scholarly journals Affective bias as a rational response to the statistics of rewards and punishments

eLife ◽  
2017 ◽  
Vol 6 ◽  
Author(s):  
Erdem Pulcu ◽  
Michael Browning
Keyword(s):  
Information Content ◽  
Learning Task ◽  
Negative Outcomes ◽  
Computational Framework ◽  
Positive Events ◽  
Treatment Development ◽  
Learning Rates ◽  
Novel Approach ◽  
Affective Bias ◽  
Human Participants

Affective bias, the tendency to differentially prioritise the processing of negative relative to positive events, is commonly observed in clinical and non-clinical populations. However, why such biases develop is not known. Using a computational framework, we investigated whether affective biases may reflect individuals’ estimates of the information content of negative relative to positive events. During a reinforcement learning task, the information content of positive and negative outcomes was manipulated independently by varying the volatility of their occurrence. Human participants altered the learning rates used for the outcomes selectively, preferentially learning from the most informative. This behaviour was associated with activity of the central norepinephrine system, estimated using pupilometry, for loss outcomes. Humans maintain independent estimates of the information content of distinct positive and negative outcomes which may bias their processing of affective events. Normalising affective biases using computationally inspired interventions may represent a novel approach to treatment development.

scholarly journals Affective Bias as a Rational Response to the Statistics of Rewards and Punishments

2017 ◽  
Author(s):  
Erdem Pulcu ◽  
Michael Browning
Keyword(s):  
Information Content ◽  
Learning Task ◽  
Negative Outcomes ◽  
Computational Framework ◽  
Positive Events ◽  
Novel Treatment ◽  
Learning Rates ◽  
Norepinephrine System ◽  
Affective Bias ◽  
Human Participants

AbstractAffective bias, the tendency to prioritise the processing of negative relative to positive events, is causally linked to clinical depression. However, why such biases develop or how they may best be ameliorated is not known. Using a computational framework, we investigated whether affective biases may reflect an individual’s estimates of the information content of negative and positive events. During a reinforcement learning task, the information content of positive and negative outcomes was manipulated independently by varying the volatility of their occurrence. Human participants altered the learning rates used for the outcomes selectively, preferentially learning from the most informative. This behaviour was associated with activity of the central norepinephrine system, estimated using pupilometry, for loss outcomes. Humans maintain independent estimates of the information content of positive and negative outcomes which bias their processing of affective events. Normalising affective biases using computationally inspired interventions may represent a novel treatment approach for depression.

scholarly journals Dopaminergic medication reduces striatal sensitivity to negative outcomes in Parkinson’s disease

2018 ◽  
Author(s):  
Brónagh McCoy ◽  
Sara Jahfari ◽  
Gwenda Engels ◽  
Tomas Knapen ◽  
Jan Theeuwes
Keyword(s):  
Parkinson’S Disease ◽  
Parkinson's Disease ◽  
Learning Task ◽  
Dopamine Neurons ◽  
Negative Outcomes ◽  
Learning Differences ◽  
Dopaminergic Medication ◽  
Learning Rates ◽  
Approach Avoidance ◽  
Midbrain Dopamine

AbstractReduced levels of dopamine in Parkinson’s disease (PD) contribute to changes in learning, resulting from the loss of midbrain dopamine neurons that transmit a teaching signal to the striatum. Dopamine medication used by PD patients has previously been linked to either behavioral changes during learning itself or adjustments in approach and avoidance behavior after learning. To date, however, very little is known about the specific relationship between dopaminergic medication-driven differences during learning and subsequent changes in approach/avoidance tendencies in individual patients. We assessed 24 PD patients on and off dopaminergic medication and 24 healthy controls (HC) performing a probabilistic reinforcement learning task, while undergoing functional magnetic resonance imaging. During learning, medication in PD reduced an overemphasis on negative outcomes. When patients were on medication, learning rates were lower for negative (but not positive) outcomes and concurrent striatal BOLD responses showed reduced prediction error sensitivity. Medication-induced shifts in negative learning rates were predictive of changes in approach/avoidance choice patterns after learning, and these changes were accompanied by striatal BOLD response alterations. These findings highlight dopamine-driven learning differences in PD and provide new insight into how changes in learning impact the transfer of learned value to approach/avoidance responses in novel contexts.

Knowledge Acquired from Learning: New Evidence of Hierarchical Conceptualization

1994 ◽  
Vol 79 (2) ◽  
pp. 975-993 ◽  
Author(s):  
Alberto Montare
Keyword(s):  
Sex Differences ◽  
Knowledge Acquisition ◽  
Reversal Learning ◽  
Concept Formation ◽  
Learning Task ◽  
Learning Rates ◽  
Level 3 ◽  
Level 1 ◽  
New Evidence ◽  
Level 2

Following successful inductive acquisition of procedural cognition of a discrimination-reversal learning task, 50 female and 50 male undergraduates articulated declarative cognizance of knowledge acquired from learning. Tests of four hypotheses showed that (1) increasingly higher levels of declarative cognizance were associated with faster learning rates, (2) six new cases of cognition-without-cognizance were observed, (3) students presumably using secondary signalization learned faster than those presumably using primary signalization, and (4) no sex differences in learning rates or declarative cognizance were observed. The notion that explicit levels of declarative cognizance may represent implicit hierarchical conceptualization comprised of four systems of knowledge acquisition led to the conclusions that primary signalization may account for inductive senscept formation at Level 1 and for inductive percept formation at Level 2, whereas emergent secondary signalization may account for inductive precept formation at Level 3 and for inductive concept formation at Level 4.

To Choose or to Avoid: Age Differences in Learning from Positive and Negative Feedback

2011 ◽  
Vol 23 (1) ◽  
pp. 41-52 ◽  
Author(s):  
Ben Eppinger ◽  
Jutta Kray
Keyword(s):  
Older Adults ◽  
Learning Task ◽  
Learning Preferences ◽  
Negative Outcomes ◽  
Learning Condition ◽  
Younger Adults ◽  
Stimulus Response ◽  
Learning Impairments ◽  
Error Related Negativity ◽  
Reversal Phase

In this study, we investigated whether older adults learn more from bad than good choices than younger adults and whether this is reflected in the error-related negativity (ERN). We applied a feedback-based learning task with two learning conditions. In the positive learning condition, participants could learn to choose responses that lead to monetary gains, whereas in the negative learning condition, they could learn to avoid responses that lead to monetary losses. To test the stability of learning preferences, the task involved a reversal phase in which stimulus–response assignments were inverted. Negative learners were defined as individuals that performed better in the negative than in the positive learning condition (and vice versa for positive learners). The behavioral data showed strong individual differences in learning from positive and negative outcomes that persisted throughout the reversal phase and were more pronounced for older than younger adults. Older negative learners showed a stronger tendency to avoid negative outcomes than younger negative learners. However, contrary to younger adults, this negative learning bias was not associated with a larger ERN, suggesting that avoidance learning in older negative learners might be decoupled from error processing. Furthermore, older adults showed learning impairments compared to younger adults. The ERP analyses suggest that these impairments reflect deficits in the ability to build up relational representations of ambiguous outcomes.

scholarly journals Learning “How to Learn”: Super Declarative Motor Learning Is Impaired in Parkinson’s Disease

2017 ◽  
Vol 2017 ◽  
pp. 1-8 ◽  
Author(s):  
Lucio Marinelli ◽  
Carlo Trompetto ◽  
Stefania Canneva ◽  
Laura Mori ◽  
Flavio Nobili ◽  
...  
Keyword(s):  
Parkinson’S Disease ◽  
Parkinson's Disease ◽  
Performance Optimization ◽  
Learning Task ◽  
Daily Activities ◽  
Learning Capacity ◽  
Learning Rates ◽  
New Information ◽  
Drug Naïve ◽  
Better Than

Learning new information is crucial in daily activities and occurs continuously during a subject’s lifetime. Retention of learned material is required for later recall and reuse, although learning capacity is limited and interference between consecutively learned information may occur. Learning processes are impaired in Parkinson’s disease (PD); however, little is known about the processes related to retention and interference. The aim of this study is to investigate the retention and anterograde interference using a declarative sequence learning task in drug-naive patients in the disease’s early stages. Eleven patients with PD and eleven age-matched controls learned a visuomotor sequence, SEQ1, during Day1; the following day, retention of SEQ1 was assessed and, immediately after, a new sequence of comparable complexity, SEQ2, was learned. The comparison of the learning rates of SEQ1 on Day1 and SEQ2 on Day2 assessed the anterograde interference of SEQ1 on SEQ2. We found that SEQ1 performance improved in both patients and controls on Day2. Surprisingly, controls learned SEQ2 better than SEQ1, suggesting the absence of anterograde interference and the occurrence of learning optimization, a process that we defined as “learning how to learn.” Patients with PD lacked such improvement, suggesting defective performance optimization processes.

scholarly journals Trading off the cost of conflict against expected rewards

2018 ◽  
Author(s):  
Nura Sidarus ◽  
Stefano Palminteri ◽  
Valérian Chambon
Keyword(s):  
Decision Making ◽  
Reinforcement Learning ◽  
Cognitive Control ◽  
Reversal Learning ◽  
Computational Models ◽  
Learning Task ◽  
Learning Rates ◽  
Different Types ◽  
Irrelevant Distractors ◽  
The Cost

AbstractValue-based decision-making involves trading off the cost associated with an action against its expected reward. Research has shown that both physical and mental effort constitute such subjective costs, biasing choices away from effortful actions, and discounting the value of obtained rewards. Facing conflicts between competing action alternatives is considered aversive, as recruiting cognitive control to overcome conflict is effortful. Yet, it remains unclear whether conflict is also perceived as a cost in value-based decisions. The present study investigated this question by embedding irrelevant distractors (flanker arrows) within a reversal-learning task, with intermixed free and instructed trials. Results showed that participants learned to adapt their choices to maximize rewards, but were nevertheless biased to follow the suggestions of irrelevant distractors. Thus, the perceived cost of being in conflict with an external suggestion could sometimes trump internal value representations. By adapting computational models of reinforcement learning, we assessed the influence of conflict at both the decision and learning stages. Modelling the decision showed that conflict was avoided when evidence for either action alternative was weak, demonstrating that the cost of conflict was traded off against expected rewards. During the learning phase, we found that learning rates were reduced in instructed, relative to free, choices. Learning rates were further reduced by conflict between an instruction and subjective action values, whereas learning was not robustly influenced by conflict between one’s actions and external distractors. Our results show that the subjective cost of conflict factors into value-based decision-making, and highlights that different types of conflict may have different effects on learning about action outcomes.

scholarly journals Monetary feedback modulates performance and electrophysiological indices of belief updating in reward learning

2017 ◽  
Author(s):  
Daniel Bennett ◽  
Karen Sasmita ◽  
Ryan T. Maloney ◽  
Carsten Murawski ◽  
Stefan Bode
Keyword(s):  
Bayesian Inference ◽  
Cognitive Strategies ◽  
Learning Task ◽  
Feedback Condition ◽  
Event Related Potential ◽  
Reward Learning ◽  
Instructive Feedback ◽  
Motivational State ◽  
Belief Updating ◽  
Learning Rates

AbstractBelief updating entails the incorporation of new information about the environment into internal models of the world. Bayesian inference is the statistically optimal strategy for performing belief updating in the presence of uncertainty. An important open question is whether the use of cognitive strategies that implement Bayesian inference is dependent upon motivational state and, if so, how this is reflected in electrophysiological signatures of belief updating in the brain. Here we recorded the electroencephalogram of participants performing a simple reward learning task with both monetary and non-monetary instructive feedback conditions. Our aim was to distinguish the influence of the rewarding properties of feedback on belief updating from the information content of the feedback itself. A Bayesian updating model allowed us to quantify different aspects of belief updating across trials, including the size of belief updates and the uncertainty of beliefs. Faster learning rates were observed in the monetary feedback condition compared to the instructive feedback condition, while belief updates were generally larger, and belief uncertainty smaller, with monetary compared to instructive feedback. Larger amplitudes in the monetary feedback condition were found for three event-related potential components: the P3a, the feedback-related negativity (FRN) and the late positive potential (LPP). These findings suggest that motivational state influences inference strategies in reward learning, and this is reflected in the electrophysiological correlates of belief updating.

scholarly journals Choosing and learning: outcome valence differentially affects learning from free versus forced choices

2019 ◽  
Author(s):  
Valérian Chambon ◽  
Héloïse Théro ◽  
Marie Vidal ◽  
Henri Vandendriessche ◽  
Patrick Haggard ◽  
...  
Keyword(s):  
Confirmation Bias ◽  
Forced Choice ◽  
Prediction Errors ◽  
Negative Events ◽  
Positive Events ◽  
Learning Rates ◽  
Outcome Valence ◽  
Special Value ◽  
Computational Analyses ◽  
Robust Evidence

AbstractPositivity bias refers to learning more from positive than negative events. This learning asymmetry could either reflect a preference for positive events in general, or be the upshot of a more general, and perhaps, ubiquitous, “choice-confirmation” bias, whereby agents preferentially integrate information that confirms their previous decision. We systematically compared these two theories with 3 experiments mixing free- and forced-choice conditions, featuring factual and counterfactual learning and varying action requirements across “go” and “no-go” trials. Computational analyses of learning rates showed clear and robust evidence in favour of the “choice-confirmation” theory: participants amplified positive prediction errors in free-choice conditions while being valence-neutral on forced-choice conditions. We suggest that a choice-confirmation bias is adaptive to the extent that it reinforces actions that are most likely to meet an individual’s needs, i.e. freely chosen actions. In contrast, outcomes from unchosen actions are more likely to be treated impartially, i.e. to be assigned no special value in self-determined decisions.

Sign in / Sign up

Export Citation Format

Share Document