Search

Abstract Working memory (WM) and reinforcement learning (RL) both influence decision-making, but how they interact to affect behaviour remains unclear. We assessed whether RL is influenced by the format of visual stimuli held in WM, either feature-based or unified, object-based representations. In a pre-registered paradigm, participants learned stimulus-action combinations that provided reward through 80% probabilistic feedback. In parallel, participants retained the RL stimulus in WM and were asked to recall this stimulus after each RL choice. Crucially, the format of representation probed in WM was manipulated, with blocks encouraging either separate features or bound objects to be remembered. Incentivising a feature-based WM representation facilitated feature-based learning, shown by an improved choice strategy. This reveals a role of WM in providing sustained internal representations that are harnessed by RL, providing a framework by which these two cognitive processes cooperate.

Abstract People differ in their levels of impulsivity and patience, and these preferences are heavily influenced by others. Previous research suggests that susceptibility to social influence may vary with age, but the mechanisms and whether people are more influenced by patience or impulsivity remain unknown. Here, using a delegated inter-temporal choice task and Bayesian computational models, we tested susceptibility to social influence in young (aged 18–36, N = 76) and older (aged 60–80, N = 78) adults. Participants completed a temporal discounting task and then learnt the preferences of two other people (one more impulsive and one more patient) before making their choices again. We used the signed Kullback-Leibler divergence to quantify the magnitude and direction of social influence. We found that, compared to young adults, older adults were relatively more susceptible to impulsive social influence. Factor analyses showed that older adults with higher self-reported levels of affective empathy and emotional motivation were particularly susceptible to impulsive influence. Importantly, older and young adults showed similar learning accuracy about others’ preferences, and their baseline impulsivity did not differ. Together, these findings suggest highly affectively empathetic and emotionally motivated older adults may be at higher risk for impulsive decisions, due to their susceptibility to social influence.

Abstract Automated online cognitive assessments are set to revolutionise clinical research and healthcare. However, their applicability for Parkinson’s Disease (PD) and REM Sleep Behavioural Disorder (RBD), a strong PD precursor, is underexplored. Here, we developed an online battery to measure early cognitive changes in PD and RBD. Evaluating 19 candidate tasks showed significant global accuracy deficits in PD (0.65 SD, p = 0.003) and RBD (0.45 SD, p = 0.027), driven by memory, language, attention and executive underperformance, and global reaction time deficits in PD (0.61 SD, p = 0.001). We identified a brief 20-min battery that had sensitivity to deficits across these cognitive domains while being robust to the device used. This battery was more sensitive to early-stage and prodromal deficits than the supervised neuropsychological scales. It also diverged from those scales, capturing additional cognitive factors sensitive to PD and RBD. This technology offers an economical and scalable method for assessing these populations that can complement standard supervised practices.

Abstract Apathy and impulsivity are expressed in a wide range of neuropsychiatric disorders, and, to a less severe extent, in healthy people too. Although traditionally considered to be opposite extremes of a single motivational spectrum, recent epidemiological questionnaire-based data suggest that both traits can in fact co-exist within the same individual. Here, we sought to investigate the relationship between these constructs in healthy people within a controlled task environment that examines the ability to make a decision under temporal uncertainty and measures the vigour of the response. Sixty participants performed a new version of the Traffic Light Task and completed self-report questionnaire measures of apathy and impulsivity. The task required individuals to make rapid decision-making for time-sensitive reward by squeezing a hand-held dynamometer as quickly as possible after a predictable event occurred (a traffic light turning green). Although apathy and impulsivity were positively correlated in questionnaire assessments, the two traits were associated with distinct behavioural signatures on the task. Impulsivity was expressed as an inflexible tendency to generate rapid anticipatory responses, regardless of cost-benefit information. Apathy, on the other hand, was associated with a blunted effect of reward on response vigour. These findings reveal how apathy and impulsivity are related to distinct dimensions of goal-directed behaviour, explaining how these traits might co-exist in the same individuals.

Abstract Reinforcement learning (RL) is widely regarded as divisible into two distinct computational strategies. Model-free learning is a simple RL process in which a value is associated with actions, whereas model-based learning relies on the formation of internal models of the environment to maximise reward. Recently, theoretical and animal work has suggested that such models might be used to train model-free behaviour, reducing the burden of costly forward planning. Here we devised a way to probe this possibility in human behaviour. We adapted a two-stage decision task and found evidence that model-based processes at the time of learning can alter model-free valuation in healthy individuals. We asked people to rate subjective value of an irrelevant feature that was seen at the time a model-based decision would have been made. These irrelevant feature value ratings were updated by rewards, but in a way that accounted for whether the selected action retrospectively ought to have been taken. This model-based influence on model-free value ratings was best accounted for by a reward prediction error that was calculated relative to the decision path that would most likely have led to the reward. This effect occurred independently of attention and was not present when participants were not explicitly told about the structure of the environment. These findings suggest that current conceptions of model-based and model-free learning require updating in favour of a more integrated approach. Our task provides an empirical handle for further study of the dialogue between these two learning systems in the future.

The willingness to exert effort into demanding tasks often declines over time through fatigue. Here the authors provide a computational account of the moment-to-moment dynamics of fatigue and its impact on effort-based choices, and reveal the neural mechanisms that underlie such computations.