Using the Implicit Association Test as a Measure of Causal Learning does not Eliminate Effects of Rule Learning

Abstract

Given that human causal judgments may be based on propositional reasoning processes rather than reflecting the strength of associations between cause and outcome representations (e.g., De Houwer, 2009, for a review), the question arises whether there are measures of causal learning that are sensitive only to the strength of associations. We investigated the potential of the Implicit Association Test (IAT) in this context. The results showed evidence for rule learning even when IAT effects were used as an index of learning. This suggests that the IAT does not allow one to measure causal learning in a manner that excludes an impact of conscious propositional knowledge.