We present a framework for the rational analysis of elemental causal induction – learning about the existence of a relationship between a single cause and effect – based upon causal graphical models. This framework makes precise the distinction between causal structure and causal strength: the difference between asking whether a causal relationship exists and asking how strong that causal relationship might be. We show that two leading rational models of elemental causal induction, ∆P and causal power, both estimate causal strength, and introduce a new rational model, causal support, that assesses causal structure. Causal support predicts several key phenomena of causal induction that cannot be accounted for by other rational models, which we explore through a series of experiments. These phenomena include the complex interaction between ∆P and the base-rate probability of the effect in the absence of the cause, sample size effects, inferences from incomplete contingency tables, and causal learning from rates. Causal support also provides a better account of a number of existing datasets than either ∆P or causal power.

This article presents a theory of categorization that accounts for the effects of causal knowledge that relates the features of categories. According to causal-model theory, people explicitly represent the probabilistic causal mechanisms that link category features and classify objects by evaluating whether they were likely to have been generated by those mechanisms. In 3 experiments, participants were taught causal knowledge that related the features of a novel category. Causal-model theory provided a good quantitative account of the effect of this knowledge on the importance of both individual features and interfeature correlations to classification. By enabling precise model fits and interpretable parameter estimates, causal-model theory helps place the theory-based approach to conceptual representation on equal footing with the well-known similarity-based approaches. For the last several decades, research on the topic of categorization has focused on the problem of learning new categories via examples of category members, that is, from empirical observations. The result has been a host of categorization models that are based on representational ideas such as central prototypes, stored exemplars, and variabilized rules, and on processing principles such as similarity, that have considerable explanatory power and experimental support. More recently, the influence of the prior “theoretical ” knowledge that learners often contribute to their representations of categories has also been a topic of study (Carey,

Abstract not found

This article offers a plausible domain-general explanation for why some concepts of processes are resistant to instructional remediation although other, apparently similar concepts are more easily understood. The explanation assumes that processes may differ in ontological ways: that some processes (such as the apparent flow in diffusion of dye in water) are emergent and other processes (such as the flow of blood in human circulation) are direct. Although precise definition of the two kinds of processes are probably impossible, attributes of direct and emergent processes are described that distinguish them in a domain-general way. Circulation and diffusion, which are used as examples of direct and emergent processes, are associated with different kinds of misconceptions. The claim is that stu-Do Not Copy dents ’ misconceptions for direct kinds of processes, such as blood circulation, are of the same ontological kind as the correct conception, suggesting that misconceptions of direct processes may be nonrobust. However, students ’ misconceptions of emergent processes are robust because they misinterpret emergent processes as a kind of commonsense direct processes. To correct such a misconception requires a re-representation or a conceptual shift across ontological kinds. Therefore, misconceptions of emergent processes are robust because such a shift requires that students know about the emergent kind and can overcome their (perhaps even innate) predisposition to conceive of all processes as a direct kind. Such a domain-general explanation suggests that teaching students the causal structure underlying emergent processes may enable them to recognize and understand a variety of emergent processes for which they have robust misconceptions, such as concepts of electricity, heat and temperature, and evolution. Correspondence and requests for reprints should be sent to Michelene T. H. Chi, Learning Research

People learn quickly when reasoning about causal relationships, making inferences from limited data and avoiding spurious inferences. Efficient learning depends on abstract knowledge, which is often domain or context specific, and much of it must be learned. While such knowledge effects are well documented, little is known about exactly how we acquire knowledge that constrains learning. This work focuses on knowledge of the functional form of causal relationships; there are many kinds of relationships that can apply between causes and their effects, and knowledge of the form such a relationship takes is important in order to quickly identify the real causes of an observed effect. We developed a hierarchical Bayesian model of the acquisition of knowledge of the functional form of causal relationships and tested it in five experimental studies, considering disjunctive and conjunctive relationships, failure rates, and cross-domain effects. The Bayesian model accurately predicted human judgments and outperformed several alternative models.

An ongoing goal in the field of categorization has been to determine how objects ’ features provide evidence of membership in one category versus another. Well-known findings include that feature diagnosticity is a function of how often the feature appears in category members versus nonmembers, their perceptual salience, how features are used in support of inferences, and how observable features are related to other observable features. We tested how diagnosticity is affected by causal relations between observable and unobserved features. Consistent with our view of classification as diagnostic reasoning, we found that observable features are more diagnostic to the extent that they are caused by underlying features that define category membership, because the presence of the latter can be (causally) inferred from the former. Implications of these results for current views of conceptual structure and models of categorization are discussed. It is generally accepted that people’s concepts include not only the features and attributes of the entity being represented, but also the ways in which those features are related to one another. For example, we know that hormones can alter a person’s behavior, that chemical structure can affect a substance’s hardness, and that processor

One of the most important requirements for accurate causal inference is that there be no confounds; the cause being evaluated must occur independently of all other causes. When this requirement is not met, causal inferences are likely to be incorrect. The current study asks participants to judge how informative various situations are with respect to drawing causal inferences. Contrary to normative principles, participants appear to believe that many confounded situations are just as informative as unconfounded situations.

A central question in cognitive research concerns how new properties are generalized to categories. This article introduces a model of how generalizations involve a process of causal inference in which people estimate the likely presence of the new property in individual category exemplars and then the prevalence of the property among all category members. Evidence in favor of this causalbased generalization (CBG) view included effects of an existing feature’s base rate (Experiment 1), the direction of the causal relations (Experiments 2 and 4), the number of those relations (Experiment 3), and the distribution of features among category members (Experiments 4 and 5). The results provided no support for an alternative view that generalizations are promoted by the centrality of the to-be-generalized feature. However, there was evidence that a minority of participants based their judgments on simpler associative reasoning processes. Keywords: Causal-based induction; Generalization; Causal reasoning 1.

This deliverable concerns the third objective of WP5, which is to develop a learning and recognition algorithm for complex actions. We have extended the factored-HMM representation of manipulation actions developed in Year 2 by another layer of dynamic models of goals, in which details of manipulation movements are abstracted away. The focus of work in Year 3 is to model the dynamics of human intention that results in the observed human complex behaviour. This report is a revised and extended version of the report submitted in Year 2, and contains new analysis results. • Two key capabilities: We identify two key capabilities for a system to understand a complex human behaviour. The first capability is to learn normal behaviour that allows induction of the agent’s goals from partial or noisy observation. The second is to be able to attribute any deviations from normal behaviour to the background knowledge (hidden) or other differentiating scene features (observable). Achieving these capabilities requires a representation that not only encodes the dynamics of the scene, but also allows information about other agents ’ beliefs, plans and intentions to be integrated. • Extension of action representation:

We use graphical models to explore the question of how people learn simple causal relationships from data. The two leading psychological theories can both be seen as estimating the parameters of a fixed graph. We argue that a complete account of causal induction should also consider how people learn the underlying causal graph structure, and we propose to model this inductive process as a Bayesian inference. Our argument is supported through the discussion of three data sets. 1 Introduction Causality plays a central role in human mental life. Our behavior depends upon our understanding of the causal structure of our environment, and we are remarkably good at inferring causation from mere observation. Constructing formal models of causal induction is currently a major focus of attention in computer science [6], psychology [2,5], and philosophy [4]. This paper attempts to connect these literatures, by framing the debate between two major psychological theories in the computationa...