents the participant-average visual shift time as estimated on the lateralized alpha modulation time course, and the green circle represents the motor shift time estimated on the lateralized mu/beta modulation time course. The lines on the left and bottom of the green and burgundy circles, respectively, depict the participant-average minimum motor and visual shift times. The lines on right and top the of the green and burgundy circles, respectively, depict the participant-averaged maximum motor and visual shift times. The colors of the scattered dots depict individual participants. (b) Pearson’s r values of the correlation between motor and visual average shift times across the 100 bootstrapping iterations for each participant (N = 30). Colored dots represent individual participants’ correlation values, the white line inside the grey box represents the mean, and the edges of the grey box represent the first (left) and third (right) quartiles. The data in this figure can be found in OSF under data/eeg/trf [52]. The numerical Pearson’s r values displayed in (b) can be found in S4 Data.
Fig 5.
Relation between lateralized alpha and mu/beta modulations and behavior.
Alpha and mu/beta activity contralateral vs. ipsilateral to the cued location/hand in fast vs. slow and precise vs. imprecise trials. (a–d) Average alpha (8–12 Hz; burgundy) activity difference between contralateral and ipsilateral occipital sensors to the cued location in trials categorized as fast (dark) or slow (light) based on a median split of RT (a, b) and in trials categorized as precise (dark) or imprecise (light) based on a median split of report error (c, d). The left column depicts short trials (a, c) and the right column shows long trials (b, d). (e–h) Average mu/beta (8–30 Hz; green) activity difference between contralateral and ipsilateral central sensors to the cued action in trials categorized as fast (dark) or slow (light) based on a median split of RT (e, f) and in trials categorized as precise (dark) or imprecise (light) based on a median split of report error (g, h). The left column depicts short trials (e, g) and the right column shows long trials (f, h). Cluster-based permutation significant time points of the contrast between the displayed time courses are indicated with horizontal grey lines (N = 30). Shaded areas represent the SEM. The data in this figure can be found in OSF under data/eeg/trf [52].
Fig 6.
Time–frequency activity difference in informative vs. noninformative (neutral) trials in sensors contralateral to the first- and second-prioritized locations and actions.
(a, c) Difference in time–frequency activity in occipital (PO7/PO8) channels contralateral to the first-prioritized location (a) and second-prioritized location (c) in informative trials vs. noninformative trials (expressed as a percentage). (b, d) Difference in time–frequency activity in central (C3/C4) channels contralateral to the first-prioritized action (b) and second-prioritized action (d) in informative trials vs. noninformative trials (expressed as a percentage). (e) Cross-participant average alpha (8−12 Hz) activity difference between informative and noninformative trials (%). Burgundy: channels contralateral to the first-prioritized location. Pink: channels contralateral to second-prioritized location. (f) Cross-participant average mu/beta (8−30 Hz) activity difference between informative and noninformative trials (%). Dark green: channels contralateral to the first-prioritized action. Light green: channels contralateral to second-prioritized action. Black outline in time–frequency spectra indicates statistically significant clusters (a–d). Shaded areas represent the SEM and cluster-permutation corrected significant time points are indicated with horizontal lines in e and f (N = 30). The first part of the time–frequency spectra in panels a–d and of the time course in e and f (−0.2–1.2 s) corresponds to the average of short and long trials, and the second part (1.2–3.2 s) corresponds to long trials only. The vertical dotted lines represent (from left to right) the onset (0 s) and offset (0.2 s) of the retro-cue and the time of probe appearance in early trials (1.2 s). The data in this figure can be found in OSF under data/eeg/trf [52].