The neural basis of belief updating and rational decision making

Sleep restriction alters the integration of multiple information sources in probabilistic decision-making

The detrimental effects of sleep loss on overall decision-making have been well described. Due to the complex nature of decisions, there remains a need for studies to identify specific mechanisms of decision-making vulnerable to sleep loss. Bayesian perspectives of decision-making posit judgement formation during decision-making occurs via a process of integrating knowledge gleaned from past experiences (priors) with new information from current observations (likelihoods). We investigated the effects of sleep loss on the ability to integrate multiple sources of information during decision-making by reporting results from two experiments: the first implementing both sleep restriction (SR) and total sleep deprivation (TSD) protocols, and the second implementing an SR protocol. In both experiments, participants were administered the Bayes Decisions Task on which optimal performance requires the integration of Bayesian prior and likelihood information. Participants in Experiment 1 showed reduced reliance on both information sources after SR, while no significant change was observed after TSD. Participants in Experiment 2 showed reduced reliance on likelihood after SR, especially during morning testing sessions. No accuracy-related impairments resulting from SR and TSD were observed in both experiments. Our findings show SR affects decision-making through altering the way individuals integrate available sources of information. Additionally, the ability to integrate information during SR may be influenced by time of day. Broadly, our findings carry implications for working professionals who are required to make high-stakes decisions on the job, yet consistently receive insufficient sleep due to work schedule demands.

To Each Their Own: The Impact of Regulatory Focus on Consumers’ Response to Online Information Load

Contrary to the common view that more information is always better, too much information can damage decision quality. Building on existing literature, this study identified regulatory focus as a critical factor influencing the effect of information load (IL) on online consumer decisions and used event-related potentials (ERPs) to uncover its underlying neural mechanism. Behavioral data showed that promotion-focused participants would spend less time making purchasing decisions in the low IL condition than in the high IL condition. However, no significant difference was found for prevention-focused participants. In contrast to the high IL condition, ERP data indicated that the low IL condition recruited more attentional resources at the early stage of rapid automated processing (larger P2 component), leading to reduced long-term memory conflict (smaller N2 component), and resulting in enhanced decision confidence (larger P3 component) for those with a promotion focus. However, we observed either weakened or even opposite outcomes for those with a prevention focus. These findings generally shed light on when e-retailers should provide large/small amounts of product information in online environments.

Causal Role of the Dorsolateral Prefrontal Cortex in Belief Updating under Uncertainty

Adaptive performance in uncertain environments depends on the ability to continuously update internal beliefs about environmental states. Recent correlative evidence suggests that a frontoparietal network including the dorsolateral prefrontal cortex (dlPFC) supports belief updating under uncertainty, but whether the dlPFC serves a "causal" role in this process is currently not clear. To elucidate its contribution, we leveraged transcranial direct current stimulation (tDCS) over the right dlPFC, while 91 participants performed an incentivized belief-updating task. Participants also underwent a psychosocial stress or control manipulation to investigate the role of stress, which is known to modulate dlPFC functioning. We observed enhanced monetary value updating after anodal tDCS when it was normatively expected from a Bayesian perspective. A model-based analysis indicates that this effect was driven by belief updating. However, we also observed enhanced non-normative value updating, which might have been driven instead by expectancy violation. Enhanced normative and non-normative value updating reflected increased vs. decreased Bayesian rationality, respectively. Furthermore, cortisol increases were associated with enhanced positive, but not with negative, value updating. The present study thereby sheds light on the causal role of the right dlPFC in the remarkable human ability to navigate uncertain environments by continuously updating prior knowledge following new evidence.

The Bayesian Approach to Decision Making and Analysis in Nutrition Research and Practice

This is part of a series of monographs on research design and analysis. The purpose of this article is to describe the purposes of and approach to conducting Bayesian decision making and analysis. Bayesian decision making involves basing decisions on the probability of a successful outcome, where this probability is informed by both prior information and new evidence the decision maker obtains. The statistical analysis that underlies the calculation of these probabilities is Bayesian analysis. In recent years, the Bayesian approach has been applied more commonly in both nutrition research and clinical decision making, and registered dietitian nutritionists would benefit from gaining a deeper understanding of this approach. This article provides a background of Bayesian decision making and analysis, and then presents applications of the approach in two different areas-medical diagnoses and nutrition policy research. It concludes with a description of how Bayesian decision making may be used in everyday life to allow each of us to appropriately weigh established beliefs and prior knowledge with new data and information in order to make well-informed and wise decisions.

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

Belief 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 EEG of participants performing a simple reward learning task with both monetary and nonmonetary 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 ERP components: the P3a, the feedback-related negativity, and the late positive potential. These findings suggest that motivational state influences inference strategies in reward learning, and this is reflected in the electrophysiological correlates of belief updating.

Motivation alters impression formation and related neural systems

Observers frequently form impressions of other people based on complex or conflicting information. Rather than being objective, these impressions are often biased by observers’ motives. For instance, observers often downplay negative information they learn about ingroup members. Here, we characterize the neural systems associated with biased impression formation. Participants learned positive and negative information about ingroup and outgroup social targets. Following this information, participants worsened their impressions of outgroup, but not ingroup, targets. This tendency was associated with a failure to engage neural structures including lateral prefrontal cortex, dorsal anterior cingulate cortex, temporoparietal junction, Insula and Precuneus when processing negative information about ingroup (but not outgroup) targets. To the extent that participants engaged these regions while learning negative information about ingroup members, they exhibited less ingroup bias in their impressions. These data are consistent with a model of ‘effortless bias’, under which perceivers fail to process goal-inconsistent information in order to maintain desired conclusions.

Neural Signatures of Rational and Heuristic Choice Strategies: A Single Trial ERP Analysis

In multi-attribute choice, people use heuristics to simplify decision problems. We studied the use of heuristic and rational strategies and their electrophysiological correlates. Since previous work linked the P3 ERP component to attention and decision making, we were interested whether the amplitude of this component is associated with decision strategy use. To this end, we recorded EEG when participants performed a two-alternative choice task, where they could acquire decision cues in a sequential manner and use them to make choices. We classified participants’ choices as consistent with a rational Weighted Additive rule (WADD) or a simple heuristic Take The Best (TTB). Participants differed in their preference for WADD and TTB. Using a permutation-based single trial approach, we analyzed EEG responses to consecutive decision cues and their relation to the individual strategy preference. The preference for WADD over TTB was associated with overall higher signal amplitudes to decision cues in the P3 time window. Moreover, the preference for WADD was associated with similar P3 amplitudes to consecutive cues, whereas the preference for TTB was associated with substantial decreases in P3 amplitudes to consecutive cues. We also found that the preference for TTB was associated with enhanced N1 component to cues that discriminated decision alternatives, suggesting very early attention allocation to such cues by TTB users. Our results suggest that preference for either WADD or TTB has an early neural signature reflecting differences in attentional weighting of decision cues. In light of recent findings and hypotheses regarding P3, we interpret these results as indicating the involvement of catecholamine arousal systems in shaping predecisional information processing and strategy selection.

Attentional Filter Training but Not Memory Training Improves Decision-Making

Decision-making has a high practical relevance for daily performance. Its relation to other cognitive abilities such as executive control and memory is not fully understood. Here we asked whether training of either attentional filtering or memory storage would influence decision-making as indexed by repetitive assessments of the Iowa Gambling Task (IGT). The IGT was developed to assess and simulate real-life decision-making (Bechara et al., 2005). In this task, participants gain or lose money by developing advantageous or disadvantageous decision strategies. On five consecutive days we trained 29 healthy young adults (20–30 years) either in working memory (WM) storage or attentional filtering and measured their IGT scores after each training session. During memory training (MT) subjects performed a computerized delayed match-to-sample task where two displays of bars were presented in succession. During filter training (FT) participants had to indicate whether two simultaneously presented displays of bars matched or not. Whereas in MT the relevant target stimuli stood alone, in FT the targets were embedded within irrelevant distractors (bars in a different color). All subjects within each group improved their performance in the trained cognitive task. For the IGT, we observed an increase over time in the amount of money gained in the FT group only. Decision-making seems to be influenced more by training to filter out irrelevant distractors than by training to store items in WM. Selective attention could be responsible for the previously noted relationship between IGT performance and WM and is therefore more important for enhancing efficiency in decision-making.

Cognitive Reflection, Decision Biases, and Response Times

We present novel evidence on response times and personality traits in standard questions from the decision-making literature where responses are relatively slow (medians around half a minute or above). To this end, we measured response times in a number of incentivized, framed items (decisions from description) including the Cognitive Reflection Test, two additional questions following the same logic, and a number of classic questions used to study decision biases in probability judgments (base-rate neglect, the conjunction fallacy, and the ratio bias). All questions create a conflict between an intuitive process and more deliberative thinking. For each item, we then created a non-conflict version by either making the intuitive impulse correct (resulting in an alignment question), shutting it down (creating a neutral question), or making it dominant (creating a heuristic question). For CRT questions, the differences in response times are as predicted by dual-process theories, with alignment and heuristic variants leading to faster responses and neutral questions to slower responses than the original, conflict questions. For decision biases (where responses are slower), evidence is mixed. To explore the possible influence of personality factors on both choices and response times, we used standard personality scales including the Rational-Experiential Inventory and the Big Five, and used them as controls in regression analysis.

Changing health behavior motivation from I-must to I-want

In the past, medicine was dominated by acute diseases. Since treatments were unknown to patients they followed their medical doctors´ directives-at least for the duration of the disease. Behavior was thus largely motivated by avoiding expected costs associated with alternative behaviors (I-must). The health challenges prevailing today are chronic conditions resulting from the way we chose to live. Traditional directive communication has not been successful in eliciting and maintaining appropriate lifestyle changes. An approach successful in other fields is to motivate behavior by increasing expected rewards (I-want). Drawing on neuroeconomic and marketing research, we outline strategies including simplification, repeated exposure, default framing, social comparisons, and consumer friendliness to foster sustained changes in preference. We further show how these measures could be integrated into the health care system.

Measures of Bayesian Reasoning Performance on 'Normal' and 'Natural' Frequency Tasks

While the majority of similar studies examining Bayesian reasoning investigate how participants avoid common errors such as base-rate neglect, the current research also examines whether different formats (frequency and probability) lead to a difference in levels of absolute accuracy. In Study One, older (≥60 years) and younger (18 to 29 years) participants completed tasks in probability and normalized frequency formats. In Study 2, participants completed tasks in probability and natural frequency formats. Findings are that frequencies lead to less over-estimation, particularly in natural frequency tasks, which also reveal an interaction between age and task format whereby older adults seem unaffected by format. There was no association found between format and the avoidance of errors such as base-rate neglect. Findings are discussed in the light of dual and multi-process theories of reasoning, having failed to support the theory that frequency formats elicit System 2 reasoning processes.

Prior probabilities modulate cortical surprise responses: A study of event-related potentials

The human brain predicts events in its environment based on expectations, and unexpected events are surprising. When probabilistic contingencies in the environment are precisely instructed, the individual can form expectations based on quantitative probabilistic information ('inference-based learning'). In contrast, when probabilistic contingencies are imprecisely instructed, expectations are formed based on the individual's cumulative experience ('experience-based learning'). Here, we used the urn-ball paradigm to investigate how variations in prior probabilities and in the precision of information about these priors modulate choice behavior and event-related potential (ERP) correlates of surprise. In the urn-ball paradigm, participants are repeatedly forced to infer hidden states responsible for generating observable events, given small samples of factual observations. We manipulated prior probabilities of the states, and we rendered the priors calculable or incalculable, respectively. The analysis of choice behavior revealed that the tendency to consider prior probabilities when making decisions about hidden states was stronger when prior probabilities were calculable, at least in some of our participants. Surprise-related P3b amplitudes were observed in both the calculable and the incalculable prior probability condition. In contrast, calculability of prior probabilities modulated anteriorly distributed ERP amplitudes: when prior probabilities were calculable, surprising events elicited enhanced P3a amplitudes. However, when prior probabilities were incalculable, surprise was associated with enhanced N2 amplitudes. Furthermore, interindividual variability in reliance on prior probabilities was associated with attenuated P3b surprise responses under calculable in comparison to incalculable prior probabilities. Our results suggest two distinct neural systems for probabilistic learning that are recruited depending on contextual cues such as the precision of probabilistic information. Individuals with stronger tendencies to rely on calculable prior probabilities seem to have better adapted expectations at their disposal, as indicated by an attenuation of their P3b surprise responses when prior probabilities are calculable.

Probabilistic Inference: Task Dependency and Individual Differences of Probability Weighting Revealed by Hierarchical Bayesian Modeling

Cognitive determinants of probabilistic inference were examined using hierarchical Bayesian modeling techniques. A classic urn-ball paradigm served as experimental strategy, involving a factorial two (prior probabilities) by two (likelihoods) design. Five computational models of cognitive processes were compared with the observed behavior. Parameter-free Bayesian posterior probabilities and parameter-free base rate neglect provided inadequate models of probabilistic inference. The introduction of distorted subjective probabilities yielded more robust and generalizable results. A general class of (inverted) S-shaped probability weighting functions had been proposed; however, the possibility of large differences in probability distortions not only across experimental conditions, but also across individuals, seems critical for the model's success. It also seems advantageous to consider individual differences in parameters of probability weighting as being sampled from weakly informative prior distributions of individual parameter values. Thus, the results from hierarchical Bayesian modeling converge with previous results in revealing that probability weighting parameters show considerable task dependency and individual differences. Methodologically, this work exemplifies the usefulness of hierarchical Bayesian modeling techniques for cognitive psychology. Theoretically, human probabilistic inference might be best described as the application of individualized strategic policies for Bayesian belief revision.

Single-Trial Event-Related Potential Correlates of Belief Updating(1,2,3)

Belief updating—the process by which an agent alters an internal model of its environment—is a core function of the CNS. Recent theory has proposed broad principles by which belief updating might operate, but more precise details of its implementation in the human brain remain unclear. In order to address this question, we studied how two components of the human event-related potential encoded different aspects of belief updating. Participants completed a novel perceptual learning task while electroencephalography was recorded. Participants learned the mapping between the contrast of a dynamic visual stimulus and a monetary reward and updated their beliefs about a target contrast on each trial. A Bayesian computational model was formulated to estimate belief states at each trial and was used to quantify the following two variables: belief update size and belief uncertainty. Robust single-trial regression was used to assess how these model-derived variables were related to the amplitudes of the P3 and the stimulus-preceding negativity (SPN), respectively. Results showed a positive relationship between belief update size and P3 amplitude at one fronto-central electrode, and a negative relationship between SPN amplitude and belief uncertainty at a left central and a right parietal electrode. These results provide evidence that belief update size and belief uncertainty have distinct neural signatures that can be tracked in single trials in specific ERP components. This, in turn, provides evidence that the cognitive mechanisms underlying belief updating in humans can be described well within a Bayesian framework.

P300 and Decision Making under Risk and Ambiguity

Our study aims to contrast the neural temporal features of early stage of decision making in the context of risk and ambiguity. In monetary gambles under ambiguous or risky conditions, 12 participants were asked to make a decision to bet or not, with the event-related potentials (ERPs) recorded meantime. The proportion of choosing to bet in ambiguous condition was significantly lower than that in risky condition. An ERP component identified as P300 was found. The P300 amplitude elicited in risky condition was significantly larger than that in ambiguous condition. The lower bet rate in ambiguous condition and the smaller P300 amplitude elicited by ambiguous stimuli revealed that people showed much more aversion in the ambiguous condition than in the risky condition. The ERP results may suggest that decision making under ambiguity occupies higher working memory and recalls more past experience while decision making under risk mainly mobilizes attentional resources to calculate current information. These findings extended the current understanding of underlying mechanism for early assessment stage of decision making and explored the difference between the decision making under risk and ambiguity.

Higher incentives can impair performance: neural evidence on reinforcement and rationality

Standard economic thinking postulates that increased monetary incentives should increase performance. Human decision makers, however, frequently focus on past performance, a form of reinforcement learning occasionally at odds with rational decision making. We used an incentivized belief-updating task from economics to investigate this conflict through measurements of neural correlates of reward processing. We found that higher incentives fail to improve performance when immediate feedback on decision outcomes is provided. Subsequent analysis of the feedback-related negativity, an early event-related potential following feedback, revealed the mechanism behind this paradoxical effect. As incentives increase, the win/lose feedback becomes more prominent, leading to an increased reliance on reinforcement and more errors. This mechanism is relevant for economic decision making and the debate on performance-based payment.

A computational analysis of the neural bases of Bayesian inference

Empirical support for the Bayesian brain hypothesis, although of major theoretical importance for cognitive neuroscience, is surprisingly scarce. This hypothesis posits simply that neural activities code and compute Bayesian probabilities. Here, we introduce an urn-ball paradigm to relate event-related potentials (ERPs) such as the P300 wave to Bayesian inference. Bayesian model comparison is conducted to compare various models in terms of their ability to explain trial-by-trial variation in ERP responses at different points in time and over different regions of the scalp. Specifically, we are interested in dissociating specific ERP responses in terms of Bayesian updating and predictive surprise. Bayesian updating refers to changes in probability distributions given new observations, while predictive surprise equals the surprise about observations under current probability distributions. Components of the late positive complex (P3a, P3b, Slow Wave) provide dissociable measures of Bayesian updating and predictive surprise. Specifically, the updating of beliefs about hidden states yields the best fit for the anteriorly distributed P3a, whereas the updating of predictions of observations accounts best for the posteriorly distributed Slow Wave. In addition, parietally distributed P3b responses are best fit by predictive surprise. These results indicate that the three components of the late positive complex reflect distinct neural computations. As such they are consistent with the Bayesian brain hypothesis, but these neural computations seem to be subject to nonlinear probability weighting. We integrate these findings with the free-energy principle that instantiates the Bayesian brain hypothesis.