Periodic Reporting for period 3 - Ctrl-ImpAct (Control of impulsive action)
Berichtszeitraum: 2021-06-01 bis 2022-11-30
Adaptive behaviour is typically attributed to an executive-control system that allows people to regulate impulsive actions and to fulfil long-term goals instead. Failures to regulate impulsive actions have been associated with a variety of clinical and behavioural disorders. Therefore, establishing a good understanding of impulse-control mechanisms and how to improve them could be hugely beneficial for both individuals and society at large. Yet many fundamental questions remain unanswered. This stems from a narrow focus on reactive inhibitory control and well-practiced actions. To make significant progress, we need to develop new models that integrate different aspects of impulsive action and executive control. The proposed research program aims to answer five fundamental questions. (1) Can novel impulsive actions arise during task-preparation stages?; (2) What is the role of negative events in the origin and control of impulsive actions?; (3) How does learning modulate impulsive behaviour?; (4) When are impulsive actions (dys)functional?; and (5) How is variation in state impulsivity associated with trait impulsivity?
To answer these questions, we will use carefully designed behavioural paradigms, cognitive neuroscience techniques (TMS & EEG), physiological measures (e.g. force), mathematical modelling of decision-making, and animal models to specify the origin and control of impulsive actions. Our ultimate goal is to transform the impulsive action field by replacing the currently dominant ‘inhibitory control’ models of impulsive action with detailed multifaceted models that can explain impulsivity and control across time and space. Developing a new behavioural model of impulsive action will also contribute to a better understanding of the causes of individual differences in impulsivity and the many disorders associated with impulse-control deficits.
To answer these questions, we will use carefully designed behavioural paradigms, cognitive neuroscience techniques (TMS & EEG), physiological measures (e.g. force), mathematical modelling of decision-making, and animal models to specify the origin and control of impulsive actions. Our ultimate goal is to transform the impulsive action field by replacing the currently dominant ‘inhibitory control’ models of impulsive action with detailed multifaceted models that can explain impulsivity and control across time and space. Developing a new behavioural model of impulsive action will also contribute to a better understanding of the causes of individual differences in impulsivity and the many disorders associated with impulse-control deficits.
In the first 30 months of the project, my team and I focused primarily on the first two questions. Specifically, we studied how (and to what extent) impulsive actions arise (1) during the task-preparation stage, and (2) in response to negative events.
With regards to the first main question, we have first shown that (prolonged) self-paced preparation is a good predicator of accurate task performance, contradicting previous findings. Interestingly, this correlation holds for different age groups, including young children (4-5 year olds). Note that the latter indicates that even young children can proactively prepare for novel tasks when sufficiently motivated (again, contradicting some previous findings). In a second step, we have shown that adult participants adjust their preparation strategies (at least to some extent) when the overall task context changes, suggesting that advance preparation (and the urge to get started) is under strategic/voluntary control. Finally, the same line of research suggests that advance preparation can also come with costs; and that when these costs start to outweigh the benefits, less preparation (a key aspect of impulsivity) is more 'functional' (or beneficial for overall task performance) than more preparation.
With regards to the second question, we have carefully demonstrated that people speed up in gambling tasks after losing points or money (as long as the task is sufficiently engaging) or when progress is thwarted in decision-making tasks. This is particularly the case when the expectancy to win on a particular trial is increased. These findings are consistent with appraisal theories and the idea that negative events can invigorate subsequent behaviors ('frustration'). Similar behaviors have been observed in various non-human animals, opening up new avenues for comparative research, and more generally, a comprehensive account of the origin of impulsive action.
With regards to the first main question, we have first shown that (prolonged) self-paced preparation is a good predicator of accurate task performance, contradicting previous findings. Interestingly, this correlation holds for different age groups, including young children (4-5 year olds). Note that the latter indicates that even young children can proactively prepare for novel tasks when sufficiently motivated (again, contradicting some previous findings). In a second step, we have shown that adult participants adjust their preparation strategies (at least to some extent) when the overall task context changes, suggesting that advance preparation (and the urge to get started) is under strategic/voluntary control. Finally, the same line of research suggests that advance preparation can also come with costs; and that when these costs start to outweigh the benefits, less preparation (a key aspect of impulsivity) is more 'functional' (or beneficial for overall task performance) than more preparation.
With regards to the second question, we have carefully demonstrated that people speed up in gambling tasks after losing points or money (as long as the task is sufficiently engaging) or when progress is thwarted in decision-making tasks. This is particularly the case when the expectancy to win on a particular trial is increased. These findings are consistent with appraisal theories and the idea that negative events can invigorate subsequent behaviors ('frustration'). Similar behaviors have been observed in various non-human animals, opening up new avenues for comparative research, and more generally, a comprehensive account of the origin of impulsive action.
Most research on impulse control has focused on well-practiced actions that can be inhibited when they are contextually inappropriate. Often it seems that there is an (implicit) assumption that novel actions cannot be impulsive. We disagree and have shown that in case of novel actions, impulsivity can arise at a preparation stage instead of a post-stimulus processing stage. Starting too quickly could lead to inappropriate actions (e.g. pressing the wrong buttons), with possible negative consequences. This type of behaviour (i.e. starting novel actions too quickly) differs from other forms of action impulsivity as it seems to arise from a response-unspecific urge to get started, and has been largely ignored in cognitive and neurological research on action impulsivity (as demonstrated by the absence of this topic in recent reviews). Thus, this work on advance preparation sheds new light on a – so far largely neglected – aspect of action impulsivity.
Second, many psychological theories assume that the executive-control system alters the settings of lower-level systems when people make an error or when outcomes are otherwise less desirable than anticipated. For example, people often slow down after they make an error (‘post-error slowing’). Such slowing has been observed in a variety of tasks, and is usually attributed to performance-related adjustments in task-relevant processing pathways. Such adjustments usually increase response latencies (i.e. people become more cautious) and reduce the likelihood of further negative or suboptimal outcomes on subsequent trials. However, we have now clearly demonstrated that post-error slowing may not be a general phenomenon after all. In gambling situations, people often take more risk after a loss (‘loss chasing’; note that we find this in some of our studies as well), and similar effects of negative outcomes on choice behaviour have been observed in other situations (e.g. many stock market investors tend to take more risk when they have lost money). We have now extended these findings to impulsive action in various self-paced gambling games (including gambling in the real world). We consistently found that after a gambled loss, subjects were faster to initiate the next game, indicated their choice faster. These findings undermine the dominant post-error slowing account. We propose that errors or suboptimal outcomes in (motivationally-salient) free-choice tasks can energise and intensify actions and approach tendencies in some situations. In the literature, such behavioral effects are often linked to 'frustration' or 'regret'. In other words, our work shows how frustration or regret could trigger post-error impulsivity instead of post-error caution if not properly regulated. Failures to adjust response strategies after a loss or a suboptimal outcome may contribute to the development of problematic behaviours. Thus, it is important to determine how and under what circumstances emotion and outcome processing modulate executive control and impulsive action.
To date, we have focused primarily on impulsive states (i.e. short-term and transient changes in behaviour in response to specific events). In the next phase, we will also study how such impulsive behaviors can become reinforced (learning), and eventually lead to a transition from state impulsivity to trait impulsivity (i.e. a stable personality characteristic that is influenced by learning, environment, and genetic factors).
Second, many psychological theories assume that the executive-control system alters the settings of lower-level systems when people make an error or when outcomes are otherwise less desirable than anticipated. For example, people often slow down after they make an error (‘post-error slowing’). Such slowing has been observed in a variety of tasks, and is usually attributed to performance-related adjustments in task-relevant processing pathways. Such adjustments usually increase response latencies (i.e. people become more cautious) and reduce the likelihood of further negative or suboptimal outcomes on subsequent trials. However, we have now clearly demonstrated that post-error slowing may not be a general phenomenon after all. In gambling situations, people often take more risk after a loss (‘loss chasing’; note that we find this in some of our studies as well), and similar effects of negative outcomes on choice behaviour have been observed in other situations (e.g. many stock market investors tend to take more risk when they have lost money). We have now extended these findings to impulsive action in various self-paced gambling games (including gambling in the real world). We consistently found that after a gambled loss, subjects were faster to initiate the next game, indicated their choice faster. These findings undermine the dominant post-error slowing account. We propose that errors or suboptimal outcomes in (motivationally-salient) free-choice tasks can energise and intensify actions and approach tendencies in some situations. In the literature, such behavioral effects are often linked to 'frustration' or 'regret'. In other words, our work shows how frustration or regret could trigger post-error impulsivity instead of post-error caution if not properly regulated. Failures to adjust response strategies after a loss or a suboptimal outcome may contribute to the development of problematic behaviours. Thus, it is important to determine how and under what circumstances emotion and outcome processing modulate executive control and impulsive action.
To date, we have focused primarily on impulsive states (i.e. short-term and transient changes in behaviour in response to specific events). In the next phase, we will also study how such impulsive behaviors can become reinforced (learning), and eventually lead to a transition from state impulsivity to trait impulsivity (i.e. a stable personality characteristic that is influenced by learning, environment, and genetic factors).