Choosing optimal causal backgrounds for causal discovery

In two experiments, we studied the strategies that people use to discover causal relationships. According to inferential approaches to causal discovery, if people attempt to discover the power of a cause, then they should naturally select the most informative and unambiguous context. For generative causes this would be a context with a low base rate of effects generated by other causes and for preventive causes a context with a high base rate. In the following experiments, we used probabilistic and/or deterministic target causes and contexts. In each experiment, participants observed several contexts in which the effect occurred with different probabilities. After this training, the participants were presented with different target causes whose causal status was unknown. In order to discover the influence of each cause, participants were allowed, on each trial, to choose the context in which the cause would be tested. As expected by inferential theories, the participants preferred to test generative causes in low base rate contexts and preventative causes in high base rate contexts. The participants, however, persisted in choosing the less informative contexts on a substantial minority of trials long after they had discovered the power of the cause. We discuss the matching law from operant conditioning as an alternative explanation of the findings.     

Positive and negative mediation as a function of whether the absent cue was previously associated with the outcome

After presenting two cues, A and B, together, later pairings of one of the cues alone with an outcome can generate changes in the associative value of the absent cue. These changes can be in the same direction as the present cue (i.e., positive mediation) or in the opposite direction to the present cue (i.e., negative mediation). We found both mediational effects in a human contingency task. In addition, we found that the direction of the change was determined by the existence of a prior association between the absent cue and the outcome. When a prior association exists, the absent cue tends to change its value in the opposite direction to the present cue, whereas when there is no prior association, the absent cue tends to change its value in the same direction as the present cue. Recent associative models (Stout & Miller, 2007 Stout, S. C. and Miller, R. R. 2007. Sometimes-competing retrieval (SOCR): A formalization of the comparator hypothesis. Psychological Review, 114: 759–783.  [Google Scholar]) can explain our results.     

Age differences in contingency judgement linked to perceptual segregation

We demonstrate large differences in judging positive and null contingencies between younger and older adults with a task commonly used to explore cue competition in both contingency and causality judgements. The one-phase blocking task uses two cues, with separate contingencies with the same outcome. The age differences persisted even when participants knew in advance which of the two contingencies to judge. The age differences disappeared, however, when the stimulus display contained markers aiding perceptual segregation. We suggest that the age differences elicited in the one-phase blocking task are linked to decrements in perceptual segregation.     

A positive patterning advantage with complex but not simple patterning: A cue constellation approach

Determinants of a positive patterning advantage (that is, an advantage for positive patterning over negative patterning) in human causal reasoning were examined in an experiment that compared simple patterning discriminations (A, B vs. AB) to complex patterning discriminations (AB, BC, AC vs. ABC). As predicted by a cue constellation analysis of complex discriminations, a positive patterning advantage was found with complex patterning but not with simple patterning discriminations. This result may explain why some recent studies have found a positive patterning advantage where earlier studies had failed to find one. The interaction of patterning complexity with the positive patterning advantage appears to pose problems for rule-based accounts of patterning. The results support the view that associative theories of human causal reasoning are more easily distinguished from rule-based approaches when applied to conditions that make simple rules difficult to identify or implement.     

Model-based recursive partitioning of extended redundancy analysis with an application to nicotine dependence among US adults

Extended redundancy analysis (ERA) is used to reduce multiple sets of predictors to a smaller number of components and examine the effects of these components on a response variable. In various social and behavioural studies, auxiliary covariates (e.g., gender, ethnicity) can often lead to heterogeneous subgroups of observations, each of which involves distinctive relationships between predictor and response variables. ERA is currently unable to consider such covariate-dependent heterogeneity to examine whether the model parameters vary across subgroups differentiated by covariates. To address this issue, we combine ERA with model-based recursive partitioning in a single framework. This combined method, MOB-ERA, aims to partition observations into heterogeneous subgroups recursively based on a set of covariates while fitting a specified ERA model to data. Upon the completion of the partitioning procedure, one can easily examine the difference in the estimated ERA parameters across covariate-dependent subgroups. Moreover, it produces a tree diagram that aids in visualizing a hierarchy of partitioning covariates, as well as interpreting their interactions. In the analysis of public data concerning nicotine dependence among US adults, the method uncovered heterogeneous subgroups characterized by several sociodemographic covariates, each of which yielded different directional relationships between three predictor sets and nicotine dependence.     

Online measurement of mental representations of complex spatial decision problems: Comparison of CNET and hard laddering

This paper introduces the online Causal Network Elicitation Technique (CNET), as a technique for measuring components of mental representations of choice tasks and compares it with the more common technique of online ‘hard’ laddering (HL). While CNET works in basically two phases, one in open question format and one as guided linking of attributes and benefits, HL works completely structured with revealed attributes and benefits. Mental representations of two activity travel tasks were collected with both techniques among members of a nationwide Dutch household panel. The results confirm the hypothesis that the revealed format of variables in HL has an effect on the indication of variables as the elicited mental representations are almost twice as big for HL than for CNET. Furthermore, it turned out that CNET is more sensitive in measuring shifts among attributes in the mental representations for situational changes of the activity-travel task.     

Do people reason rationally about causally related events? Markov violations,weak inferences,and failures of explaining away

Making judgments by relying on beliefs about the causal relationships between events is a fundamental capacity of everyday cognition. In the last decade, Causal Bayesian Networks have been proposed as a framework for modeling causal reasoning. Two experiments were conducted to provide comprehensive data sets with which to evaluate a variety of different types of judgments in comparison to the standard Bayesian networks calculations. Participants were introduced to a fictional system of three events and observed a set of learning trials that instantiated the multivariate distribution relating the three variables. We tested inferences on chains X₁ → Y → X₂, common cause structures X₁ ← Y → X₂, and common effect structures X₁ → Y ← X₂, on binary and numerical variables, and with high and intermediate causal strengths. We tested transitive inferences, inferences when one variable is irrelevant because it is blocked by an intervening variable (Markov Assumption), inferences from two variables to a middle variable, and inferences about the presence of one cause when the alternative cause was known to have occurred (the normative “explaining away” pattern). Compared to the normative account, in general, when the judgments should change, they change in the normative direction. However, we also discuss a few persistent violations of the standard normative model. In addition, we evaluate the relative success of 12 theoretical explanations for these deviations.     

A Regression Framework for Causal Mediation Analysis with Applications to Behavioral Science

We introduce and extend the classical regression framework for conducting mediation analysis from the fit of only one model. Using the essential mediation components (EMCs) allows us to estimate causal mediation effects and their analytical variance. This single-equation approach reduces computation time and permits the use of a rich suite of regression tools that are not easily implemented on a system of three equations. Additionally, we extend this framework to non-nested mediation systems, provide a joint measure of mediation for complex mediation hypotheses, propose new visualizations for mediation effects, and explain why estimates of the total effect may differ depending on the approach used. Using data from social science studies, we also provide extensive illustrations of the usefulness of this framework and its advantages over traditional approaches to mediation analysis. The example data are freely available for download online and we include the R code necessary to reproduce our results.     

Causal judgments about empirical information in an interrupted time series design

Empirical information available for causal judgment in everyday life tends to take the form of quasi-experimental designs, lacking control groups, more than the form of contingency information that is usually presented in experiments. Stimuli were presented in which values of an outcome variable for a single individual were recorded over six time periods, and an intervention was introduced between the fifth and sixth time periods. Participants judged whether and how much the intervention affected the outcome. With numerical stimulus information, judgments were higher for a pre-intervention profile in which all values were the same than for pre-intervention profiles with any other kind of trend. With graphical stimulus information, judgments were more sensitive to trends, tending to be higher when an increase after the intervention was preceded by a decreasing series than when it was preceded by an increasing series ending on the same value at the fifth time period. It is suggested that a feature-analytic model, in which the salience of different features of information varies between presentation formats, may provide the best prospect of explaining the results.     

Contiguity and the causal theory of memory

In Memory: A Philosophical Study, Bernecker argues for an account of contiguity. This Contiguity View is meant to solve relearning and prompting, wayward causation problems plaguing the causal theory of memory. I argue that Bernecker’s Contiguity View fails in this task. Contiguity is too weak to prevent relearning and too strong to allow prompting. These failures illustrate a problem inherent in accounts of memory causation. Relearning and prompting are both causal relations, wayward only with respect to our interest in specifying remembering’s requirements. Solving them requires saying more about remembering, not causation. I conclude by sketching such an account.