Use of prior beliefs in the assignment of causal roles: causal powers versus regularity-based accounts

There is a tradition of models of causal judgment in which causes and other causal roles are defined and identified in terms of empirical patterns of association with effects. In the present experiments, results conflicting with the predictions of such models were obtained. In one experiment, subjects judged that an interpretation in which a factor constantly present was identified as the cause was more likely than was an interpretation in which a perfect positive covariate was identified as the cause. In a second experiment, possible effects of prior beliefs about covariation were controlled and similar findings were obtained in two out of three scenarios. These results favor the idea that people make causal judgments by applying preexisting beliefs framed in terms of causal concepts, such as causal powers, and in ways that cannot be accounted for by models in the empiricist tradition.