York Hagmayer

Estimating causal strength: the role of structural knowledge and processing effort

The strength of causal relations typically must be inferred on the basis of statistical relations between observable events. This article focuses on the problem that there are multiple ways of extracting statistical information from a set of events. In causal structures involving a potential cause, an effect and a third related event, the assumed causal role of this third event crucially determines whether it is appropriate to control for this event when making causal assessments between the potential cause and the effect. Three experiments show that prior assumptions about the causal roles of the learning events affect the way contingencies are assessed with otherwise identical learning input. However, prior assumptions about causal roles is only one factor influencing contingency estimation. The experiments also demonstrate that processing effort affects the way statistical information is processed. These findings provide further evidence for the interaction between bottom-up and top-down influences in the acquisition of causal knowledge. They show that, apart from covariation information or knowledge about mechanisms, abstract assumptions about causal structures also may affect the learning process.

Beyond the Information Given: Causal Models in Learning and Reasoning

The philosopher David Humes conclusion that causal induction is solely based on observed associations still presents a puzzle to psychology. If we only acquired knowledge about statistical covariations between observed events without accessing deeper information about causality, we would be unable to understand the differences between causal and spurious relations, between prediction and diagnosis, and between observational and interventional inferences. All these distinctions require a deep understanding of causality that goes beyond the information given. We report a number of recent studies that demonstrate that people and rats do not stick to the superficial level of event covariations but reason and learn on the basis of deeper causal representations. Causal-model theory provides a unified account of this remarkable competence.

How temporal assumptions influence causal judgments

Causal learning typically entails the problem of being confronted with a large number of potentially relevant statistical relations. One type of constraint that may guide the choice of appropriate statistical indicators of causality are assumptions about temporal delays between causes and effects. There have been a few previous studies in which the role of temporal relations in the learning of events that are experienced in real time have been investigated. However, human causal reasoning may also be based on verbally described events, rather than on direct experiences of the events to which the descriptions refer. The aim of this paper is to investigate whether assumptions about the temporal characteristics of the events that are being described also affect causal judgment. Three experiments are presented that demonstrate that different temporal assumptions about causal delays may lead to dramatically different causal judgments, despite identical learning inputs. In particular, the experiments show that temporal assumptions guide the choice of appropriate statistical indicators of causality by structuring the event stream (Experiment 1), by selecting the potential causes among a set of competing candidates (Experiment 2), and by influencing the level of aggregation of events (Experiment 3).

Decision makers conceive of their choices as interventions.

Causal considerations must be relevant for those making decisions. Whether to bring an umbrella or leave it at home depends on the causal consequences of these options. However, most current decision theories do not address causal reasoning. Here, the authors propose a causal model theory of choice based on causal Bayes nets. The critical ideas are (a) that people decide using causal models of the decision situation and (b) that people conceive of their own choice as an intervention. Four corroborating experiments are reported. The first 2 experiments showed that participants chose on the basis of the causal structure underlying a choice scenario rather than the statistical relation among actions and outcomes. Experiments 3 and 4 showed that participants treated choices and interventions similarly. They also suggest that decision makers use causal models to derive inferences about expected outcomes. Boundary conditions on causal decision making and examples of faulty causal inferences in choice (e.g., self-deception) are discussed. (PsycINFO Database Record (c) 2009 APA, all rights reserved).

The causal psycho-logic of choice

Choices do not merely identify one option among a set of possibilities; choosing is an intervention, an action that changes the world. As a result, good decision making generally requires a model specifying how actions are causally related to outcomes. Interventions license different inferences than observations because an event whose state has been determined by intervention is not diagnostic of the normal causes of that event. We integrate these ideas into a causal framework for decision making based on causal Bayes nets theory, and suggest that deliberate decision making is based on simplified causal models and imaginary interventions. The framework is consistent with what we know so far about how people make decisions.

The nature and perception of fluctuations in human musical rhythms

Although human musical performances represent one of the most valuable achievements of mankind, the best musicians perform imperfectly. Musical rhythms are not entirely accurate and thus inevitably deviate from the ideal beat pattern. Nevertheless, computer generated perfect beat patterns are frequently devalued by listeners due to a perceived lack of human touch. Professional audio editing software therefore offers a humanizing feature which artificially generates rhythmic fluctuations. However, the built-in humanizing units are essentially random number generators producing only simple uncorrelated fluctuations. Here, for the first time, we establish long-range fluctuations as an inevitable natural companion of both simple and complex human rhythmic performances. Moreover, we demonstrate that listeners strongly prefer long-range correlated fluctuations in musical rhythms. Thus, the favorable fluctuation type for humanizing interbeat intervals coincides with the one generically inherent in human musical performances.

Categories and causality: the neglected direction.

The standard approach guiding research on the relationship between categories and causality views categories as reflecting causal relations in the world. We provide evidence that the opposite direction also holds: categories that have been acquired in previous learning contexts may influence subsequent causal learning. In three experiments we show that identical causal learning input yields different attributions of causal capacity depending on the pre-existing categories to which the learning exemplars are assigned. There is a strong tendency to continue to use old conceptual schemes rather than switch to new ones even when the old categories are not optimal for predicting the new effect, and when they were motivated by goals that differed from the present context of causal discovery. However, we also found that the use of prior categories is dependent on the match between categories and causal effect. Whenever the category labels suggest natural kinds which can be plausibly related to the causal effects, transfer was observed. When the categories were arbitrary, or could not be plausibly related to the causal effect learners abandoned the categories, and used different categories to predict the causal effect.

Ecological rationality or nested sets? Individual differences in cognitive processing predict Bayesian reasoning

The presentation of a Bayesian inference problem in terms of natural frequencies rather than probabilities has been shown to enhance performance. The effect of individual differences in cognitive processing on Bayesian reasoning has rarely been studied, despite enabling us to test process-oriented variants of the two main accounts of the facilitative effect of natural frequencies: The ecological rationality account (ERA), which postulates an evolutionarily shaped ease of natural frequency automatic processing, and the nested sets account (NSA), which posits analytical processing of nested sets. In two experiments, we found that cognitive reflection abilities predicted normative performance equally well in tasks featuring whole and arbitrarily parsed objects (Experiment 1) and that cognitive abilities and thinking dispositions (analytical vs. intuitive) predicted performance with single-event probabilities, as well as natural frequencies (Experiment 2). Since these individual differences indicate that analytical processing improves Bayesian reasoning, our findings provide stronger support for the NSA than for the ERA.

Self-deception requires vagueness.

The paper sets out to reveal conditions enabling diagnostic self-deception, peoples tendency to deceive themselves about the diagnostic value of their own actions. We characterize different types of self-deception in terms of the distinction between intervention and observation in causal reasoning. One type arises when people intervene but choose to view their actions as observations in order to find support for a self-serving diagnosis. We hypothesized that such self-deception depends on imprecision in the environment that allows leeway to represent ones own actions as either observations or interventions. Four experiments tested this idea using a dot-tracking task. Participants were told to go as quickly as they could and that going fast indicated either above-average or below-average intelligence. Precision was manipulated by varying the vagueness in feedback about performance. As predicted, self-deception was observed only when feedback on the task used vague terms rather than precise values. The diagnosticity of the feedback did not matter.

Inferences about unobserved causes in human contingency learning

Estimates of the causal efficacy of an event need to take into account the possible presence and influence of other unobserved causes that might have contributed to the occurrence of the effect. Current theoretical approaches deal differently with this problem. Associative theories assume that at least one unobserved cause is always present. In contrast, causal Bayes net theories (including Power PC theory) hypothesize that unobserved causes may be present or absent. These theories generally assume independence of different causes of the same event, which greatly simplifies modelling learning and inference. In two experiments participants were requested to learn about the causal relation between a single cause and an effect by observing their co-occurrence (Experiment 1) or by actively intervening in the cause (Experiment 2). Participants’ assumptions about the presence of an unobserved cause were assessed either after each learning trial or at the end of the learning phase. The results show an interesting dissociation. Whereas there was a tendency to assume interdependence of the causes in the online judgements during learning, the final judgements tended to be more in the direction of an independence assumption. Possible explanations and implications of these findings are discussed.