Modeling the Role of Unobserved Causes in Causal Learning

Current theories suggest that causal learning is based on covariation information. However, information about the presence/absence of events (particularly causes) is frequently unavailable, rendering them unobserved. The current paper presents a new model of causal learning, BUCKLE (Bidirectional Unobserved Cause LEarning), which extends existing models of causal learning by dynamically inferring information about unobserved causes. During the course of causal learning, BUCKLE continually computes the probability that an unobserved cause is present on each occasion and uses the results of these inferences to adjust the strengths of the unobserved, as well as observed, causes.