Elsevier

Cortex

Volume 133, December 2020, Pages 188-200
Cortex

Special Issue “The Brain’s Brake”: Research Report
Cognitive brakes in interference resolution: A mouse-tracking and EEG co-registration study

https://doi.org/10.1016/j.cortex.2020.09.024Get rights and content

Abstract

Cognitive control is particularly challenged when it is necessary to resolve interference and correct our behavior on-the-fly. To do this, it is necessary to inhibit the ongoing wrong action and reprogram a new motor plan as appropriate for the current task. This ability requires a complex interaction between cognitive and motor control. Here, we aimed at shedding light on this interplay. To do this, we administered a spatial version of the Stroop task comprising blocks with different Proportion Congruency (PC) manipulations (i.e., manipulating the percentage of congruent trials at 25%, 50% or 75%), to elicit different cognitive control demands. Moreover, we used two techniques with high-temporal resolution, as we simultaneously recorded EEG and mouse trajectories, that can be considered the real-time kinematic correlates of the ongoing cognitive processing. Specifically, we analyzed the Event Related Potentials (ERPs) locked to the peak deceleration time, which marks the suppression of ongoing erroneous trajectories, and we estimated their neural sources. We found three PC-dependent ERP components engaging distinct neural regions, which showed a reduction of the Stroop effect for low-PC blocks. By using a novel co-registration of mouse-trajectories and EEG, we suggest that the observed components may reflect different mechanisms engaged by reactive cognitive control to resolve the interference, including the suppression of an ongoing but no longer appropriate response, the selection of the new motor plan and its actual updating.

Introduction

The ability to adapt our behavior according to contexts and internal goals is one of the most fascinating aspects of human cognition. This ability, referred to as cognitive control, is particularly engaged when we need to select task-relevant information in the presence of non-target and distracting information (i.e., interference resolution: Miller & Cohen, 2001; Norman & Shallice, 1986). A paradigm that has been often adopted to understand how cognitive control works in the presence of distracting information is the Stroop task (Stroop, 1935). In this task, participants need to ignore a pre-potent distracting feature (color-word meaning), in order to attend another task-relevant one (ink color). In the spatial version of the Stroop task, participants are asked to attend the direction of an arrow (i.e., the task-relevant feature) while ignoring the location in which it appears (i.e., the to-be-ignored feature). Therefore, the stimuli are characterized by two features and they are defined as congruent when these features match, or incongruent when they mismatch. The cost in responding to the incongruent stimuli, compared to the facilitation reported for the congruent ones, is referred to as Stroop effect.

Despite the large interest in cognitive control and interference resolution, the precise mechanisms through which interference resolution is implemented are still not entirely clear. According to the Dual Mechanisms of Control model (Braver, 2012) cognitive control may work by means of two different modes: proactive and reactive control. Proactive control works by anticipating the interference, maintaining goal-relevant information in a sustained manner to impose an attentional bias toward the task-relevant information, whereas reactive control works as a late correction mechanism that is engaged on-demand and only after the detection of conflicts or interfering events. Some of the previous Stroop studies manipulated the proportion of congruency (PC) to modulate control demands and differently elicit proactive or reactive control, reporting a general reduction of the Stroop effect in low-PC conditions, that is, the high presence of incongruent trials engaged more control, improving the performance (see Bugg & Crump, 2012, for a review).

An important aspect to consider in the study of proactive and reactive control mechanisms concerns their different temporal dynamics, as they are thought to act in an anticipatory and corrective way, respectively (Braver, 2012). Surprisingly, however, very little is known on this issue up to date, as the specific mechanisms through which proactive and reactive control work are still unclear. For instance, even though proactive control has been defined as an anticipatory attentional bias and reactive control as a late correction mechanism, there is some evidence reporting that also reactive control may work as a fast stimulus–attention association (i.e., item-specific proportion congruency; Bugg, 2017; Bugg & Crump, 2012). This is in line with evidence suggesting that the Stroop interference especially arises at the stimulus processing stage, due to the mismatch of the two stimulus features (i.e., stimulus–stimulus competition; De Houwer, 2003; Zhang & Kornblum, 1998). However, there is also evidence suggesting that the Stroop interference also arises later, at the response selection stage, as due to the elicitation of a competing response associated with the to-be-ignored feature (i.e., stimulus-response competition; De Houwer, 2003). Indeed, paradigms where no stimulus-response competition is presented (e.g., color-word Stroop with manual keypress responses) elicit weaker Stroop effects as compared to paradigms including both types of competition (Augustinova, Parris, & Ferrand, 2019; Fennell & Ratcliff, 2019; MacLeod, 1991).

This body of evidence seems to suggest that Stroop interference may arise at different loci and may require different mechanisms to be resolved. For instance, according to the Cascade of Control model (Banich, 2009), multiple steps are engaged in interference resolution, working in a cascade-like manner. Early steps are carried out by different portions of the dorso-lateral prefrontal cortex and are involved in setting attentional biases toward the relevant information. Later steps are carried by different portions of the dorso-medial prefrontal cortex (DMPFC) and are involved in more response-related stages to select the appropriate response and evaluate it. A prediction of this model is that when the earlier steps fail (or cannot be implemented), the later steps are recruited to a greater extent, meaning that cognitive control would be required at response-related steps to resolve interference and, if possible, correct the responses. In line with the Dual Mechanism of Control model, this control mechanism would thus be particularly relevant in situations where the likelihood of interference occurrence is low (i.e., in conditions with high PC) and, thus, cognitive control is not engaged proactively but reactively, in response to the detection of the Stroop interference in incongruent trials (De Pisapia & Braver, 2006; Grandjean et al., 2012; West & Bailey, 2012).

Importantly, standard response outputs, such as keypresses or vocal responses, constitute only the end point of cognitive processes and cannot usually be corrected, making it hard to study the dynamic correction processes that may be implemented by this type of reactive control. To shed light on this aspect, the recording of mouse trajectories can result particularly helpful, because it can be considered as a window on the online cognitive processes while they are ongoing, capturing the updating and the correction of the responses. The use of this technique can provide important advantages in understanding how interference effects evolve over time. For example Bundt, Ruitenberg, Abrahamse, and Notebaert (2018), using a Stroop task recording mouse trajectories, suggest that cognitive control can affect the response also after the initiation of the movement, against the traditional view that serial distinct processing steps are required (Buc Calderon, Verguts, & Gevers, 2015). Moreover, another previous study, using frequency tagged EEG in a Flanker task, showed that cognitive control readjustments occur continuously in the same trial (Scherbaum, Fischer, Dshemuchadse, & Goschke, 2011). Other recent studies similarly suggest that cognitive and sensorimotor systems dynamically interact and may work in parallel in a continuous way (Cisek & Kalaska, 2005, 2010; Song & Nakayama, 2006). For these reasons, the kinematic correlates (i.e., mouse trajectories) of interference resolution can provide important insights on the control exerted during response execution, shedding light on how a wrong response is stopped, updated and modified in order to provide the correct one. The recording of mouse trajectories has been incrementally attracting attention in the last years, but only few previous studies focused on the study of cognitive control, specifically adopting a PC manipulation at the item-level (Bundt et al., 2018; Ruitenberg, Braem, Du Cheyne, & Notebaert, 2019).

In the study of the temporal dynamics of interference resolution, the recording of the electrophysiological (EEG) activity can also provide important benefits. Prior works investigated interference resolution process in the Stroop task through the study of the event-related potential (ERP) components. However, the precise cognitive mechanisms that these components reflect in interference resolution (e.g., conflict detection, response selection or conflict resolution) are still a matter of debate. Also, previous studies in the framework of the Dual Mechanism of Control model tried to dissociate proactive and reactive ERP components (Appelbaum, Boehler, Davis, Won, & Woldorff, 2014; Tillman & Wiens, 2011; West & Bailey, 2012), but it is still difficult to draw firm conclusions about the specific neural correlates of cognitive control, especially due to the differences of the paradigms and PC manipulations used. Finally, to date no study investigated the electrophysiological correlates of the reactive control processes that are required to stop ongoing erroneous action plans and correct them in order to provide the correct response.

In the present study, we tackled these issues by investigating the impact of different cognitive control demands in interference resolution. Specifically, we aimed to shed light on the interaction between motor and cognitive control processes in interference resolution to reveal for the first time the neural correlates of the control mechanisms required to stop and correct the execution of prepotent but erroneous actions during the resolution of Stroop interference. To this aim, we recorded both EEG activity and mouse trajectories while participants were engaged in a 4-choice spatial Stroop task requiring mouse responses, where we manipulated the PC by presenting blocks with high or low PC. We combined EEG recording and mouse tracking to link behavioral and neural dynamics to specific control mechanisms in interference resolution. Based on trial-level mouse kinematics, we analyzed ERPs locked to the time of the peak deceleration of mouse trajectories, which we assumed to mark the suppression of ongoing erroneous responses. Indeed, the Stroop interference in incongruent trials causes an attraction of the mouse trajectory towards the location of the arrow (i.e., the to-be-ignored feature). Once this interference has been detected, cognitive control will thus “pull the brake” of the initiated mouse movement, decelerating the ongoing movement and updating the motor plan to provide the correct answer.

We expected to find significant PC-dependent modulations of deceleration-locked ERPs in a time window preceding the time of the peak deceleration. This would reflect the control mechanisms required to detect the interference and resolve it at response-related stages by selecting the appropriate response (Banich, 2009). Based on an extensive body of evidence, we expected this effect to be localized over midfrontal channels to reflect the involvement of DMPFC in these processes (Cavanagh & Frank, 2014; Milham et al., 2001). Moreover, we predicted significant PC-dependent modulations of deceleration-locked ERPs soon after the time of the peak deceleration, which would reflect the updating of the motor plan and the correction of the ongoing mouse trajectories. All these effects should be especially implicated in incongruent trials, where the suppression and correction of ongoing erroneous responses are mainly required, and should be larger in conditions with higher PC, where the Stroop interference is stronger.

Section snippets

Material and methods

We report how we determined our sample size, all data exclusions, all inclusion/exclusion criteria, whether inclusion/exclusion criteria were established prior to data analysis, all manipulations, and all measures in the study.

The raw, pre-processed, and summary EEG and mouse-tracking data are available from our project repository on the Open Science Framework (https://osf.io/sg4f6/). The code for the experimental task and for the analysis is also available at the same link. No part of the

Behavioral measures

Table 1 shows descriptive statistics for the behavioral results. The percentage of correct trials was very high (99.31%), so we did not analyze accuracy. On average, all participants reported faster Initiation Time, Movement Time, RTs, and times of peak deceleration, as well as weaker peak decelerations and smaller deviations, for congruent compared to incongruent trials, as demonstrated by a significant main effect of Trial Type [respectively, F (1,43) = 63.75, 114.01, 216.41, 255.35, 6.92,

Discussion

In the present study, we investigated the neural correlates of cognitive control in interference resolution. In particular, we wanted to shed light on the interaction between motor and cognitive processes to unveil how cognitive control continuously monitors and adjust motor responses, selecting the appropriate one and updating the motor plan if the ongoing action is no longer appropriate. To this aim, we recorded mouse trajectories and EEG during the execution of a spatial Stroop task with

Conclusions

To sum up, with this study we tried to shed light on the complex interaction between cognitive and motor control in overcoming Stroop interference, especially when it is required to stop and correct a motor response. To the best of our knowledge, this is the first study investigating the electrophysiological correlates (ERP) of the movement inhibition and adjustments in a mouse-tracking paradigm. Our results revealed three main deceleration-locked ERP components, with clear PC-related

Open practices

The study in this article earned Open Materials and Open Data badges for transparent practices. Materials and data for the study are available at https://osf.io/sg4f6/.

Acknowledgments

The authors would like to thank Dr. Antonino Visalli, who helped with the development of the mouse-tracking paradigm. This work was funded by the European Research Council Starting Grant LEX-MEA n° 313692 (FP7/2007–2013) to A.V.

References (76)

  • A. Delorme et al.

    EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis

    Journal of Neuroscience Methods

    (2004)
  • E. Filevich et al.

    Negative motor phenomena in cortical stimulation: Implications for inhibitory control of human action

    Cortex

    (2012)
  • A. Hampshire et al.

    The role of the right inferior frontal gyrus: Inhibition and attentional control

    Neuroimage

    (2010)
  • C. Meckler et al.

    Motor inhibition and response expectancy: A Laplacian ERP study

    Biological Psychology

    (2010)
  • M.P. Milham et al.

    Competition for priority in processing increases prefrontal cortex's involvement in top-down control: An event-related fMRI study of the stroop task

    Cognitive Brain Research

    (2003)
  • M. Milham et al.

    The relative involvement of anterior cingulate and prefrontal cortex in attentional control depends on nature of conflict

    Cognitive Brain Research

    (2001)
  • R.C. Oldfield

    The assessment and analysis of handedness: The Edinburgh inventory

    Neuropsychologia

    (1971)
  • F. Perrin et al.

    Spherical splines for scalp potential and current density mapping

    Electroencephalography and Clinical Neurophysiology

    (1989)
  • M. Rushworth et al.

    Action sets and decisions in the medial frontal cortex

    Trends in Cognitive Sciences

    (2004)
  • S.M. Smith et al.

    Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference

    Neuroimage

    (2009)
  • N.C. Swann et al.

    Roles for the pre-supplementary motor area and the right inferior frontal gyrus in stopping action: Electrophysiological responses and functional and structural connectivity

    Neuroimage

    (2012)
  • D. Swick et al.

    Are the neural correlates of stopping and not going identical? Quantitative meta-analysis of two response inhibition tasks

    Neuroimage

    (2011)
  • L.G. Appelbaum et al.

    The dynamics of proactive and reactive cognitive control processes in the human brain

    Journal of Cognitive Neuroscience

    (2014)
  • A.R. Aron et al.

    Triangulating a cognitive control network using diffusion-weighted magnetic resonance imaging (MRI) and functional MRI

    Journal of Neuroscience

    (2007)
  • M. Augustinova et al.

    The loci of stroop interference and facilitation effects with manual and vocal responses

    Frontiers in Psychology

    (2019)
  • S. Baillet et al.

    Evaluation of inverse methods and head models for EEG source localization using a human skull phantom

    Physics in Medicine & Biology

    (2001)
  • M.T. Banich

    Executive function: The search for an integrated account

    Current Directions in Psychological Science

    (2009)
  • C. Buc Calderon et al.

    Losing the boundary: Cognition biases action well after action selection

    Journal of Experimental Psychology: General

    (2015)
  • J.M. Bugg

    Context, conflict, and control

    (2017)
  • J.M. Bugg et al.

    In support of a distinction between voluntary and stimulus-driven control: A review of the literature on proportion congruent effects

    Frontiers in Psychology

    (2012)
  • J.M. Bugg et al.

    Revealing list-level control in the Stroop task by uncovering its benefits and a cost

    Journal of Experimental Psychology: Human Perception and Performance

    (2011)
  • C. Bundt et al.

    Early and late indications of item-specific control in a Stroop mouse tracking study

    Plos One

    (2018)
  • B. Burle et al.

    Preventing (impulsive) errors: Electrophysiological evidence for online inhibitory control over incorrect responses

    Psychophysiology

    (2016)
  • C.H. Chatham et al.

    Cognitive control reflects context monitoring, not motoric stopping, in response inhibition

    Plos One

    (2012)
  • P. Cisek et al.

    Neural mechanisms for interacting with a world full of action choices

    Annual Review of Neuroscience

    (2010)
  • M. Corbetta et al.

    Control of goal-directed and stimulus-driven attention in the brain

    Nature Reviews Neuroscience

    (2002)
  • D. Czernochowski et al.

    On why not to rush older adults—relying on reactive cognitive control can effectively reduce errors at the expense of slowed responses

    Psychophysiology

    (2010)
  • J. De Houwer

    On the role of stimulus-response and stimulus-stimulus compatibility in the Stroop effect

    Memory & Cognition

    (2003)
  • Cited by (10)

    • Impaired cognitive control in patients with brain tumors

      2022, Neuropsychologia
      Citation Excerpt :

      We did not find significant associations between behavioral indices of the Stroop task and lesioned voxels. This null result could be in part explained by the fact that the task used here has not a fine-grained structure suitable to capture distinct cognitive control processes (see Gonthier et al., 2016a; Tafuro et al., 2020). Remarkably, all neuroimaging results were obtained after controlling for the total lesion volume, therefore they are independent of the lesion size.

    View all citing articles on Scopus
    View full text