Error-induced adaptability: Behavioral and neural dynamics of response-stimulus interval modulations on posterror slowing.

Abstract

Response errors often cause individuals to slow down their subsequent reactions (posterror slowing [PES]). Despite intensive investigations on PES, the adaptive nature of PES remains unresolved. Here, we systematically examined this issue by manipulating response-stimulus intervals (RSIs) and examining their influence on behaviors and neural dynamics of PES. Behavioral and electrophysiological (EEG) measures were recorded while male and female human participants performed a four-choice flanker task as RSIs were manipulated. Behaviorally, PES showed maladaptive features at short RSIs but some adaptive features at long RSIs. EEG results indicated that RSIs did not affect basic error-related processing, indexed by the same pattern in the contrasts between flanker errors and correct responses on the error-related negativity (ERN), error positivity (Pe), or theta band, no matter at short or long RSIs. However, RSIs significantly influenced postflanker error attentional adjustment, motor inhibition, and sensory sensitivity. At short RSIs, compared with postcorrect trials, postflanker error trials elicited larger beta band power and smaller P1 amplitude but did not affect alpha band power, suggesting that motor processing was inhibited, and subsequent sensory processing was impaired, but no attentional adjustment occurred. By contrast, at long RSIs, postflanker error trials led to smaller alpha and beta band power but did not affect P1 amplitude, indicating that attentional adjustment but not motor inhibition occurred, and sensory processing was not impaired. Together with behavioral results, the current study demonstrated that PES was adaptive at long RSIs but maladaptive at short RSIs. We further discuss the role of central resources in the adaptability of PES.

Learning from mistakes: Incidental encoding reveals a time-dependent enhancement of posterror target processing.

“…At even longer RSIs, the cascade is not interrupted, such that a PIA, PERI, and PES should be observed, and in this case, these would in part reflect a strategic change in processing, such as retuning attention to the task at hand. A recent study by Li et al (2021) used EEG to investigate posterror processing at a short (200 ms) versus long RSI (1,000 ms). At the short RSI, EEG markers indicated dampened sensory processing, enhanced motor inhibition, and an absence of attentional adjustments on posterror compared with postcorrect trials.…”

Post-error adjustments depend causally on executive attention: Evidence from an intervention study

“…Consequently, the first aim of this study was to demonstrate the trainability of EA, the second aim was to test the causal link from EA to error processing, and the third aim was to clarify the underlying cognitive mechanisms to determine if there was a causal link between the two. Theoretically, we assumed that the cognitive root of error processing is EA, which would be consistent with some theories (Jentzsch and Dudschig, 2009;Li et al, 2021), and hypothesized that EA intervention improves post-error performance by effectively diminishing the attention bottleneck induced by error monitoring, and supplying more attention resources for the post-error adjustments. To measure post-error performance before and after the intervention, we employed the modified flanker task (the four-choice flanker task; Maier et al, 2008) which was more difficult than typical flanker tasks (Eriksen and Eriksen, 1974), in order to obtain adequate post-error trials for analysis.…”

“…Likewise, maladaptive theories propose that error commission momentarily impairs ongoing processing, resulting in decreased PEA ( Notebaert et al, 2009 ; Houtman and Notebaert, 2013 ). Significantly, the recent research suggests that available attention resources play an important role in the adaptability of error processing ( Li et al, 2021 ). This seems to indicate that attention is the core cognitive ability determining PEA.…”

Binding Error-Induced Control States

“…We hypothesized that correct and erroneous responses induce different control states that carry over to the subsequent behavioral instance (we will refer to the behavioral episode that instigates a certain control state as trial n–2). For errors, increased cognitive control manifests robustly as post-error slowing, especially for the response immediately following the error ( Jentzsch & Dudschig, 2009 ; Li et al, 2021 ; Notebaert et al, 2009 ; Pfister & Foerster, 2022 ; Wessel, 2018 ). 1 Empirical evidence on changes in accuracy after errors does not provide a systematic pattern although more cognitive control should lead to a post-error increase in accuracy (see the General Discussion for a more detailed argument).…”