Neural and behavioral traces of error awareness

Beyond peaks and troughs: Multiplexed performance monitoring signals in the EEG

With the discovery of event-related potentials elicited by errors more than 30 years ago, a new avenue of research on performance monitoring, cognitive control, and decision making emerged. Since then, the field has developed and expanded fulminantly. After a brief overview on the EEG correlates of performance monitoring, this article reviews recent advancements based on single-trial analyses using independent component analysis, multiple regression, and multivariate pattern classification. Given the close interconnection between performance monitoring and reinforcement learning, computational modeling and model-based EEG analyses have made a particularly strong impact. The reviewed findings demonstrate that error- and feedback-related EEG dynamics represent variables reflecting how performance-monitoring signals are weighted and transformed into an adaptation signal that guides future decisions and actions. The model-based single-trial analysis approach goes far beyond conventional peak-and-trough analyses of event-related potentials and enables testing mechanistic theories of performance monitoring, cognitive control, and decision making.

The neuroscience of active learning and direct instruction

Throughout the educational system, students experiencing active learning pedagogy perform better and fail less than those taught through direct instruction. Can this be ascribed to differences in learning from a neuroscientific perspective? This review examines mechanistic, neuroscientific evidence that might explain differences in cognitive engagement contributing to learning outcomes between these instructional approaches. In classrooms, direct instruction comprehensively describes academic content, while active learning provides structured opportunities for learners to explore, apply, and manipulate content. Synaptic plasticity and its modulation by arousal or novelty are central to all learning and both approaches. As a form of social learning, direct instruction relies upon working memory. The reinforcement learning circuit, associated agency, curiosity, and peer-to-peer social interactions combine to enhance motivation, improve retention, and build higher-order-thinking skills in active learning environments. When working memory becomes overwhelmed, additionally engaging the reinforcement learning circuit improves retention, providing an explanation for the benefits of active learning. This analysis provides a mechanistic examination of how emerging neuroscience principles might inform pedagogical choices at all educational levels.

The Development of Cognitive Control in Preschoolers and Kindergarteners: The Case of Post-Error Slowing and Delayed Disinhibition

This study aimed to investigate two specific behavioral manifestations of the executive attention systems in preschoolers and kindergarteners, beyond the unique contribution of intelligence. We tested post-error slowing [RT¯Post-error trial-RT¯Not post-error trial] as a marker of reactive control and delayed disinhibition as a novel marker for proactive control. One hundred and eighty preschool- and kindergarten-aged children, as well as their mothers (final sample: 155 children and 174 mothers), performed an adapted task based on Go/NoGo and Stroop-like paradigms-the emotional day-night task. The children showed reliable post-error slowing and delayed disinhibition (mean size effects of 238.18 ms and 58.31 ms, respectively), while the adult size effects were 40-50% smaller. The post-error slowing effect was present for both sexes in all the tested ages, while the delayed disinhibition effect was present only for girls. Both effects showed large individual differences that became smaller in adulthood. Our findings emphasize the earlier maturation of reactive control compared to proactive control, and the earlier maturation of proactive cognitive control in girls compared to boys.

Transdiagnostic inflexible learning dynamics explain deficits in depression and schizophrenia

Deficits in reward learning are core symptoms across many mental disorders. Recent work suggests that such learning impairments arise by a diminished ability to use reward history to guide behaviour, but the neuro-computational mechanisms through which these impairments emerge remain unclear. Moreover, limited work has taken a transdiagnostic approach to investigate whether the psychological and neural mechanisms that give rise to learning deficits are shared across forms of psychopathology. To provide insight into this issue, we explored probabilistic reward learning in patients diagnosed with major depressive disorder (n = 33) or schizophrenia (n = 24) and 33 matched healthy controls by combining computational modelling and single-trial EEG regression. In our task, participants had to integrate the reward history of a stimulus to decide whether it is worthwhile to gamble on it. Adaptive learning in this task is achieved through dynamic learning rates that are maximal on the first encounters with a given stimulus and decay with increasing stimulus repetitions. Hence, over the course of learning, choice preferences would ideally stabilize and be less susceptible to misleading information. We show evidence of reduced learning dynamics, whereby both patient groups demonstrated hypersensitive learning (i.e. less decaying learning rates), rendering their choices more susceptible to misleading feedback. Moreover, there was a schizophrenia-specific approach bias and a depression-specific heightened sensitivity to disconfirmational feedback (factual losses and counterfactual wins). The inflexible learning in both patient groups was accompanied by altered neural processing, including no tracking of expected values in either patient group. Taken together, our results thus provide evidence that reduced trial-by-trial learning dynamics reflect a convergent deficit across depression and schizophrenia. Moreover, we identified disorder distinct learning deficits.

Success versus failure in cognitive control: Meta-analytic evidence from neuroimaging studies on error processing

Brain mechanisms of error processing have often been investigated using response interference tasks and focusing on the posterior medial frontal cortex, which is also implicated in resolving response conflict in general. Thereby, the role other brain regions may play has remained undervalued. Here, activation likelihood estimation meta-analyses were used to synthesize the neuroimaging literature on brain activity related to committing errors versus responding successfully in interference tasks and to test for commonalities and differences. The salience network and the temporoparietal junction were commonly recruited irrespective of whether responses were correct or incorrect, pointing towards a general involvement in coping with situations that call for increased cognitive control. The dorsal posterior cingulate cortex, posterior thalamus, and left superior frontal gyrus showed error-specific convergence, which underscores their consistent involvement when performance goals are not met. In contrast, successful responding revealed stronger convergence in the dorsal attention network and lateral prefrontal regions. Underrecruiting these regions in error trials may reflect failures in activating the task-appropriate stimulus-response contingencies necessary for successful response execution.

Can a failure in the error-monitoring system explain unawareness of memory deficits in Alzheimer's disease?

Unawareness of memory deficits is an early manifestation in patients with Alzheimer's disease (AD), which often delays diagnosis. This intriguing behavior constitutes a form of anosognosia, whose neural mechanisms remain largely unknown. We hypothesized that anosognosia may depend on a critical synaptic failure in the error-monitoring system, which would prevent AD patients from being aware of their own memory impairment. To investigate, we measured event-related potentials (ERPs) evoked by erroneous responses during a word memory recognition task in two groups of amyloid positive individuals with only subjective memory complaints at study entry: those who progressed to AD within the five-year study period (PROG group), and those who remained cognitively normal (CTRL group). A significant reduction in the amplitude of the positivity error (Pe), an ERP related to error awareness, was observed in the PROG group at the time of AD diagnosis (vs study entry) in intra-group analysis, as well as when compared with the CTRL group in inter-group analysis, based on the last EEG acquisition for all subjects. Importantly, at the time of AD diagnosis, the PROG group exhibited clinical signs of anosognosia, overestimating their cognitive abilities, as evidenced by the discrepancy scores obtained from caregiver/informant vs participant reports on the cognitive subscale of the Healthy Aging Brain Care Monitor. To our knowledge, this is the first study to reveal the emergence of a failure in the error-monitoring system during a word memory recognition task at the early stages of AD. This finding, along with the decline of awareness for cognitive impairment observed in the PROG group, strongly suggests that a synaptic dysfunction in the error-monitoring system may be the critical neural mechanism at the origin of unawareness of deficits in AD.

Success versus failure in cognitive control: meta-analytic evidence from neuroimaging studies on error processing

Brain mechanisms of error processing have often been investigated using response interference tasks and focusing on the posterior medial frontal cortex, which is also implicated in resolving response conflict in general. Thereby, the role other brain regions may play has remained undervalued. Here, activation likelihood estimation meta-analyses were used to synthesize the neuroimaging literature on brain activity related to committing errors versus responding successfully in interference tasks and to test for commonalities and differences. The salience network and the temporoparietal junction were commonly recruited irrespective of whether responses were correct or incorrect, pointing towards a general involvement in coping with situations that call for increased cognitive control. The dorsal posterior cingulate cortex, posterior thalamus, and left superior frontal gyrus showed error-specific convergence, which underscores their consistent involvement when performance goals are not met. In contrast, successful responding revealed stronger convergence in the dorsal attention network and lateral prefrontal regions. Underrecruiting these regions in error trials may reflect failures in activating the task-appropriate stimulus-response contingencies necessary for successful response execution.

Updating the relationship of the Ne/ERN to task-related behavior: A brief review and suggestions for future research

The error negativity/error-related negativity (Ne/ERN) is one of the most well-studied event-related potential (ERP) components in the electroencephalography (EEG) literature. Peaking about 50 ms after the commission of an error, the Ne/ERN is a negative deflection in the ERP waveform that is thought to reflect error processing in the brain. While its relationships to trait constructs such as anxiety are well-documented, there is still little known about how the Ne/ERN may subsequently influence task-related behavior. In other words, does the occurrence of the Ne/ERN trigger any sort of error corrective process, or any other behavioral adaptation to avoid errors? Several theories have emerged to explain how the Ne/ERN may implement or affect behavior on a task, but evidence supporting each has been mixed. In the following manuscript, we review these theories, and then systematically discuss the reasons that there may be discrepancies in the literature. We review both the inherent biological factors of the neural regions that underlie error-processing in the brain, and some of the researcher-induced factors in analytic and experimental choices that may be exacerbating these discrepancies. We end with a table of recommendations for future researchers who aim to understand the relationship between the Ne/ERN and behavior.

Learning from errors: Distinct neural networks for monitoring errors and maintaining corrects through repeated practice and feedback

How memory representations are eventually established and maintained in the brain is one of central issues in memory research. Although the hippocampus and various brain regions have been shown to be involved in learning and memory, how they coordinate to support successful memory through errors is unclear. In this study, a retrieval practice (RP) - feedback (FB) paradigm was adopted to address this issue. Fifty-six participants (27 in the behavioral group, and 29 in the fMRI group) learned 120 Swahili-Chinese words associations and underwent two RP-answer FB cycles (i.e., RP1, FB1, RP2, FB2). The responses of the fMRI group were recorded in the fMRI scanner. The trials were divided based on participant's performance (correct or incorrect, C or I) during the two RPs and the final test (i.e., trial type, CCC, ICC, IIC III). The results showed that the regions in the salience and executive control networks (S-ECN) during RP, but not during FB, was strongly predictive of final successful memory. Their activation was just before the errors were corrected (i.e., RP1 in ICC trials and RP2 in IIC trials). The anterior insula (AI) is a core region in monitoring repeated errors, and it had differential connectivity with the default mode network (DMN) regions and the hippocampus during the RP and FB phases to inhibit incorrect answers and update memory. In contrast, maintaining corrected memory representation requires repeated RP and FB, which was associated with the DMN activation. Our study clarified how different brain regions support error monitoring and memory maintenance through repeated RP and FB, and emphasized the role of the insula in learning from errors.

Post-error behavioral adjustments under reactive control among older adults

This study analyzed the effects of aging on post-error behavioral adjustments from the perspective of cognitive control. A modified error awareness task was administered to young (n = 50) and older (n = 50) adults. In this task, two buttons were placed on the left and right sides in front of the participants, who were instructed to use the right button to perform a go/no-go task, and were notified if they made an error. There were three experimental conditions (A, B, and C): participants had to push the right button once in Condition A and twice in Condition B and C when a go-stimulus was presented. Conversely, participants were asked to withhold their response when a no-go stimulus was presented. Response inhibition differed depending on the experimental condition. The participants were asked to push the left button as quickly as possible when an error occurred. The results showed relatively longer reaction times to sudden errors among older adults compared with young adults. Furthermore, the difference in the error responses (i.e., accidentally pushing the right button once or twice when a no-go stimulus was presented) strongly influenced older adults' response time after an error. These results suggest that the shift from proactive to reactive control may significantly influence post-error behavioral adjustments in older adults.

Are items actively removed from working memory during free time in children with developmental language disorder?

BACKGROUND

Children with developmental language disorder (DLD) have deficits in verbal and non-verbal processing relative to typically developing (TD) peers, potentially reflecting difficulties in working memory, processing speed and inhibition of interference. We examined working memory in children with DLD using the serial-order-in-a-box-complex span (SOB-CS) interference-based model, which posits a time-based mechanism, free time, that governs how interference affects processing performance.

AIMS

(1) To determine the degree to which children with DLD and TD children differ in the amount of free time available during working memory tasks, and whether potential group differences in free time differ depending on the domain of task demands? (2) To determine the relationship between free time and interference effects on working memory accuracy in children with DLD relative to TD peers.

METHODS & PROCEDURES

We examined the relationship between free time and working memory in children aged 9-13 years with DLD relative to age-matched TD peers. Working memory tasks involved five conditions that varied verbal versus non-verbal task demands in an interference processing phase relative to a recall test phase. Free time was the time between response on the interference processing task and onset of the recall test phase.

OUTCOMES & RESULTS

DLD and TD groups did not differ in total free time in any condition. Results indicated group differences in the relationship between free time and accuracy in the conditions involving verbal recall, but not non-verbal recall. In the verbal-only condition, relatively more free time was associated with worse accuracy for the DLD group, but with better accuracy for the TD group. In the condition with verbal recall paired with non-verbal interference processing, relatively more free time was associated with better accuracy for the DLD group, but not for the TD group.

CONCLUSIONS & IMPLICATIONS

The overall findings suggest that free time between cognitive operations is positively associated with working memory for both verbal and non-verbal recall, except in the presence of high verbal interference for the DLD group (i.e., verbal interference paired with verbal recall). This finding may reflect poor encoding and attention under particularly challenging verbal processing demands for the DLD group. This study also demonstrates the importance of considering the interrelationships between processing speed and interference in working memory performance.

WHAT THIS PAPER ADDS

What is already known on the subject DLD is characterized by core deficits in verbal processing, but also deficits in non-verbal processing. Processing-based hypotheses of DLD-limited verbal working memory, slowed processing speed and inefficient inhibition-do not fully account for behavioural profiles in DLD when considered separately, yet there is evidence suggesting interrelationships among these factors. What this paper adds to existing knowledge The current study tests the key mechanism posited by a theoretical framework that has the potential to integrate these processing-based hypotheses. Our findings indicate that the effect of this mechanism differed in DLD relative to TD peers in the presence of high verbal task demands. Our findings also demonstrate the importance of considering the interrelationships among cognitive processes in children with DLD. What are the potential or actual clinical implications of this work? In practice, results from the current study suggest that children with DLD may benefit from supplementing verbal information with non-verbal information and from pauses between successive presentations of verbal information. These strategies may support their ability to maintain and act on information during verbal processing.

Anxiety disrupts performance monitoring: integrating behavioral, event-related potential, EEG microstate, and sLORETA evidence

Anxiety impacts performance monitoring, though theory and past research are split on how and for whom. However, past research has often examined either trait anxiety in isolation or task-dependent state anxiety and has indexed event-related potential components, such as the error-related negativity or post-error positivity (Pe), calculated at a single node during a limited window of time. We introduced 2 key novelties to this electroencephalography research to examine the link between anxiety and performance monitoring: (i) we manipulated antecedent, task-independent, state anxiety to better establish the causal effect; (ii) we conducted moderation analyses to determine how state and trait anxiety interact to impact performance monitoring processes. Additionally, we extended upon previous work by using a microstate analysis approach to isolate and sequence the neural networks and rapid mental processes in response to error commission. Results showed that state anxiety disrupts response accuracy in the Stroop task and error-related neural processes, primarily during a Pe-related microstate. Source localization shows that this disruption involves reduced activation in the dorsal anterior cingulate cortex and compensatory activation in the right lateral prefrontal cortex, particularly among people high in trait anxiety. We conclude that antecedent anxiety is largely disruptive to performance monitoring.

Neuroscience from the comfort of your home: Repeated, self-administered wireless dry EEG measures brain function with high fidelity

Recent advances have enabled the creation of wireless, “dry” electroencephalography (EEG) recording systems, and easy-to-use engaging tasks, that can be operated repeatedly by naïve users, unsupervised in the home. Here, we evaluated the validity of dry-EEG, cognitive task gamification, and unsupervised home-based recordings used in combination. Two separate cohorts of participants—older and younger adults—collected data at home over several weeks using a wireless dry EEG system interfaced with a tablet for task presentation. Older adults (n = 50; 25 females; mean age = 67.8 years) collected data over a 6-week period. Younger male adults (n = 30; mean age = 25.6 years) collected data over a 4-week period. All participants were asked to complete gamified versions of a visual Oddball task and Flanker task 5–7 days per week. Usability of the EEG system was evaluated via participant adherence, percentage of sessions successfully completed, and quantitative feedback using the System Usability Scale. In total, 1,449 EEG sessions from older adults (mean = 28.9; SD = 6.64) and 684 sessions from younger adults (mean = 22.87; SD = 1.92) were collected. Older adults successfully completed 93% of sessions requested and reported a mean usability score of 84.5. Younger adults successfully completed 96% of sessions and reported a mean usability score of 88.3. Characteristic event-related potential (ERP) components—the P300 and error-related negativity—were observed in the Oddball and Flanker tasks, respectively. Using a conservative threshold for inclusion of artifact-free data, 50% of trials were rejected per at-home session. Aggregation of ERPs across sessions (2–4, depending on task) resulted in grand average signal quality with similar Standard Measurement Error values to those of single-session wet EEG data collected by experts in a laboratory setting from a young adult sample. Our results indicate that easy-to-use task-driven EEG can enable large-scale investigations in cognitive neuroscience. In future, this approach may be useful in clinical applications such as screening and tracking of treatment response.

Post-error Slowing Reflects the Joint Impact of Adaptive and Maladaptive Processes During Decision Making

Errors and their consequences are typically studied by investigating changes in decision speed and accuracy in trials that follow an error, commonly referred to as “post-error adjustments”. Many studies have reported that subjects slow down following an error, a phenomenon called “post-error slowing” (PES). However, the functional significance of PES is still a matter of debate as it is not always adaptive. That is, it is not always associated with a gain in performance and can even occur with a decline in accuracy. Here, we hypothesized that the nature of PES is influenced by one’s speed-accuracy tradeoff policy, which determines the overall level of choice accuracy in the task at hand. To test this hypothesis, we had subjects performing a task in two distinct contexts (separate days), which either promoted speed (hasty context) or cautiousness (cautious context), allowing us to consider post-error adjustments according to whether subjects performed choices with a low or high accuracy level, respectively. Accordingly, our data indicate that post-error adjustments varied according to the context in which subjects performed the task, with PES being solely significant in the hasty context (low accuracy). In addition, we only observed a gain in performance after errors in a specific trial type, suggesting that post-error adjustments depend on a complex combination of processes that affect the speed of ensuing actions as well as the degree to which such PES comes with a gain in performance.

Early correlates of error-related brain activity predict subjective timing of error awareness

Humans are remarkably reliable in detecting errors in their behavior. Whereas error awareness has been assumed to emerge not until 200-400 ms after an error, the so-called early error sensations refer to the subjective feeling of having detected an error even before the erroneous response was executed. Here, we collected electroencephalogram (EEG) to track how early error sensations are reflected in neural correlates of performance monitoring. Participants first had to perform a task, and then had to indicate whether an error has occurred and whether this error was detected before or after response execution. EEG results showed that early error sensations were associated with an earlier peak of the error-related negativity (Ne/ERN), a component of error-related brain activity that occurs briefly after the error response. This demonstrates that early error-related activity influences metacognitive judgments on the time course of error awareness, and thus contributes to error awareness.

The influence of error detection and error significance on neural and behavioral correlates of error processing in a complex choice task

Error detection and error significance form essential mechanisms that influence error processing and action adaptation. Error detection often is assessed by an immediate self-evaluation of accuracy. Our study used cognitive neuroscience methods to elucidate whether self-evaluation itself influences error processing by increasing error significance in the context of a complex response selection process. In a novel eight-alternative response task, our participants responded to eight symbol stimuli with eight different response keys and a specific stimulus-response assignment. In the first part of the experiment, the participants merely performed the task. In the second part, they also evaluated their response accuracy on each trial. We replicated variations in early and later stages of error processing and action adaptation as a function of error detection. The additional self-evaluation enhanced error processing on later stages, probably reflecting error evidence accumulation, whereas earlier error monitoring processes were not amplified. Implementing multivariate pattern analysis revealed that self-evaluation influenced brain activity patterns preceding and following the response onset, independent of response accuracy. The classifier successfully differentiated between responses from the self- and the no-self-evaluation condition several hundred milliseconds before response onset. Subsequent exploratory analyses indicated that both self-evaluation and the time on task contributed to these differences in brain activity patterns. This suggests that in addition to its effect on error processing, self-evaluation in a complex choice task seems to have an influence on early and general processing mechanisms (e.g., the quality of attention and stimulus encoding), which is amplified by the time on task.

Meta-control: From psychology to computational neuroscience

Research in the past decades shed light on the different mechanisms that underlie our capacity for cognitive control. However, the meta-level processes that regulate cognitive control itself remain poorly understood. Following the terminology from artificial intelligence, meta-control can be defined as a collection of mechanisms that (a) monitor the progress of controlled processing and (b) regulate the underlying control parameters in the service of current task goals and in response to internal or external constraints. From a psychological perspective, meta-control is an important concept because it may help explain and predict how and when human agents select different types of behavioral strategies. From a cognitive neuroscience viewpoint, meta-control is a useful concept for understanding the complex networks in the prefrontal cortex that guide higher-level behavior as well as their interactions with neuromodulatory systems (such as the dopamine or norepinephrine system). The purpose of the special issue is to integrate hitherto segregated strands of research across three different perspectives: 1) a psychological perspective that specifies meta-control processes on a functional level and aims to operationalize them in experimental tasks; 2) a computational perspective that builds on ideas from artificial intelligence to formalize normative solutions to meta-control problems; and 3) a cognitive neuroscience perspective that identifies neural correlates of and mechanisms underlying meta-control.