Abstract

This EEG study investigates the electrophysiological activity underlying processes of stimulus and response selection, and their executive orchestration via long-range functional connectivity under conflict condition, in order to shed more light on how these brain dynamics shape individual behavioral performance. Participants (n = 91) performed a modified flanker task, in which bilateral visual stimulation and a bimanual response pattern were employed to isolate the stimulus and response selection-related lateralized activity. First, we identified conflict-related markers of task-relevant processes; most importantly, the stimulus and response selection were evidenced by contra–ipsilateral differences in visual and motor activity, respectively, and executive control was evidenced by modulations of midfrontal activity. Second, we identified conflict-related functional connectivity between midfrontal and other task-relevant areas. The results showed that interregional phase synchronization in theta band was centered at the midfrontal site, interpreted here as a “hub” of executive communication. Importantly, the theta functional connectivity was more robust under the condition of increased demands for stimulus and response selection, including connectivity between the medial frontal cortex and the lateral frontal and motor areas, as well as cross-frequency theta–alpha coupling between the medial frontal cortex and contralateral visual areas. Third, we showed that individual differences in the measured conflict-related EEG activity, particularly the midfrontal N2, theta power, and global theta connectivity, predict the behavioral efficiency in conflict resolution.

INTRODUCTION

Stimulus and Response Selection

Goal-oriented behavior and quick response to relevant environmental changes require a coordinated set of selection-for-action mechanisms enabling effective linking of sensory information with suitable response programs (Allport, 1987). Such a selection for action may be considered as a two-stage or two-level process. One is the mechanism of stimulus selection that acts on perceptual representations; the other is the mechanism of response selection acting on motor representations. Both mechanisms operate via simultaneous signal enhancements and distractor suppressions (Thiele & Bellgrove, 2018). Within the perceptual system, neuronal activity underlying sensory representation of a relevant object is increased, whereas the activity underlying representation of distractors is inhibited (Moore & Zirnsak, 2017; Corbetta & Shulman, 2002). Similarly, within the motor system, neuronal activity underlying the representation of a proper motor program is increased, whereas activity underlying inappropriate, reflex-like, or incorrect motor programs is inhibited, which allows us to deal with conflicting response tendencies (Carter & Krug, 2012; Ridderinkhof, Ullsperger, Crone, & Nieuwenhuis, 2004).

The two selection mechanisms, for stimulus and response, are therefore functionally and anatomically separate, yet they have to work together in a harmonious, coordinated manner during complex tasks of everyday life. A plausible candidate for a functional hub orchestrating these selection processes is the medial frontal cortex (MFC), including the midcingulate cortex and ACC (Cohen & Ridderinkhof, 2013; Cohen, 2011). Consistently with such a role, the MFC has long-range reciprocal structural connections with numerous cortical and subcortical brain regions (Cavanagh & Frank, 2014). A key component of this communication is an ongoing theta band oscillatory activity. It has been proposed that the MFC-generated interregional theta phase synchronization reflects the transmission of top–down control signals through which the MFC may entrain distinct task-relevant regions (Cohen, 2014a). It has also been argued that the theta band activity is a fundamental property of the stimulus selection mechanism, determining sensory sampling and reweighting functional connections between cortical and subcortical areas involved in the selection process (Fiebelkorn & Kastner, 2019). Finally, robust evidence shows that an increase in midfrontal theta power is instrumental for response selection under conditions of response conflict (Cavanagh & Frank, 2014; Cohen, 2014a).

To investigate the brain dynamics underlying such an executive orchestration of the selection processes, an experimental tool is needed that would allow for simultaneous measurement of stimulus and response selection under a condition demanding this kind of supervisory control. An example of such a procedure is the flanker task (Eriksen & Eriksen, 1974). In a typical arrow version of this choice-reaction paradigm, a target stimulus is either a left- or right-pointing arrow, and responses are defined as key presses with the left or right hand, respectively, to the target arrow. This establishes two fixed and ready-to-use stimulus–response (S–R) links or “event-files”: left arrow = left-hand response, and right arrow = right-hand response (cf. Verleger, Metzner, Ouyang, Śmigasiewicz, & Zhou, 2014; Hommel, 2004). In each trial of the task, the target arrow is flanked by four other arrows, which are either congruent or incongruent with the target. In the congruent condition, the target and flanker arrows point in the same direction, so that activation of the correct S–R link is easy, and responses are quick and accurate. In the incongruent condition, however, the flanker arrows point in the direction opposite to the target arrow, which causes a perceptual interference similar to the effects of visual crowding (Whitney & Levi, 2011) or grouping (Luo & Proctor, 2016). Thus, the involvement of the stimulus selection mechanism is needed for target identification. Moreover, both the target and the incongruent flankers simultaneously activate their respective S–R links (as both the left- and right-pointing arrows have established fixed links with the corresponding left- and right-hand responses), thereby triggering a response conflict between two motor programs (Egner, 2008; Gratton, Coles, & Donchin, 1992). The involvement of the response selection mechanism is therefore required. As a result, RT is longer and error rate is larger in the incongruent condition than in the congruent condition. This incongruent–congruent difference, that is, the behavioral flanker effect, reflects an accumulated, presumably additive (noninteracting; cf. Maniscalco, Bang, Iravani, Camps-Febrer, & Lau, 2012) cost of dealing with the perceptual interference and the response conflict. The cost is assumed to be inversely proportional to the efficiency of the stimulus and response selection mechanisms—the higher their efficiency, the lower is the cost. The overall performance level should therefore be, obviously, the resultant of the cumulated efficiency of both the primary S–R processing chain and its higher-order executive orchestrations.

This Study

In this study, we aimed to further investigate how the brain dynamics underlying the selection processes and their executive orchestration shape behavioral performance on an individual basis. To this end, we analyzed EEG data from 91 participants performing a modified version of the arrow flanker task. Our strategy was, first, to identify conflict-related electrophysiological markers of task-relevant processes; second, to identify conflict-related functional connectivity between the MFC and other task-relevant areas; and third, to examine which of and to what extent the identified markers of conflict-related brain activity and connectivity are related to the behavioral efficiency of conflict resolution.

Selective sensorimotor processing during the performance of conflict tasks is expected to involve predominantly the occipital areas—where visual selective processing is implemented (cf. Luck & Gold, 2008), the posterior parietal areas—where S–R translation processing occurs (cf. Gulbinaite, van Rijn, & Cohen, 2014; Freund, 2001), the premotor and motor areas—where response selection is implemented (cf. van Driel, Sligte, Linders, Elport, & Cohen, 2015; Gratton, Coles, Sirevaag, Eriksen, & Donchin, 1988), and finally the MFC, along with the lateral frontal cortex (LFC; predominantly the dorsolateral prefrontal cortex)—from which we gain the ability to exert executive control over perceptual interference and response conflict (cf. van Driel et al., 2015; Carter & Krug, 2012). On that account, we proceeded to measure EEG markers of the listed processes at the specified sites. To discern EEG activity related specifically to stimulus and response selection, we introduced bilateral visual stimulation to the task and used a bimanual response pattern. This allowed us, by means of a double subtraction method (Coles, 1989), to isolate a lateralized (contralateral–ipsilateral) EEG activity relative to the target hemifield and responding hand—that is to obtain a simultaneous assessment of stimulus and response selection in one task. To assess the former, we measured the posterior contralateral negativity (PCN; including the N2pc and sustained posterior contralateral negativity [SPCN] components) of the ERPs, and posterior contralateral time–frequency power and phase coherence in the alpha band. To assess the latter, we measured the lateralized readiness potential (LRPs) and lateralized beta band power over motor areas. Furthermore, to estimate the executive functioning-related midfrontal activity, we measured the N2 component of the ERP and theta band power. In addition, we measured the parietal P3b component of the ERP, to identify possible conflict-related modulations of the S–R translation process. Finally, to assess the hypothesized executive orchestration processes, we measured conflict-related functional connectivity in theta band between the MFC and other task-relevant sites, that is, the visual, motor, centro-parietal, and lateral-frontal areas. If any of the identified incongruent–congruent difference in the EEG activity is behaviorally relevant, it should correlate with the behavioral flanker effect. A thorough rationale for each measurement and specific performance-related hypotheses are detailed in the following paragraphs.

Assessment of Stimulus Selection

To obtain a comprehensive picture of the process of stimulus selection, we measured three complementary markers of selection-related lateralized EEG activity: PCN, and posterior contralateral time–frequency power and phase coherence in the alpha band. The PCN is event-related lateralization (ERL; Wascher & Wauschkuhn, 1996) of the ERPs, measured as a contra–ipsilateral difference relative to the target visual hemifield at sites located above the occipito-temporal cortex (usually PO7 and PO8). The most prominent component of the PCN is the N2pc, occurring at about 200–250 msec after target onset at the contralateral hemisphere, and thought to reflect target selection (Eimer, 1996; Luck, Fan, & Hillyard, 1993). Another PCN component that often follows N2pc is an SPCN (Vogel & Machizawa, 2004). As suggested by Schneider, Hoffmann, and Wascher (2014), SPCN may reflect creating a stable percept of the selected stimuli “that can serve as the perceptual basis for response selection and the initiation of goal-directed behavior” (p. 1). Therefore, the PCN components should present a higher amplitude in the incongruent trials than in the congruent trials because of increased stimulus selection and discrimination demands. Assuming that larger N2pc and SPCN indicate a stronger involvement of the selection mechanism (cf. Luck, Girelli, McDermott, & Ford, 1997), the magnitude of their amplitudes should be inversely related to the behavioral flanker effect.

The lateralized power spectra (LPS) in alpha band provide an index of ipsilateral–contralateral differences calculated by the double-subtraction (as the ERLs), so that all activity unrelated to spatial selection and general hemispheric differences are subtracted out (van der Lubbe, de Kleine, & Rataj, 2019; van der Lubbe & Utzerath, 2013; see also Zhigalov & Jensen, 2020). Thus far, the LPS index has been used for pretarget cue-evoked activity related to preparatory orienting and showed alpha suppression over the hemisphere contralateral to the cued hemifield, indicating inhibition of the irrelevant side and/or disinhibition of the relevant side (van der Lubbe et al., 2019; van der Lubbe & Utzerath, 2013), in line with other similar methods (Thut, Nietzel, Brandt, & Pascual-Leone, 2006; Worden, Foxe, Wang, & Simpson, 2000). Here, however, we calculated the LPS for a target-evoked activity to examine the role of alpha suppression in stimulus selection. Recently, Bacigalupo and Luck (2019) have reported a contralateral alpha suppression evoked by uncued targets (i.e., unrelated to preparatory orienting) suggesting that “alpha also serves as an active mechanism of target processing” (p. 900). Of importance, they also found a trend for enhanced alpha suppression with increased perceptual crowding by reducing the distance between target and flanker letters. Accordingly, we expected that the LPS index would confirm the role of alpha desynchronization in target selection and show a stronger alpha suppression in the incongruent trials than in the congruent trials, as well as a negative correlation of the alpha suppression with the behavioral cost. Of interest was also the relationship between the LPS and the PCN components of the ERP, because Bacigalupo and Luck's (2019) results have suggested that the alpha lateralization and N2pc reflect related but distinct aspects of visual selection.

Further information that is unobservable from ERPs and time–frequency power may be obtained from event-related phase coherence analysis. Indeed, it has been observed that processing of laterally presented targets is associated with a contralateral increase of phase coherence in alpha band over visual cortex, which presumably reflects the mechanism of temporal gating of visual stimuli (Keitel et al., 2019; Spaak, de Lange, & Jensen, 2014). Thus, assuming that increased involvement of the stimulus selection mechanism would entail an increase of the contralateral alpha phase coherence, we expected to observe a larger contra–ipsilateral difference in the incongruent than in the congruent flanker condition, as well as a correlation of such a flanker effect on the contra–ipsilateral alpha phase difference with the behavioral flanker effect.

Assessment of Response Selection

To isolate response-related motor activity, the LRPs and LPS were calculated from the ERPs and time–frequency power, respectively. The LRP is ERL measured over the right and left motor cortex relative to the responding hands (Smulders & Miller, 2011). Typically, a negative LRP emerges about 200–400 msec after target onset, indicating selection and activation of the correct response program. In the incongruent condition of the flanker task, the selection of the correct response is delayed because of the simultaneous activation of the incorrect response (Asanowicz, Wołoszyn, Panek, & Wronka, 2019; Grent-'t-Jong, Oostenveld, Jensen, Medendorp, & Praamstra, 2013). Inhibition of the incorrect response is thought to be a result of the implementation of executive control within motor representations (Klein, Petitjean, Olivier, & Duque, 2014; Verleger, Kuniecki, Möller, Fritzmannova, & Siebner, 2009). Assuming that the conflict is resolved prior to the execution of the correct response program, the motor processes should be delayed but not lengthened by the conflict processing. Thus, the flanker effect on the LRP latencies should only be observed relative to stimulus onset and not relative to the response. Specifically, in the incongruent trials, the motor processes should begin later—which is expected to be evidenced in the stimulus-locked LRPs—but once the correct program is selected, its execution should be as fast as in the congruent trials—hence, no latency difference should be found in the response-locked LRPs (cf. Hackley & Valle-Inclán, 1998). The same rationale and predictions apply to our measurement of the hand-movement-related oscillatory activity in beta band, which we assessed by calculating lateralized beta power spectra (beta LPS), using the same method as for the LPS in alpha band. Positive LPS will indicate beta desynchronization in motor areas contralaterally to the responding hand (cf. Siegel & Donner, 2010; Crone et al., 1998), and the time course of the beta desynchronization indicates the timing of response selection (Tzagarakis, Ince, Leuthold, & Pellizzer, 2010; Kaiser, Birbaumer, & Lutzenberger, 2001).

In addition, we measured the P3b, which is a positive ERP deflection at about 300–700 msec after target onset with a maximum at the parietal midline (usually Pz). In the flanker task, P3b usually displays a smaller amplitude and a longer latency in the incongruent condition than in the congruent one (Asanowicz et al., 2019; Kałamała, Szewczyk, Senderecka, & Wodniecka, 2017; Neuhaus et al., 2010). Recently, P3b has been shown to be related to sensorimotor decision-making, that is, an accumulation of perceptual evidence in translating sensory information into appropriate actions (Twomey, Murphy, Kelly, & O'Connell, 2015; O'Connell, Dockree, & Kelly, 2012) and to activation of established S–R links (Asanowicz et al., 2020; Verleger et al., 2014), which are a necessary prerequisite for correct response selection. The flanker effect on P3b may therefore reflect the interference between two S–R links (evoked simultaneously by the target and incongruent flankers), which delays the sensorimotor decision and activation of the correct S–R link. Accordingly, the smaller the incongruence cost, the smaller the flanker effect on P3b should be observed.

Assessment of Conflict-related Midfrontal Activity

The involvement of midfrontal activity in conflict processing was evaluated by measuring its two complementary signatures: the midfrontal N2 component of the ERP and midfrontal theta power. The N2 is a negative deflection occurring at about 200–400 msec after target onset with a maximum at the frontal midline (usually FCz). In the conflict trials of the flanker task, the N2 tends to display an increased amplitude (Heil, Osman, Wiegelmann, Rolke, & Hennighausen, 2000; Kopp, Mattler, Goertz, & Rist, 1996) originating from the MFC (particularly the ACC; van Veen & Carter, 2005; Yeung, Botvinick, & Cohen, 2004). This conflict-related N2 is interpreted as a signature of the involvement of the executive system in detection and resolution of response conflict, which enables the selection and execution of the correct response (Yeung et al., 2004; van Veen & Carter, 2002). Assuming that a larger N2 amplitude indicates stronger involvement of the response selection control, we should observe a negative correlation between the N2 and behavioral flanker effects. However, this relationship has remained unclear, because Cohen and Donner (2013) found no correlation between the N2 and RT in a response conflict task (specifically, in their versions of the Simon task; see Egner, 2008, for a description of conflict tasks) and suggested that the phase-locked midfrontal activity may not be related to conflict resolution. Moreover, some studies even questioned the interpretation of the N2 as a proper index of response conflict processing in the flanker task (Kałamała et al., 2017; Tillman & Wiens, 2011). Hence, we further examined and compared the behavioral relevance of the phase-locked (the N2) and non-phase-locked (theta modulations, see below) midfrontal activities.

The conflict-related theta power is an increase of the midfrontal theta observed during response conflict, typically at about 300–600 msec after stimulus onset at frontal midline sites (usually FCz; Cohen & Cavanagh, 2011; Nigbur, Ivanova, & Stürmer, 2011). It is thought to reflect the neurophysiological oscillatory mechanism underpinning conflict detection and resolution (Cavanagh & Frank, 2014; Cohen, 2014a). As such, the theta band oscillations have been argued to have a causal role in the efficiency of conflict processing (van Driel et al., 2015; Cohen & Donner, 2013). Assuming that a larger theta power indicates increased involvement of the mechanism, we should observe an inverse relationship between the magnitude of the conflict-theta effect and the behavioral cost. Of importance, in their Simon task study, Cohen and Donner (2013) have shown that the conflict-related theta is time-locked but not phase-locked to the onset of conflict-eliciting stimuli, which indicates that this activity and the conflict-related N2 component of the ERP, despite their spatiotemporal overlap, are two distinct physiological phenomena, possibly functionally separate cogs in the executive control machine. Thus, here, we measured specifically non-phase-locked theta power to dissociate this activity from the N2. Alternatively, if the theta power and N2 reflect the same neuronal mechanism, not only they should be similarly related to the behavioral flanker effect but also be highly correlated with each other.

Assessment of Interregional Functional Connectivity

Understanding the interregional functional connectivity is pivotal for answering the question on how the selection processes are orchestrated within an organically connected and functionally coherent system. Such long-range connectivity seems to be mechanistically implemented by phase-locking of oscillations in the communicating neuronal groups (Siegel, Donner, & Engel, 2012; Fries, 2005). Thus, to assess the hypothesized midfrontal-centered executive orchestration, we measured the so-called intersite phase coherence (ISPC), which is an estimation of frequency-specific phase synchrony between two sources of oscillations and is interpreted as an index of functional connectivity between two brain areas.

First, we examined whether the MFC may indeed be considered as a “hub” of the executive communication in the theta band during response conflict, as predicted by the hypothesis that the MFC spreads top–down control signals through theta oscillations. Such a role of the MFC has been demonstrated for error-related processing by Cohen (2011). He used a graph theory-based method to quantify the number of functional connections of each node (electrode) with every other node over the entire scalp (i.e., all-to-all ISPC) and found out that long-range phase synchronization in theta band was stronger at the MFC (centered at FCz) than at any other site, and this effect was larger in error trials than in the correct trials. Moreover, individual differences in this theta band functional connectivity were correlated with the strength of structural connectivity in the corpus callosum and dorsomedial prefrontal white matter pathways. Here, we used the same approach as Cohen (2011), expecting to find that (a) the MFC-centered theta phase synchronization degree will be significantly higher under conflict condition, reflecting a larger number of long-range functional connections between the MFC and other brain areas involved under conflict condition (cf. Cohen, 2014a), and (b) that individual differences in the degree of this MFC synchronization will be related to the efficiency of conflict resolution that is negatively correlated with the behavioral flanker effect.

Second, we examined direct functional connectivity between the MFC and specific ROIs. As mentioned before, the cortical structures functionally involved in sensorimotor conflict tasks form a network of several areas recurrently interconnected with the MFC. Theta band phase synchrony has been observed between the MFC and extrastriate visual areas, posterio parietal sensorimotor areas, motor cortex, and LFC (Cohen & Ridderinkhof, 2013; Nigbur, Cohen, Ridderinkhof, & Stürmer, 2012; van Driel, Ridderinkhof, & Cohen, 2012; Cohen & Cavanagh, 2011), which presumably reflect functional connectivity through which the MFC coordinates sensory gain control and stimulus selection, S–R integration, response selection, and action control, respectively. As alpha band oscillations play a key role in the implementation of stimulus selection in visual areas (Foster & Awh, 2019; van Diepen, Foxe, & Mazaheri, 2019), the executive communication between the MFC and occipital cortex may be carried out by coupling theta signals from MFC with alpha activity on the occipital areas (Fiebelkorn & Kastner, 2020). Thus, to assess theta–alpha coupling, we computed a cross-frequency version of ISPC (Varela, Lachaux, Rodriguez, & Martinerie, 2001; Tass et al., 1998) between midfrontal theta and occipital alpha oscillations. To access connectivity between the MFC and the other task-related regions, namely, the lateral motor areas, posterior parietal area, and LFC, we measured within-frequency coupling, that is, the theta–theta ISPC. If the MFC is an executive hub for stimulus and response selection, the task-relevant brain regions should become more strongly phase-locked to ongoing MFC theta band oscillations under the incongruent condition of the flanker task, indicating an increase of the task-related communication between MFC and these ROIs. Finally, assuming that such executive communication determines the efficiency of task performance, we should observe negative correlations between the flanker effects on connectivity strength and the behavioral flanker effect.

METHODS

Participants

Ninety-one students (62 women) of the Jagiellonian University took part in the study in return for course credits. Their average age was 20.4 years (SD = 1.5 years). The data were collected in three separate sessions (n = 27, 31, and 33; see below for further details). All participants had normal or corrected-to-normal vision and no history of neurological disorders. Informed written consent was obtained from each participant before the experiment.

Stimuli, Apparatus, and Procedure

The task is illustrated in Figure 1. Each trial of the task began with a fixation point displayed at the center of the computer screen. The time of the initial fixation varied randomly between 1000 and 2000 msec (in 150-msec steps, distributed uniformly). The fixation point was continuously displayed, except when it was replaced by visual cues (see below). The target stimulus was an arrow pointing either up or down, presented in the left or right visual field (50/50). The target arrow was flanked by four additional arrows pointing in either the same or the opposite direction, that is, the congruent and incongruent flankers (50/50). A corresponding distractor set consisted of five vertical lines without arrowheads and was simultaneously presented in the visual field opposite to the target. All these stimuli were presented for 180 msec.

Figure 1.

The stimuli used in the task and sequence of events in a trial. The example shows the target surrounded by incongruent flankers and the corresponding distractor set of five vertical lines (without arrowheads) simultaneously presented in the visual field opposite to the target. See Methods section for details.

Figure 1.

The stimuli used in the task and sequence of events in a trial. The example shows the target surrounded by incongruent flankers and the corresponding distractor set of five vertical lines (without arrowheads) simultaneously presented in the visual field opposite to the target. See Methods section for details.

For reasons not of interest to the present research questions, the target was preceded by visual cues. In one group of participants (n = 27), a predictive or neutral cue was presented 1000 msec before target onset and displayed until target offset. The cue was an increase of the fixation cross (by 0.2°) and change of its color from black to red (indicating the incongruent condition) or green (indicating the congruent condition). In the neutral cue condition, the cross remained black. In another group (n = 31), a predictive cue or no cue was presented 1000 msec before target onset. Here, the cue was a color change of the fixation cross from black to either red or green (indicating the incongruent and congruent flanker conditions, respectively), and it was also displayed until the target offset. In the third group (n = 33), the target was preceded in each trial by a spatially valid or neutral orienting cue, displayed at the center of the screen for 100 msec, with 600 msec of cue–target onset asynchrony. The valid cue was a double arrow sign pointing left or right: “<<” or “>>,” whereas the neutral cue was a square composed of the four arms of the two arrows. The valid cue always correctly indicated the target location. All these variables were counterbalanced. Condition order was randomized per participant. In the present set of analyses, only incongruent–congruent differences, that is, the flanker effects, were examined.

The target arrow and the flankers were each 8 mm (0.8°) long. The target and flanker arrowheads were 4 mm wide (0.4°). The length of all five (target and flanker) arrows in the display was 44 mm (4.4°). The arrows' midpoints were displayed 18.5 mm (1.85°) to the left or right of the center of the screen. The fixation cross was 4 mm (0.4°), or 2 mm wide (0.2°). The height and width of the predictive cue was 4 mm (0.4°). The height and width of the orienting cue were 5 mm (0.5°). All stimuli were black and were presented on a light gray background (red, green, blue: 245, 245, 245). PsychoPy software (www.psychopy.org) was used for experimental control.

The participants' task was to identify the direction of the target (middle) arrow and respond by pressing the left or right Ctrl key on the computer keyboard with the left or right index finger, respectively, to the up- and down-pointing target arrows. Speed and accuracy of responses were measured. A new trial began automatically after the response, or after 2000 msec if the participant did not respond. Participants were given written and then verbal instructions describing the task. They were asked to respond to target stimuli as quickly and accurately as possible. They were also carefully instructed to keep central fixation and explained why proper fixation is necessary during EEG measurement. The task began with two practice blocks consisting of 32 trials in total in which participants received accuracy feedback after each response. The experimental session consisted of 640 trials in five blocks (the first group), or 648 trials in three blocks (the second group), or 576 trials in four blocks (the third group). No feedback was given during the experimental trials. Between the blocks, participants were asked to take breaks to rest their eyes. The task lasted up to 1 hr. The whole session lasted up to 90 min.

EEG Data Recording and Preprocessing

EEG was recorded using a BioSemiActiveTwo system with Ag-AgCl electrodes on 64 monopolar locations according to the extended 10–20 system, and two additional electrodes, the common mode sense active electrode and the driven right leg passive electrode, used as reference and ground electrodes, respectively (www.biosemi.com/faq/cms&drl.htm). The vertical EOG was recorded from above and below the left eye, and horizontal EOG (hEOG) was recorded from the external canthi of both eyes. The data were stored at a sampling rate of 256 Hz. Brain-Vision Analyzer software (Version 2) was used for off-line data preprocessing. Data were filtered with a 0.1- to 50-Hz band-pass and a 50-Hz band-rejection filter (Butterworth zero-phase filters, attenuation of 12 dB/octave), rereferenced to linked mastoids, and split into segments from 1000 msec before target onset to 900 msec afterward. Segmented data were referred to the window from −200 to 0 msec of the segment as a baseline. Trials with an incorrect response, or with RT shorter than 200 msec or longer than 1200 msec, were automatically excluded. For response-locked analysis of the lateralized motor activity, data were segmented from 1000 msec before RT to 100 msec afterward. The same trials entered stimulus- and response-locked averages.

To exclude trials with horizontal eye movements, segments were marked if the horizontal EOG activity (right minus left EOG) exceeded ± 40 μV (which corresponds with an eye movement of approximately 2°; van der Lubbe & Woestenburg, 1997) and/or voltage steps between adjacent data points exceeded 5 μV. The marked trials were subsequently removed after inspection if a distinctive “boxcar” shape, characteristic for saccadic eye movements, was found in these marked segments within 0–800 msec after target onset. On average, 18% of segments were removed because of horizontal eye movements. Next, we calculated averaged horizontal EOG waveforms contralateral minus ipsilateral to the target hemifield, to assess whether any residual hEOG activity was present in the remaining data. The average peak amplitude of this hEOG index in the final data sample was 1.7 μV, showing that the exclusion criteria were effective and did not exceed 3 μV in individual averages of 83 participants, indicating that the residual eye movements in those participants averaged less than about 0.1° (Lins, Picton, Berg, & Scherg, 1993). In the remaining eight participants, the peak amplitude was between 3 and 5 μV. Importantly, we found no correlations between the hEOG index and any relevant contra–ipsilateral EEG difference, r ≤ .09, p ≥ .37, indicating that the EEG lateralizations were not contaminated by the residual eye-movement artifacts.

Before ocular correction, gross artifacts at the scalp electrodes were first removed by rejecting segments with overall minimum–maximum voltage differences ≥ 200 μV or with voltage steps between adjacent data points ≥ 80 μV (except AF3, AFz, AF4, Fp1, Fpz, and Fp2, so that fewer segments would be rejected because of blinks). Next, the data from remaining trials were corrected for eye blinks and artifacts that had a noncortical origin using independent component analysis (Jung et al., 2000), and then, after resetting the baseline, edited for other artifacts by rejecting trials with overall minimum–maximum voltage differences ≥ 120 μV, voltage steps between adjacent data points ≥ 20 μV, and absolute amplitudes ≥ 120 μV. On average, 7% of segments (range: 1%–21%) were excluded because of artifacts. The average of accepted segments was 226 (range: 145–290) for the congruent condition and 214 (range: 131–292) for the incongruent condition.

ERP Measurement and Analysis

Before ERP analysis, artifact-free segments were averaged over each condition separately for each participant. All the ERP components of interest had fairly typical topographies and were measured at the sites at which their peaks were largest. The effects of flanker manipulation on the ERP measures were statistically evaluated by repeated-measures ANOVAs with the factor Flanker (congruent, incongruent).

To measure PCN, we calculated ERLs (Wascher & Wauschkuhn, 1996) from activity recorded over visual cortex at PO7 and PO8, as the average of contra–ipsilateral differences for trials with the LVF and RVF targets, by the formula: ((PO8LVF − PO7LVF) + (PO7RVF − PO8RVF)) / 2. Thus, a negative ERL value indicates larger negativity at the hemisphere contralateral to the target visual field. Based on previous studies and inspection of the obtained grand averages, we measured the N2pc component of the PCN as mean amplitude 230–270 msec after target onset, and the sustained contralateral negativity (SPCN) as mean amplitude 360–600 msec after target onset. The latencies of the ERLs did not differ between the conditions (see Figure 3A and 3B); thus, their analysis is not reported here, for brevity. To obtain topographies of the ERLs, we subtracted all symmetrical electrodes and plotted the averaged contra–ipsilateral differences on the left hemisphere (Figure 3B).

To measure the LRP we calculated ERLs from activity recorded over motor cortex at C3 and C4, as the average of the difference contra–ipsilateral relative to the responding hand: ((C4Left Hand − C3Left Hand) + (C3Right Hand − C4Right Hand)) / 2 (Coles, 1989), so that a negative value reflects activation of the correct response. In the stimulus-locked averages, the LRP amplitudes were measured as mean amplitudes 260–560 msec for congruent condition and 350–650 msec for the incongruent condition. In the response-locked averages, the amplitudes were measured as mean activity from −130 to −70 msec before the response. The LRP latencies were calculated by means of the 50% area under the curve measure, that is, by assessing the time point that divides the negative area under the LRP waveforms into two halves (Kiesel, Miller, Jolicœur, & Brisson, 2008).

The midfrontal N2 component was measured at FCz as a mean amplitude 260–320 msec after target onset. We also measured a positive component at FCz that preceded the N2, which we classified as P2, as mean amplitude 200–240 msec after target onset. The P3b component was measured at Pz. As seen in Figure 6, the P3b waves differed between the flanker conditions in terms of both its amplitude and its latency; thus, we measured both parameters. The P3b amplitude was measured as mean voltage 300–700 msec after target onset, and the P3b latency was determined using the 50% area measure in the time-window 200–800 msec after cue onset.

Time–Frequency Analysis

Power and Phase Decomposition

To obtain phase and power of ongoing EEG oscillations, the segmented artifact-free single-trial EEG data were decomposed into their time–frequency representation using custom-written MATLAB 2018a scripts based on Cohen (2014b, 2017). The power spectrum of the EEG was multiplied by the power spectrum of complex Morlet wavelets, defined as follows:
ei2πftet2/2σ2
where i is the complex operator, t is time, f is frequency (increased from 1 to 40 Hz in 40 logarithmically spaced steps), and σ is the width of each frequency band defined as n/(2πf), where n is a number of wavelet cycles. The wavelet cycles varied from four to eight in logarithmically spaced steps, to obtain an optimal tradeoff between temporal and frequency resolution (Trujillo & Allen, 2007). To avoid edge artifacts, we applied a signal reflection procedure on the segmented time-series data (Cohen, 2017). The convolution was calculated in frequency domain for each electrode and relevant experimental condition per participant. After the convolution, from the resulting analytic signal, we calculated frequency-band-specific phase and power. Phase angle values of the trial were defined as follows:
Φkfnt=argReykfntImykfnt
where arg{…} stands for arc-tangent, Re for the real part of the signal, and Im for the imaginary part of the signal.
Power (i.e., the magnitude) of the signal yk was estimated by averaging Wk over k-th EEG trial:
Wkfnt=Reykfnt2+Imykfnt2

Intertrial Phase Coherence

To measure the stimulus selection-related contra–ipsilateral modulations of phase synchronization in the alpha band, we calculated frequency-band-specific intertrial phase coherence (ITPC; Benwell, Keitel, Harvey, Gross, & Thut, 2018; Cavanagh, Cohen, & Allen, 2009; or intertrial coherence, Busch, Dubois, & VanRullen, 2009; also called as intertrial phase-locked value, Wang, Goschl, Friese, Konig, & Engel, 2019; Lachaux, Rodriguez, Martinerie, & Varela, 1999). The ITPC is an estimation of consistency between phase angle values at a given electrode site for a specific frequency and time point across trials (or realizations), according to the following formula:
ITPCkfnt=1kt=1keiΦkfnt

The ITPC values vary from 0 to 1, where 0 indicates random phase and 1 indicates identical phase values, which is the perfect phase coherence, at a given time–frequency point across trials. For statistical analysis, ITPC was measured from the PO7–8 electrode pair within a 50- to 300-msec 9- to 14-Hz time–frequency window (see Figure 3D and 3E). The ITPC was then submitted to a repeated-measures ANOVA with Flanker (congruent, incongruent) and Hemisphere (contralateral, ipsilateral to the target hemifield) as within-subject factors.

LPS in Alpha and Beta Bands

To measure stimulus selection-related contra–ipsilateral modulations of alpha power, we calculated the LPS index (van der Lubbe et al., 2019; van der Lubbe & Utzerath, 2013). First, the wavelet analysis described above was carried out for two flanker conditions, and hemispheres contralateral and ipsilateral to the target hemifield. Next, the LPS index was calculated for all symmetrical electrode pairs from the condition averaged power by a double subtraction similar to the ERL calculations. That is, a single subtraction (ipsilateral − contralateral) was calculated separately for the left and right target trials, scaled by the sum of activation of both hemispheres (ipsilateral + contralateral), and then averaged, which may be expressed by the following formula:
LPSkfnt=xkfntykfntykfnt+xkfnt
where xk is trial-averaged power from the ipsilateral channel and yk is power from the contralateral side, both measured within the same frequency band (fn = fn) and time-window. The LPS values vary from −1 to +1. A positive LPS value indicates a smaller power at the contralateral site relative to the ipsilateral site (which may be interpreted as desynchronization of a given frequency at the contralateral site), and zero indicates no hemispheric difference. To measure hand-movement-related contra–ipsilateral modulations of beta power, we computed LPS in the same way as the alpha LPS, but for contralateral and ipsilateral sites relative to the responding hand. For statistical analysis, the amplitudes of LPS in alpha band (9–14 Hz) were measured as mean activity 400–900 msec after target onset from the PO7–8 electrode pair, where the peaks were largest (see Figure 3G). The amplitude of motor-related LPS in beta band (18–25 Hz) was measured from the C3–4 electrode pair as the mean activity in a 300- to 900-msec time-window in the stimulus-locked averages, and in a −300- to 0-msec time-window in the response-locked averages. The LPS latencies were estimated using the 50% area measure. Flanker effects on the LPS were tested by repeated-measures ANOVAs. To improve the signal-to-noise ratio, the latencies were determined using the jackknife method: The parameters were measured in one-leave-out grand means, and the diminished error variance was corrected by dividing F values by (n − 1)2 (Kiesel et al., 2008; Ulrich & Miller, 2001).

Non-phase-locked Midfrontal Theta Power

To obtain the non-phase-locked power, the phase-locked activity (i.e., the ERP, calculated as the time-domain trial average) was subtracted from the time-domain single-trial EEG signal, before the described above time–frequency decomposition (Cohen & Donner, 2013). The subtractions and then wavelet convolutions were done for each electrode and flanker condition per participant. For statistical analysis, the non-phase-locked power from the FCz site was averaged over a 400- to 600-msec, 4- to 8-Hz time–frequency window, specified based on both previous studies and inspection of the current data, and submitted into a repeated-measures ANOVA with Flanker (congruent, incongruent) as a within-subject factor.

Time–Frequency Connectivity Calculations

ISPC

Before connectivity analysis, a surface Laplacian filter (also called current source density or current scalp density; Cohen & Donner, 2013; Srinivasan, Winter, Ding, & Nunez, 2007) was applied to the artifact-free single-trial data (we used a 10th-order Legendre polynomial, and lambda was set at 1e−5) to increase spatial selectivity and minimize volume conduction. To calculate estimations of functional connectivity between the brain areas, we used the phase angles obtained from the complex wavelet convolution (as described above) and computed frequency-band-specific ISPC (Cohen, Ridderinkhof, Haupt, Elger, & Fell, 2008; also called intersite phase connectivity, Cohen, 2015; intersite phase clustering, van Driel et al., 2015; Gulbinaite et al., 2014; interchannel phase synchrony, Cavanagh et al., 2009; and between-region phase-locking value or phase-locked value, Fiebelkorn, Pinsk, & Kastner, 2018). The ISPC provides a measure of a difference in time–frequency phase angles between two sites (electrodes or sources of oscillations) xk (fn, t) and yk (fn, t) within the same specific frequency band (fn = fn) at each time point across trials, according to the following formula:
ISPCxykfnt=1kt=1keiΦxkfntΦykfnt

The ISPC value varies from 0 to 1, where 0 indicates no phase synchrony between two sites (i.e., random phases), and 1 indicates identical phase values at two sites (i.e., a perfect phase coherence between the two sites).

Graph Theory-based Estimation of All-to-all Theta ISPC

To test the hypothesis that the MFC is a hub for theta-based connectivity, we calculated the ISPC in theta (4–8 Hz) band for each pair of electrodes (all-to-all) and applied the mathematical framework of graph theory to identify the electrodes that had most robust synchronization degree with other electrodes (Cohen & van Gaal, 2013; Cohen, 2011). Graph theory has been successfully used to quantify various aspects of local and global network characteristics (Stam et al., 2009). Here, the network consisted of 64 nodes (electrodes) and 63 vertices per node (connections of each node to all other nodes). Synchronization degree (or strength) of a given node is defined as a number of other nodes with whom this node is strongly phase synchronized (i.e., the number of connections exceeding a threshold). In detail, first, we created weighted links (connections) based on pairwise ISPC connectivity vector. Second, all links have been binarized by thresholding: If a link exceeded the threshold, the value was replaced by 1; if it was below the threshold, the value was replaced by 0. Thus, the synchronization degree may be expressed as follows:
SDinfn=iN,j=1naijfn
where n is the total number of relevant connections between electrodes and aij refers to the connectivity between electrodes i and electrode j. For thresholding, we employed a subject-specific data-based approach and set the threshold at 1 SD above the median of each participant's ISPC (cf. Cohen, 2011). For statistical analysis, synchronization degree was calculated separately for each theta band frequency (4–8 Hz) in a 400- to 600-msec time-window, electrode, and flanker condition, per participants, and submitted into a repeated-measures ANOVA with the factor Flanker.

MFC-to-ROI Theta Band Connectivity

To examine theta band connectivity between MFC and lateral frontal, centro-parietal, and lateral motor areas, we specified three ROIs—all based on both previous findings and data inspection—and computed ISPC between the midfrontal FCz electrode used here as “seed” and these ROIs. The AF3, AF4, F3, and F4 sites were used as lateral frontal ROI (cf. van Driel et al., 2015; Gulbinaite et al., 2014), the CPz and Pz sites were used as centro-parietal ROI (cf. Gulbinaite et al., 2014), and the C5–6 and CP5–6 electrode pairs were used as lateral motor ROI (cf. van Driel et al., 2015; Nigbur et al., 2012). After inspection of time–frequency plots (Figure 7), we decided to measure ISPC between FCz and lateral frontal and centro-parietal ROIs over two subsequent time–frequency windows: 100–400 msec 5–6 Hz, and 400–700 msec 5–6 Hz. For statistical analysis, the data were pooled across the ROI electrodes, and submitted into a repeated-measures ANOVA with Flanker and Time-window as within-subject factors. For ISPC between FCz and motor ROI, a time–frequency window of 500–800 msec and 4–6 Hz was used (see Figure 8A), the data were pooled across the ROI electrodes and submitted into a repeated-measures ANOVA with Flanker (congruent, incongruent) and Hemisphere (contralateral, ipsilateral) as within-subject factors. Although all our statistical analyses were generally specified a priori, the details of the ISPC analyses were further determined after inspection of the data. Therefore, to reduce the chance of false positives, all ISPC effects were additionally evaluated by nonparametric permutation tests with cluster-based corrections for multiple comparisons across all time–frequency points (Maris & Oostenveld, 2007). One thousand iterations of randomization were performed for each analysis. At each iteration, the ISPC values were randomly shuffled across the conditions. Next, a two-tailed t test was performed for each time–frequency point against a null-distribution of the condition difference. After all iterations, a null-hypothesis distribution was created, against which the observed t-test results were compared at p < .001. Finally, the obtained t value map was corrected at cluster level and thresholded at p < .01, so that any cluster in the real data that was equal or greater than 99% of the null distribution of clusters was considered statistically significant.

Cross-frequency Coupling

Finally, to investigate connectivity between the MFC and visual areas, we measured cross-frequency ISPC (cf-ISPC) between MFC theta and stimulus selection-related occipital alpha. To quantify cf-ISPC, we estimated a nonlinear phase synchronization between the lower (theta) frequency in the seed channel and the higher (alpha) frequency in the target ROI (Varela et al., 2001; Tass et al., 1998). The FCz electrode was used as “seed,” as before, and PO7–8 and O1–2 electrode pairs were used as occipital ROI (cf. Cohen & van Gaal, 2013; Cohen, van Gaal, Ridderinkhof, & Lamme, 2009). Specifically, we quantified the cross-frequency version of ISPC for each of 16 pairs of theta (4–7 Hz) and alpha (9–12 Hz) frequency bands (4–9 Hz, 4–10 Hz,…, 7–12 Hz), according to the following formula:
cfISPCxykf1f2t=1kt=1keinΦxkf1tmΦykf2t
where f1 and f2 are two different frequency bands (f1f2) of which ratio is given by integers n and m (nf1 = mf2, n = 1). Whereas xk indicates seed region at the theta (4−7 Hz) frequency, and theta phase values are normalized by an integer n to match the temporal resolution of the periodic cycles of the target region yk, which operates on the alpha (9–12 Hz) frequency band (Palva, Palva, & Kaila, 2005). As for ISPC, the cf-ISPC values vary from 0 to 1, where 0 indicates no phase synchrony between two sites (i.e., random phases), and 1 indicates identical phase consistency between the two sites. For statistical analysis, cf-ISPC was computed as percentage change relative to a 200-msec pretarget time-window, by the formula: [100 × ((cfISPCxycfISPCxybs)/cfISPCxybs)], where bs is the frequency and condition-specific average of cf-ISPC over the −300- to −100-msec time-window relative to target onset. Furthermore, cf-ISPC was measured over the 50- to 300-msec time-window (see Figure 9B) separately for each of the 16theta–alpha frequency pairs, two flanker conditions, and two hemisphere conditions (contralateral, ipsilateral to the target hemifield), per participant. The data were pooled across the ROI electrodes and submitted into a 16 × 2 × 2 repeated-measures ANOVA with Frequency Pair, Flanker, and Hemisphere as within-subject factors.

Correlation Analysis

Finally, to assess the relationship between the behavioral and electrophysiological effects of the flanker manipulation, individual differences in the behavioral performance were correlated with individual differences in the electrophysiological activity. To this end, the incongruent minus congruent difference was calculated for each ERP, time–frequency, and connectivity measure, as indices of the EEG flanker effects (in shorthand notation, these differences will be referred to as the N2pc flanker effect, the P3 flanker effect, etc.), and the Pearson correlation coefficient was calculated to test for linear associations between these indices and the RT flanker effect. All reported p values are two-tailed.

RESULTS

Behavioral Results

The average RT of correct responses was 580 msec (SD = 65) in the congruent condition and 678 msec (SD = 87) in the incongruent condition, F(1, 90) = 319.08, p < .001, ηp2 = .78, which gives a RT flanker effect of 98 msec (SD = 52). This incongruent–congruent difference correlated with the average RT in the incongruent condition, r = .67, p < .001, and not with the RT in the congruent condition, r = .09, p = .41 (Figure 2). To estimate the reliability of the RT flanker effect for individual participants, we calculated a split-half correlation between the first and second halves of the task, and an odd–even correlation between the odd-numbered and even-numbered trials (see Figure 2). The Pearson coefficient was r = .82, p < .001, for the split-half reliability, and r = .92, p < .001, for the odd–even reliability.

Figure 2.

The accuracy and reliability of the RT flanker effect. The top shows scatter plots for correlations between the RT flanker effect (incongruent minus congruent) and the average RT in two flanker conditions. The RT flanker effect did correlate with the average RT in the incongruent condition and did not correlate significantly with the RT in the congruent condition, proving to be an accurate index of processes triggered specifically by the incongruent flankers. The bottom shows scatter plots for correlations of the RT flanker effect between the odd-numbered and even-numbered trials, and between the first and second halves of the task. Both the odd–even and the split-half correlations show satisfactory reliability estimations for the RT flanker effect.

Figure 2.

The accuracy and reliability of the RT flanker effect. The top shows scatter plots for correlations between the RT flanker effect (incongruent minus congruent) and the average RT in two flanker conditions. The RT flanker effect did correlate with the average RT in the incongruent condition and did not correlate significantly with the RT in the congruent condition, proving to be an accurate index of processes triggered specifically by the incongruent flankers. The bottom shows scatter plots for correlations of the RT flanker effect between the odd-numbered and even-numbered trials, and between the first and second halves of the task. Both the odd–even and the split-half correlations show satisfactory reliability estimations for the RT flanker effect.

Overall response accuracy was near ceiling, but there was still a significant difference of 3% between the congruent (97%, SD = 2) and incongruent (94%, SD = 5) conditions, F(1, 90) = 38.00, p < .001, ηp2 = .30. There was also a positive correlation between the flanker effects in accuracy and RT, r = .31, p = .003.

Lastly, we examined the effect of spatial S–R correspondence, that is, the Simon effect (Simon, 1969), because it might interact with the flanker effect. Typically, left-hand responses are faster if the stimulus also appears on the left rather than on the right side, and vice versa. Nevertheless, no Simon effect has been observed in this study, F(1, 90) = 1.91, p = .17, ηp2 = .02, which replicates results from previous studies combining the flanker and Simon tasks (Fan et al., 2009).

Stimulus Selection-related Activity at the Visual Cortex

Introducing bilateral stimulation to the flanker task and applying the double subtraction method enabled us to isolate the stimulus selection-related lateralized signal from the overall visual activity. To thoroughly examine the effects of flanker interference on visual selection, and their relationship to behavioral performance, we analyzed three complementary features of the signal: ERPs, time–frequency power, and ITPC.

PCN (N2pc and SPCN)

Figure 3A shows grand averages of the ERPs recorded from electrodes over the visual cortex (PO7, PO8) at the hemispheres contralateral and ipsilateral to the target hemifield in two flanker conditions. The ERPs formed typical waveforms with P1 and N1 peaks, which tend to have slightly larger amplitudes at the contralateral site,1 followed by a conspicuous negative deflection at the contralateral site at about 250 msec after target onset, which we identified as the N2pc component. This contralateral negativity continued onward, forming an SPCN. Grand averages of the contra–ipsilateral differences in which the ERLs are isolated from the rest of the ERPs are shown in Figure 3B. The N2pc is evident as the largest negative peak, followed by the SPCN wave. All the ERLs had very similar occipito-temporal topographies centered at PO7–8. Of importance, the N2pc amplitude was significantly larger in the incongruent condition, F(1, 90) = 12.87, p < .001, ηp2 = .12, and so was the amplitude of the SPCN, F(1, 90) = 77.88, p < .001, ηp2 = .46. Moreover, we found a positive correlation between the N2pc flanker effect and the RT flanker effect, r = .29, p = .006 (instead of the expected negative correlation; see Figure 3C), as well as between the N2pc flanker effect and the average RT in the incongruent trials, r = .23, p = .030. In contrast, the SPCN flanker effect did not correlate with the RT flanker effect nor the average RTs, r ≤ −.02, ns.

Figure 3.

Stimulus selection-related activity recorded over visual cortex (PO7–8). For all panels: Time point zero is the onset of target and flankers. Colors of lines and frames denote the conditions. (B, D, and F) The topographic maps were obtained by subtracting all the symmetrical electrodes (contra–ipsilateral) and plotting the difference on the left hemisphere. The maps are min–max scaled, with positive polarity in red and negative polarity in blue. The head view is from above. (A) Visual-evoked potentials. Grand means of the ERPs from the hemispheres contralateral and ipsilateral to the target visual field in the congruent and incongruent flanker conditions, pooled across the occipito-temporal sites PO7 and PO8. Negative voltage points upward. The gray areas indicate the time-windows in which the N2pc and SPCN components were measured. (B) ERLs. Grand means of the contra–ipsilateral differences from the PO7–8 electrode pair. Negative values (plotted upward) indicate a larger posterior negativity contralateral to the target visual field. The gray areas indicate the time-windows in which the N2pc and SPCN components were measured. The head maps depict topographies of these components within the indicated time-windows. (C) Scatter plot for correlations between the RT flanker effect and the N2pc flanker effect. The flanker effects are calculated as a difference between the incongruent and congruent conditions. (D) ITPC. Grand means of ITPC in alpha band (9–14 Hz) from the hemispheres contralateral and ipsilateral to the target hemifield in the congruent and incongruent flanker conditions, pooled across the occipito-temporal sites PO7 and PO8. The plot reveals a contralateral advantage of the alpha phase coherence. The gray area indicates the time-window in which the ITPC means were measured. (E) Lateralized ITPC. Time–frequency representations of the contra–ipsilateral difference of the ITPC from the PO7–8 electrode pair in two flanker conditions reveal the contralateral ITPC increase in the alpha band (9–14 Hz). Topographical maps show the localization of the contra–ipsilateral difference of the alpha ITPC for relevant time-windows. (F) Lateralized power over the PO7–8 electrode pair. The plots show time–frequency power representations of the contra–ipsilateral difference in two flanker conditions. Note the reduced contralateral alpha activity (9–14 Hz) peaking at about 600–700 msec after stimulus onset. (G) LPS. Grand means of the LPS estimations for the alpha band (9–14 Hz) from the PO7–8 electrode pair in two flanker conditions. The positive LPS deflection (plotted upward) indicates that the contralateral alpha power was reduced as compared to the ipsilateral power. The gray area indicates the time-window in which the alpha LPS were measured. Topographical maps show the alpha LPS in the indicated time-window.

Figure 3.

Stimulus selection-related activity recorded over visual cortex (PO7–8). For all panels: Time point zero is the onset of target and flankers. Colors of lines and frames denote the conditions. (B, D, and F) The topographic maps were obtained by subtracting all the symmetrical electrodes (contra–ipsilateral) and plotting the difference on the left hemisphere. The maps are min–max scaled, with positive polarity in red and negative polarity in blue. The head view is from above. (A) Visual-evoked potentials. Grand means of the ERPs from the hemispheres contralateral and ipsilateral to the target visual field in the congruent and incongruent flanker conditions, pooled across the occipito-temporal sites PO7 and PO8. Negative voltage points upward. The gray areas indicate the time-windows in which the N2pc and SPCN components were measured. (B) ERLs. Grand means of the contra–ipsilateral differences from the PO7–8 electrode pair. Negative values (plotted upward) indicate a larger posterior negativity contralateral to the target visual field. The gray areas indicate the time-windows in which the N2pc and SPCN components were measured. The head maps depict topographies of these components within the indicated time-windows. (C) Scatter plot for correlations between the RT flanker effect and the N2pc flanker effect. The flanker effects are calculated as a difference between the incongruent and congruent conditions. (D) ITPC. Grand means of ITPC in alpha band (9–14 Hz) from the hemispheres contralateral and ipsilateral to the target hemifield in the congruent and incongruent flanker conditions, pooled across the occipito-temporal sites PO7 and PO8. The plot reveals a contralateral advantage of the alpha phase coherence. The gray area indicates the time-window in which the ITPC means were measured. (E) Lateralized ITPC. Time–frequency representations of the contra–ipsilateral difference of the ITPC from the PO7–8 electrode pair in two flanker conditions reveal the contralateral ITPC increase in the alpha band (9–14 Hz). Topographical maps show the localization of the contra–ipsilateral difference of the alpha ITPC for relevant time-windows. (F) Lateralized power over the PO7–8 electrode pair. The plots show time–frequency power representations of the contra–ipsilateral difference in two flanker conditions. Note the reduced contralateral alpha activity (9–14 Hz) peaking at about 600–700 msec after stimulus onset. (G) LPS. Grand means of the LPS estimations for the alpha band (9–14 Hz) from the PO7–8 electrode pair in two flanker conditions. The positive LPS deflection (plotted upward) indicates that the contralateral alpha power was reduced as compared to the ipsilateral power. The gray area indicates the time-window in which the alpha LPS were measured. Topographical maps show the alpha LPS in the indicated time-window.

Lateralized Alpha Power

Time–frequency power representations of the contra–ipsilateral difference from PO7–8 electrode pair in two flanker conditions are shown in Figure 3F, and grand averages of LPS estimations for the alpha band (9–14 Hz) from PO7–8, along with their topographical maps for the relevant time-window, are shown in Figure 3G. As seen in the figures, the alpha LPS formed an early negative peak at about 140 msec after target onset and then reversed its polarity forming a positive deflection that peaked between 500 and 800 msec at the PO7–8 site. This positive LPS deflection (9–14 Hz, 400–900 msec) was significantly larger than 0 in both flanker conditions: congruent, t(90) = 4.44, p < .001, and incongruent, t(90) = 9.09, p < .001, indicating that the contralateral alpha power was significantly reduced as compared to the ipsilateral power. Of importance, this late positive LPS component was larger and had longer latency in the incongruent condition than in the congruent condition; F(1, 90) = 37.98, p < .001, ηp2 = .30, and F(1, 90) = 13.55, p < .001, ηp2 = .13, respectively. The correlation analysis showed no significant relationship between the alpha LPS flanker effect and the RT flanker effect, r ≤ .11, p ≥ .28. Still, we found a negative correlation between the alpha LPS flanker effect and the SPCN flanker effect, r = −.27, p = .010. Correlations with the flanker effects on the other ERLs were not significant r ≤ −.05, ns.2

Lateralized Alpha Phase

Figure 3D shows grand averages of ITPC in the alpha band (9–14 Hz) recorded from the PO7 and PO8 sites at the hemispheres contralateral and ipsilateral to the target hemifield in two flanker conditions. Figure 3E depicts time–frequency power representations of the contra–ipsilateral difference of the ITPC from the PO7–8 electrode pair in two flanker conditions and topographical maps of the alpha ITPC for relevant time-windows. As evident in the figures, the occipito-temporal alpha (9–14 Hz) phase coherence peaked at about 160 msec after target onset and was significantly stronger at the contralateral site than at the ipsilateral site, F(1, 90) = 159.70, p < .001, ηp2 = .64. Neither the main flanker effect nor the Flanker × Hemisphere interaction was significant, F ≤ 2.26, p ≥ .13. However, we found that the contra–ipsilateral ITPC difference (9–14 Hz, 50–300 msec) in the incongruent condition was negatively correlated with the RT flanker effect, r = −.21, p = .042, indicating that an increase of the contralateral phase coherence was related to a lower conflict cost in RTs. The correlation of the contra–ipsilateral ITPC difference in the congruent condition with the RT flanker effect was weaker and not significant, r = −.18, p = .09.

Response-related Motor Activity

This section describes the effects of flanker manipulation on the lateralized motor activity, measured as LRPs and LPS in the beta band. Both were calculated separately for stimulus-locked and response-locked analyses.

Hand-movement-related LRP

Figure 4 shows grand averages and topographies of the stimulus-locked LRPs (upper left panel) and response-locked LRPs (upper right panel) recorded over the motor cortex (C3–4) in two flanker conditions. As expected, the stimulus-locked LRPs showed a significant flanker effect on latencies, whereas the response-locked LRPs did not. Still, the response-locked LRPs had a larger amplitude in the incongruent condition. In detail, the stimulus-locked results showed that, in the congruent trials, the LRP formed a negative peak reflecting activation of the correct response about 90 msec earlier than in the incongruent trials, which corresponds with the similar delay of behavioral responses. This latency difference was significant, F(1, 90) = 119.78, p < .001, ηp2 = .57, and so was, obviously, the correlation between this LRP flanker effect and the RT flanker, r = .31, p = .003.3 The amplitudes of the stimulus-locked LRPs did not differ between flanker conditions, F(1, 90) = 2.07, p = .15, ηp2 = .02, and did not correlatewith the RT index, r = .06, p = .60. In the response-locked averages, the LRP latencies did not differ, F < 1.0, whereas the LRP amplitude was significantly larger in the incongruent condition, F(1, 90) = 5.06, p = .028, ηp2 = .05, indicating that, although the timing of response execution was not affected by response conflict, the strength of the activation of the correct response program was affected. Of importance, this flanker effect on the LRP amplitudes was positively correlated with the RT flanker effect, r = .20, p = .05.

Figure 4.

Hand-movement-related activity recorded over motor cortex (C3–4). Upper panels: LRPs. Negative voltage (plotted upward) represents activation of the correct response (contralateral to the responding hand). Bottom parts: LPS in the beta band (18–25 Hz). Positive values (plotted upward) represent a contralateral beta power decrease relative to the responding hand. In the left panels, time point zero is the onset of target and flankers. In the right panels, time point zero indicates the time of the response. For both panels: Colors of lines and frames denote the conditions. The head maps show topographic distribution of the LRP and LPS maxima. The topographic maps were obtained by subtracting all the symmetrical electrodes (contra–ipsilateral) and plotted the difference on the left hemisphere. The maps are min–max scaled, with positive polarity in red and negative polarity in blue. The head view is from above.

Figure 4.

Hand-movement-related activity recorded over motor cortex (C3–4). Upper panels: LRPs. Negative voltage (plotted upward) represents activation of the correct response (contralateral to the responding hand). Bottom parts: LPS in the beta band (18–25 Hz). Positive values (plotted upward) represent a contralateral beta power decrease relative to the responding hand. In the left panels, time point zero is the onset of target and flankers. In the right panels, time point zero indicates the time of the response. For both panels: Colors of lines and frames denote the conditions. The head maps show topographic distribution of the LRP and LPS maxima. The topographic maps were obtained by subtracting all the symmetrical electrodes (contra–ipsilateral) and plotted the difference on the left hemisphere. The maps are min–max scaled, with positive polarity in red and negative polarity in blue. The head view is from above.

Hand-movement-related Lateralized Beta Power

The bottom parts of Figure 4 show LPS estimates for the beta band (18–25 Hz) from the C3–4 electrode pair in two flanker conditions and their topographical maps for relevant time-windows. In the stimulus-locked averages, the beta band LPS formed a positive deflection that peaked at about 400–500 msec in the congruent condition and about 550–650 msec in the incongruent condition. Statistical analysis showed that the stimulus-locked LPS (400–900 msec) was significantly larger than 0 in both flanker conditions, t(90) ≥ 4.05, p < .001; its latency was significantly longer in the incongruent condition, F(1, 90) = 14.35, p < .001, ηp2 = .14; and its amplitude was marginally larger in the congruent condition, F(1, 90) = 3.78, p = .055, ηp2 = .04. In the response-locked averages, the beta band LPS peaked at about 80 msec before response execution was significantly larger than 0 in both conditions, t(90) ≥ 2.83, p ≤ .006, and was significantly larger in the congruent than in the incongruent condition, F(1, 90) = 5.10, p = .026, ηp2 = .05. Unlike the stimulus-locked LPS, however, the latencies of the response-locked LPS did not differ, F < 1.0. Still, Figure 4 (lower right panel) suggests that their onsets might differ significantly. Thus, we estimated the onset latencies by means of the 50% of peak latency measure, which defines the onset latency as the time point at which the wave reaches 50% of the peak value (Kiesel et al., 2008), and confirmed that there was no significant difference there, F(1, 90) = 1.52, p = .22, ηp2 = .02. The beta LPS flanker effects did not correlate significantly with the RT flanker effect, r ≤ −.08, ns2.

Conflict-related Midfrontal Activity

Here, we present conflict-related and behaviorally relevant modulations of the midfrontal activity, measured as the flanker effects on the N2 component of the ERPs, non-phase-locked power, and midfrontal-centered global theta phase synchronization.

Conflict-related N2 Component of the ERP

Grand-average waveforms from FCz are depicted in Figure 5A. The second negative peak reached its maximum at about 290 msec after target onset and had the expected midfrontal topography centered at FCz, thus was classified as the N2. Importantly, the N2 amplitude was larger in the incongruent condition than in the congruent condition, F(1, 90) = 11.93, p < .001, ηp2 = .12, and this difference, that is, the N2 flanker effect, was negatively correlated with the RT flanker effect, r = −.38, p < .001 (see Figure 5A). We also quantified the positive peak that preceded the N2, which we classified as P2 (200–240 msec), and found out that its amplitude did not differ significantly between the flanker conditions, F(1, 90) = 1.97, p = .16, ηp2 = .02.

Figure 5.

Conflict-related midfrontal activity. (A) Conflict-related N2 component of the ERPs. Left panel: grand means of the ERPs from FCz in two flanker conditions. Negative voltage points upward. Time point zero is the onset of target and flankers. The gray area indicates the time-window within the N2 mean amplitude was measured. Middle panel: topographic distribution of the incongruent–congruent difference (i.e., the flanker effect) in the indicated N2 time-window. The map is min–max scaled, with positive polarity in red, negative polarity in blue. The head view is from above. Right panel: a scatter plot for correlation between the N2 flanker effect and the RT flanker effect. As seen on the figure, the N2 amplitude was significantly larger in the incongruent condition, and this conflict-related N2 peaked at FCz and its amplitude was negatively correlated with the behavioral conflict cost. (B) Conflict-related theta power modulation. Left panel: Time–frequency plot for the incongruent–congruent difference of the non-phase-locked power (total power minus phase-locked power, see Methods section for details) shows an increase of theta power in the incongruent condition within a 400–600 msec, 4- to 8-Hz time–frequency window. Time point zero is the onset of target and flankers. Middle panel: The topographical map for this window displays the location of the effect at FCz. Right panel: The scatter plot depicts a significant negative correlation between the theta power and RT flanker effects. (C) Conflict-related theta phase synchronization. Left panel: Topographical distributions of theta band (4–8 Hz) phase synchronization degree in the relevant time–frequency window for two flanker conditions. “Hotter” colors indicate a greater number of cortico-cortical functional connections. The head view is from above. Colors of frames denote flanker conditions. Synchronization degree is a graph theory-based quantification of the number of functional connections of each node of a network (electrode) with every other node of this network (see Methods section for details). The obtained result suggests that the MFC is a “hub” of global connectivity in the theta band. Middle panel. The topographical map of the incongruent–congruent difference for the relevant time–frequency window shows that the flanker conflict entailed an increase of the global synchronization in the theta band, indicating a greater number of the MFC-centered functional connections in the conflict trials. Right panel: The scatter plot depicts a correlation between the flanker effect on the conflict-related theta band connectivity and the flanker effect on the RTs. The result shows that individual differences in the degree of the MFC connectivity are negatively correlated with the RT flanker effect, that is, with the efficiency of conflict resolution.

Figure 5.

Conflict-related midfrontal activity. (A) Conflict-related N2 component of the ERPs. Left panel: grand means of the ERPs from FCz in two flanker conditions. Negative voltage points upward. Time point zero is the onset of target and flankers. The gray area indicates the time-window within the N2 mean amplitude was measured. Middle panel: topographic distribution of the incongruent–congruent difference (i.e., the flanker effect) in the indicated N2 time-window. The map is min–max scaled, with positive polarity in red, negative polarity in blue. The head view is from above. Right panel: a scatter plot for correlation between the N2 flanker effect and the RT flanker effect. As seen on the figure, the N2 amplitude was significantly larger in the incongruent condition, and this conflict-related N2 peaked at FCz and its amplitude was negatively correlated with the behavioral conflict cost. (B) Conflict-related theta power modulation. Left panel: Time–frequency plot for the incongruent–congruent difference of the non-phase-locked power (total power minus phase-locked power, see Methods section for details) shows an increase of theta power in the incongruent condition within a 400–600 msec, 4- to 8-Hz time–frequency window. Time point zero is the onset of target and flankers. Middle panel: The topographical map for this window displays the location of the effect at FCz. Right panel: The scatter plot depicts a significant negative correlation between the theta power and RT flanker effects. (C) Conflict-related theta phase synchronization. Left panel: Topographical distributions of theta band (4–8 Hz) phase synchronization degree in the relevant time–frequency window for two flanker conditions. “Hotter” colors indicate a greater number of cortico-cortical functional connections. The head view is from above. Colors of frames denote flanker conditions. Synchronization degree is a graph theory-based quantification of the number of functional connections of each node of a network (electrode) with every other node of this network (see Methods section for details). The obtained result suggests that the MFC is a “hub” of global connectivity in the theta band. Middle panel. The topographical map of the incongruent–congruent difference for the relevant time–frequency window shows that the flanker conflict entailed an increase of the global synchronization in the theta band, indicating a greater number of the MFC-centered functional connections in the conflict trials. Right panel: The scatter plot depicts a correlation between the flanker effect on the conflict-related theta band connectivity and the flanker effect on the RTs. The result shows that individual differences in the degree of the MFC connectivity are negatively correlated with the RT flanker effect, that is, with the efficiency of conflict resolution.

Conflict-related Theta Power

Figure 5B shows a time–frequency plot for the incongruent–congruent difference of non-phase-locked power (total power minus phase-locked power; cf. Cohen & Donner, 2013) from FCz along with topographical maps for relevant time-window and a scatter plot for correlation between the theta and RT flanker effects. As expected, we found a significant increase of non-phase-locked theta (4–8 Hz) power in the incongruent condition, as compared to the congruent condition, F(1, 90) = 34.21, p < .001, ηp2 = .27. This conflict-related theta peaked at about 400–600 msec after target onset and had the typical midfrontal topography centered at FCz. Of importance, we found a significant negative correlation between the theta flanker effect and the RT flanker effect, r = −.35, p = .001 (the correlation with the average RT in the incongruent trials was even stronger, r = −.54, p < .001), and no significant correlation between the conflict theta (non-phase-locked activity) and the flanker N2 effect (phase-locked activity), r = −.05, p = .6.

Conflict-related Intersite Theta Phase Synchronization

Graph theory-based quantification of the intersite (all-to-all) phase synchronization degree showed that theta band (4–8 Hz) synchronization was strongest over the MFC (centered around FCz), as expected. This effect indicates that the midfrontal site carries out more functional connections in the theta band with all other sites than the rest of the electrodes over the scalp. Of importance, the theta synchronization degree was significantly higher in the incongruent condition than in the congruent condition, F(1, 90) = 37.86, p < .001, ηp2 = .30 (see Figure 5C), indicating a greater number of the MFC-centered functional connections in the conflict trials. Moreover, similar to the power results, this incongruent–congruence difference in synchronization degree was negatively correlated with the RT flanker effect, r = −.33, p = .002 (as well as with the average RT in the incongruent trials, r = −.43, p < .001).

Conflict-related Centro-parietal Activity

Figure 6A plots grand-average waveforms for the congruent and incongruent conditions at Pz, along with P3b topography in two flanker conditions. As evident, P3b peaked earlier in the congruent condition, F(1, 90) = 40.06, p < .001, ηp2 = .31, and this P3b latency flanker effect was modestly correlated with the RT flanker effect, r = .24, p = .02. Flanker congruency also had an impact on the P3b amplitude (measured as mean voltage 300–700 msec after target onset), which was in overall larger in the congruent than in the incongruent condition, F(1, 90) = 7.03, p = .009, ηp2 = .07, and this amplitude difference also correlated positively with the RT flanker effect, r = .33, p = .001 (Figure 6B).

Figure 6.

Conflict-related parietal activity. (A) Grand means of the ERPs from Pz in two flanker conditions. Negative voltage points upward. Time point zero is the stimulus onset. The gray area indicates the time-window in which the P3b mean amplitude was measured. The head maps show topographies in the indicated P3b time-window in two flanker conditions (denoted by the color of frames). The maps are min–max scaled, with positive polarity in red. The head view is from above. (B) A scatter plot for correlation between the flanker effect on the P3b amplitude and the RT flanker effect.

Figure 6.

Conflict-related parietal activity. (A) Grand means of the ERPs from Pz in two flanker conditions. Negative voltage points upward. Time point zero is the stimulus onset. The gray area indicates the time-window in which the P3b mean amplitude was measured. The head maps show topographies in the indicated P3b time-window in two flanker conditions (denoted by the color of frames). The maps are min–max scaled, with positive polarity in red. The head view is from above. (B) A scatter plot for correlation between the flanker effect on the P3b amplitude and the RT flanker effect.

Conflict-related Connectivity between MFC and Task-related ROIs

Theta Phase Synchronization between MFC and Lateral Frontal Areas

Figure 7A shows time–frequency plots of ISPC between FCz (“seed”) and lateral frontal ROI (AF3, AF4, F3, F4) for congruent and incongruent conditions, the incongruent–congruent difference along with theta ISPC topography for relevant time–frequency window, and a scatter plot for correlation between the theta ISPC flanker effect and the RT flanker effect. Black contours in the right panel outline time–frequency regions in which the condition difference was significant at p < .01, corrected for multiple comparisons using cluster-based permutation testing (also in Figures 7B and 8A). The figure reveals that theta band phase synchronization between FCz and the lateral frontal sites emerged in two stages. The first burst of synchronization occurred within 100–400 msec after target onset and the second within 400–700 msec. Although there is no difference between flanker conditions at the first stage, that second ISPC burst seems to occur only in the incongruent condition. This suggests that only the second stage reflects conflict-specific communication. Therefore, the theta ISPC between FCz and lateral prefrontal sites (AF3–4, F3–4) was evaluated with an additional factor: time-window (100–400 msec and 400–700 msec; see Methods section for details). The ANOVA yielded a significant main effect of Time-Window, F(1, 90) = 76.05, p < .001, ηp2 = .46, and a Time-Window × Flanker interaction, F(1, 90) = 52.63, p < .001, ηp2 = .37. The flanker effect was indeed present only in the second time-Window, F(1, 90) = 47.74, p < .001, ηp2 = .35, and not in the first one, F < 1.0, ns. This was confirmed by the permutation test (see the right panel of Figure 7A). Moreover, the flanker effect on midfrontal—lateral frontal theta band connectivity within the second time-window (400–700 msec) correlated negatively with the RT flanker effect, r = −.24, p = .025 (Figure 7A), indicating that the participants who showed a stronger conflict-related increase of the theta ISPC tended to demonstrate a smaller conflict-related slowing of their RTs.

Figure 7.

Conflict-related midfrontal connectivity. (A) ISPC between midfrontal “seed” (FCz) and lateral frontal ROI (AF3–4 and F3–4). (B) ISPC between midfrontal “seed” (FCz) and centro-parietal ROI (CPz, Pz). The left panels show time–frequency plots of the ISPC in two flanker conditions, along with time plots depicting grand means of the ISPC in the theta band (5–6 Hz). The right panels show time–frequency plots of the incongruent–congruent differences, that is, the ISPC flanker effects, and topographies of the effects' time–frequency maxima. Black contours outline time–frequency regions in which the condition difference was significant at p > .01, corrected for multiple comparisons. The scatter plots depict correlations between the presented ISPC flanker effects and the RT flanker effect. For all panels: Time point zero is the onset of target and flankers. “Hotter” colors on the time–frequency plots and head maps indicate more robust phase synchronization. The black on white dot on head maps denotes the midfrontal “seed” at FCz, and the channels marked in black denote the selected ROI sites. ISPC is an estimation of frequency-specific phase synchrony between two sources of oscillations (sites and electrodes) and is interpreted as an index of functional connectivity between two brain areas. The depicted results show conflict-related bursts of theta communication between the medial frontal and lateral frontal cortices (A), and between the medial frontal and parietal areas (B). In both cases, the conflict-related communication occurred within 400–700 msec after target onset, and individuals with stronger conflict-related midfrontal ISPC showed a smaller behavioral cost of conflict (i.e., a smaller RT flanker effect).

Figure 7.

Conflict-related midfrontal connectivity. (A) ISPC between midfrontal “seed” (FCz) and lateral frontal ROI (AF3–4 and F3–4). (B) ISPC between midfrontal “seed” (FCz) and centro-parietal ROI (CPz, Pz). The left panels show time–frequency plots of the ISPC in two flanker conditions, along with time plots depicting grand means of the ISPC in the theta band (5–6 Hz). The right panels show time–frequency plots of the incongruent–congruent differences, that is, the ISPC flanker effects, and topographies of the effects' time–frequency maxima. Black contours outline time–frequency regions in which the condition difference was significant at p > .01, corrected for multiple comparisons. The scatter plots depict correlations between the presented ISPC flanker effects and the RT flanker effect. For all panels: Time point zero is the onset of target and flankers. “Hotter” colors on the time–frequency plots and head maps indicate more robust phase synchronization. The black on white dot on head maps denotes the midfrontal “seed” at FCz, and the channels marked in black denote the selected ROI sites. ISPC is an estimation of frequency-specific phase synchrony between two sources of oscillations (sites and electrodes) and is interpreted as an index of functional connectivity between two brain areas. The depicted results show conflict-related bursts of theta communication between the medial frontal and lateral frontal cortices (A), and between the medial frontal and parietal areas (B). In both cases, the conflict-related communication occurred within 400–700 msec after target onset, and individuals with stronger conflict-related midfrontal ISPC showed a smaller behavioral cost of conflict (i.e., a smaller RT flanker effect).

Figure 8.

Conflict-related midfrontal connectivity. (A) ISPC between midfrontal “seed” (FCz) and lateral motor ROI (C5–6 and CP5–6 electrode pairs). The upper left panel shows time–frequency plots of contra–ipsilateral differences of the ISPC in two flanker conditions, along with topographies of these differences in a 500–800 msec, 4- to 6-Hz time–frequency window. The lower left panel shows time plot for grand means of the ISPC in the theta band (4–6 Hz) from the hemispheres contralateral and ipsilateral to the responding hand in two flanker conditions. The upper right panel shows a time–frequency plot of the flanker effect on the contra–ipsilateral ISPC difference (i.e., contra–ipsiincongruentminus contra–ipsicongruent), and topography of this difference in the specified time–frequency window. The scatter plot depicts a correlation between the conflict-related lateralized theta ISPC and the RT flanker effect. Head maps were obtained by subtracting all symmetrical electrodes and plotting the averaged contra–ipsilateral differences on the left hemisphere. (B) cf-ISPC between midfrontal theta “seed” (FCz, 4–7 Hz) and visual alpha ROI (PO7–8 and O1–2 electrode pairs, 9–12 Hz). The left panel shows grand means of the cross-frequency theta–alpha ISPC from the hemispheres contralateral and ipsilateral to the target hemifield in two flanker conditions. The right panel shows topographies of the theta–alpha cf-ISPC within 50–300 msec after target onset, separately for the hemispheres contralateral and ipsilateral to the target visual field, and for the contra–ipsilateral difference. Head maps for the contralateral averages were obtained by averaging all lateral electrodes from the left and right contralateral (to the target hemifield) hemispheres and plotting the averages on the left hemisphere. The same was done for the ipsilateral activity. Head maps for the contra–ipsilateral difference were obtained by subtracting all symmetrical electrodes and plotting the averaged contra–ipsilateral differences on the left hemisphere. The results show a generally stronger cross-frequency connectivity between the MFC and visual areas at the contralateral site than at the ipsilateral site, and in the incongruent condition than in the congruent condition (although the figure suggests otherwise, a hemisphere by flanker interaction did not reach significance). For all panels: Time point zero is the onset of target and flankers. “Hotter” colors on time–frequency plots and head maps indicate more robust phase synchronization. The black on white dot on head maps denotes the midfrontal “seed” at FCz, and the channels marked in black denote the selected ROI sites. Topographies are presented on min–max scaled maps with positive values in red and negative values in blue. The head view is from above. Colors of frames denote the conditions.

Figure 8.

Conflict-related midfrontal connectivity. (A) ISPC between midfrontal “seed” (FCz) and lateral motor ROI (C5–6 and CP5–6 electrode pairs). The upper left panel shows time–frequency plots of contra–ipsilateral differences of the ISPC in two flanker conditions, along with topographies of these differences in a 500–800 msec, 4- to 6-Hz time–frequency window. The lower left panel shows time plot for grand means of the ISPC in the theta band (4–6 Hz) from the hemispheres contralateral and ipsilateral to the responding hand in two flanker conditions. The upper right panel shows a time–frequency plot of the flanker effect on the contra–ipsilateral ISPC difference (i.e., contra–ipsiincongruentminus contra–ipsicongruent), and topography of this difference in the specified time–frequency window. The scatter plot depicts a correlation between the conflict-related lateralized theta ISPC and the RT flanker effect. Head maps were obtained by subtracting all symmetrical electrodes and plotting the averaged contra–ipsilateral differences on the left hemisphere. (B) cf-ISPC between midfrontal theta “seed” (FCz, 4–7 Hz) and visual alpha ROI (PO7–8 and O1–2 electrode pairs, 9–12 Hz). The left panel shows grand means of the cross-frequency theta–alpha ISPC from the hemispheres contralateral and ipsilateral to the target hemifield in two flanker conditions. The right panel shows topographies of the theta–alpha cf-ISPC within 50–300 msec after target onset, separately for the hemispheres contralateral and ipsilateral to the target visual field, and for the contra–ipsilateral difference. Head maps for the contralateral averages were obtained by averaging all lateral electrodes from the left and right contralateral (to the target hemifield) hemispheres and plotting the averages on the left hemisphere. The same was done for the ipsilateral activity. Head maps for the contra–ipsilateral difference were obtained by subtracting all symmetrical electrodes and plotting the averaged contra–ipsilateral differences on the left hemisphere. The results show a generally stronger cross-frequency connectivity between the MFC and visual areas at the contralateral site than at the ipsilateral site, and in the incongruent condition than in the congruent condition (although the figure suggests otherwise, a hemisphere by flanker interaction did not reach significance). For all panels: Time point zero is the onset of target and flankers. “Hotter” colors on time–frequency plots and head maps indicate more robust phase synchronization. The black on white dot on head maps denotes the midfrontal “seed” at FCz, and the channels marked in black denote the selected ROI sites. Topographies are presented on min–max scaled maps with positive values in red and negative values in blue. The head view is from above. Colors of frames denote the conditions.

Theta Phase Synchronization between MFC and Centro-parietal Area

ISPC between FCz (“seed”) and centro-parietal ROI (CPz, Pz) is shown in Figure 7B. Similar to the midfrontal-to-lateral-frontal ISPC, theta connectivity occurred in two stages, with a notably stronger pronounced incongruent-congruent difference at the second stage. Thus, the analysis was also carried out with the time-window factor (100–300 msec and 400–700 msec). Both the main effect of Time-Window and the interaction between Time-Window and Flanker were significant, F(1, 90) = 27.47, p < .001, ηp2 = .23, and F(1, 90) = 11.51, p = .001, ηp2 = .11, respectively. Again, the flanker effect exceeded the level of significance in the second time-window, F(1, 90) = 33.89, p < .001, ηp2 = .27, but not in the first one, F(1, 90) = 3.41, p = .07, ηp2 = .04, which was further confirmed by the permutation test (see the right panel of Figure 7B). Also, the RT flanker effect correlated negatively with the flanker effect on the midfrontal-to-centro-parietal theta ISPC in the second time-window, r = −.27, p = .010 (Figure 7B), but not in the first time-window, r = .17, p = .11.

Theta Phase Synchronization between MFC and Motor Areas

The left panel of Figure 8A shows grand averages of theta (4–6 Hz) ISPC between FCz (“seed”) and lateral motor ROI (C5–6, CP5–6) for the hemispheres contralateral and ipsilateral to the responding hand in two flanker conditions, and time–frequency plots for contra–ipsilateral differences of the ISPC, along with topographies of these contra–ipsilateral differences in relevant time–frequency window (500–800 msec, 4–6 Hz). As seen in the figure, connectivity between FCz and the contralateral motor areas was stronger and lasted longer than the connectivity between FCz and the ipsilateral areas, which corresponds with the flanker effects on the RTs and the motor-related activity indexed by the LRPs and beta LPS. More interestingly, the results show that although the contralateral connectivity was almost equally strong in both flanker conditions, the ipsilateral connectivity was conspicuously stronger in the incongruent condition than in the congruent condition. Thereby, the contra–ipsilateral difference is larger in the congruent condition than in the incongruent condition, and the flanker effect on the contra–ipsilateral difference (i.e., contra–ipsiincongruentminus contra–ipsicongruent) is negative (see the time–frequency plots and topography maps in Figure 8A). The ANOVA yielded a significant main effect of Hemisphere, F(1, 90) = 96.18, p < .001, ηp2 = .52, and flanker, F(1, 90) = 15.88, < .001, ηp2 = .15, as well as a significant Hemisphere × Flanker interaction, F(1, 90) = 4.51, p = .036, ηp2 = .05, confirming that the contra–ipsilateral effect in the incongruent condition, F(1, 90) = 68.34, < .001, ηp2 = .43, was indeed smaller than in the congruent condition, F(1, 90) = 85.50, < .001, ηp2 = .49 (see also the permutation test result in Figure 8A). Moreover, this flanker effect on the ISPC contra–ipsilateral difference (Figure 8A, the right panel) correlated with the RT flanker effect, r = −.22, p = .035 (as well as with the average RT in the incongruent condition, r = −.25, p = .016).

Cross-frequency (Theta–Alpha) Phase Synchronization between MFC and Visual Areas

Figure 8B depicts results for cross-frequency phase synchronization between midfrontal theta (4–7 Hz at FCz “seed”) and occipital alpha (9–12 Hz at PO7–8/O1–2) oscillations. The left panel depicts grand averages of cf-ISPC (averaged across all 16 theta–alpha pairs and the two occipital electrode pairs) for the hemispheres contralateral and ipsilateral to target visual field in two flanker conditions, and the right panel shows head topographies of the cf-ISPC in a relevant time-window. The theta–alpha phase synchronization between FCz and the visual areas reached its maximum over the visual areas within 100–200 msec after target onset and was generally stronger at the contralateral site than at the ipsilateral site, F(1, 90) = 13.86, p < .001, ηp2 = .13. As evident in Figure 8B, this contra–ipsilateral difference tended to be larger in the incongruent condition than in the congruent condition, although the Flanker × Hemisphere interaction did not reach significance, F(1, 90) = 2.19, p = .14, ηp2 = .02. Of importance, the main effect of Flanker was significant, F(1, 90) = 5.45, p = .02, ηp2 = .06, indicating generally stronger connectivity between MFC and visual areas in the incongruent trials. The ANOVA also showed a significant main effect of Frequency Pair, F(15, 1350) = 20.20, p < .001, ηp2 = .18, but none of the remained interactions was significant, F ≤ 1.9. Finally, we found no correlation of the overall cf-ISPC flanker effect with the RT flanker effect, r = .08, ns. However, such a correlation was found when we calculated the cf-ISPC flanker effect (i.e., incongruent–congruent) separately for the contralateral and ipsilateral sites: The RT flanker effect was negatively correlated with the cf-ISPC flanker effect at the contralateral site, r = −.24, p = .021, but not at the ipsilateral site, r = −.12, p = .25, suggesting that better behavioral performance may be partially related to the increased strength of cross-frequency connectivity between the MFC and contralateral visual areas.

DISCUSSION

Behavioral Results

The obtained behavioral results showed a typical flanker effect: Responses were longer and more error-prone in the incongruent condition than in the congruent condition. Importantly, the incongruent–congruent difference in RTs, that is, the RT flanker effect, proved to be an accurate and reliable measure of the behavioral performance. First, the RT flanker effect was strongly positively correlated with the average RT in the incongruent condition, and not with the RT in the congruent condition, indicating that the effect indexes specifically the processes triggered by incongruent flankers, as assumed. Second, we obtained fairly high odd–even and split-half reliability estimations, which exceeded previously reported split-half reliabilities of the RT flanker effects (Asanowicz, Marzecová, Jaśkowski, & Wolski, 2012; Macleod, Lawrence, Mcconnell, & Eskes, 2010; Fan, McCandliss, Sommer, Raz, & Posner, 2002).

Stimulus Selection- and Visual Interference-related Activity

To assess the electrophysiological markers of visuospatial stimulus selection, we introduced bilateral stimulation into the arrow flanker task and measured selection-related lateralizations of the EEG activity (contra–ipsilateral to the target hemifield): posterior contralateral lateralization of the ERPs, and posterior contralateral time–frequency phase coherence and power in the alpha band. After the contra–ipsilateral subtraction of the visual-evoked ERPs, we obtained a prominent N2pc component reflecting the act of stimulus selection within sensory representations in the visual cortex (Luck, 2012; Eimer, 1996). As expected, the N2pc amplitude was larger in the incongruent flanker condition. Based on the assumption that a larger N2pc would indicate a stronger involvement of the selection mechanism (cf. Śmigasiewicz, Asanowicz, Westphal, & Verleger, 2015; Luck et al., 1997), we predicted that the N2pc amplitude would be inversely related to the RT flanker effect. Contrary to this hypothesis, we found a positive correlation between the N2pc and RT flanker effects. This result suggests that a larger N2pc does not reflect a more efficient but rather more effortful processing: The more difficult the selection is for a participant, the more effort it takes, and, despite the effort, the larger is the behavioral cost of the flanker interference. In other words, individuals who experienced more difficulty in target selection would have a larger N2pc. Interestingly, inspection of the ERP grand averages suggests that the incongruent–congruent difference in N2pc is more because of increased positivity at the ipsilateral site, than because of increased negativity at the contralateral site (see Figure 3A). Previous studies have seemed to rule out the hypothesis that N2pc reflects suppression of spatial distractors, favoring the interpretation of N2pc as an index of target selection per se (Mazza, Turatto, & Caramazza, 2009; Eimer, 1996). Nevertheless, the present results suggest that in particular circumstances like the current task (bilateral stimulation plus incongruent flankers), N2pc might reflect both the contralateral signal enhancement and, at least partially, the ipsilateral distractor suppression. This interpretation of N2pc would be in agreement with some previous findings (Luck et al., 1997; Luck & Hillyard, 1994) and an older hypothesis according to which N2pc does indeed reflect a distractor-suppression process (Luck, 2005).

The N2pc was followed by the SPCN, which also had a significantly larger amplitude in the incongruent condition. Unlike the N2pc flanker effect, however, the SPCN flanker effect did not correlate with the RTs, but it was negatively correlated with the (discussed below) lateralization of alpha power (which in turn also did not correlate with the RTs). The SPCN (also called the contralateral delay activity; Vogel & Machizawa, 2004) is interpreted most often as a neuronal manifestation of visual working memory capacity (Luria, Balaban, Awh, & Vogel, 2016). Plausibly, the flanker task performance, not particularly relying on visual working memory capacity, was not significantly influenced by individual differences in terms of this ability.

As mentioned in the Introduction section, Bacigalupo and Luck (2019) have recently reported a target-evoked contralateral alpha suppression, suggesting that alpha modulations may be a part of a mechanism of visuospatial target processing. They also inquired into the relationships of the alpha suppression with visual interference (induced by changing target–flanker distance) and the target selection-related N2pc. Here, we calculated the LPS to assess the target-related contra–ipsilateral difference in the alpha band (van der Lubbe & Utzerath, 2013). The obtained LPS estimates showed that a contralateral (vs. ipsilateral) reduction of alpha power was present from 400 msec onward, and it was larger and lasted longer in the incongruent trials than in the congruent trials. This indicates that the contralateral alpha suppression is indeed a part of the mechanisms underlying target processing, and these mechanisms are engaged to a stronger extent and for a longer time under visual interference condition. Our result confirms therefore the nonsignificant trend for enhancing alpha suppression with increased perceptual crowding found by Bacigalupo and Luck (2019). However, we neither found a correlation of the alpha LPS flanker effect with the RT flanker effect nor with the N2pc flanker effect. Still, the alpha LPS flanker effect was negatively correlated with the SPCN flanker effect (as mentioned above). Taking into account that, first, the N2pc is a marker of target selection and the N2pc flanker effect did correlate with the RTs, and second, the SPCN is a marker of visual working memory and the SPCN flanker effect did not correlate with the RTs but did correlate with the LPS flanker effect, we may tentatively conclude that the target-elicited contralateral alpha suppression reflects the same mechanisms of target processing as the SPCN, that is, sustained processing within visual working memory. This would be in line with other evidence for the role of alpha oscillations in the maintenance of visual representations over time (e.g., VanRullen & Macdonald, 2012).

Another electrophysiological aspect of stimulus selection was revealed by the difference between contralateral and ipsilateral phase coherence in the alpha band over the visual cortex. The stimulus-related alpha phase coherence peaked within 50–300 msec after stimulus onset, and the effect was notably stronger at the hemisphere contralateral to the target hemifield. This result is in line with some previous findings (e.g., Keitel et al., 2019; Spaak et al., 2014). Interestingly, the selection-related lateralization of alpha phase coherence did not differ between the flanker conditions. We did find, however, a modest negative correlation between the individual differences in the magnitude of the contra–ipsilateral bias in the incongruent condition and the individual flanker costs in RTs. This result suggests that an increase of the contralateral alpha phase may boost our selection capacity. In conclusion, phase coherence seems to reflect another detail of the mechanism of visuospatial processing that adds to the overall efficiency of target selection under flanker-induced visual interference.

Response Selection- and Response Conflict-related Activity

Response selection begins with the translation of perceptual codes into response codes and activation of the established S–R link or “event-file” (Verleger et al., 2014; Hommel, 2004). This sensorimotor integration has been localized predominantly within association areas of the posterior parietal cortex (Caspers, Amunts, & Zilles, 2012; Freund, 2001). Recent studies have shown that modulations of the parietal P3b may reflect the activity underlying sensorimotor decisions and activation of S–R links (Twomey et al., 2015; Verleger et al., 2014). In the present results, the P3b component had a longer latency and smaller amplitude in the incongruent condition than in the congruent condition (replicating previous findings; Asanowicz et al., 2019; Kałamała et al., 2017; Neuhaus et al., 2010; Osman et al., 2000), which possibly reflects the delay of the S–R translation and activation of the correct S–R link caused by the flanker-induced response conflict. These P3b flanker effects did correlate with the delay of behavioral responses in the conflict trials, as expected.

To translate an S–R link into overt behavior, the proper response program has to be selected and activated in the premotor and motor cortex, and passed on through action control. The time course of this motor-related activity is shown in measurements of the LRPs and beta band LPS. The observed flanker effect on the stimulus-locked LRP is in line with the RT result showing that the selection of the correct response in the conflict condition was slowed down nearly 100 msec, compared to the congruent condition. The LPS complemented the LRP result, showing a conflict-elicited delay of beta suppression over motor areas contralaterally to the responding hand (note that the suppression is represented by positive LPS). This suppression is thought to reflect desynchronization of beta oscillations triggered by the functional cortical activation in motor-related brain areas during preparation and execution of voluntary movements (Siegel & Donner, 2010; Crone et al., 1998). The delay of the suppression reflects the timing of response selection (Kaiser et al., 2001; Tzagarakis et al., 2010). Once the conflict is resolved and the proper response program is selected, the response execution runs with no further flanker interference, as indicated by the latencies of response-locked LRPs and LPS. In other words, the temporal dynamic of the process of response execution is independent from the processes of conflict resolution. Still, a trace of the conflict appeared in the amplitudes of both measures, as a slightly larger LRP and a smaller LPS in the incongruent condition. These amplitude effects may reflect a rebound-like modulation of response activation after winning the “fight” with the incorrect response program. It is unclear, though, how to parsimoniously explain the simultaneous conflict-related increase of the LRP and decrease of the LPS.

A successful S–R translation and response selection under the response-conflict condition is assumed to be orchestrated by an executive hub localized in the MFC (Cavanagh & Frank, 2014; Cohen, 2014a). Conflict-related modulations of midfrontal EEG were expected to be observed in the phase-locked activity, as an increased amplitude of the N2 component of the ERP (Heil et al., 2000; Kopp et al., 1996), and in the non-phase-locked-activity, as an increased theta power (Cohen & Donner, 2013; Cavanagh et al., 2009). Our results showed that the N2 peaked at the FCz site in both flanker conditions, and its amplitude was larger in the conflict trials, as expected. Similarly, theta band power was centered at FCz and was larger in the conflict trials. Of importance, Cohen and Donner (2013), in their version of the Simon task, have found correlations of RTs with the non-phase-locked theta power, but not with the N2. Here, however, we found that the RT flanker effect was negatively correlated with both the theta flanker effect and the N2 flanker effect. In other words, both the non-phase-locked and the phase-locked components of the conflict-related midfrontal signal proved to be similarly related to the behavioral efficiency of conflict resolution: A relatively larger theta in the conflict trials resulted in participants' better performance, reflected by a smaller cost of conflict in RTs, and similarly, the participants who had a relatively larger conflict-related N2 performed better in conflict trials. Moreover, we found no correlation between the N2 and theta power flanker effects, which suggests that they reflect two distinct processes or subsystems of the physiological mechanism of executive control.

Of importance, while the reported correlation here between the RT and theta flanker effects was negative, the aforementioned previously observed correlations between average RTs and theta power were positive (Cohen & van Gaal, 2014; Cohen & Donner, 2013). Note, however, that here, we report correlations between condition differences (incongruent–congruent) at the subject level, whereas the latter were single-trial correlations (condition-specific and/or across conditions). When we calculated the within-subject single-trial correlation on the present data set using the same method as Cohen and Donner (2013), we did obtain a positive correlation between the average RTs and theta power, closely conforming to the previous findings. The result is shown in Figure 9, and the method is described in the figure legend. This difference between the correlation results highlights two aspects of the theta–RT relationship. The across-trial positive correlation has shown that trial-to-trial fluctuations in midfrontal theta predict trial-to-trial fluctuations in behavioral task performance. This indicates that the executive system responds stronger when the task at hand becomes more demanding—either because of exogenous factors like the experimental conditions, or endogenous factors like fluctuations of the states of the visual and motor systems. Whereas the subject-level negative correlation between the condition differences shows that, generally, a more efficient conflict resolution (indicated by a smaller conflict cost) is related to a smaller involvement of the conflict-related midfrontal activity. This plausibly indicates that individuals with a more efficient executive system perform the task at hand not only more efficiently, but also with a smaller involvement of the resources, as compared to individuals with less efficient executive control. In other words, on the one hand, there is a simple relationship—the more demanding the situation, the more activated the system is. On the other hand, there is the matter of individual differences—the more efficient the system is, the less activated it needs to be in a given situation.

Figure 9.

Within-subject cross-trial correlations between RT and midfrontal power (at FCz). “Hotter” colors indicate stronger positive correlations. Black contour outlines regions that are statistically significant at p < .001, corrected for multiple comparisons. Significant positive correlations are present in the theta band about 400–700 msec after target onset. The correlations ware calculated using the same method as Cohen and Donner (2013). Spearman's rho between single-trial RTs and single-trial time–frequency power was calculated for each time–frequency point, per participant. This produced a time–frequency map of correlation coefficients for each participant, which were then averaged across subjects, Fisher z-transformed, and compared to trial-baseline interval (−100 to 0 msec) at the group level using the cluster-based permutation testing.

Figure 9.

Within-subject cross-trial correlations between RT and midfrontal power (at FCz). “Hotter” colors indicate stronger positive correlations. Black contour outlines regions that are statistically significant at p < .001, corrected for multiple comparisons. Significant positive correlations are present in the theta band about 400–700 msec after target onset. The correlations ware calculated using the same method as Cohen and Donner (2013). Spearman's rho between single-trial RTs and single-trial time–frequency power was calculated for each time–frequency point, per participant. This produced a time–frequency map of correlation coefficients for each participant, which were then averaged across subjects, Fisher z-transformed, and compared to trial-baseline interval (−100 to 0 msec) at the group level using the cluster-based permutation testing.

The question now is, what are the functions of the two conflict-related midfrontal EEG components, N2 and theta power? Effective conflict processing must entail (a) the detection of interference or conflict between simultaneously active, competing response programs, and (b) the broadcasting of the information about the conflict to other areas involved in the task performance, including executive, sensory, and motor areas (Cohen, 2014a; Carter & van Veen, 2007). According to a popular account of executive control, these functions are performed by a monitoring mechanism localized in the MFC, particularly in ACC (Carter & Krug, 2012; Botvinick, Cohen, & Carter, 2004). Although the N2 and theta power effects often overlap in time (Cohen & Donner, 2013), which might hinder their interpretation, in the present results, they are unambiguously separated in time. The N2 flanker effect peaked at about 290 msec after target onset, whereas the theta flanker effect emerged within 400–600 msec after target onset. Therefore, it seems that the N2 reflects the detection of conflict, as proposed by Yeung et al. (2004), whereas the theta reflects the mechanism of “the realization of the need for control” (Cavanagh & Frank, 2014, p.414). Further support for this view comes from our connectivity results, described in the following section.

Conflict-related Theta Band Functional Connectivity

The hypothesized midfrontal broadcasting mechanism has been said to operate via the MFC-generated theta phase top–down synchronization that entrains distinct task-relevant areas (Cavanagh & Frank, 2014; Cohen, 2014a). Here, we have confirmed that long-range theta-phase connectivity is indeed strongly centered at the MFC area, and shown that this midfrontal theta connectivity is more robust under the conflict condition. Moreover, for participants who presented a relatively stronger theta synchronization degree in the conflict trials, the behavioral cost of the conflict was smaller, which suggests they were in overall better in conflict resolution. A similar increase of the long-range theta connectivity was previously reported for motor errors (Cohen, 2011) and perceptual errors (Cohen & van Gaal, 2013), interpreted as an engagement of posterror adaptation mechanisms. The present result extends those findings in line with the idea that the MFC is an executive hub of long-range functional connectivity.

Furthermore, the present results confirm that this MFC-centered long-range executive communication includes an exchange of information with distinct task-relevant areas, that is, functional connectivity through which the MFC is thought to coordinate stimulus selection (in visual areas), sensorimotor integration (in parietal areas), response program activation (in premotor and motor areas), and executive control over response selection and execution (in lateral frontal areas). Importantly, all these MFC-centered connectivity were more robust in the incongruent condition than in the congruent one, and all those condition differences showed an inverse relationship with the behavioral cost of the flanker incongruence: The stronger the connectivity, the more efficient was the conflict resolution (although the correlations were modest).

First of all, we observed cross-frequency communication between the MFC and occipital areas. Previously, the phase-amplitude coupling between midfrontal theta phase and posterior alpha power has been demonstrated, for instance, during visual perception (Demiralp et al., 2007) and perceptual errors (Cohen & van Gaal, 2013; Mazaheri, Nieuwenhuis, van Dijk, & Jensen, 2009). More recently, it has been shown that the theta phase coordinates neural activity in the fronto-parietal selection network (Fiebelkorn et al., 2018). Here, we measured interregional connectivity by computing cross-frequency synchrony between theta-phase at the MFC and alpha-phase over the visual cortex. The results showed that the theta–alpha phase coupling was stronger for communication with occipital areas at the hemisphere contralateral to the target hemifield, compared to the ipsilateral hemisphere, reflecting presumably a modulatory feedback loop from the MFC to visual cortex consolidating target processing. The theta–alpha phase coupling was also notably stronger in the incongruent condition than in the congruent condition, that is, when the target selection was more difficult because of visual interference caused by the incongruent flankers. Such theta–alpha coupling has been proposed to be one of the crucial oscillatory mechanisms of top–down control over visuospatial selection (Fiebelkorn & Kastner, 2020).

Second, the present results showed a task-related increase of the theta band synchrony between the MFC and centro-parietal area, probably reflecting processes of sensorimotor integration. This communication occurred in two bursts peaking at about 200 and 500 msec after target onset. The first one was notably stronger than the second one, but the difference between the flanker conditions was more pronounced during the second burst, reflecting enhanced connectivity in the conflict trials. Importantly, the flanker effect on the connectivity within the second burst correlated significantly with the RT flanker effect. The observed condition difference corresponds with the flanker effect on the centro-parietal P3b component of the ERPs. Similar communication between the MFC and some parietal areas has been previously reported in the Simon task (Gulbinaite et al., 2014) and error-detection task (Zhang, Chavarriaga, & Millán, 2015).

Third, our LRP and beta LPS results, showing the flanker effect on activation of correct response programs in motor areas, are complemented by observing the theta band synchronization between the MFC and motor areas contralateral and ipsilateral to responding hand. Specifically, communication with the contralateral areas was expectedly generally stronger than with the ipsilateral areas, and lasted longer in the incongruent condition. Interestingly, whereas the contralateral connectivity was almost equally strong in both flanker conditions, the ipsilateral connectivity was stronger in the incongruent condition than in the congruent one. This is in line with the hypothesis that the MFC not only coordinates the selection of the correct response program, but also orchestrates an inhibition of the incorrect response program that was triggered by the incongruent flankers, that is, the inhibitory projections to the ipsilateral motor site.

Finally, we observed an interesting two-phase communication between the medial and lateral frontal cortices (MFC and LFC). The first phase of the MFC-LFC communication occurred in the time-window of 100–400 msec, and here, the connectivity strength was similar in both flanker conditions. After this first burst, in the congruent condition, the communication began to fade out. In the incongruent condition, however, there was a second, even stronger burst of the theta synchronization, peaking at about 600–700 msec, plausibly reflecting the exchange of information on conflict resolution. Importantly, the increase of the MFC-LFC connectivity strength in the incongruent condition (relative to the congruent condition) within this second phase was correlated with the efficiency of conflict resolution exhibited by the speed of correct responses under the conflict condition. There is a general agreement that action control involves communication between the MFC and LFC (Petersen & Posner, 2012; Banich, 2009). According to the conflict–control loop theory (Carter & van Veen, 2007), in case of response conflict, the MFC informs the LFC about the need for control, and the LFC implements the top–down control functions to resolve the conflict and to prevent future performance decrements (Botvinick, Braver, Barch, Carter, & Cohen, 2001; Miller & Cohen, 2001). The theta band synchrony between the MFC and LFC reflects a plausible oscillatory mechanism of information transmission between action-monitoring and action-control areas (Nigbur et al., 2012; Cavanagh et al., 2009). The current result adds another data point to this line of research on the communication within the executive network.

Conclusions

First, this study provides a comprehensive overview of the electrophysiological activity underlying selective processing under sensorimotor conflict. This includes activity at the occipital visual areas—where stimulus selection occurs, the posterior parietal area—where visual information is translated into S–R links, the motor areas—where response programs are selected and activated, and the midfrontal area—whereas supervisory processing control is computed. Second, the obtained results show that the task-related areas are functionally connected via interregional oscillatory synchrony in the theta band with a central executive hub located in the MFC. Importantly, this long-range connectivity was more robust under the condition of increased demands for stimulus and response selection. Third, individual differences in the measured conflict-related EEG activity, particularly the midfrontal N2, theta power, and global theta connectivity, predict the behavioral efficiency in conflict resolution. Taken together, the present findings help to understand how the brain dynamics underlying sensorimotor processing and its executive orchestration shape individual behavioral performance.

Acknowledgments

The present work was supported by grant 2016/22/E/HS6/00139 awarded from the National Science Centre of Poland to D. A.

Reprint requests should be sent to Dariusz Asanowicz, Institute of Psychology, Jagiellonian University, Ingardena 6, 30-060 Krakow, Poland, or via e-mail: d.asanowicz@uj.edu.pl.

Funding Information

Dariusz Asanowicz, Narodowe Centrum Nauki (https://dx.doi.org/10.13039/501100004281), grant number: 2016/22/E/HS6/00139.

Diversity in Citation Practices

A retrospective analysis of the citations in every article published in this journal from 2010 to 2020 has revealed a persistent pattern of gender imbalance: Although the proportions of authorship teams (categorized by estimated gender identification of first author/last author) publishing in the Journal of Cognitive Neuroscience (JoCN) during this period were M(an)/M = .408, W(oman)/M = .335, M/W = .108, and W/W = .149, the comparable proportions for the articles that these authorship teams cited were M/M = .579, W/M = .243, M/W = .102, and W/W = .076 (Fulvio et al., JoCN, 33:1, pp. 3–7). Consequently, JoCN encourages all authors to consider gender balance explicitly when selecting which articles to cite and gives them the opportunity to report their article's gender citation balance.

Notes

1. 

The P1 and N1 asymmetry, present in the difference waves in Figure 3B as a small early positivity at 110 msec and negativity at 160 msec, did not differ significantly between the flanker conditions. These ERL deflections are caused by a slightly asymmetrical stimulation (arrows vs. vertical lines); thus, being a stimulus-driven perceptual asymmetry has been considered not relevant for the current topic and is not further discussed.

2. 

Note that the LPS correlations were calculated for power estimations only because the LPS latencies could not be reliably measured for individual participants; the jackknife method was used instead (see Methods section for details).

3. 

Given that the LRP latencies (as seen in the grand averages) correspond closely to the average RTs in two flanker conditions, one could expect a rather stronger correlation. However, the signal-to-noise ratio of LRPs is typically relatively low, so that LRP latencies are generally more difficult to measure reliably for individual participants.

REFERENCES

Allport
,
D. A.
(
1987
).
Selection for action: Some behavioral and neurophysiological considerations of attention
. In
H.
Heuer
&
A. F.
Sanders
(Eds.),
Perspectives on perception and action
(1st ed., pp.
395
419
).
Hillsdale, NJ
:
Erlbaum
.
Asanowicz
,
D.
,
Gociewicz
,
K.
,
Koculak
,
M.
,
Finc
,
K.
,
Bonna
,
K.
,
Cleeremans
,
A.
, et al
(
2020
).
The response relevance of visual stimuli modulates the P3 component and the underlying sensorimotor network
.
Scientific Reports
,
10
,
3818
.
Asanowicz
,
D.
,
Marzecová
,
A.
,
Jaśkowski
,
P.
, &
Wolski
,
P.
(
2012
).
Hemispheric asymmetry in the efficiency of attentional networks
.
Brain and Cognition
,
79
,
117
128
.
Asanowicz
,
D.
,
Wołoszyn
,
K.
,
Panek
,
B.
, &
Wronka
,
E.
(
2019
).
On the locus of the effect of alerting on response conflict: An event-related EEG study with a speed-accuracy tradeoff manipulation
.
Biological Psychology
,
145
,
62
75
.
Bacigalupo
,
F.
, &
Luck
,
S. J.
(
2019
).
Lateralized suppression of alpha band EEG activity as a mechanism of target processing
.
Journal of Neuroscience
,
39
,
900
917
.
Banich
,
M. T.
(
2009
).
Executive function: The search for an integrated account
.
Current Directions in Psychological Science
,
18
,
89
94
.
Benwell
,
C. S. Y.
,
Keitel
,
C.
,
Harvey
,
M.
,
Gross
,
J.
, &
Thut
,
G.
(
2018
).
Trial-by-trial co-variation of pre-stimulus EEG alpha power and visuospatial bias reflects a mixture of stochastic and deterministic effects
.
European Journal of Neuroscience
,
48
,
2566
2584
.
Botvinick
,
M. M.
,
Braver
,
T. S.
,
Barch
,
D. M.
,
Carter
,
C. S.
, &
Cohen
,
J. D.
(
2001
).
Conflict monitoring and cognitive control
.
Psychological Review
,
108
,
624
652
.
Botvinick
,
M. M.
,
Cohen
,
J. D.
, &
Carter
,
C. S.
(
2004
).
Conflict monitoring and anterior cingulate cortex: An update
.
Trends in Cognitive Sciences
,
8
,
539
546
.
Busch
,
N. A.
,
Dubois
,
J.
, &
VanRullen
,
R.
(
2009
).
The phase of ongoing EEG oscillations predicts visual perception
.
Journal of Neuroscience
,
29
,
7869
7876
.
Carter
,
C. S.
, &
Krug
,
M. K.
(
2012
).
Dynamic cognitive control and frontal—Cingulate interactions
. In
M. I.
Posner
(Ed.),
Cognitive neuroscience of attention
(pp.
89
98
).
New York
:
Guilford Press
.
Carter
,
C. S.
, &
van Veen
,
V.
(
2007
).
Anterior cingulate and conflict detection: An update of theory and data
.
Cognitive, Affective, & Behavioral Neuroscience
,
7
,
367
379
.
Caspers
,
S.
,
Amunts
,
K.
, &
Zilles
,
K.
(
2012
).
Posterior parietal cortex: Multimodal association cortex
. In
K. J.
Mai
&
G.
Paxinos
(Eds.),
The human nervous system
(3rd ed., pp.
1036
1055
).
Elsevier
.
Cavanagh
,
J. F.
,
Cohen
,
M. X.
, &
Allen
,
J. J. B.
(
2009
).
Prelude to and resolution of an error: EEG phase synchrony reveals cognitive control dynamics during action monitoring
.
Journal of Neuroscience
,
29
,
98
105
.
Cavanagh
,
J. F.
, &
Frank
,
M. J.
(
2014
).
Frontal theta as a mechanism for cognitive control
.
Trends in Cognitive Sciences
,
18
,
414
421
.
Cohen
,
M. X.
(
2011
).
Error-related medial frontal theta activity predicts cingulate-related structural connectivity
.
Neuroimage
,
55
,
1373
1383
.
Cohen
,
M. X.
(
2014a
).
A neural microcircuit for cognitive conflict detection and signaling
.
Trends in Neurosciences
,
37
,
480
490
.
Cohen
,
M. X.
(
2014b
).
Analyzing neural time series data: Theory and practice
.
Cambridge, MA
:
MIT Press
.
Cohen
,
M. X.
(
2015
).
Effects of time lag and frequency matching on phase-based connectivity
.
Journal of Neuroscience Methods
,
250
,
137
146
.
Cohen
,
M. X.
(
2017
).
MATLAB for brain and cognitive scientists
.
Cambridge, MA
:
MIT Press
.
Cohen
,
M. X.
, &
Cavanagh
,
J. F.
(
2011
).
Single-trial regression elucidates the role of prefrontal theta oscillations in response conflict
.
Frontiers in Psychology
,
2
,
30
.
Cohen
,
M. X.
, &
Donner
,
T. H.
(
2013
).
Midfrontal conflict-related theta band power reflects neural oscillations that predict behavior
.
Journal of Neurophysiology
,
110
,
2752
2763
.
Cohen
,
M. X.
, &
Ridderinkhof
,
K. R.
(
2013
).
EEG source reconstruction reveals frontal-parietal dynamics of spatial conflict processing
.
PLoS One
,
8
,
e57293
.
Cohen
,
M. X.
,
Ridderinkhof
,
K. R.
,
Haupt
,
S.
,
Elger
,
C. E.
, &
Fell
,
J.
(
2008
).
Medial frontal cortex and response conflict: Evidence from human intracranial EEG and medial frontal cortex lesion
.
Brain Research
,
1238
,
127
142
.
Cohen
,
M. X.
, &
van Gaal
,
S.
(
2013
).
Dynamic interactions between large-scale brain networks predict behavioral adaptation after perceptual errors
.
Cerebral Cortex
,
23
,
1061
1072
.
Cohen
,
M. X.
, &
van Gaal
,
S.
(
2014
).
Subthreshold muscle twitches dissociate oscillatory neural signatures of conflicts from errors
.
Neuroimage
,
86
,
503
513
.
Cohen
,
M. X.
,
van Gaal
,
S.
,
Ridderinkhof
,
K. R.
, &
Lamme
,
V. A. F.
(
2009
).
Unconscious errors enhance prefrontal-occipital oscillatory synchrony
.
Frontiers in Human Neuroscience
,
3
,
54
.
Coles
,
M. G. H.
(
1989
).
Modern mind-brain reading: Psychophysiology, physiology, and cognition
.
Psychophysiology
,
26
,
251
269
.
Corbetta
,
M.
, &
Shulman
,
G. L.
(
2002
).
Control of goal-directed and stimulus-driven attention in the brain
.
Nature Reviews Neuroscience
,
3
,
201
215
.
Crone
,
N. E.
,
Miglioretti
,
D. L.
,
Gordon
,
B.
,
Sieracki
,
J. M.
,
Wilson
,
M. T.
,
Uematsu
,
S.
, et al
(
1998
).
Functional mapping of human sensorimotor cortex with electrocorticographic spectral analysis I. Alpha and beta event-related desynchronization
.
Brain
,
121
,
2271
2299
.
Demiralp
,
T.
,
Bayraktaroglu
,
Z.
,
Lenz
,
D.
,
Junge
,
S.
,
Busch
,
N. A.
,
Maess
,
B.
, et al
(
2007
).
Gamma amplitudes are coupled to theta phase in human EEG during visual perception
.
International Journal of Psychophysiology
,
64
,
24
30
.
Egner
,
T.
(
2008
).
Multiple conflict-driven control mechanisms in the human brain
.
Trends in Cognitive Sciences
,
12
,
374
380
.
Eimer
,
M.
(
1996
).
The N2pc as an indicator of attentional selectivity
.
Electroencephalography and Clinical Neurophysiology
,
99
,
225
234
.
Eriksen
,
B.
, &
Eriksen
,
C. W.
(
1974
).
Effects of noise letters upon the identification of a target letter in a nonsearch task
.
Perception & Psychophysics
,
16
,
143
149
.
Fan
,
J.
,
Gu
,
X.
,
Guise
,
K. G.
,
Liu
,
X.
,
Fossella
,
J.
,
Wang
,
H.
, et al
(
2009
).
Testing the behavioral interaction and integration of attentional networks
.
Brain and Cognition
,
70
,
209
220
.
Fan
,
J.
,
McCandliss
,
B. D.
,
Sommer
,
T.
,
Raz
,
A.
, &
Posner
,
M. I.
(
2002
).
Testing the efficiency and independence of attentional networks
.
Journal of Cognitive Neuroscience
,
14
,
340
347
.
Fiebelkorn
,
I. C.
, &
Kastner
,
S.
(
2019
).
A rhythmic theory of attention
.
Trends in Cognitive Sciences
,
23
,
87
101
.
Fiebelkorn
,
I. C.
, &
Kastner
,
S.
(
2020
).
Functional specialization in the attention network
.
Annual Review of Psychology
,
71
,
221
249
.
Fiebelkorn
,
I. C.
,
Pinsk
,
M. A.
, &
Kastner
,
S.
(
2018
).
A dynamic interplay within the frontoparietal network underlies rhythmic spatial attention
.
Neuron
,
99
,
842
853
.
Foster
,
J. J.
, &
Awh
,
E.
(
2019
).
The role of alpha oscillations in spatial attention: Limited evidence for a suppression account
.
Current Opinion in Psychology
,
29
,
34
40
.
Freund
,
H.-J.
(
2001
).
The parietal lobe as a sensorimotor interface: A perspective from clinical and neuroimaging data
.
Neuroimage
,
14
,
142
146
.
Fries
,
P.
(
2005
).
A mechanism for cognitive dynamics: Neuronal communication through neuronal coherence
.
Trends in Cognitive Sciences
,
9
,
474
480
.
Gratton
,
G.
,
Coles
,
M. G. H.
, &
Donchin
,
E.
(
1992
).
Optimizing the use of information: Strategic control of activation of responses
.
Journal of Experimental Psychology: General
,
121
,
480
506
.
Gratton
,
G.
,
Coles
,
M. G. H.
,
Sirevaag
,
E. J.
,
Eriksen
,
C. W.
, &
Donchin
,
E.
(
1988
).
Pre- and poststimulus activation of response channels: A psychophysiological analysis
.
Journal of Experimental Psychology: Human Perception and Performance
,
14
,
331
344
.
Grent-'t-Jong
,
T.
,
Oostenveld
,
R.
,
Jensen
,
O.
,
Medendorp
,
W. P.
, &
Praamstra
,
P.
(
2013
).
Oscillatory dynamics of response competition in human sensorimotor cortex
.
Neuroimage
,
83
,
27
34
.
Gulbinaite
,
R.
,
van Rijn
,
H.
, &
Cohen
,
M. X.
(
2014
).
Fronto-parietal network oscillations reveal relationship between working memory capacity and cognitive control
.
Frontiers in Human Neuroscience
,
8
,
761
.
Hackley
,
S. A.
, &
Valle-Inclán
,
F.
(
1998
).
Automatic alerting does not speed late motoric processes in a reaction-time task
.
Nature
,
391
,
786
788
.
Heil
,
M.
,
Osman
,
A.
,
Wiegelmann
,
J.
,
Rolke
,
B.
, &
Hennighausen
,
E.
(
2000
).
N200 in the Eriksen-task: Inhibitory executive processes?
Journal of Psychophysiology
,
14
,
218
225
.
Hommel
,
B.
(
2004
).
Event files: Feature binding in and across perception and action
.
Trends in Cognitive Sciences
,
8
,
494
500
.
Jung
,
T.-P.
,
Makeig
,
S.
,
Humphries
,
C.
,
Lee
,
T.-W.
,
McKeown
,
M. J.
,
Iragui
,
V.
, et al
(
2000
).
Removing electroencephalographic artifacts by blind source separation
.
Psychophysiology
,
37
,
163
178
.
Kaiser
,
J.
,
Birbaumer
,
N.
, &
Lutzenberger
,
W.
(
2001
).
Event-related beta desynchronization indicates timing of response selection in a delayed-response paradigm in humans
.
Neuroscience Letters
,
312
,
149
152
.
Kałamała
,
P.
,
Szewczyk
,
J.
,
Senderecka
,
M.
, &
Wodniecka
,
Z.
(
2017
).
Flanker task with equiprobable congruent and incongruent conditions does not elicit the conflict N2
.
Psychophysiology
,
55
,
e12980
.
Keitel
,
C.
,
Keitel
,
A.
,
Benwell
,
C.
,
Daube
,
C.
,
Thut
,
G.
, &
Gross
,
J.
(
2019
).
Stimulus-driven brain rhythms within the alpha band: The attentional-modulation conundrum
.
Journal of Neuroscience
,
39
,
3119
3129
.
Kiesel
,
A.
,
Miller
,
J.
,
Jolicœur
,
P.
, &
Brisson
,
B.
(
2008
).
Measurement of ERP latency differences: A comparison of single-participant and jackknife-based scoring methods
.
Psychophysiology
,
45
,
250
274
.
Klein
,
P.-A.
,
Petitjean
,
C.
,
Olivier
,
E.
, &
Duque
,
J.
(
2014
).
Top–down suppression of incompatible motor activations during response selection under conflict
.
Neuroimage
,
86
,
138
149
.
Kopp
,
B.
,
Mattler
,
U.
,
Goertz
,
R.
, &
Rist
,
F.
(
1996
).
N2, P3 and the lateralized readiness potential in a nogo task involving selective response priming
.
Electroencephalography and Clinical Neurophysiology
,
99
,
19
27
.
Lachaux
,
J.
,
Rodriguez
,
E.
,
Martinerie
,
J.
, &
Varela
,
F. J.
(
1999
).
Measuring phase synchrony in brain signals
.
Human Brain Mapping
,
8
,
194
208
.
Lins
,
O. G.
,
Picton
,
T. W.
,
Berg
,
P.
, &
Scherg
,
M.
(
1993
).
Ocular artifacts in EEG and event-related potentials I: Scalp topography
.
Brain Topography
,
6
,
51
63
.
Luck
,
S. J.
(
2005
).
The operation of attention, millisecond by millisecond, over the first half second
. In
H.
Ogmen
&
B. G.
Breitmeyer
(Eds.),
The first half second: The microgenesis and temporal dynamics of unconscious and conscious visual processes
.
Cambridge, MA
:
MIT Press
.
Luck
,
S. J.
(
2012
).
Electrophysiological correlates of the focusing of attention within complex visual scenes: N2pc and related ERP components
. In
S. J.
Luck
&
E. S.
Kappenman
(Eds.),
The Oxford handbook of event-related potential components
(pp.
329
360
).
New York
:
Oxford University Press
.
Luck
,
S. J.
,
Fan
,
S.
, &
Hillyard
,
S. A.
(
1993
).
Attention-related modulation of sensory-evoked brain activity in a visual search task
.
Journal of Cognitive Neuroscience
,
5
,
188
195
.
Luck
,
S. J.
,
Girelli
,
M.
,
McDermott
,
M.
, &
Ford
,
M.
(
1997
).
Bridging the gap between monkey neurophysiology and human perception: An ambiguity resolution theory of visual selective attention
.
Cognitive Psychology
,
33
,
64
87
.
Luck
,
S. J.
, &
Gold
,
J. M.
(
2008
).
The construct of attention in schizophrenia
.
Biological Psychiatry
,
64
,
34
39
.
Luck
,
S. J.
, &
Hillyard
,
S. A.
(
1994
).
Spatial filtering during visual search: Evidence from human electrophysiology
.
Journal of Experimental Psychology: Human Perception and Performance
,
20
,
1000
1014
.
Luo
,
C.
, &
Proctor
,
R.
(
2016
).
Perceptual grouping of objects occupied by target and flankers affects target–flanker interference
.
Attention, Perception, & Psychophysics
,
78
,
251
263
.
Luria
,
R.
,
Balaban
,
H.
,
Awh
,
E.
, &
Vogel
,
E. K.
(
2016
).
The contralateral delay activity as a neural measure of visual working memory
.
Neuroscience and Biobehavioral Reviews
,
62
,
100
108
.
Macleod
,
J. W.
,
Lawrence
,
M.
,
Mcconnell
,
M.
, &
Eskes
,
G. A.
(
2010
).
Appraising the ANT: psychometric and theoretical considerations of the attention network test
.
Neuropsychology
,
24
,
637
651
.
Maniscalco
,
B.
,
Bang
,
J. W.
,
Iravani
,
L.
,
Camps-Febrer
,
F.
, &
Lau
,
H.
(
2012
).
Does response interference depend on the subjective visibility of flanker distractors?
Attention, Perception, & Psychophysics
,
74
,
841
851
.
Maris
,
E.
, &
Oostenveld
,
R.
(
2007
).
Nonparametric statistical testing of EEG- and MEG data
.
Journal of Neuroscience Methods
,
164
,
177
190
.
Mazaheri
,
A.
,
Nieuwenhuis
,
I. L. C.
,
van Dijk
,
H.
, &
Jensen
,
O.
(
2009
).
Prestimulus alpha and mu activity predicts failure to inhibit motor responses
.
Human Brain Mapping
,
30
,
1791
1800
.
Mazza
,
V.
,
Turatto
,
M.
, &
Caramazza
,
A.
(
2009
).
Attention selection, distractor suppression and N2pc
.
Cortex
,
45
,
879
890
.
Miller
,
E. K.
, &
Cohen
,
J. D.
(
2001
).
An integrative theory of prefrontal cortex function
.
Annual Review of Neuroscience
,
24
,
167
202
.
Moore
,
T.
, &
Zirnsak
,
M.
(
2017
).
Neural mechanisms of selective visual attention
.
Annual Review of Psychology
,
68
,
47
72
.
Neuhaus
,
A. H.
,
Urbanek
,
C.
,
Opgen-Rhein
,
C.
,
Hahn
,
E.
,
Ta
,
T. M. T.
,
Koehler
,
S.
, et al
(
2010
).
Event-related potentials associated with attention network test
.
International Journal of Psychophysiology
,
76
,
72
79
.
Nigbur
,
R.
,
Cohen
,
M. X.
,
Ridderinkhof
,
K. R.
, &
Stürmer
,
B.
(
2012
).
Theta dynamics reveal domain-specific control over stimulus and response conflict
.
Journal of Cognitive Neuroscience
,
24
,
1264
1274
.
Nigbur
,
R.
,
Ivanova
,
G.
, &
Stürmer
,
B.
(
2011
).
Theta power as a marker for cognitive interference
.
Clinical Neurophysiology
,
122
,
2185
2194
.
O'Connell
,
R. G.
,
Dockree
,
P. M.
, &
Kelly
,
S. P.
(
2012
).
A supramodal accumulation-to-bound signal that determines perceptual decisions in humans
.
Nature Neuroscience
,
15
,
1729
1735
.
Osman
,
A.
,
Lou
,
L.
,
Muller-Gethmann
,
H.
,
Rinkenauer
,
G.
,
Mattes
,
S.
, &
Ulrich
,
R.
(
2000
).
Mechanisms of speed-accuracy tradeoff: Evidence from covert motor processes
.
Biological Psychology
,
51
,
173
199
.
Palva
,
J. M.
,
Palva
,
S.
, &
Kaila
,
K.
(
2005
).
Phase synchrony among neuronal oscillations in the human cortex
.
Journal of Neuroscience
,
25
,
3962
3972
.
Petersen
,
S. E.
, &
Posner
,
M. I.
(
2012
).
The attention system of the human brain: 20 years after
.
Annual Review of Neuroscience
,
35
,
73
89
.
Ridderinkhof
,
K. R.
,
Ullsperger
,
M.
,
Crone
,
E. A.
, &
Nieuwenhuis
,
S.
(
2004
).
The role of the medial frontal cortex in cognitive control
.
Science
,
306
,
443
447
.
Schneider
,
D.
,
Hoffmann
,
S.
, &
Wascher
,
E.
(
2014
).
Sustained posterior contralateral activity indicates re-entrant target processing in visual change detection: An EEG study
.
Frontiers in Human Neuroscience
,
8
,
247
.
Siegel
,
M.
, &
Donner
,
T. H.
(
2010
).
Linking band-limited cortical activity to fMRI and behavior
. In
M.
Ullsperger&
S.
Debener
(Eds.),
Simultaneous EEG and fMRI: Recording, analysis, and application
(pp.
1
23
).
Oxford University Press
.
Siegel
,
M.
,
Donner
,
T. H.
, &
Engel
,
A. K.
(
2012
).
Spectral analysis spectral fingerprints of large-scale neuronal interactions
.
Nature Reviews Neuroscience
,
13
,
121
134
.
Simon
,
J. R.
(
1969
).
Reactions toward the source of stimulation
.
Journal of Experimental Psychology
,
81
,
174
176
.
Śmigasiewicz
,
K.
,
Asanowicz
,
D.
,
Westphal
,
N.
, &
Verleger
,
R.
(
2015
).
Bias for the left visual field in rapid serial visual presentation: Effects of additional salient cues suggest a critical role of attention
.
Journal of Cognitive Neuroscience
,
27
,
266
279
.
Smulders
,
F. T. Y.
, &
Miller
,
J. O.
(
2011
).
The lateralized readiness potential
. In
S. J.
Luck
&
E. S.
Kappenman
(Eds.),
The Oxford handbook of event-related potential components
. (pp.
209
230
).
New York
:
Oxford University Press
.
Spaak
,
E.
,
de Lange
,
F. P.
, &
Jensen
,
O.
(
2014
).
Local entrainment of alpha oscillations by visual stimuli causes cyclic modulation of perception
.
Journal of Neuroscience
,
34
,
3536
3544
.
Srinivasan
,
R.
,
Winter
,
W. R.
,
Ding
,
J.
, &
Nunez
,
P. L.
(
2007
).
EEG and MEG coherence: Measures of functional connectivity at distinct spatial scales of neocortical dynamics
.
Journal of Neuroscience Methods
,
166
,
41
52
.
Stam
,
C. J.
,
de Haan
,
W.
,
Daffertshofer
,
A.
,
Jones
,
B. F.
,
Manshanden
,
I.
,
van Cappellen van Walsum
,
A. M.
, et al
(
2009
).
Graph theoretical analysis of magnetoencephalographic functional connectivity in Alzheimer's disease
.
Brain
,
132
,
213
224
.
Tass
,
P.
,
Rosenblum
,
M. G.
,
Weule
,
J.
,
Kurths
,
J.
,
Pikovsky
,
A.
,
Volkmann
,
J.
, et al
(
1998
).
Detection of n:m phase locking from noisy data: Application to Magnetoencephalography
.
Physical Review Letters
,
81
,
3291
3294
.
Thiele
,
A.
, &
Bellgrove
,
M.
(
2018
).
Neuromodulation of Attention
.
Neuron
,
97
,
769
785
.
Thut
,
G.
,
Nietzel
,
A.
,
Brandt
,
S. A.
, &
Pascual-Leone
,
A.
(
2006
).
α-Band electroencephalographic activity over occipital cortex indexes visuospatial attention bias and predicts visual target detection
.
Journal of Neuroscience
,
26
,
9494
9502
.
Tillman
,
C. M.
, &
Wiens
,
S.
(
2011
).
Behavioral and ERP indices of response conflict in Stroop and flanker tasks
.
Psychophysiology
,
48
,
1405
1411
.
Trujillo
,
L. T.
, &
Allen
,
J. J. B.
(
2007
).
Theta EEG dynamics of the error-related negativity
.
Clinical Neurophysiology
,
118
,
645
668
.
Twomey
,
D. M.
,
Murphy
,
P. R.
,
Kelly
,
S. P.
, &
O'Connell
,
R. G.
(
2015
).
The classic P300 encodes a build-to-threshold decision variable
.
European Journal of Neuroscience
,
42
,
1636
1643
.
Tzagarakis
,
C.
,
Ince
,
N. F.
,
Leuthold
,
A. C.
, &
Pellizzer
,
G.
(
2010
).
Beta band activity during motor planning reflects response uncertainty
.
Journal of Neuroscience
,
30
,
11270
11277
.
Ulrich
,
R.
, &
Miller
,
J.
(
2001
).
Using the jackknife-based scoring method for measuring LRP onset effects in factorial designs
.
Psychophysiology
,
38
,
816
827
.
Valle-Inclán
,
F.
(
1996
).
The locus of interference in the Simon effect: An ERP study
.
Biological Psychology
,
43
,
147
162
.
van Diepen
,
R. M.
,
Foxe
,
J. J.
, &
Mazaheri
,
A.
(
2019
).
The functional role of alpha band activity in attentional processing: The current zeitgeist and future outlook
.
Current Opinion in Psychology
,
29
,
229
238
.
van Driel
,
J.
,
Ridderinkhof
,
R. K.
, &
Cohen
,
M. X.
(
2012
).
Not all errors are alike: Theta and alpha EEG dynamics relate to differences in error-processing dynamics
.
Journal of Neuroscience
,
32
,
16795
16806
.
van Driel
,
J.
,
Sligte
,
I. G.
,
Linders
,
J.
,
Elport
,
D.
, &
Cohen
,
M. X.
(
2015
).
Frequency band-specific electrical brain stimulation modulates cognitive control processes
.
PLoS One
,
10
,
e0138984
.
van der Lubbe
,
R. H. J.
,
de Kleine
,
E.
, &
Rataj
,
K.
(
2019
).
Dyslexic individuals orient but do not sustain visual attention: Electrophysiological support from the lower and upper alpha bands
.
Neuropsychologia
,
125
,
30
41
.
van der Lubbe
,
R. H. J.
, &
Utzerath
,
C.
(
2013
).
Lateralized power spectra of the EEG as an index of visuospatial attention
.
Advances in Cognitive Psychology
,
9
,
184
201
.
van der Lubbe
,
R. H. J.
, &
Woestenburg
,
J. C.
(
1997
).
Modulation of early ERP components with peripheral precues: A trend analysis
.
Biological Psychology
,
45
,
143
158
.
VanRullen
,
R.
, &
Macdonald
,
J. S. P.
(
2012
).
Perceptual echoes at 10 Hz in the human brain
.
Current Biology
,
22
,
995
999
.
van Veen
,
V.
, &
Carter
,
C. S.
(
2002
).
The anterior cingulate as a conflict monitor: fMRI and ERP studies
.
Physiology and Behavior
,
77
,
477
482
.
van Veen
,
V.
, &
Carter
,
C. S.
(
2005
).
Separating semantic conflict and response conflict in the Stroop task: A functional MRI study
.
Neuroimage
,
27
,
497
504
.
Varela
,
F.
,
Lachaux
,
J.
,
Rodriguez
,
E.
, &
Martinerie
,
J.
(
2001
).
The Brainweb: Phase large-scale integration
.
Nature Reviews Neuroscience
,
2
,
229
239
.
Verleger
,
R.
,
Kuniecki
,
M.
,
Möller
,
F.
,
Fritzmannova
,
M.
, &
Siebner
,
H. R.
(
2009
).
On how the motor cortices resolve an inter-hemispheric response conflict: An event-related EEG potential-guided TMS study of the flankers task
.
European Journal of Neuroscience
,
30
,
318
326
.
Verleger
,
R.
,
Metzner
,
M. F.
,
Ouyang
,
G.
,
Śmigasiewicz
,
K.
, &
Zhou
,
C.
(
2014
).
Testing the stimulus-to-response bridging function of the oddball-P3 by delayed response signals and residue iteration decomposition (RIDE)
.
Neuroimage
,
100
,
271
280
.
Vogel
,
E. K.
, &
Machizawa
,
M. G.
(
2004
).
Neural activity predicts individual differences in visual working memory capacity
.
Nature
,
428
,
748
751
.
Wang
,
P.
,
Goschl
,
F.
,
Friese
,
U.
,
Konig
,
P.
, &
Engel
,
A. K.
(
2019
).
Long-range functional coupling predicts performance: Oscillatory EEG networks in multisensory processing
.
Neuroimage
,
196
,
114
125
.
Wascher
,
E.
, &
Wauschkuhn
,
B.
(
1996
).
The interaction of stimulus- and response-related processes measured by event-related lateralizations of the EEG
.
Electroencephalography and Clinical Neurophysiology
,
99
,
149
162
.
Whitney
,
D.
, &
Levi
,
D. M.
(
2011
).
Visual crowding: A fundamental limit on conscious perception and object recognition
.
Trends in Cognitive Sciences
,
15
,
160
168
.
Worden
,
M. S.
,
Foxe
,
J. J.
,
Wang
,
N.
, &
Simpson
,
G. V.
(
2000
).
Anticipatory biasing of visuospatial attention indexed by retinotopically specific alpha band electroencephalography increases over occipital cortex
.
Journal of Neuroscience
,
20
,
RC63
.
Yeung
,
N.
,
Botvinick
,
M. M.
, &
Cohen
,
J. D.
(
2004
).
The neural basis of error detection: Conflict monitoring and the error-related negativity
.
Psychological Review
,
111
,
931
959
.
Zhang
,
H.
,
Chavarriaga
,
R.
, &
Millán
,
J. d. R.
(
2015
).
Discriminant brain connectivity patterns of performance monitoring at average and single-trial levels
.
Neuroimage
,
120
,
64
74
.
Zhigalov
,
A.
, &
Jensen
,
O.
(
2020
).
Alpha oscillations do not implement gain control in early visual cortex but rather gating in parieto-occipital regions
.
Human Brain Mapping
,
41
,
5176
5186
.