Search

The current research proposes a novel approach to identity-based choice that focuses on consumers' representations of the self-concept, as captured by the perceived cause–effect relationships among features of an individual consumer's self-concept. More specifically, the studies reported here test the proposal that the causal centrality of an identity—the number of other features of a consumer's self-concept that the consumer believes influenced or were influenced by the identity—underlies identity importance and is a determinant of identity-based consumer behaviors. Across seven studies, using both measured and manipulated causal centrality, the current research provides evidence for the role of causal centrality in identity-based choice. Among consumers who share an identity (belong to the same social category), those who believe that the identity is more causally central perceive the identity as more important and are more likely to engage in behaviors consistent with the norms of the social category. [ABSTRACT FROM AUTHOR]

Background: The primary cause of end-stage renal disease (ESRD) is diabetic nephropathy (DN), and a growing body of research indicates that immunology plays a part in how DN develops into ESRD. Our objective is to identify causal relationships between various immune invading cells and DN to identify possible targets for immunotherapy. Methods: This study used a complete Mendelian randomization (MR) analysis with two samples to identify the underlying mechanism linking immune cell characteristics with DN. Using publicly available genetic data, we investigated the causal link between 731 immune cell profiles and DN risk. Included were four different types of immune systems: morphological parameters (MP), absolute cell (AC), relative cell (RC), and median fluorescence intensities (MFI). The results' robustness, heterogeneity, and horizontal pleiotropy were confirmed through extensive sensitivity analysis. Results: Following FDR (False Discovery Rate correction method) correction, no statistically significant differences were observed; however, six immunophenotypes were shown to be significantly associated with DN risk at the 0.25 level. Only CD28+ CD4− CD8− T cells were identified as the protective immunophenotype (OR = 0.588, 95% CI 0.437–0.792, P = 4.71 × 10−4). Moreover, DN had no discernible impact on immunophenotyping after FDR correction. Surprisingly, three unadjusted phenotypes with low P values were discovered to be positively correlated with the risk of DN: CD20 on IgD− CD27− B cells (OR = 1.263, 95% CI 1.076–1.482, P = 4.22 × 10−3), CD8 on naive CD8 + T cells with Effector Memory (OR = 1.107, 95% CI 1.013–1.209, P = 2.40 × 10−2), and CD8 on Effector Memory CD8 + T cells (OR = 1.126, 95% CI 1.024–1.239, P = 1.46 × 10−2). Conclusions: Our findings provide a genetic basis for the association between immune cells and DN and should inform future clinical research. [ABSTRACT FROM AUTHOR]

This study examined how Puerto Rican parents living in the United States reason about illnesses and remedies. Special focus is placed on the influence of Puerto Rican cultural, religious, and spiritual beliefs due to the influence of Taino, West African, Spanish, and U.S. beliefs and customs. Parents of 3.5 to 12-year-old children from the Rochester, NY area were asked to complete a questionnaire centered on their beliefs and an open-ended interview that explored how one could get sick and treated for biological illnesses (e.g., the common cold, COVID-19, cancer). Although participants reported religious, spiritual, and mystical beliefs in the questionnaire, they provided scientific explanations for illness and remedies in the interview. This study provides novel findings on how individuals of Puerto Rican descent understand illnesses and remedies and the variation across individuals. This study provides strong evidence for the variation in coexistence of beliefs within this population. [ABSTRACT FROM AUTHOR]

This article investigates people's judgments of actual causation in the context of a previously neglected property of causal structures—their reversibility, that is, whether an effect persists or returns to its original state if its causes are removed. Causal reversibility, and its potential impact on causal judgment, was recently analyzed theoretically by Ross and Woodward (2022). They hypothesized that reversibility might affect people's evaluation of causes in late-preemption scenarios. The typical finding in preemption scenarios is that events happening earlier are considered to be actual causes, while events happening later are regarded as noncauses. The hypothesis is that this robust intuition depends on causal reversibility and that in reversible structures later events are regarded as actual causes. Across three main experiments and one supplementary study (N = 590), it is shown that reversibility has the predicted effect: later causes are perceived to make an actual causal contribution to the effect. It is also shown that Henne et al. (2023), in a first study, did not find evidence for Ross and Woodward's hypothesis because they did not test whether people regard later causes in preemption-like sequences of reversible structures as maintainers and not as triggers of their effect. Because they used test questions that asked explicitly for triggering rather than maintaining or were at least ambiguous, their results seemed to show that people think that later events have no causal impact. Maintaining is a relevant causal concept deserving more attention in both philosophical theories and psychological studies on causal cognition. [ABSTRACT FROM AUTHOR]

The literature shows that an individual's moral valence affects the causal judgements of social observers. The effect is considered by some to be a cognitive bias. This article challenges this notion by investigating the relationship between how facts are presented to participants and their interpretation of causal judgment tasks, whether as a search for moral responsibility or a strict causal link. To do this, we proposed the first categorisation of the causal models of the facts described (i.e. the arrangement of causes presented to the participants). Through a systematic review, our analysis suggests an alternative hypothesis to the cognitive bias view: the arrangement of the facts presented to the participants could influence the relevant interpretation they make of the causal judgement task. Specifically, the causal models predominantly used in the literature may prompt an interpretation focused on attributing blame rather than identifying a narrow causal link. [ABSTRACT FROM AUTHOR]

Large language models (LLMs) have shown their capabilities in numerical and logical reasoning, yet their capabilities in higher-order cognitive tasks, particularly causal reasoning, remain less explored. Current research on LLMs in causal reasoning has focused primarily on tasks such as identifying simple cause-effect relationships, answering basic "what-if" questions, and generating plausible causal explanations. However, these models often struggle with complex causal structures, confounding variables, and distinguishing correlation from causation. This work addresses these limitations by systematically evaluating LLMs' causal reasoning abilities across three representative scenarios, namely analyzing causation from effects, tracing effects back to causes, and assessing the impact of interventions on causal relationships. These scenarios are designed to challenge LLMs beyond simple associative reasoning and test their ability to handle more nuanced causal problems. For each scenario, we construct four paradigms and employ three types of prompt scheme, namely zero-shot prompting, few-shot prompting, and Chain-of-Thought (CoT) prompting in a set of 36 test cases. Our findings reveal that most LLMs encounter challenges in causal cognition across all prompt schemes, which underscore the need to enhance the cognitive reasoning capabilities of LLMs to better support complex causal reasoning tasks. By identifying these limitations, our study contributes to guiding future research and development efforts in improving LLMs' higher-order reasoning abilities. [ABSTRACT FROM AUTHOR]

Analogical studies demonstrate that participants often fail to retrieve a well-learned base analog during the subsequent processing of a semantically-distant target analog. We evaluated whether presenting the target analog before the base analog increases analogical retrieval during hypothesis-generation. Experiment 1 revealed a higher rate of analogical retrieval when the target analog preceded the base analog, as compared to the traditional “base-target” sequence. Using a factorial design, Experiment 2 assessed whether spontaneously acknowledging the relevance of a subsequently encountered explanation for resuming a failed explanatory attempt requires the presence of structural similarities between the base and target situations. Results demonstrated that the primary contributor to spontaneous reactivation of a failed explanatory attempt is the presentation of an analogous phenomenon, while the presence of a useful explanation alone did not yield a significant impact. These findings contribute valuable insights to the dynamics of analogical retrieval and offer relevant implications for educational strategies.

Reducing inequality, mitigating climate change, and responding to public health crises are large-scale goals that require the cooperation and coordination of many individuals. These goals cannot be achieved by one individual alone, and contributing is not always beneficial to each individual. And yet, individuals must contribute in order to make a difference. How do we hold individuals and groups responsible for collective action?

Outward differences between cultures are very salient, with Western and East Asian cultures as a prominent comparison pair. A large literature describes cross-cultural variation in cognition, but relatively less research has explored the developmental origins of this variation. This study helps to fill the empirical gap by replicating four prominent findings documenting cross-cultural differences in children's reasoning, visual attention, and social cognition in a cross-sectional sample of 240 3-12-year-olds from the US and China. We observe cross-cultural differences in three of the four tasks and describe the distinct developmental trajectory that each task follows throughout early and middle childhood.

Comparison of visibility inferences across congenitally blind and sighted people provides insight into the contribution of first-person sensory experience to intuitive theories. We hypothesized that both groups understand others' visual experiences via an intuitive theory incorporating variables known to influence visual psychophysics (distance, looking duration, and feature size). Adults born blind (n=20) and sighted (n=40) listened to short scenarios that described an observer looking at another person from different distances and for varying durations. Participants rated how likely the observer would perceive appearance features of the person that varied in size (e.g., eye color vs. hat). A probabilistic formalization of intuitive visibility fit the ratings with high accuracy across scenarios and features. Model parameters were qualitatively identical across groups but blind adults weighted distance and size less. A quantitative and generative intuitive theory of vision develops without first-person sensory access, possibly through linguistic communication, and is fine-tuned by visual experience.

As AI systems like language models are increasingly integrated into decision-making processes affecting people's lives, it's critical to ensure that these systems have sound moral reasoning. To test whether they do, we need to develop systematic evaluations. We provide a framework that uses a language model to translate causal graphs that capture key aspects of moral dilemmas into prompt templates. With this framework, we procedurally generated a large and diverse set of moral dilemmas---the OffTheRails benchmark---consisting of 50 scenarios and 400 unique test items. We collected moral permissibility and intention judgments from human participants for a subset of our items and compared these judgments to those from two language models (GPT-4 and Claude-2) across eight conditions. We find that moral dilemmas in which the harm is a necessary means (as compared to a side effect) resulted in lower permissibility and higher intention ratings for both participants and language models. The same pattern was observed for evitable versus inevitable harmful outcomes. However, there was no clear effect of whether the harm resulted from an agent's action versus from having omitted to act. We discuss limitations of our prompt generation pipeline and opportunities for improving scenarios to increase the strength of experimental effects.

Functional relations are prevalent in everyday life and science. Do children have intuitive knowledge of functional relations, and can they learn these relations by active information gathering (i.e., choosing a few input values and observing the corresponding outputs)? We found that 6- to 9-year-olds can learn different families of functions (linear, Gaussian, and exponential) through both informative data provided by an experimenter and data they gather from the environment for themselves. Overall, children learn linear functions more accurately than non-linear functions. When choosing data points to learn about, some children select highly similar points that only shed light on a narrow region of a function, while others choose more variable inputs and gain a more holistic view of a function. Children who use this latter, globally informative strategy have higher learning accuracy, particularly for non-linear functions. Results suggest that children are in the process of developing effective strategies for active function learning.

Prior studies of Theory of Mind have primarily asked observers to predict others' actions given their beliefs and desires, or to infer agents' beliefs and desires given observed actions. However, if Theory of Mind is genuinely a causal theory, people should also be able to plan interventions on others' mental states to change their behavior. The intuitive causal model of Theory of Mind predicts an asymmetry: one has to instill both the relevant belief and desire to cause an agent to act; however, to prevent a likely action, it suffices to remove either the relevant belief or desire. Here, we use these asymmetric causal interventions to probe the structure of Theory of Mind. In Experiments 1 and 2, both adults (N=80) and older children (N=42, 8-10 years) distinguished generative and preventativecases: selecting interventions on both mental states (both belief and desire) to induce an agent to act and just one of the mental states (either belief or desire) to prevent an action. However, younger children (N =42, 5-7 years) did not. To probe this age difference, in Experiment 3, we asked younger children(N=42, 5-7 years) just to predict the outcome of others' mental state interventions. Children predicted that interventions were more likely to prevent actions than to cause them, but failed to predict that intervening on both the relevant beliefs and desires is more likely to generate a novel action than intervening oneither alone. These findings suggest that by eight to ten years old, people represent the causal structure of Theory of Mind and can selectively intervene on beliefs and desires to induce and prevent others' actions.

In this paper, we explored the attribution of causal responsibility in a causal chain of events, where an agent A instructs an intermediate agent B to execute some harmful action which leads to a bad outcome. In Study 1, participants judged B to be more causally responsible, more blameworthy, and more deserving of punishment than A. In Study 2, we explored the effect of proximity on judgments of the two agents by adding a third, subsequent contributing cause, such that B's action no longer directly caused the final outcome. Participants judged both agents A and B to be less causally responsible and deserving of punishment (but not less blameworthy) when they were less proximal to the outcome, and there were no differences in judgments between the two agents. In Study 3, we varied whether each of the two agents (A and B) intended for the final outcome to occur. We find an interaction between role and intent, where participants only mitigated judgments for A when A did not intend for the outcome to occur – regardless of B's intent. We discuss possible explanations for our findings and its implications for moral and legal decision-making.

Adults often hold different goals for children's achievement: Sometimes adults want children to learn as much as possible, while at other times adults discount children's learning in favor of high performance. How do children reason about the achievement goals adults have for them? Across 3 preregistered studies (n = 120), we asked whether 5- and 6-year-old children understand the causal relationship between adults' achievement goals, their task choices, and children's competence. In Experiment 1, we found adults are more likely to give harder tasks to children when they hold learning versus performance goals and when the child is more competent. In Experiment 2, we found that children make similar inferences about adults' task selections given the adult's achievement goal and the receiving child's competence. Finally, in Experiment 3, children inferred that adults would pick harder tasks for them when they possessed a learning goal versus a performance goal, which matched their own task choice given the same achievement goals. Thus, young children can infer the relationship between adults' child-directed achievement goals and actions and may use this information to learn about what adults prioritize for children across contexts.

Independent pieces of corroborating evidence should provide stronger support to a hypothesis than dependent pieces of evidence. Overlooking the inferiority of dependent relative to independent items of evidence can lead to a chain reaction of double-counting evidence, over-estimating the probability that the fact under consideration is true, and making wrongful decisions. Within fictitious scenarios, we investigate people's sensitivity to the independency advantage. We assess their ability to integrate multiple items of evidence that come from (in)dependent sources who differ in reliability. We find that participants properly perceive dependencies when explicitly asked but fail to distinguish the probative value of dependent versus independent evidence in their belief updating. Still, individuals who perceive a strong dependence between sources treat the evidence as being more redundant. We find no dependency-related effects on participants' individual Bayesian network model predictions. Potential reasons why participants perceive (in)dependencies and yet (mostly) fail to discount for them are discussed.

People often select only a few events when explaining what happened. What drives people's explanation selection? Prior research argued that people's explanation choices are affected by event normality and causal structure. Here, we propose a new model of these existing findings and test its predictions in a novel experiment. The model predicts that speakers value accuracy and relevance. They choose explanations that are true, and that communicate useful information to the listener. We test the model's predictions empirically by manipulating what goals a listener has and what actions they can take. Across twelve experimental conditions, we find that our model accurately predicts that people like to choose explanations that communicate optimal interventions

When a human and an AI agent collaborate to complete a task and something goes wrong, who is responsible? Prior work has developed theories to describe how people assign responsibility to individuals in teams. However, there has been little work studying the cognitive processes that underlie responsibility judgments in human-AI collaborations, especially for tasks comprising a sequence of interdependent actions. In this work, we take a step towards filling this gap. Using semi-autonomous driving as a paradigm, we develop an environment that simulates stylized cases of human-AI collaboration using a generative model of agent behavior. We propose a model of responsibility that considers how unexpected an agent's action was, and what would have happened had they acted differently. We test the model's predictions empirically and find that in addition to action expectations and counterfactual considerations, participants' responsibility judgments are also affected by how much each agent actually contributed to the outcome.

The control-of-variables strategy is often considered to be the superior strategy when children learn from experiments. However, by simulating Bayesian likelihoods of outcomes from a water displacement task, we show that certain confounded comparisons may support belief revision better than controlled comparisons. We tested this assumption by experimentally varying the types of comparisons that participants observed in a learning task involving balls of different sizes and materials (N = 90, age range 6- to 9-yrs). In the Size, Material, and Mixed conditions we presented controlled comparisons. In the Confounded Incongruent Condition, we presented confounded comparisons in which the larger ball was made of the heavier material. In line with our hypotheses, children in the Confounded Incongruent Condition revised their beliefs more than children in the other conditions, as indicated by higher transfer test scores. These findings suggest that confounded comparisons may in fact sometimes provide more optimal information for learning.

Attribution theory aims to explain people's judgments about the cause of some behavior or outcome, often involving other people. The theory has proven to be broadly applicable and points towards important aspects of human cognition. This relevance is perhaps unsurprising given that attribution theory is a type of causal inference. However, there has been relatively little work on attribution theory in relation to causal learning. More specifically, previous literature has mostly examined attributions and their behavioral and motivational outcomes following a single observation, rather than capturing the dynamics of causal attribution (i.e., how those judgments shift as people observe more vignettes and thereby learn about the situation). We thus ran an exploratory study using a vignette design to investigate whether attributions and their outcomes change across multiple instances of observation and behavior adaptation.

Building on counterfactual theories of causal-selection, according to which humans intuitively evaluate the causal responsibility of events, we developed an experimental paradigm to examine the effect of causal-selection explanations on abductive causal inference. In our experiment, participants attempted to infer the rule responsible for winning outcomes of random draws from urns with varying sampling probabilities.Participants who were provided with causal-selection judgments as explanations for the outcomes made significantly closer inferences to the rule than those relying on observations alone, or on other explanations of causal relevance.We mirror these empirical results with a computational model of inference from explanation leveraging the theories of causal selection.

We propose that kinds relate to particular things either constitutively or contingently. Taxonomic categories of animals and artifacts constitutively relate their members: DOG and CAR group things by aspects of the forms of their matter; the forms that make them things instead of stuff. Categories of things in roles or with diseases contingently relate to their members: LAWYER and DIABETIC group things by forms other than the forms that make them things. We confirm this distinction in five experiments with American adults.

Recent years have seen an explosion of theoretical interest, as well as increasingly fraught real-world debate, around issues to do with discourse participation. For example, marginalised groups may find themselves excluded or may exclude themselves from discourse contexts that are hostile. This not only has ethical implications, but likely impacts epistemic outcomes. The nature and scale of such outcomes remain difficult to estimate in practice. In this paper, we use agent-based modelling to explore the implications of a tendency toward `agreeableness' whereby agents might shape their communication so as to reduce direct conflict. Our simulations show that even mild tendencies to avoid disagreement can have significant consequences for information exchange and the resultant beliefs within a population.

Humans are remarkably adept at inferring the causes of events in their environment; doing so often requires incorporating information from multiple sensory modalities. For instance, if a car slows down in front of us, inferences about why they did so are rapidly revised if we also hear sirens in the distance. Here, we investigate the ability to reconstruct others' actions and events from the past by integrating multimodal information. Participants were asked to infer which of two agents performed an action in a household setting given either visual evidence, auditory evidence, or both. We develop a computational model that makes inferences by generating multimodal simulations, and also evaluate our task on a large language model (GPT-4) and a large multimodal model (GPT-4V). We find that humans are relatively accurate overall and perform best when given multimodal evidence. GPT-4 and GPT-4V performance comes close overall, but is very weakly correlated with participants across individual trials. Meanwhile, the simulation model captures the pattern of human responses well. Multimodal event reconstruction represents a challenge for current AI systems, and frameworks that draw on the cognitive processes underlying people's ability to reconstruct events offer a promising avenue forward.

Although well-intentioned diversity, equity, and inclusion (DEI) initiatives aim to increase minority representation in elite groups, they can sometimes backfire by causing candidates to question whether they were selected for merit. Prior work in social psychology suggests that this effect is driven mainly by stereotype threat. Here, we propose a novel cognitive framework: DEI initiatives backfire due to causal inference. Specifically, when candidates hear that they were selected based on a DEI initiative and/or enter a group where they are a minority, they may hypothesize that their selection was based more on their identity and less on their merit. Across two pre-registered experiments manipulating selection messages (DEI vs. merit) and statistical gender representation (represented or under-represented in the selected group), we find evidence in favor of our hypothesis. DEI messages and under-representation independently caused successful candidates to attribute their selection more to their identity and less to their merit but did not directly impact perceptions of competence. A third pre-registered experiment revealed that women selectively rated themselves as less competent in DEI contexts when selection tasks were more difficult. Taken together, this work shows that people make different causal hypotheses about their selection into elite groups based on DEI messages and group composition in conjunction with selection task difficulty and their social identity. Importantly, this work paves the way for designing DEI-based initiatives that license more helpful causal inferences about success to ensure that minority candidates thrive in their positions.

Causal reasoning enables us to explain the past, predict the future, and intervene in the present. Does the brain allocate specialized cortical regions to causal reasoning? And if so, are they involved in reasoning about both physical and social causal relationships, or are they domain-specific? In a pre-registered experiment (Exp 1) we scanned adults using fMRI while they matched physical and social causes to effects (e.g., ‘The car swerved to avoid a crash' -> ‘Coffee spilled all over the car seat'; ‘He was late for work' -> ‘Tom was scolded by his boss') or physical and social descriptions of the same entity matched for difficulty and linguistic variables to the causal conditions (e.g., ‘The brightest object in the sky'-> ‘The closest star to earth'; ‘She works at a hotel' -> ‘She brings in guests' luggage'). A region in the left lateral prefrontal cortex (LPFC) responded significantly more strongly to causal than descriptive conditions in most subjects individually. Responses in this region in held-out data were high for both social and physical causal conditions, yet no greater than baseline for the two descriptive (non-causal) conditions. In a follow-up exploratory experiment (Exp 2), we tested a different task (answering causal versus non-causal questions about physical and social narratives, matched for linguistic variables). Again, we found that both the physical and social causal stimuli selectively engaged the LPFC region. Finally, in both experiments, we found that brain regions previously implicated in intuitive physical reasoning responded more to the physical causal than the physical non-causal stimuli. Collectively, these results suggest that a) a region in the LPFC is selectively engaged in causal reasoning independent of content domain and b) the hypothesized physics network (hPN) is selectively involved in physical causal reasoning across modalities (visual vs. linguistic).

Our remarkable ability to infer complex cause-effect relationships is thought to distinguish humans from all other species. Despite that causal inferencing pervades human cognition, it remains unclear whether this fundamental cognitive ability is supported by a unified, domain-general mechanism or multiple domain-specific mechanisms. Both the language and logical reasoning systems have been described as possible unified substrates of causal inferencing. The current study uses neuroimaging to offer insight into this debate. We specifically focus on the culturally universal and highly motivationally relevant case of inferring illness causes. Participants read causal and noncausal vignettes about illness and mechanical failure while undergoing fMRI. We find that inferring the causes of illness selectively activates the brain's ‘animacy network,' particularly the precuneus. By contrast, a domain-general (i.e., ‘content-invariant') preference for causal inferencing did not emerge, including in the language and logical reasoning networks. Together, this evidence suggests that domain-specific mechanisms enable causal inferencing.

The mechanisms of learning stimulus-stimulus relationships are a longstanding research subject in psychology and neuroscience. Although traditional computational models provide valuable insights into learning processes, they often focus on the average behavior of a population. Individual learning trajectories, however, exhibit a diverse range of behaviors not captured by these models. In this paper, we compare sampling-based process-level models (i.e., particle filters) to representative associative and causal models (i.e., augmented Rescorla-Wagner and PowerPC) in their ability to capture individual learning behavior. We use likelihood-free inference incorporating machine-learned summary statistics for model estimation. We conduct a simulation study to demonstrate high model identifiability and test the models on an existing dataset and a newly conducted experiment which replicates and extends previous studies. We find that most participants are best explained by a particle filtering account, but more targeted experimental designs are required to estimate the best-fitting sub-type of these particle filter models.

Why do beliefs that gender differences are innate (i.e., gender essentialism) sometimes lead to normative judgments about how individual people ought to be? In the current study, we propose that a missing premise linking gender essentialism and normativity rests on the common folk-biological assumption that biological features serve a biological function. When participants (N = 289) learned that a novel feature of the gender category “mothers” was common and innate, they overwhelmingly assumed that it must have served some function across human history. When they learned that it served a historical function, they assumed that it must still be beneficial in today's environment. When participants learned that the feature was beneficial, they judged that contemporary mothers ought to have it, and they were more willing to intervene to ensure that they would by constraining the choices of individual mothers. Thus, we suggest that essentialist assumptions can shape normative social judgments via the explanations people tend to generate about why certain features of natural kind categories become common to begin with. This finding articulates one manifestation of the naturalistic fallacy, with implications for policy debates about bodily autonomy and choice.

Investigating how people evaluate more or less complex causal theories has been a focal point of research. However, previous studies have either focused on token-level causation or restricted themselves to very small sets of explanatory variables. We provide a new approach for modeling theory selection that foregrounds the balance between observed and latent structure in the mechanism being explained. We combine a Bayesian framework with program induction, allowing an unbounded and partially observable model space through sampling, and reflecting how a preference for simplicity emerges naturally in this setting. Through simulation, we identify two rational principles: (1) Simpler explanations should be favored as latent uncertainty (the number of hidden variables) increases; (2) latent structure is attributed a larger role when the observable patterns become less compressible. We conducted a behavioral experiment and found that human judgments tended to reflect these principles, indicating that people are sensitive to latent uncertainty when selecting between explanations.

This work provides an alternative account for deviations in human causal reasoning from normative predictions based on Causal Bayesian Networks (CBNs). We highlight violations of the Markov condition (Screening Off) and insufficient Explaining Away. Different from other accounts, our model does not assume that people fail to honor normative predictions due to reliance on heuristics, hidden nodes and links or cognitive limitations. Instead, we propose that people are rationally uncertain about the received causal model they are asked to reason with. We fitted the model to published data from two experiments where people were asked to make probability estimates on inferences of interest within a causal model. We find that the model is able to i) reproduce deviations from normative predictions, and ii) predict changes in the magnitude of these deviations across contexts. We conclude that assuming that people, in order to be rational, will always fully believe in the information they receive about a causal model may be too strong an assumption.

Although research in the area of belief updating has flourished in the last two decades, most studies do not treat beliefs as part of a complex and interactive network. In this study, we investigate humans' use of auxiliary hypotheses as a mechanism to avoid belief updating in light of conflicting information. In Experiment 1, we replicate an unpublished study by Kahneman and Tversky, introducing two additional domain conditions (N=119). Participants construct an initial model, express a prior belief, and face conflicting information. They are then prompted to provide an explanation. Across three domains, only 37% of responses demonstrate belief updating, by attributing the information conflict to the original report being unreliable or invalid. In Experiment 2 (N=29), a within-participants manipulation of credibility shows no effect on generating auxiliary hypotheses. Even in the presence of credibility cues to explain away information conflicts by invoking the reliability of either source, participants instead generated auxiliary hypotheses to resolve them in 27% of the cases.

Practitioners of mentalism can perform apparently impossible feats, but when performing for an audience these feats are attributed to pseudoscientific explanations such as advanced psychological skills. Research that has investigated the psychological foundations of mentalism has found a strong tendency for people to believe these explanations. In three experiments, we investigated the strength of this belief by comparing apparently impossible effects relating to mental phenomena with physical phenomena. We observed that mental magic tricks are readily explained in terms of advanced psychological skills, whereas physical tricks are not. This was true: i) even when alternative feasible explanations are explicitly presented; ii) when they are presented as mentalism effects but the effects themselves are classical card tricks; iii) regardless of the context in which the effects are observed (a research laboratory vs. a theater). We interpreted the tendency to appeal to this pseudo-explanation (and the changes in narratives employed by mentalists across the decades) in terms of the community of knowledge framework.

Humans are resistant to changing their beliefs even in the face of disconfirming evidence. The Bayesian brain theory suggests that we should update our beliefs optimally in light of new evidence, but recent research indicates that belief formation is far from the Bayesian ideal. Individuals can exhibit "stronger-than-rational" updating or be resistant to revising their beliefs. The present study proposes a novel paradigm to explore perceptions and preferences for belief updating patterns in stable and dynamic stochastic environments, using an advice-taking paradigm. In an experiment (N=567) based on a fishing task, we introduce three advisor characters representing formal updating models: Bayesian, Volatile and Rigid. We find that participants exhibit higher trust for the Bayesian advisor than the Rigid advisor, in the stable but not changeable environment conditions. In the changeable environment, participants exhibit higher trust for the Volatile advisor, compared to both the Bayesian and Rigid advisors. The findings also suggest that participants' own learning closely mimics the pattern of the Volatile model. This study illustrates that people can differentiate between Bayesian updating, and its "stronger-than" and "weaker-than" variations, and exhibit preferences for these updating patterns, in different environment structures.

Explanations are social, and when people try to explain something, they usually seek input from others. We present a simple theory of how people use the explanations they encounter as clues to the broader landscape of possible explanations, informing their decision to exploit what has been found or explore new possibilities. The challenge of coming up with novel explanations draws people to exploit or imitate appealing ones (information cascades); this draw increases as less appealing alternatives become more distant (the ``strawman'' effect). Conversely, pairs of low-quality explanations promote exploratory behavior or long-leaps away from observed attempts, and pairs of divergent high-quality explanations can lead to merging and syncretism. We use a transmission-chain experiment to test, and confirm, these predictions. Intriguingly, we also find that while people imitate good explanations, their imitations often fall short in quality. Our work provides new insight into how collective exploration can be promoted, or stalled, by implicit information about what is yet to be discovered.

Language profiles of individuals with Down syndrome (DS) reveal a pattern of heterogeneous abilities, with receptive vocabulary exhibiting strengths over receptive grammar, and expressive language lagging behind. Little is known about inferential abilities in this population, in either children or adults, despite inferencing playing a pivotal role in language comprehension. Inferential abilities are particularly relevant to the successful understanding of narratives, as story plots combine explicit (factual) and implicit (inferential) information.This study investigated inferential abilities in 26 English-speaking adolescents and adults with DS (age: 13-43, M=22.9 years) compared to 23 young vocabulary-matched typical controls (age: 4-11, M=6.96 years). Inferencing was assessed through a narrative comprehension task, which targeted understanding of story characters' goals and internal states (ISs). Participants with DS showed poorer comprehension of inferential questions, across both goals and ISs, with vocabulary level and receptive grammar positively contributing to the comprehension of inferences. Working memory showed a positive albeit non-significant relationship with inferencing ability, while executive functioning skills had no effect. Our results suggest that difficulties understanding, and potentially expressing, inferential information relating to story characters' goals and ISs persevere into adulthood in individuals with DS. Such difficulties are moderated by general verbal abilities and seem driven by poor grammatical skills. We discuss the contributions of verbal and nonverbal abilities to inference-making in Down syndrome, and potential implications for future research.

Recent literature often presents memory as ultimately dealing with the future–helping the organism to anticipate events and increase its adaptive success. Yet, the distinct contribution of episodic (as opposed to semantic) memory to future-oriented simulations remains unclear. We claim that episodic memory yields adaptive success because of its crucial role in singular counterfactual causal reasoning, which thus far has been mostly ignored in the literature. Our paper presents a causal inference model based on the predictive processing framework and the minimal trace account of episodic memory. According to our model, evaluating the cause of an event involves (i) generating an episodic memory related to the said potential cause, (ii) constructing a counterfactual scenario through inhibition of the relevant part of the past episode, and (iii) temporal evolution followed by alternative model evaluation.

In three experiments, we investigated the effect of repeated exposure and social interaction on adults' tendency to make sense of novel events. Specifically, we examined whether, across trials, participants' observations shifted from descriptive to explanatory, from specific to generic, became more inclined to reference causes, and more evaluative. We found that while there was an effect of repeated exposure on generalization and of social interaction on both explanation and generalization, the intervention that was most likely to shift adults' sense-making behavior was a communicative context of small groups in which each participant had partial and different knowledge. We suggest that this is because social contexts inherently motivate individuals to integrate new information, reconcile discrepancies, and forge efficient, generalizable concepts.

We observe causal relationships naturally and quickly in events that we experience in our life. The current research investigates if causal events like collisions attract our attention to other changes in objects involved in the causal event. Participants reported colour changes in two objects, one involved in a causal event (collision) and the other independent. Aligning with our expectation, we observed that participants are more likely to report the colour change involved in the causal event when it happened at the same time as the collision. Against our prediction however, we observed a similar effect when colour changes happened before the collision, while the difference was less strong when the colour changes happened after the collision. One possible explanation is that the effect stems from participants anticipating causal events, leading them to pay extra attention to objects potentially involved in collisions. This focused attention makes participants more likely to notice colour changes during the anticipation period, which means people are actively devoting more cognitive resources anticipating and confirming causal interactions. This finding suggests that people prioritise causal observations in visual search tasks.

We compare two perspectives on base-rate neglect (Kahneman& Tversky, 1973) in probabilistic judgment. The evidentialimpact perspective derives it from humans' focus on theimpact of evidence on belief, rather than conditional probabilities.The Causal Models perspective derives it from humans'inability to integrate information that is causally opaque,as base-rates often are in such experiments. Because causaland evidential-impact relations are often concomitant and confounded,we designed an experiment that specifically teasesapart their respective influence on probabilistic judgment. Ourresults support a combination of the two perspectives, withcausal transparency influencing the degree to which one engagesin evidential impact reasoning strategies.

When someone realizes they do not actually know how a can opener works, do they think it is just a one-time bout of overconfidence? Or, do they assume they lack understanding of all the devices in their home? Causal knowledge is a fundamental part of both daily functioning and long-term learning. Previous studies have shown that writing out a causal explanation has the ability to induce knowledge reassessment and decrease inflated perceptions of knowledge specific to the concept being explained. However, the generalization of this knowledge reassessment has only recently been explored. In this preregistered experiment, we used the Illusion of Explanatory Depth (IOED) paradigm to see whether a decrease in perceived understanding of an explained item affects the perceived understanding of an item that was not asked to be explained. We also assessed the effect of explanation quality on this transfer of knowledge. Results showed that knowledge reassessment for explained items led to an even greater reassessment for unexplained items, suggesting possible overgeneralization. While explanation quality influenced knowledge reassessment for explained items, it did not for unexplained items. We discuss the possible reasons for these results as well as future studies to help understand the boundaries of knowledge reassessment.

How do people's causal knowledge influence how they seek information? The current work tasks participants with choosing to observe disease symptoms in a setting where they know a disease's etiology and related symptoms. We use causal graphical models (CGMs) to formalize their causal knowledge of the disease, and find that people tend to use their expected information gain, computed over their CGM-generated probability beliefs, to search for information in causal settings.

Classic base-rate neglect studies have been consistently criticised for lacking ecological validity. A study by Welsh & Navarro (2012) found this heuristic was significantly reduced when participants perceived the base rate as more relevant. The present study aims to study this phenomenon through a more realistic scenario while simultaneously capturing participants' written reasoning. Using mixed-methods, participants (N = 2,052) read an engaging scenario regarding a person who committed infidelity and containing a base-rate and specific information where the contextual information regarding the base-rate was manipulated. They were then asked to provide an estimate of the person's likelihood to cheat in the future. Results show that each of our three manipulations to the context of the base rate are significant in affecting participants' estimates, supporting Welsh and Navarro's findings. Analysis of participants' written reasoning demonstrates the sophistication and nuance of participants' engagement with the base-rate, challenging the original view of this supposed heuristic.

Humans possess knowledge of causal systems with deep compositional structures. For example, we know that a good soccer team needs players to fill different roles, with each role demanding a configuration of skills from the player. These causal systems operate on multiple object types (player roles) that are defined by features within objects (skills). This study explores how human learners perform on novel causal learning problems in which they need to infer multiple object types in a bottom-up manner, using empirical information as a cue for their existence. We model subjects' learning process with Bayesian models, drawing hypotheses from different spaces of logical expressions. We found that although subjects exhibited partial success on tasks that required learning one object type, they mostly failed at those that required learning multiple types. Our result identifies the learning of object types as a major obstacle for human acquisition of complex causal systems.

An open question in the developmental causal learning literature concerns how children's beliefs about causal systems impact their inferences. This study investigated how 4- and 5-year-olds' causal beliefs related to their “backwards blocking” abilities, as well as whether associative learning or Bayesian inference better explained their judgements. Children were taught either that two causes together produced a larger effect than that produced by each individually or that they produced the same size effect as that produced by either one. A third group received no training. Results indicated that 4-year-olds engaged in backwards blocking only after additivity training and that their inferences mainly matched an associative model. In contrast, 5-year-olds consistently engaged in backwards blocking and produced responses that largely matched a Bayesian model. These findings suggest that the effect of children's beliefs about causal systems on their inferences undergoes a developmental progression and implicate the role of multiple cognitive mechanisms.

Developmental psychologists have often turned to children to clarify understanding of functional and mechanistic cognition. Here, we investigate children's epistemic inferences of function – what a thing is for – and mechanism – how a thing works. Children, like adults, believe a mechanism-knower knows more than a function-knower (Study 1). Yet, unlike adults, children do not expect that a mechanism-knower is also more likely to know function than a function-knower is to know mechanism (Study 2). Children's experience of learning function and mechanism of complex systems sheds light on this asymmetry; Children who are taught just mechanism can infer the complementary function, but, interestingly, children who are taught just function can likewise infer the complementary mechanism (Study 3). This paper considers the nature of children's epistemic intuitions and whether those beliefs are reflective of children's learning experience.

Function – what a thing is for – and mechanism – how a thing's parts interact to make it work – are considered by cognitive psychologists and philosophers of science to be integrally related despite people's acute sensitivity to their differences. Here, we set out to better characterize lay adults' intuitions about functional and mechanistic knowledge (Study 1). Then, we use learning studies to investigate to what degree these intuitions accurately capture functional and mechanistic cognition (Studies 2, 3). While some intuitions (e.g., that mechanism is more difficult to learn than function) are supported by these learning studies, others (e.g., that function should precede mechanism in explanations) are not. Possible reasons for matches and mismatches are explored.

Decades of work exploring the development of children's analogical reasoning illustrates that 3- and 4-year-old children struggle with reasoning by analogy (i.e. glove:hand::sock:___), almost always preferring superficially related “object matches” (:shoe) over “relational matches” (:foot). However, one recent study demonstrated preschoolers' ability to choose relational matches when a traditional relational-match-to-sample task is embedded in causal scaffolding, framing the target abstract relation as one between beginning and ending states of a causal transformation. Current work aims to discover which factors of causal framing facilitate this boost in early abstract reasoning. In Study 1, we replicate this effect while adapting the transformation to involve two objects, showing that preservation of identity is not necessary for analogical reasoning in a causal context. In Study 2, we explore the replicated effect in a case of non-agentive causation, finding that the causal boost, while still present, is significantly weaker when scaffolding involves a machine vs. an agent. These findings demonstrate that causal framing can be a powerful tool in bolstering children's early abstract reasoning capabilities and show that this enhancing effect is even stronger when an agent holds causal power.