Originally established at UCL in the early 1970’s as a weekly Cognition and Reasoning seminar, it later became an intercollegiate seminar on Language and Cognition in the early 1980’s. The name LJDM was finally coined in 1990, and the group has been running seminars under this name ever since, with lecturers and researchers in and around the UK meeting on a regular basis to discuss judgment and decision-making, judgments of likelihood, reasoning, thinking, problem solving, forecasting, risk perception and communication, and other related topics.
The LJDM seminar series is supported by:
University College London
City, University of London
Kings College London
Unless specified otherwise, all seminars take place on Wednesdays from 5:15-6:15 PM UK time. The seminar series will be held in a hybrid format via Zoom as well as at University College London, unless stated otherwise.
To get updates on the current schedule and weekly reminders of the seminars, please subscribe to the Risk and Decision mailing list and follow our X account. All are welcome to attend.
Titles, abstracts and recordings (where available) of previous seminars can be found here.
If you would like to present your research to the group or to suggest a speaker, please contact the organizers:
- Hadeel Haj Ali (hadil.ali.22@ucl.ac.uk)
- Calvin Deans-Browne (calvin.deans-browne.20@ucl.ac.uk)
Audrey Zhang (zhongyao.zhang.19@ucl.ac.uk)
Academic Year 2024/25
Term 2
Wednesday January 15, 2025
Speaker: Prof David Comerford, Stirling Management School.
Title: Overthinking
Abstract: Conceptually, there can be three types of people: those who overrely on their gut feelings; those who overrely on deliberation; and those who are well calibrated. The first type has been identified by the heuristics and biases research program. The current research investigates the second. I start my search from the following premise: evolutionary psychology posits that gut feelings have evolved to help us navigate the complexities of living in social groups. If this is so, then suppressing these instincts in favour of deliberation ought to be especially costly in the social domain. I hypothesize that individuals with high numeracy scores—reflecting a tendency to engage cognitive effort—will report lower social satisfaction. Analyses of five large national datasets (one preregistered) find that respondents who score in the top decile for numeracy report lower social satisfaction than those who scored lower on numeracy. This effect is robust to controls for sociodemographic characteristics, Big-5 personality traits and verbal intelligence. A preregistered analysis demonstrates the effect is driven by items measuring cognitive reflection. In sum, respondents who demonstrate a tendency to stop and think experience less satisfying social lives.
Time & Location: 17:15-18:15 UK time. Room 541, IoE, 20 Bedford Way.
- Prof. David Comerford will be presenting remotely via zoom (which you can access at the time of the presentation via the Zoom Link below), but all are welcome to come to the assigned room to watch the seminar as usual.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday January 22, 2025
Speaker: Prof Renato Frey, University of Zurich.
Title: Decision making in a world full of risk and uncertainty: From the lab to real life with an integrated modeling framework
Abstract: One of the key goals of the behavioral sciences has been to model how people make decisions under risk and uncertainty, to thus understand and predict consequential real-life behaviors and, if necessary, help people make better choices through pre- and interventions. In this overview talk, I will argue that previous research with this goal might have been limited by adopting (1) a strong focus on measuring and modeling stable interindividual differences in people's risk preferences rather than modeling dynamic psychological mechanisms, and (2) primarily conducting behavioral research in the lab as opposed to in real-life settings. I will illustrate our ongoing efforts to overcome these hurdles, such as mapping the current "ecology of risk", developing a taxonomy of state-specific psychological mechanism, and deploying ecological momentary assessments paired with large language modeling.
Time & Location: 17:15-18:15 UK time. Room G03 in Bedford Way (26).
Zoom: A Zoom option will not be available for this talk.
Wednesday January 29, 2025
Speaker: Prof Pedro Bordalo, University of Oxford.
Title: A cognitive theory of reasoning and choice
Abstract: We offer a theory of decisions in which attention to the features of choice options is determined by the DM's categorization of the current choice problem in a set of problems she solved in the past. Categorization depends on goal-relevant as well as contextual problem-level features. The model yields systematic heterogeneity in attention and choice in a given problem based on different past experiences, rigidity of choices when categorization does not change despite new data, and discontinuous shifts when changes in bottom-up salient features cause re-categorization. The model unifies major puzzles and framing effects in riskless, statistical, and lottery choice based on heterogenous and unstable mental representations.
Time & Location: 17:15-18:15 UK time. Room G.03, 26 Bedford Way.
Zoom Link: A Zoom option will not be available for this talk.
Wednesday February 5, 2025
Speaker: Dr Manos Konstantinidis, University of Warwick.
Title: Decision under risk: A tale of two sources of information
Abstract: Information about risky or uncertain situations typically comes in two forms. In some cases, we have access to complete descriptive summaries of risky outcomes and their associated probabilities (e.g., a physician may rely on published literature and success likelihoods to recommend a risky course of treatment). In other situations, we rely on our own experiences to form estimates of outcomes and probabilities (e.g., a physician drawing on their own prior experience with a particular treatment). Real-life decision-making often involves integrating both descriptive summaries and personal experience. However, previous research in risky choice has largely studied these sources separately. In this talk, I will present empirical and computational modelling results that explore how descriptive and experiential information are combined in studies of risky decision making. I will also discuss ongoing empirical work investigating whether people naturally prefer one type of information over the other (description vs. experience) and the factors that moderate this preference. Finally, I will address the theoretical, practical, and computational implications of these findings.
Time & Location: 17:15-18:15 UK time. Room 728, IoE, 20 Bedford Way.
Zoom Link: A Zoom option will not be available for this talk.
Wednesday February 12, 2025
Speaker: Dr Greta Mohr, University College London.
Title: Biases in Causal Reasoning: The Perceived Influence of Good and Bad Actions in Everyday Medical Decision-Making
Abstract: Causal counterfactual reasoning is fundamental to everyday medical decision-making, where individuals must evaluate whether specific actions (e.g., gentle yoga to alleviate muscle soreness) contribute to health outcomes. This study investigates how performing good or bad actions, their consistency with each other, and alignment with the outcome influence causal strength judgments. Participants were presented with everyday medical scenarios where agents took 2 actions (good or bad) and judged how much each action contributed to a health outcome (improvement or worsening). They then made a series of counterfactual judgements, were then asked to rate how the outcome may have differed if the actions taken had been different.
Results revealed systematic biases in causal attribution. Actions consistent with the outcome but inconsistent with the other action were perceived as having stronger causal strength, than those with consistent other actions. Additionally, bad actions (e.g., skipping hydration to alleviate a headache) were judged as having greater causal strength than good actions, though this effect was more pronounced for the second action. Notably, when both actions were taken bad actions were perceived as contributing more to the outcome than good actions. Overall, these findings suggest a bias in causal reasoning, where harmful or unexpected actions are perceived as more influential in determining health outcomes. This bias could lead people to overemphasise the negative consequences of bad actions while underestimating the benefits of good decisions. Recognising these biases can help improve public health messaging and encourage more accurate causal reasoning in everyday healthcare decisions.
Time & Location: 17:15-18:15 UK time. Room C3.12, IoE, 20 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday February 19, 2025 - Reading Week - No seminar
Wednesday February 26, 2025
Speaker: Dr Kevin O'Neill, University College London.
Title: TBA
Abstract: TBA
Time & Location: 17:15-18:15 UK time. Room G.03, 26 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday March 5, 2025
Speaker: Dr Jason Burton, Copenhagen Business School.
Title: TBA
Abstract: TBA
Time & Location: 17:15-18:15 UK time. Room 822, IoE, 20 Bedford Way.
- Dr Jason Burton will be presenting remotely via zoom (which you can access at the time of the presentation via the Zoom Link below), but all are welcome to come to the assigned room to watch the seminar as usual.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday March 12, 2025
Speaker: Dr Jens Koed Madsen, London School of Economics.
Title: A view of polarisation built on dependencies
Abstract: Issue polarisation describes a state where citizens in a society move away from each other on a particular issue – for example, people may harden their views in both directions on abortion rights, economic policy, and more. In this talk, we first consider whether issue polarisation is occurring, as this is a subject of debate within the field. Across 247 samples from 105 countries, we show that issue polarisation is a contextual challenge. That is, while there is no global increase from 1999 to 2022, we see an increase in issue polarisation in some countries, such as the USA (typically driven by disagreement between culturally-conservative majorities and culturally-liberal minorities).
Having set the scene, we present a Bayesian model that builds on perceived dependencies. According to the model, when presented with conflicting testimony from two source groups, Bayesian agents should update towards the position of the group they deem to be more independent in terms of the factors that influence their testimony, meaning those who disagree about which group that is should polarize. In a pre-registered experiment, we find support for this model, as we can experimentally create polarisation by manipulating dependency perceptions.
To see if the intuitions of the model are appropriate, we survey citizens from the UK and the USA to gauge their dependency perceptions of in- and out-group members. In both countries, we find that, using a novel scale instrument, real-world partisans (Labour, Conservative, Republican, and Democrat) perceive their party’s supporters to be more independent than the opposing party’s supporters, with large average effect sizes (d = 0.87 UK, d = 0.82 US), suggesting the conditions are in place for such polarization to occur in the real world. We conclude the talk by considering limitations to the current work and future directions for the polarisation project.
Time & Location: 17:15-18:15 UK time. Room G03 in Bedford Way (26).
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday March 19, 2025
Speaker: TBA
Title: TBA
Abstract: TBA
Time & Location: 17:15-18:15 UK time. Room 822, IoE, 20 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday March 26, 2025
Speaker: Dr Elinor Abado, University of Göttingen
Title: TBA
Abstract: TBA
Time & Location: 17:15-18:15 UK time. Room 780, IoE, 20 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498
Term 1
Wednesday September 25, 2024
Speaker: Prof Benedetto De Martino, University College London.
Title: Value, Goals and Abstractions in Decision Making
Abstract: In the fields of neuroscience and machine learning, “value” has often been used interchangeably with “reward,” emphasising its hedonic aspect while neglecting its functional, concept-like nature. However, in real life, we rarely receive discrete numerical rewards at the end of our actions and need to build our own values. In this talk, I will present some work from our lab that challenges the canonical view that equates value and reward by showing how the brain is able to construct the value of an option or action on-the-fly by building abstractions that flexibly adapt to changing goals. I will demonstrate how this mechanism can facilitate learning and foster generalisation. Overall, I will discuss how rethinking the true nature of value is both feasible and necessary.
Time & Location: 17:15-18:15 UK time. Room 421, Roberts Building.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday October 2, 2024
No seminar this week. Our next seminar will be next week on Wednesday October 9th 2024.
Wednesday October 9, 2024
Speaker: Dr Konstantinos Tsetsos, University of Bristol.
Title: The timescale and form of context-dependence during human value-learning
Abstract: The way humans and other animals represent the values of alternatives can be systematically distorted by the presence of inferior or unavailable alternatives in the immediate choice-set (immediate context); or by the values of alternatives that were encountered in previous episodes (temporal context). Yet, the extent to which the immediate and temporal context (co-) shape such context-dependent value coding remains unclear. I will present data from experiments where we asked human participants to learn the reward values associated with three alternatives and to explicitly report their learned values before making binary and ternary choices. Context-dependent value distortions appeared in the pre-choice reports and were indistinguishable in binary (inferior decoy absent) and ternary (inferior decoy present) choice trials. These findings suggest that value representations are modulated by the temporal, rather than the immediate, context. Interestingly, the form of distortion cannot be explained by existing value normalization theories but is best captured by a mechanism that constructs value through memory-based binary comparisons.
Time & Location: 17:15-18:15 UK time. Room 421, Roberts Building.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday October 16, 2024
Speaker: Dr Tim Mullett, Warwick Business School.
Title: Attentional Dilution in Multi-Attribute Choice
Abstract: Many choices require comparing complex options with multiple attributes. In such choices, integrating the information together requires attending to each attribute in turn, and distributing the total attention time in an efficient manner. One potential approach is attentional dilution. This describes a mechanism whereby any attributes with the same value for all available options are non-differentiating and therefore receive zero attention other than that to determine they are equal. Attention is instead split between only those attributes that differentiate between the available options in a given choice. The consequence of this is that in choices with only one, or few differentiating attributes, attention is concentrated, and in choices with many differentiating attributes, attention is diluted. A series of experimental results show that the influence of an attribute upon choice is significantly larger when it is the only, or one of few differentiating attributes. This influence reduces as there are more differentiating attributes, in proportion to that predicted by attentional dilution. Eye-tracking results also show that this pattern is mirrored in the proportion of visual attention given to each attribute as the number of differentiating attributes changes.
Time & Location: 17:15-18:15 UK time. Room 421, Roberts Building.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday October 23, 2024
Speaker: Dr Tom Reader, London School of Economics.
Title: Stakeholder safety action: patients and families intervening to prevent medical errors in hospitals.
Abstract: Research in healthcare finds that patients and families frequently intercede in the decision-making and work of clinicians in order to prevent adverse events (e.g., wrong diagnosis). This is significant for research on safety and risk, because it suggests that external stakeholders (e.g., service-users) can act as form of ‘safety net’ for catching errors and addressing problems in decision-making within organisations. I explored this idea by theorizing the concept of ‘stakeholder safety action’: this relates to stakeholders engaging in voicing (e.g., speaking-up) and correcting (e.g., fixing mistakes) behaviours in order to resolve perceived errors within an organisation and prevent them from causing harm. I investigated safety action by undertaking a quantitative and qualitative analysis of 1,857 written complaints sent by patients and families to UK hospitals about experiences of unsafe treatments. The analysis found patients routinely reported engaging in voicing and correcting behaviours, with these being elicited by concerns about decision-making (e.g., in making diagnoses, dispensing medicines) being missed or not resolved by staff. Through these behaviours, and depending on how staff responded to them, patients described trying to prevent accidents in three escalating ways: helping staff to avoid errors (e.g., during diagnoses), pushing into clinical work to resolve perceived mistakes (by changing medications), and bypassing teams and hospitals judged irredeemably unsafe. I surmise that stakeholder safety action contributes to organizational safety by acting as an important – yet external and often unrecognised – source of resilience within organisations.
Time & Location: 17:15-18:15 UK time. Room 421, Roberts Building.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday October 30, 2024
Speaker: Dr Nicolette Sullivan, London School of Economics.
Title: Contextual sensitivity of pro-social behavioral associated with neural response to giving
Abstract: One pillar of behavioral economics is that context matters. The key support for this pillar revolves around hundreds of choice experiments, whereby individuals make choices over economically-identical sets that only vary the context. We move this literature in a new direction by exploring whether results from fMRI neuroimaging can predict which individuals are responsive to contextual changes. Our study revolves around the motivations for prosocial behavior. We find that neural responses in several cognitive control and valuation regions, estimated during a standard dictator game task, are predictive of the magnitude of shifts in altruistic giving in modified dictator game tasks. This finding demonstrates how neural data can identify individual differences in processes that underlie observed altruism related to specific theories of altruistic behavior. Importantly, we find that activation in the DLPFC, mPFC, striatum, ACC, and insula distinguish those who continue to give across multiple contexts that differ in demands on the giver, from those whose giving is sensitive to contextual modulation and may be driven by impure altruistic motivations.
Time & Location: 17:15-18:15 UK time. Room 421, Roberts Building.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday November 6, 2024 - Reading Week - No talk
Wednesday November 13, 2024
Speaker: Dr Claire Heard, King’s College London.
Title: Understanding Scam Susceptibility: The Role of Affective Content in Persuasive Messages
Abstract:
Fraud, accounting for approximately 40% of all crime in England and Wales, presents a significant societal issue with an estimated 3.2 million offences annually, resulting in an approximate cost of £6.8 billion (Home Office, 2024). This crime not only has profound financial implications but also inflicts substantial emotional distress on victims. Despite these severe impacts, the current body of research provides an incomplete picture of the factors influencing scam susceptibility. Prior studies have produced mixed results regarding the role of demographic characteristics and personality variables (e.g., age, risk-taking propensity) in scam susceptibility.
Furthermore, the influence of specific scam design elements on an individual’s likelihood of victimisation is more limited. Initial investigations in this domain have examined persuasive elements within scams, such as scarcity cues (“don’t delay, call now to claim your prize”) and authority indicators (Wood et al., 2018). However, these studies have not conclusively established their impact, instead such studies emphasise the role of risk and benefit perceptions in likelihood of responding.
Reflecting on theoretical research that highlights the key role that affect can play in perception of risk and benefit perceptions (Slovic et al. 2011), as well as research highlighting how affect can lead to heuristic processing and suboptional decision making in persuasive contexts (Kircanski et al., 2018), this study extends the existing literature by investigating the influence of increasing the affective content of scam-style messages on individuals’ intentions to respond.
In an online experimental study involving 280 participants, we randomly assigned participants to receive either a low affect or high affect message. Participants rated their response intentions (1 question), perceptions of the message (14 items), and their perceptions of the associated risks (2 questions) and benefits (2 questions). Given that scams can evoke different emotional valences – either positive (joy/excitement) or negative (fear/worry), each participant was exposed to, and answered the above questions for two different messages presented in their assigned affect level (low or high). This consisted of a lottery scenario (e.g. you are the winner of our prize giveaway.... call to claim your winnings) for the positive valence scenario and a bank fraud scenario (e.g. Your debit card was recently used..... get in touch as soon as possible if you didn't make this purchase) for the negative valence scenario. We also collected a range of exploratory individual difference and demographic variables.
Our results indicate that exposure to high-affect messages does increase participants’ response intentions. However, this was only in the positive emotion (lottery) scenario, not in the negative emotion (bank) scenario. These findings suggest that the affect level of a scam message could play a critical role in influencing susceptibility to scams. However, the impacts may not be universal and may depend on the specific message presented. These findings underscore the need for future research into the influence of affect on susceptibility to scam messages and call for investigations considering a wide range of message types to gain a nuanced understanding of how people react to different types of messages.
Time & Location: 17:15-18:15 UK time. Room 780, IoE, 20 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday November 20, 2024
Speaker: Dr Tim Rakow, King’s College London.
Title: A programmatic investigation of the possible impulses that drive binary choices with full feedback
Abstract: Building on research exploring repeated risky choice and strategic decisions, we tested what constitutes a sufficiently satisfactory outcome to prompt people to repeat their actions. Our programmatic investigation (total N > 1400) examined incentivised choices between pairs of binary-outcome options, with immediate feedback for obtained and forgone outcomes. Across experiments, we varied the payoff structure for the possible outcomes, whereby (1) both options had identical possible outcomes; or (2) one option was safer than the other, e.g., possible outcomes {2,3} vs. {1,4}; or (3) the possible outcomes from the two options were ‘interleaved’, i.e., {4,2} vs. {3,1}. Within experiments, we manipulated the probabilities associated with each outcome, either across choice tasks (Experiments 1-3) or unannounced within a task (Experiments 4-6). These designs allowed us to compare the roles of implied disappointment (receiving the lower outcome from an option), implied regret (receiving a lower outcome than the forgone payoff), and the relative average value of the options, as possible drivers of choice. Across all experiments and most tasks and conditions, implied regret exerted a stronger influence on subsequent choices than implied disappointment, the ratio of options’ average values, or the absolute magnitude of losses or gains. Moreover, the explanatory power of ‘mere regret’ was good – it only minimally improved the prediction of choice to take account of the size of regret over and above coding which option ‘won’ on a given round. Our findings illustrate how responding to implied regret might help people choose well and can also indicate which potential inputs are more (or less) important for descriptive models of repeated choice.
Time & Location: 17:15-18:15 UK time. Room W2.06, IoE, 20 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday November 27, 2024
Speaker: Prof Ulrike Hahn, Birkbeck, University of London.
Title: What Rational Agent-Based Models Tell Us About Social Networks
Abstract: The talk will review work using agent-based models with rational actors to model communication across social networks. This illustrates the depth of the challenge posed by testimony and our attempts to gauge the reliability of our sources. It highlights ways in which the accuracy of our beliefs is fundamentally beyond our (individual) control, and draws out implications for the (re)design of online social networks.
Time & Location: 17:15-18:15 UK time. Room 731, IoE, 20 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498
Wednesday December 4, 2024
No seminar this week. Our next seminar will be next week on Wednesday December 11th 2024.
Wednesday December 11, 2024
Speaker: Prof Sophie Scott, University College London.
Title: What's the point of studying laughter
Abstract: In this talk I will consider evolutionary, psychological and neuroscientific evidence about the importance and role of laughter in humans and other animals. I will argue that studying laughter allows us to explore an intensely social positive emotional expression, and that laughter is a highly complex behaviour in humans. I will also draw some distinctions with humour.
Time & Location: 17:15-18:15 UK time. Room 639, IoE, 20 Bedford Way.
Zoom Link: https://ucl.zoom.us/j/99605138498