RESEARCH DOSSIER
Action Selection and Inhibition
Abstract. Action selection and inhibition govern how competing behavioral possibilities become executable responses and how prepared or ongoing actions are withheld or canceled. This fundamental research monograph critically synthesizes the historical experimental literature, Stroop and Simon interference methods, go/no-go and stop-signal paradigms, independent race models, evidence-accumulation accounts, proactive and reactive control, learned stimulus-stop associations, and frontal–basal-ganglia mechanisms. Original experiments, methodological consensus, lesion and stimulation evidence, comparative research, and computational analyses are examined in relation to measurement assumptions, causal inference, construct validity, and replication. The synthesis distinguishes early withholding, late cancellation, decision caution, attentional detection, conflict-related adjustment, and motor suppression. It considers development, clinical relevance, ecological validity, and human–artificial decision systems. The work presents a critical narrative synthesis of published evidence; it reports no original experiments and introduces no independent psychological theory.
Keywords: action selection; response inhibition; stop-signal task; go/no-go; Stroop interference; Simon effect; cognitive control; proactive control; reactive inhibition; independent race model; stop-signal reaction time; decision threshold; fronto-basal-ganglia circuits; subthalamic nucleus; inferior frontal cortex; trigger failure; executive functions; motor suppression; Artificial Era
Introduction: Choosing Actions and Regulating Their Execution
Action selection is the process by which an organism resolves competing possibilities for behavior into an executable response under the constraints of goals, perceptual information, learned contingencies, and available motor capacity. Inhibition is a family of processes by which an imminent, prepotent, or ongoing response is withheld, interrupted, or displaced. Their relation is constitutive: choosing one response changes the fate of alternatives, while interrupting an action returns the system to a state in which another response may be selected. This monograph examines their behavioral and neural implementation, historical experimental foundations, and methodological limits. The empirical domain is human cognitive control, informed by comparative experiments and computational models rather than by an assumption of a unitary inhibitory faculty Mostofsky & Simmonds (2008) Ridderinkhof et al. (2004).
The research questions follow the temporal structure of action. How are candidate responses activated by stimuli and habits? How do task goals bias competition before movement begins? What permits a late stop signal to cancel a prepared action? How does repeated withholding modify later choice? Finally, how should scientists distinguish a failure to detect a stop cue from failure of the stopping mechanism itself? These questions require evidence from controlled reaction-time paradigms, response-frequency manipulations, individual-difference research, lesion and stimulation studies, and neurally constrained mathematical models. Every method captures a partial causal sequence, and its interpretive power depends on the relation between its observations and its assumptions Logan et al. (1984) Verbruggen et al. (2019).
The evidential approach is a critical narrative synthesis of original experiments, theoretical analyses, methodological recommendations, and reviews. It compares convergent and opposing results without treating the existence of a task effect as automatic identification of a neural or psychological mechanism. In particular, response slowing may reflect strategy, surprise, interference, changes in decision threshold, or motor suppression, depending on design. Experimental inhibition must be described operationally, with the measured response, delay, accuracy, speed distribution, and counterfactual control condition specified. The result is a scientific account of action coordination across task, brain, and environment, with explicit boundaries for translating the account to applied and artificial decision systems Hampshire et al. (2010) Ratcliff & McKoon (2008).
The Scientific Object: Selection, Withholding, Cancellation, and Replacement
Selection operates whenever multiple actions have nonzero readiness and only a subset can be executed. A task can demand a choice between buttons, a switch from a habitual to a rule-governed response, an eye movement away from a salient target, or an adjustment to an unexpected environmental signal. Withholding is the instructed omission of an otherwise possible response, commonly studied using go/no-go tasks. Cancellation concerns interrupting an action that has already been prepared or initiated, as operationalized by stop-signal procedures. Replacement requires terminating one response plan while initiating another. These distinctions separate the observable demands before explaining whether they share processes Eagle et al. (2008) Mostofsky & Simmonds (2008).
The word inhibition occurs at several explanatory levels. At the behavioral level it names successful nonexecution under conditions favoring response. At the cognitive level it may designate a proposed control operation that suppresses an activated representation or motor plan. At the neural level inhibition can refer to inhibitory synaptic transmission, a circuit computation, or a downstream suppression of corticospinal excitability. Such meanings can be related only through explicit measurement models. A stopped button press cannot on its own establish that a dedicated suppressive command occurred; an elevated hemodynamic signal likewise cannot show which motor command was canceled Munakata et al. (2011) Hampshire et al. (2010).
Research on selection and inhibition therefore studies a coupled control problem across multiple timescales. Milliseconds determine whether a stop process outruns a go process, seconds and minutes allow post-error and post-stop adjustments, and repeated task experience builds associations between stimuli and stopping. Over longer periods, development and individual differences modify performance distributions through interacting cognitive capacities. A successful theory must explain the immediate decision and the change in future behavior caused by the decision. This framework retains precise operational distinctions while allowing that some tasks recruit overlapping goals, attention, learning, and motor circuitry Verbruggen & Logan (2008a) Bissett & Logan (2011).
Historical Foundations: Interference and Experimental Control
Early experimental psychology approached action regulation through interference between competing associations. Stroop's 1935 experiments demonstrated the difficulty of naming an ink color when a printed color word signaled a competing response, compared with more compatible conditions. The result helped establish interference as a measurable consequence of the relative strengths and automaticity of stimulus-response pathways. It did not, by itself, isolate a single inhibition process. The task combines reading experience, color naming, response selection, perceptual registration, and strategic task maintenance; the modern investigator must identify which contribution is changed by a manipulation Stroop (1935) Friedman & Miyake (2004).
The spatial compatibility effects investigated by Simon in 1969 introduced another influential way to reveal response activation that conflicts with task instructions. A stimulus can elicit a tendency to respond on the corresponding side even when spatial location is irrelevant to the explicitly required decision. Unlike the classic color-word Stroop manipulation, a Simon task places the conflict directly in the mapping between irrelevant spatial information and motor output. Convergence across the two families supports the broad existence of response competition, while differences among tasks warn against treating conflict cost as a pure measure of one inhibitory capacity Simon (1969) Wiecki & Frank (2013).
A decisive later change was to study stopping against the clock. Logan, Cowan, and Davis's 1984 experiments presented a go stimulus and, on some trials, a delayed signal instructing participants to stop. The formal race model assigned separate finishing times to going and stopping, enabling inference about the unobserved stopping latency from observed response distributions and inhibition frequencies. This converted a broad intuition about self-control into a quantitatively constrained experimental question. Its enduring influence comes from the model's falsifiable distributional assumptions and from the possibility of improving task design when those assumptions fail Logan et al. (1984) Verbruggen & Logan (2008c).
Response Competition in Stroop, Simon, and Flanker Paradigms
Interference tasks create situations in which task-relevant information favors a response while another stimulus dimension activates a competitor. In a Stroop design, the participant responds to ink color while reading processes can activate the word meaning. In Simon designs, irrelevant location biases spatially aligned responses. In flanker designs, neighboring symbols may activate the same or a different response from the target. Their common output is frequently a contrast between compatible and incompatible response time or error rates. The contrast establishes that additional processing demands occur under conflict, but it need not identify exactly how competing activation is resolved Stroop (1935) Simon (1969).
Several mechanisms can produce similar interference costs. An irrelevant stimulus can accelerate the competing response, divert attention before evidence accumulation, prolong response verification, or require threshold adjustment. Different speed-accuracy regimes can reverse the apparent ranking of conditions: a person choosing caution may take longer but make fewer errors. Consequently, reports should retain both accuracy and the shape of response-time distributions rather than rely on an isolated mean. Sequential-sampling models provide a disciplined method for distinguishing changes in evidence quality from changes in decision threshold, although parameter identification still requires suitable trial counts and model validation Ratcliff & McKoon (2008) Forstmann et al. (2016).
Inference improves when a study manipulates the timing and relative strength of competing evidence and compares the resulting distributions with explicit mechanistic predictions. For example, a theory of early automatic activation predicts a particular time course of incorrect response tendencies, whereas a late strategic adjustment predicts a different response-time pattern. Cross-paradigm correlations also demand care: weak associations can arise because tasks differ in measurement reliability, practice histories, error variance, and perceptual load. The scientific task is to infer which components are shared across paradigms after these sources of variation are measured Friedman & Miyake (2004) Miyake et al. (2000).
The Stop-Signal Paradigm and the Independent Race Model
The stop-signal task embeds occasional stop trials within a primary speeded response task. On a go trial, the participant executes the instructed choice; on a stop trial, a later signal requires cancellation. The stop-signal delay is the interval between go stimulus and stop cue. Longer delays ordinarily leave less opportunity to prevent the response. The race model posits a go finishing time and a stop finishing time, and successful inhibition occurs when the stop process completes first. Since successful stopping yields no observable latency, researchers infer stop-signal reaction time from the go-response distribution and the proportion of failed stops under specified assumptions Logan et al. (1984) Verbruggen et al. (2019).
The integration method estimates stopping latency using the quantile of the go response-time distribution corresponding to the probability of responding on stop trials, with appropriate correction for stop-signal delay. Tracking methods vary that delay adaptively to bring the probability of responding toward a target region and sustain an informative mix of successful and failed stops. This statistical construction is compelling because its logic is explicit, yet its estimate is conditioned on independence assumptions and adequate task engagement. Omission errors on go trials, deliberate waiting, and insufficient stop-trial counts can severely distort the derived measure Verbruggen & Logan (2008c) Verbruggen et al. (2019).
A crucial diagnostic implication is that failed-stop responses should generally be faster than ordinary go responses when the independent race is operating as assumed. If this ordering is violated, the investigator should consider strategic slowing, failures to trigger the stopping process, contextual dependence between going and stopping, or data and design problems. The stop-signal reaction time is a model-based latent estimate rather than a directly recorded neural event or a context-free trait score. Its value becomes informative when reported alongside the full inhibition function, go performance, failed-stop response times, and uncertainty Logan et al. (1984) Matzke et al. (2017).
Go/No-Go Performance and the Problem of Prepotency
Go/no-go tasks ask participants to respond to one category of stimuli and withhold responding to another. This is a useful assay of response restraint, particularly when go trials are frequent and strongly establish a prepotent response. The observable outcomes include omission errors on go trials, commission errors on no-go trials, and response latency on the trials that require action. Prepotency is therefore an experimental property that must be constructed and documented, not presumed from the task label. A design with equally frequent go and no-go signals may demand a qualitatively different strategy from a rare no-go signal embedded in sustained response practice Eagle et al. (2008) Nigg (2000).
The central interpretive problem is that successful nonresponding can occur because the no-go cue was discriminated early and the go process was never launched. In a stop-signal task, by comparison, the delay is deliberately arranged to confront an action already in progress or advanced preparation. Both paradigms recruit selection between acting and withholding, but they differ in timing, stimulus classification, and the behavioral evidence available for a race between competing processes. Conflating commission errors with stop-signal reaction time can therefore obscure meaningful distinctions about control initiation and late cancellation Mostofsky & Simmonds (2008) Verbruggen & Logan (2008c).
Task learning adds another layer. Experiments by Verbruggen and Logan found that repeated pairing of particular stimuli with stopping can subsequently slow responding to those stimuli, including after response rules change. This provides evidence that inhibition-related performance can acquire stimulus-specific associative components. An observed improvement may consequently reflect memory retrieval and stimulus-response history as well as any change in a flexible executive operation. The strongest research designs counterbalance stimulus mappings, examine transfer to new stimuli, and separate general capacity from learned cue-specific tendencies Verbruggen & Logan (2008a) Verbruggen & Logan (2008b).
Goal Maintenance, Preparation, and Proactive Restraint
Selection begins before an overt stimulus arrives. A participant who knows that stopping may be required can slow go responses, maintain the stopping rule in working memory, and adjust the amount of commitment to a prepared action. This prospective control changes the probability of successful stopping even if the speed of the stop process itself stays constant. Braver's dual-mechanisms framework distinguishes sustained, anticipatory use of task context from control recruited after a triggering event; these modes provide hypotheses about proactive and reactive components without assuming that each uses a separate anatomical organ Braver (2012) Bissett & Logan (2011).
In stop-signal research, strategic slowing creates a serious inferential challenge. Participants can improve stopping success simply by waiting before responding, which changes the go-response distribution and may change the stochastic independence of the race. Investigators therefore instruct participants to respond promptly, monitor response times, and use sufficiently low stop-trial frequency to preserve a genuine go task. The adaptive staircase changes stop-signal delay in response to performance, but this cannot compensate for every form of strategic waiting. Researchers should report evidence of slowing and changes in response frequency instead of attributing all between-group differences to stopping speed Verbruggen et al. (2019) Bissett & Logan (2011).
Proactive restraint has adaptive significance in environments where acting too soon is costly. A driver approaching a crowded intersection, for example, may maintain a readiness to brake before any particular hazard appears. The analogy illustrates how the probability and cost of an interruption can influence action preparation, but it supplies no measurement by itself. Experimental demonstrations must vary expected stopping demands and evaluate whether the pattern follows a predictive account. When speed-accuracy tradeoffs and strategic caution are directly modeled, an apparent inhibitory deficit can sometimes be traced to a different policy for balancing caution against rapid action Ratcliff & McKoon (2008) Forstmann et al. (2016).
Reactive Cancellation and Temporal Competition
Reactive stopping begins when a signal requires cancellation under the current policy. The go and stop processes race across a constrained interval; changing the delay modifies which process is likely to finish first. This formal temporal competition permits investigators to generate quantitative inhibition functions rather than rely on a binary description of successful control. Yet the model's mathematical clarity also exposes its limits. Selective attention to the stop cue, sensory encoding delay, and an occasional failure to initiate stopping are processes whose contribution must be measured or modeled before estimating a putative motor suppression latency Logan et al. (1984) Matzke et al. (2017).
Matzke, Love, and Heathcote's Bayesian model explicitly estimated trigger-failure probability together with the distribution of stopping times. Their analysis showed that stopping could fail because the stop process was never initiated and that ignoring this possibility could distort the inferred stop latency. Such findings change interpretation of individual and clinical differences. A longer estimated stopping time can reflect perceptual or attentional differences in reliably initiating a stop, not necessarily a slower inhibitory operation. A defensible experiment reports signal detection, go-task performance, trial-level errors, and parameter uncertainty in addition to any summary latency Matzke et al. (2017) Verbruggen et al. (2019).
The distinction between canceling one particular response and suppressing motor output more globally is another active topic. A signal may require stopping only one component of a movement while other actions continue, or it may evoke a broader transient pause. The consequences for partially prepared alternatives can be examined through electromyography, motor evoked potentials, and selective reaction measures. Broad suppression and selective stopping are mechanistic possibilities that depend on neural architecture, context, and training. Neither follows directly from a successful no-button-press trial, and both must be tested against possible attentional orienting to an unusual cue Wessel & Aron (2017) Hampshire et al. (2010).
Automaticity, Associations, and Learning to Stop
With practice, the association between a stimulus and a previous stopping episode can influence the next response. Verbruggen and Logan trained participants with consistently mapped stop cues and then changed the mapping; formerly stop-associated stimuli produced delayed responding under the new instruction. This result supplies a learning-based route to inhibitory effects. It is especially important because a participant can show action restraint without consciously re-evaluating a stopping rule on every trial. In this sense, control becomes partly embedded in cue-triggered retrieval from memory, while explicit goal maintenance remains available to handle unusual or changing situations Verbruggen & Logan (2008a) Verbruggen & Logan (2008b).
Associative stopping complements rather than replaces the temporal account of cancellation. A learned cue can slow the go process or strengthen early withholding before a late stop process would be necessary, while a fresh stop signal can still demand reactive control. Laboratory tasks therefore need to distinguish learning during repeated trials from the response to a newly introduced interruption. Repetition probabilities, stimulus identity, task-switching rules, and interval structure can all alter this relation. A model that attributes every slow response to deliberative inhibition risks missing the experimental signature of retrieved stimulus-stop associations Verbruggen & Logan (2008a) Bissett & Logan (2011).
These findings have consequences for training claims. Improving performance on repeated no-go items may mean that a specific item acquired an inhibitory association. Stronger evidence for transferable inhibitory control requires new stimuli, altered mappings, alternate tasks, and follow-up measures beyond the original training context. Transfer should be quantified rather than assumed from pretest-posttest change, particularly if baseline tasks are unreliable or participants know the repeated materials. The methodological implication is direct: a theory of action regulation must explain what is learned and under which conditions the learned policy survives changes in context Verbruggen & Logan (2008a) Friedman & Miyake (2004).
Control Monitoring, Conflict Signals, and Behavioral Adjustment
An agent needs information about when its current response policy is insufficient. Conflict-monitoring theories address this problem by proposing that simultaneous activation of incompatible response tendencies can signal demand for stronger control. Botvinick and colleagues implemented computational models in which a conflict measure could explain patterns of anterior cingulate recruitment and facilitate subsequent control adjustment. The models provide an articulated causal proposal rather than a direct demonstration that one neural region is a conflict detector. Their contribution is the link between a measurable conflict signal, an adaptive control policy, and predicted behavior on later trials Botvinick et al. (2001) Ridderinkhof et al. (2004).
Post-stop slowing illustrates the challenge of identifying an adjustment's cause. After a stop signal, the next go response can be slower following successful or failed stopping. Bissett and Logan compared explanations including error monitoring, conflict, surprise, goal priority, and memory. Their results favored a contribution of changing task priority and indicated memory-related effects, while some patterns contradicted simpler error-based accounts. Slowing is therefore an outcome that requires explanatory discrimination across conditions; a single comparison of after-error and baseline reaction times cannot determine the control mechanism Bissett & Logan (2011).
Control monitoring unfolds within ongoing decision making. The same observed slowing may represent a higher response threshold, a revised goal priority, response-specific priming, or interruption after surprising feedback. To distinguish these possibilities, an investigator can manipulate stimulus repetition, preceding trial type, stop-signal frequency, and trial-to-trial reward or cost contingencies. Model comparison then becomes a way of testing predictions about the full response-time distribution and error pattern rather than selecting an attractive brain-region narrative. Such analyses also require confidence intervals and out-of-sample checks, since a flexible model may reproduce a result for multiple incompatible reasons Forstmann et al. (2016) Bissett & Logan (2011).
Evidence Accumulation and Decision Thresholds
Action selection can be described as a dynamic accumulation of noisy evidence favoring one option over another. In the diffusion decision model, drift, boundary separation, starting point, and nondecision time affect both accuracy and response-time distributions. A higher boundary can improve accuracy while delaying the response; stronger evidence accumulation can improve performance without an independent inhibitory command. The model provides a diagnostic language for separating caution from informational quality in speeded choices. It should be used where its assumptions fit the task structure, not treated as a universal neural anatomy of action Ratcliff & McKoon (2008) Forstmann et al. (2016).
A stop-signal race model and a two-choice diffusion process address partially overlapping questions. The race model quantifies the contest between completing an already developing go response and responding to the stop instruction. A diffusion model characterizes how evidence accumulates toward a choice and how criteria determine commitment. Integrated accounts must specify what happens to accumulated evidence after a stop cue, whether the decision process pauses, and whether a separate process suppresses motor implementation. An improved fit to mean response times alone is insufficient because several parameter combinations can imitate the same pattern Logan et al. (1984) Wiecki & Frank (2013).
Computational comparisons become informative when their models are confronted with selective experimental manipulations. A threshold account predicts particular joint changes in speed and accuracy; a stimulus-quality manipulation should alter evidence quality; a motor disruption may change downstream execution without producing the same choice signature. Simulation-based recovery checks whether the dataset can identify the parameters it is supposed to estimate. Behavioral distributions and neural timing jointly constrain a hypothesis more tightly than isolated effect sizes, while model misspecification remains an explicit possibility Ratcliff & McKoon (2008) Forstmann et al. (2016).
A Taxonomy of Inhibitory Functions and Individual Differences
The construct of inhibition covers several operations that can dissociate. Resistance to distractor interference concerns intrusion from irrelevant external stimuli. Resistance to proactive interference concerns persistence of previously relevant material in memory. Prepotent response inhibition concerns the regulation of a dominant action. Friedman and Miyake's latent-variable study of 220 adults found substantial commonality between prepotent-response inhibition and distractor resistance, while resistance to proactive interference showed a different pattern of relations. This result supports a structured taxonomy rather than a single questionnaire-like inhibitory score that explains every task Friedman & Miyake (2004) Nigg (2000).
Miyake and colleagues' earlier latent-variable research treated shifting, updating, and inhibition as related yet distinguishable executive functions, drawing attention to their shared and distinct variance. Such approaches can reduce the influence of idiosyncratic task demands, but their validity still depends on reliable component tasks and an adequate measurement model. The researcher's causal interpretation cannot exceed what the latent structure establishes. A factor fitted to behavioral indicators measures shared statistical variance; its psychological nature requires convergent experimental tests and theoretical specification Miyake et al. (2000) Friedman & Miyake (2004).
For individual differences, the critical issue is whether a task's ranking of people generalizes across sessions, materials, and contexts. A robust group-level interference effect can coexist with poor reliability of individual difference scores, because subtraction scores often discard the stable variance shared across conditions. Developmental, neurological, and clinical comparisons must therefore distinguish mean performance differences, the distributions of commission and omission errors, strategy, and the reliability of estimated latent parameters. Claims about impulsivity or self-control should be tied to precisely measured behavior and not inferred from one failed response Diamond (2013) Nigg (2000).
Frontal–Basal-Ganglia Circuits in Action Regulation
Neuroscientific models of action control describe interactions among frontal cortex, basal ganglia, thalamus, and motor systems. The basal ganglia contribute to selecting and gating candidate actions, while frontal representations maintain rules and signal contextual adjustments. In computational work, competition between rapid habit-like and slower goal-sensitive policies can be paused when conflict is detected, allowing an updated action to be selected. Wiecki and Frank formalized interactions between executive frontal networks and basal-ganglia pathways, testing whether their circuit model could reproduce performance in antisaccade, Simon, and stop-signal paradigms Wiecki & Frank (2013) Ridderinkhof et al. (2004).
The subthalamic nucleus is frequently implicated in rapid adjustment of the action-selection gate. A proposed hyperdirect pathway from frontal regions can transiently influence basal-ganglia output, buying time for conflict resolution or cancellation. Evidence includes computational simulations, recordings, and interventions that alter performance when basal-ganglia circuits are disrupted. Yet different tasks involve distinct sensory inputs, movement effectors, and demands on preparation; the same anatomical node may contribute to several computations. This is why a neural circuit model must distinguish a broad response pause, a choice-specific gate, and adjustment of decision caution Wiecki & Frank (2013) Wessel & Aron (2017).
Interventional evidence provides a stronger causal constraint than coactivation. Studies examining stimulation of the subthalamic region in Parkinson's disease reported changes in both response selection and inhibition. Such findings demonstrate that perturbing a component of the relevant circuit can change performance, but they arise in a disease and treatment context. Medication state, disease-related neural changes, task design, and stimulation spread can affect interpretation. The study of healthy action selection is strengthened by triangulating these results with nonclinical behavior, computational experiments, and noninvasive measurements rather than assuming direct physiological equivalence van den Wildenberg et al. (2006) Eagle et al. (2008).
Inferior Frontal Cortex: Stopping and Salience Detection
The inferior frontal cortex has received particular attention because stop-signal tasks recruit its activity and lesions can alter response control. Aron, Robbins, and Poldrack reviewed evidence implicating right inferior frontal regions in inhibitory control and situated these regions in a larger fronto-subcortical network. This literature supports the importance of frontal circuitry, but precise functional assignment remains a question of experimental contrasts. When stopping is compared only with ordinary going, additional events such as detection of the infrequent stop cue, attentional orienting, and switching task goals are also changed Aron et al. (2014) Ridderinkhof et al. (2004).
Hampshire and colleagues addressed this problem by holding cue-related stimulation more closely matched while varying whether the cue required inhibition, a new response, or no overt response. They observed right inferior frontal engagement across important-cue conditions, supporting a more general account involving attentional control or cue significance. The finding challenged a narrow interpretation in which activation of this region uniquely demonstrates motor inhibition. The theoretical consequence is that a neural response must be interpreted relative to the contrast that generated it and the full space of functions recruited by the task Hampshire et al. (2010).
Lesion evidence adds specificity while revealing network complexity. Swick, Ashley, and Turken reported response inhibition impairments linked to left inferior frontal damage, complicating an exclusive one-region account. Differences in lesion location, task context, required motor output, and compensation may help explain variable lateralization. A robust mechanistic explanation therefore maps distributed circuit interactions and their causal contribution under defined conditions instead of using an activated region as a single-function label. A frontal area may participate in identifying a need for change, initiating a stop process, maintaining a task rule, or several of these operations Swick et al. (2008) Aron et al. (2014).
Subthalamic and Global Motor Suppression Hypotheses
An unexpected event can interrupt an action even when the event contains no explicit instruction to stop. Wessel and Aron's synthesis proposed that surprise can recruit a fronto-basal-ganglia stopping network with a transient and potentially broad effect on the motor system. This account links orienting, interruption, and working-memory disruption to a common circuit-level possibility. The hypothesis is scientifically useful because it predicts measurable changes in motor excitability and behavior across contexts and makes it possible to compare deliberate stopping with the interruption generated by an unanticipated cue Wessel & Aron (2017).
Global suppression has to be distinguished from the local withholding of a particular motor response. If all motor channels are suppressed, unaffected actions should show transient costs; if the stop signal selectively gates one planned response, another can sometimes proceed. Measurements of electromyographic activity, corticospinal excitability, response timing in remaining effectors, and selective-stop designs can discriminate these possibilities. Even then, a task may recruit an initial broad pause followed by more selective reconfiguration. The inference depends on precisely synchronized measurements of the stop signal, physiological change, and overt movement Wessel & Aron (2017) Jahfari et al. (2010).
The circuitry also presents a problem of timescale. Frontal detection, basal-ganglia communication, thalamocortical modulation, and final muscle activity unfold sequentially and partly in parallel. If a model proposes that one signal exerts control at a particular stage, it should predict what happens when the action has already passed that stage. A signal detected before motor commitment can prevent a response; the same signal after commitment may only attenuate or modify it. Thus the temporal point of causal intervention is as important as its anatomical origin for evaluating neural stopping mechanisms Wiecki & Frank (2013) Logan et al. (1984).
Electrophysiology, Motor Excitability, and Partial Responses
Reaction times and errors summarize action outcomes, whereas electrophysiological measures can reveal preparation and suppression before overt movement. Surface electromyography may detect brief partial muscle activation during a successfully canceled trial. Motor evoked potentials, when used under suitable protocols, can index momentary changes in corticospinal excitability. Electroencephalography supplies time-resolved signals related to attention, conflict, and motor preparation. Each measurement has a different relation to the latent process of interest: a muscle burst shows downstream activation, while a scalp potential is a mixture of contributing neural sources Jahfari et al. (2010) Wessel & Aron (2017).
Jahfari and colleagues combined behavioral and physiological approaches to studying restraint during action preparation. The work illustrates how candidate stopping and preparation mechanisms can be assessed through convergent physiological and task measures. Nevertheless, recording a physiological difference during a stop trial does not by itself establish that the measured signal caused the cancellation. Stronger inference comes from timing, selective perturbation, control conditions for attentional surprise, and predictions linking physiological variation to the probability of observed behavior Jahfari et al. (2010) Hampshire et al. (2010).
Partial responses are theoretically important because they expose a continuum between complete prevention and overt execution. A participant may correctly withhold the button press while producing weak, short-lived activation of the response muscle. Such cases suggest that withholding can intervene late in motor preparation and that response success can hide considerable internal dynamics. Studies should predefine physiological thresholds, distinguish measurement artifacts from genuine activation, and report how signals are related to trial-level performance. An accurate account of inhibition includes the degrees and temporal structure of aborted movement, not only the eventual response count Verbruggen et al. (2019) Wessel & Aron (2017).
Lesions, Stimulation, and the Limits of Localization
Causal localization can be investigated by comparing task performance before and after focal neural disruption, stimulation, or damage. Yet the inference must include the anatomical specificity of the perturbation, functional compensation, and the interdependence of connected regions. A response-inhibition deficit following frontal injury shows that the damaged system contributed to the assessed behavior; it does not establish that the region independently performs an isolated mental function. Converging results from lesion studies, stimulation, electrophysiology, and models can progressively constrain the causal explanation, especially when tasks separately measure initiation, selection, and cancellation Swick et al. (2008) Aron et al. (2014).
Neuroimaging studies offer spatial coverage and sensitivity to task-related changes, but the hemodynamic response is temporally indirect and does not uniquely identify neural inhibition. A subtraction between stop and go trials also includes stimulus frequency, task relevance, expectancy, and response omission. Carefully designed control conditions can separate some of these components. The Hampshire and colleagues study illustrates the benefit of matching perceptual cues while altering task demands, whereas intervention studies underscore the involvement of distributed circuits. The strongest inference arises when convergent results are tied to an independently specified computational mechanism Hampshire et al. (2010) van den Wildenberg et al. (2006).
Clinical neural evidence requires an additional distinction between mechanistic research and treatment advice. Subthalamic stimulation and disorders of movement provide valuable opportunities to examine action regulation, but their effects cannot be generalized automatically to a healthy population. Symptom severity, medication, disease stage, and neurosurgical targeting shape the observed behavior. An evidence-based monograph can use these studies to test circuit hypotheses while avoiding recommendations concerning intervention. Its central claims remain about experimentally defined psychological processes and how causal manipulations constrain their models van den Wildenberg et al. (2006) Eagle et al. (2008).
Cross-Species Evidence and Translational Boundaries
Comparative research studies action inhibition in species where response timing, reward contingencies, and neural mechanisms can be manipulated with unusual experimental control. Eagle, Bari, and Robbins reviewed the cross-species use of stop-signal and go/no-go tasks, including differences between preventing a planned response and canceling an already initiated one. Animal studies can support candidate causal mechanisms when experimental structures are homologous in their measurable demands. However, the behavioral task may also differ in training intensity, reward schedules, stimulus modality, and how the animal learns to interpret stop cues Eagle et al. (2008).
Translational validity requires more than using the same task name. Researchers should specify what response is dominant, when the stop signal appears, how stopping probability is measured, and whether the same mathematical assumptions apply in each species. Latencies are affected by the sensory and motor systems under investigation and by differences in preparatory state. A circuit manipulation producing better no-go performance does not automatically establish improved cancellation speed or a generalized ability to resist impulses. Strong comparative work seeks overlapping computational signatures while preserving the causal role of species-specific learning and movement Eagle et al. (2008) Logan et al. (1984).
Cross-species comparison is most informative when it combines several forms of evidence: behavioral dissociations, interventions, electrophysiological timing, and simulations. These can reveal whether a proposed gate or threshold operates in a functionally comparable way across preparations. At the same time, the human concepts of intention, self-control, and deliberate choice have broader social and linguistic meanings that should not be mapped directly from an animal response omission. A careful translational account identifies the experimentally shared operation and leaves wider interpretations open to appropriately designed evidence Wiecki & Frank (2013) Nigg (2000).
Development, Aging, and Variation in Action Regulation
Developmental research treats action regulation as an evolving coordination among rule representation, attention, motor preparation, motivation, and learned behavioral expectations. Diamond's synthesis of executive functions distinguishes response inhibition from interference control and describes their relation to working memory and cognitive flexibility. These capacities support successful classroom participation, everyday planning, and adaptation to novel situations, but the observed performance of a child depends on task instructions, motor skill, processing speed, and comprehension. An age difference in a go/no-go error rate therefore cannot be interpreted as a direct neural estimate of inhibitory strength without corresponding controls Diamond (2013) Nigg (2000).
Lifespan comparisons are especially sensitive to shifts in speed-accuracy priorities. Two age groups may show comparable commission-error rates because one group deliberately responds more slowly, or their mean reaction times may differ because of sensory and motor slowing unrelated to cue detection. Repeated-measures designs, reliability analysis, distributional modeling, and age-appropriate instructions help identify which component changes. Latent executive-function measures reduce some task-specific noise, yet measurement invariance must be examined before interpreting group differences as evidence of the same construct on a common scale Miyake et al. (2000) Ratcliff & McKoon (2008).
Development also changes what counts as habitual. Extensive practice can make an initially difficult response almost automatic, altering the amount of proactive preparation necessary for correct behavior. Repeated stopping cues can acquire learned associations that differ across environments and developmental histories. Consequently, developmental changes may involve greater rule maintenance, different strategic priorities, and increasing experience with task categories alongside neural maturation. Longitudinal and cross-context studies are needed to separate these contributions and to determine whether improvement on a familiar task represents a transferable capacity or task-specific adaptation Verbruggen & Logan (2008a) Braver (2012).
Clinical Relevance and Construct Validity
Failure to withhold or cancel behavior is investigated in clinical and neurodevelopmental research because it may contribute to impulsive action, compulsive repetition, and difficulties regulating behavior under changing goals. Nigg's taxonomy emphasizes that the word disinhibition has been used for different psychological functions and should be operationalized before group comparisons. The same clinical symptom can arise from attention to cues, motivational valuation, learned habits, difficulty representing rules, or motor-control disturbances. A single task score is therefore insufficient for diagnosis and should never be treated as a complete proxy for a disorder Nigg (2000) Bari & Robbins (2013).
Clinical studies of stopping depend on experimental design and measurement quality. Participants may differ in go-response speed, comprehension, omission errors, probability of detecting the stop cue, and strategic waiting; each can change the estimated stop-signal reaction time. The consensus guidelines recommend reporting and checking these quantities, emphasizing that model assumptions need evaluation before group-level inference. Clinical generalization further requires evidence about the population sampled, medication and symptom context, and whether the observed mechanism relates to meaningful functional outcomes. Cross-sectional correlations do not alone establish that an inhibitory mechanism causes symptoms Verbruggen et al. (2019) Matzke et al. (2017).
Neurosurgical and pharmacological findings offer mechanistic clues, especially when interventions influence fronto-basal-ganglia processes. Their interpretation must separate the change caused by a particular intervention from the background neural and behavioral effects of illness. Eagle and colleagues' comparative review and the basal-ganglia stimulation literature show how different stages of action control may be differentially sensitive to experimental manipulations. Such evidence can guide further causal tests, yet a scientific synthesis of inhibition should remain independent of individual prescribing decisions or unsupported promises of treatment efficacy Eagle et al. (2008) van den Wildenberg et al. (2006).
Measurement Reliability, Replication, and Reporting
A scientific account of selection and stopping depends on the validity of the behavioral measures used to test it. Commission errors, response time, stop-signal reaction time, post-stop slowing, and conflict costs each combine multiple influences. Researchers should report raw trial counts, exclusions, response-time distributions, omission and commission frequencies, stimulus probabilities, delay schedules, and all procedures used to estimate latent parameters. Verbruggen and colleagues' consensus guide supplies a comprehensive set of design and analysis recommendations for the stop-signal paradigm, including diagnostic checks for violations of race-model assumptions Verbruggen et al. (2019).
The reliability of a group-level experimental manipulation must be distinguished from the reliability of individual rankings. A stable average conflict cost can coexist with noisy difference scores because subtracting conditions removes shared stable variance and leaves measurement error. Latent-variable designs partly address this by combining several indicators, but they demand sufficiently informative tasks and justified factor structures. Studies seeking correlations with everyday self-control need test-retest reliability, uncertainty intervals, and independent replication before claiming a general individual capacity. Otherwise the interpretive distance between task performance and a broad psychological trait remains substantial Friedman & Miyake (2004) Miyake et al. (2000).
Replication and model comparison should consider both the qualitative effect and its quantitative signature. An inhibition function may replicate in direction while its precise parameters vary with signal modality, participant strategy, and sampling. Computational claims should be checked through simulations, parameter recovery, and posterior or predictive diagnostics wherever applicable. The researcher should report alternative fits and the conditions that make models distinguishable. A successful description of the available data is a starting point for scientific inference, while a discriminating experimental prediction provides stronger grounds for identifying the proposed mechanism Forstmann et al. (2016) Wiecki & Frank (2013).
Ecological Control, Multitasking, and Human–Artificial Interaction
Real-world action selection is embedded in environments containing delayed feedback, multiple goals, social coordination, and changing opportunity costs. A person deciding whether to interrupt a current activity may rely on learned cues, explicit goals, and prospective estimates of risk, often over seconds or minutes rather than the milliseconds of the stop-signal task. Laboratory paradigms offer powerful mechanistic isolation, but ecological transfer requires testing that their variables predict choices in more complex environments. The practical meaning of response regulation depends on the interaction between task demands and the policy by which an agent allocates attention and motor resources Braver (2012) Bissett & Logan (2011).
In human–artificial systems, action selection can be distributed across a person and a software controller. A system may recommend an action, rank alternatives, require authorization before execution, or interrupt an automated process when new evidence arrives. The psychological questions concern when human users anticipate interruption, how notification timing changes their readiness, and whether interface cues create learned response biases. Behavioral theories of stopping provide candidate experimental tools for these questions, provided the studies explicitly distinguish human motor inhibition from software gating and from the social meaning of delegating a decision Logan et al. (1984) Ratcliff & McKoon (2008).
Artificial decision policies can be studied through computational objectives, action-selection thresholds, stop conditions, and uncertainty-triggered revisions. These constitute operational properties of a system's behavior rather than evidence that a system possesses the same neural inhibitory mechanisms as Homo. The Artificial Era provides a new research setting for comparative study of decision and control architectures. A rigorous program specifies the system boundary, observable actions, intervention point, task objective, and validation dataset, and then tests the adequacy of the explanatory comparison. Its conclusions follow from measured function and documented architecture Wiecki & Frank (2013) Munakata et al. (2011).
Unresolved Problems and Research Priorities
A central unresolved problem is the relationship between response suppression and goal-driven competition. Some accounts posit an active suppressive operation, while others emphasize maintenance of task goals that biases processing toward relevant alternatives and thereby produces inhibitory effects. Munakata and colleagues proposed a unified approach in which active goal representations can shape multiple kinds of inhibitory behavior. The empirical challenge is to design experiments in which the accounts disagree about temporal dynamics, neural effects, or transfer to novel stimuli. Clear disagreement is more informative than a general appeal to inhibitory control Munakata et al. (2011) Wiecki & Frank (2013).
Another open issue is whether neural motor suppression recruited by surprising events is the same mechanism used in instructed stopping. A broad circuit account predicts overlapping physiological signatures, but surprise and stop cues can also differ in attention, task relevance, and response specificity. Experiments need matched rare-event controls, measures of motor excitability, and selective-stop conditions that reveal the spatial breadth and latency of suppression. Competing theories should specify when stopping is global, when it is targeted, and how action is subsequently released or reselected Wessel & Aron (2017) Hampshire et al. (2010).
Future work also needs better estimates of trigger failures, stable individual differences, and the relation between a task's latent parameter and ecologically important outcomes. Adaptive stop-delay designs, hierarchical estimation, preregistered quality criteria, and large multi-session datasets can improve reliability but cannot remove construct ambiguity on their own. Clinical, developmental, and artificial-system applications require independent validation within their own contexts. The most promising research program links experimental perturbation to distributional models and temporally precise physiology while testing whether the proposed mechanism generalizes across different action environments Matzke et al. (2017) Verbruggen et al. (2019).
Experimental Design, Data Interpretation, and Responsible Synthesis
An experimental program should begin by declaring the exact action to be selected or prevented and the task circumstances that make it likely. For a go/no-go experiment, this includes the go probability, stimulus categories, timing, and training required to establish prepotency. For a stop-signal experiment, it includes the stop-signal modality, trial frequency, stop-signal delay adjustment, instructions emphasizing the primary task, and diagnostic checks for go omissions and failed-stop speed. For an interference paradigm, it includes the compatibility manipulation and the information that is explicitly relevant to the instructed choice. This specification makes the inference about behavior reproducible and constrains the meaning of response control Verbruggen et al. (2019) Stroop (1935).
Analysis should preserve trial-level speed and accuracy, report response distributions and uncertainty, and consider whether participants adopted different control policies across experimental conditions. An adaptive stopping staircase yields useful observations only when the design produces informative responses and participants remain engaged with the go task. Statistical models should distinguish estimated latent parameters from directly observed variables. The use of sequential-sampling methods can illuminate decision caution and evidence quality, while Bayesian trigger-failure approaches can illuminate failure to engage stopping. Each model requires identifiable parameters, transparent assumptions, and predictive checks; agreement with one aggregate mean cannot establish the proposed psychological mechanism Matzke et al. (2017) Ratcliff & McKoon (2008).
Scientific synthesis should evaluate convergent evidence in the order of experimental specificity. A controlled behavioral manipulation demonstrates sensitivity of performance to a task factor. An intervention or lesion can support a causal role for affected circuitry under the relevant conditions. A formal model explains a distribution only to the degree that its assumptions survive diagnostic tests and alternatives have been compared. Longitudinal or cross-task generalization requires independent datasets and reliability evidence. By retaining this hierarchy, research on action selection and inhibition becomes a precise comparative science of how competing possibilities produce regulated behavior, with applications evaluated according to the evidence available in each domain Wiecki & Frank (2013) Friedman & Miyake (2004).
Conclusion
Action selection and inhibition form an integrated field of psychological research into how behavior is organized under competition, changing goals, and the possibility of interruption. The historical development from interference paradigms to race models, latent-variable analyses, and fronto-basal-ganglia circuitry has provided complementary methods for probing the same broad problem. The evidence supports a distinction among competing response activation, early withholding, late cancellation, prospective caution, learned stimulus-stop associations, and post-event adjustments. These operations can interact, yet their operational definitions and experimental signatures remain essential for scientific explanation Stroop (1935) Logan et al. (1984) Friedman & Miyake (2004).
The stop-signal paradigm has particular explanatory strength because the race model connects observable performance to an explicit estimate of hidden stopping time. That estimate depends on the correctness of its assumptions and the reliable initiation of the stop process. Interference and go/no-go tasks capture other aspects of response conflict and restraint, while computational and neural measurements add constraints on decision thresholds, cue processing, and the timing of motor suppression. Together, these methods show why the label inhibition must be attached to a specified task, measurement, and causal hypothesis rather than used as an undifferentiated explanation Verbruggen et al. (2019) Ratcliff & McKoon (2008).
The remaining scientific agenda is to identify when apparently similar forms of control rely on shared mechanisms, determine which neural operations are causally necessary, improve the reliability of individual-difference measurement, and test transfer beyond laboratory tasks. Advances will depend on combining original experiments, intervention evidence, computational identifiability, and independent replication. Research in the Artificial Era extends these questions to decision systems and human–artificial partnerships through explicit functional comparisons. The resulting understanding of action regulation is a cumulative scientific synthesis whose authority rests on observed behavior and testable explanatory relations Munakata et al. (2011) Forstmann et al. (2016) Wessel & Aron (2017).
References
- Aron, A. R., Robbins, T. W., & Poldrack, R. A. (2014). Inhibition and the right inferior frontal cortex: One decade on. Trends in Cognitive Sciences, 18(4), 177–185. https://doi.org/10.1016/j.tics.2013.12.003
- Bari, A., & Robbins, T. W. (2013). Inhibition and impulsivity: Behavioral and neural basis of response control. Progress in Neurobiology, 108, 44–79. https://doi.org/10.1016/j.pneurobio.2013.06.005
- Bissett, P. G., & Logan, G. D. (2011). Balancing cognitive demands: Control adjustments in the stop-signal paradigm. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(2), 392–404. https://doi.org/10.1037/a0021800
- Botvinick, M. M., Braver, T. S., Barch, D. M., Carter, C. S., & Cohen, J. D. (2001). Conflict monitoring and cognitive control. Psychological Review, 108(3), 624–652. https://doi.org/10.1037/0033-295X.108.3.624
- Braver, T. S. (2012). The variable nature of cognitive control: A dual mechanisms framework. Trends in Cognitive Sciences, 16(2), 106–113. https://doi.org/10.1016/j.tics.2011.12.010
- Diamond, A. (2013). Executive functions. Annual Review of Psychology, 64, 135–168. https://doi.org/10.1146/annurev-psych-113011-143750
- Eagle, D. M., Bari, A., & Robbins, T. W. (2008). The neuropsychopharmacology of action inhibition: Cross-species translation of the stop-signal and go/no-go tasks. Psychopharmacology, 199(3), 439–456. https://doi.org/10.1007/s00213-008-1127-6
- Forstmann, B. U., Ratcliff, R., & Wagenmakers, E.-J. (2016). Sequential sampling models in cognitive neuroscience: Advantages, applications, and extensions. Annual Review of Psychology, 67, 641–666. https://doi.org/10.1146/annurev-psych-122414-033645
- Friedman, N. P., & Miyake, A. (2004). The relations among inhibition and interference control functions: A latent-variable analysis. Journal of Experimental Psychology: General, 133(1), 101–135. https://doi.org/10.1037/0096-3445.133.1.101
- Hampshire, A., Chamberlain, S. R., Monti, M. M., Duncan, J., & Owen, A. M. (2010). The role of the right inferior frontal gyrus: Inhibition and attentional control. NeuroImage, 50(3), 1313–1319. https://doi.org/10.1016/j.neuroimage.2009.12.109
- Jahfari, S., Stinear, C. M., Claffey, M., Verbruggen, F., & Aron, A. R. (2010). Responding with restraint: What are the neurocognitive mechanisms? Journal of Cognitive Neuroscience, 22(7), 1479–1492. https://doi.org/10.1162/jocn.2009.21307
- Logan, G. D., Cowan, W. B., & Davis, K. A. (1984). On the ability to inhibit simple and choice reaction time responses: A model and a method. Journal of Experimental Psychology: Human Perception and Performance, 10(2), 276–291. https://doi.org/10.1037/0096-1523.10.2.276
- Matzke, D., Love, J., & Heathcote, A. (2017). A Bayesian approach for estimating the probability of trigger failures in the stop-signal paradigm. Behavior Research Methods, 49, 267–281. https://doi.org/10.3758/s13428-015-0695-8
- Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D. (2000). The unity and diversity of executive functions and their contributions to complex frontal lobe tasks: A latent variable analysis. Cognitive Psychology, 41(1), 49–100. https://doi.org/10.1006/cogp.1999.0734
- Mostofsky, S. H., & Simmonds, D. J. (2008). Response inhibition and response selection: Two sides of the same coin. Journal of Cognitive Neuroscience, 20(5), 751–761. https://doi.org/10.1162/jocn.2008.20500
- Munakata, Y., Herd, S. A., Chatham, C. H., Depue, B. E., Banich, M. T., & O'Reilly, R. C. (2011). A unified framework for inhibitory control. Trends in Cognitive Sciences, 15(10), 453–459. https://doi.org/10.1016/j.tics.2011.07.011
- Nigg, J. T. (2000). On inhibition/disinhibition in developmental psychopathology: Views from cognitive and personality psychology and a working inhibition taxonomy. Psychological Bulletin, 126(2), 220–246. https://doi.org/10.1037/0033-2909.126.2.220
- Ratcliff, R., & McKoon, G. (2008). The diffusion decision model: Theory and data for two-choice decision tasks. Neural Computation, 20(4), 873–922. https://doi.org/10.1162/neco.2008.12-06-420
- Ridderinkhof, K. R., van den Wildenberg, W. P. M., Segalowitz, S. J., & Carter, C. S. (2004). Neurocognitive mechanisms of cognitive control: The role of prefrontal cortex in action selection, response inhibition, performance monitoring, and reward-based learning. Brain and Cognition, 56(2), 129–140. https://doi.org/10.1016/j.bandc.2004.09.016
- Simon, J. R. (1969). Reactions toward the source of stimulation. Journal of Experimental Psychology, 81(1), 174–176. https://doi.org/10.1037/h0027448
- Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18(6), 643–662. https://doi.org/10.1037/h0054651
- Swick, D., Ashley, V., & Turken, A. U. (2008). Left inferior frontal gyrus is critical for response inhibition. BMC Neuroscience, 9, Article 102. https://doi.org/10.1186/1471-2202-9-102
- van den Wildenberg, W. P. M., van Boxtel, G. J. M., van der Molen, M. W., Bosch, D. A., Speelman, J. D., & Brunia, C. H. M. (2006). Stimulation of the subthalamic region facilitates the selection and inhibition of motor responses in Parkinson's disease. Journal of Cognitive Neuroscience, 18(4), 626–636. https://doi.org/10.1162/jocn.2006.18.4.626
- Verbruggen, F., & Logan, G. D. (2008). Automatic and controlled response inhibition: Associative learning in the go/no-go and stop-signal paradigms. Journal of Experimental Psychology: General, 137(4), 649–672. https://doi.org/10.1037/a0013170
- Verbruggen, F., & Logan, G. D. (2008). Long-term aftereffects of response inhibition: Memory retrieval, task goals, and cognitive control. Journal of Experimental Psychology: Human Perception and Performance, 34(5), 1229–1235. https://doi.org/10.1037/0096-1523.34.5.1229
- Verbruggen, F., & Logan, G. D. (2008). Response inhibition in the stop-signal paradigm. Trends in Cognitive Sciences, 12(11), 418–424. https://doi.org/10.1016/j.tics.2008.07.005
- Verbruggen, F., Aron, A. R., Band, G. P. H., Beste, C., Bissett, P. G., Brockett, A. T., Brown, J. W., Chamberlain, S. R., Chambers, C. D., Colonius, H., Colzato, L. S., Corneil, B. D., Coxon, J. P., Dupuis, A., Eagle, D. M., Garavan, H., Greenhouse, I., Heathcote, A., Huster, R. J., ... Boehler, C. N. (2019). A consensus guide to capturing the ability to inhibit actions and impulsive behaviors in the stop-signal task. eLife, 8, e46323. https://doi.org/10.7554/eLife.46323
- Wessel, J. R., & Aron, A. R. (2017). On the globality of motor suppression: Unexpected events and their influence on behavior and cognition. Neuron, 93(2), 259–280. https://doi.org/10.1016/j.neuron.2016.12.013
- Wiecki, T. V., & Frank, M. J. (2013). A computational model of inhibitory control in frontal cortex and basal ganglia. Psychological Review, 120(2), 329–355. https://doi.org/10.1037/a0031542