Beliefs & Action Selection

Believing in conspiracy theories has often been found to be associated with having incoherent beliefs, although different measures of association and of incoherence have led to somewhat varied findings. Here we formalise pairs of incompatible conspiracy beliefs from the literature as logical contraries. Contraries cannot both be true but they can both be false, the latter being the case e.g. when the official version of events holds (e.g. "Diana faked her own death vs. "Diana was killed by the British secret service"). We operationalise coherence as probabilistic consistency in a Bayesian framework, and measure extent of incoherence as the Cartesian distance to the nearest response that would have been coherent. These measures are arguably more precise and less susceptible to confounds than previous proposals. Using these measures, we reanalyse the data from 8 studies involving 8590 participants through a series of multilevel generalised linear mixed models. We complement this reanalysis with a new experiment featuring neutral in addition to conspiratorial content, and contradictions in addition to contraries. Contradictory items can neither both be true nor both be false (e.g. "Diana faked her own death" vs. "Diana did not fake her own death"). Our overall findings replicate the positive association between beliefs in conspiracy theories and incoherent beliefs - both in terms of the relative frequency of incoherence and in terms of the distance from the nearest coherent response. We discuss potential implications for ways of conceptualising and addressing conspiracy beliefs.

This is an in-person presentation on July 20, 2026 (15:00 ~ 15:20 EDT).

Human decision-making often involves conflicts between competing evaluative processes that generate distinct action preferences. In this talk, I propose self-deception as a product of such conflict and demonstrate how the emergence of self-deception provides a unified framework for explaining a wide range of everyday and psychopathological behaviors traditionally addressed by separate accounts. The core idea is that when a process’s preferred action is rejected due to conflict with other processes, it can still influence choices by promoting attentional biases that shape the agent's beliefs and thereby manipulate the agent into selecting the process’s preferred action. Consequently, over time, the process will learn to assign value to such attentional biases. Across three simulations, I illustrate how such dynamics lead agents to value biased beliefs, maintain uncertainty when it is instrumental, and suppress bias when it becomes maladaptive. I then shortly discuss the behavioral implications of the model and show how it provides a parsimonious account of phenomena such as optimism, self-handicapping, procrastination, and depression- and bipolar-like behavioral dynamics. By integrating computational theory with clinically relevant constructs, the model offers new theoretical insight into the adaptive and pathological functions of self-deception.

This is an in-person presentation on July 20, 2026 (15:20 ~ 15:40 EDT).

The field of Human Factors and Ergonomics (HFE) has traditionally focused on studying how people interact with tools and tasks and designing systems that improve usability, performance, and reliability. As currently practiced, HFE is narrowly focused on short-term business goals, product usability, and user performance while ignoring environmental and societal externalities such as resource depletion, waste, and negative effects on education, health, or labor (e.g., planned obsolescence or optimizing for sales rather than longevity or reuse). Some theorists and practitioners (e.g., Monteiro, Norman, Thatcher) have called for transforming HFE toward a more ethically responsible and inclusive discipline. The field of Mathematical and Computational Psychology (MCP) has traditionally developed formal models of how the human mind works. Some of these models have been used to inform HFE theories and practices. In turn, HFE has been used to validate MCP models. The two areas have complemented each other and co-evolved. However, they might have also constrained each other. For example, since MCP models tend to favor parsimony, experimental control, and domain-independent formalisms, they might have inadvertently contributed to the myopic focus of HFE applications. Conversely, the HFE’s neglect of societal and ethical factors might have deprived MCP of opportunities to expand its scope. Moving forward, there is potential for both MCP and HFE to synergistically expand their scope and contribute to the goals of resilience and sustainability for all of humanity, through, e.g., modeling adaptive capacity and resilience in dynamic and complex systems, designing for long-term horizons, energy efficiency, and ecosystem integrity.

This is an in-person presentation on July 20, 2026 (15:40 ~ 16:00 EDT).