Recent research in human causal learning indicates that the amount of "processing power" devoted to learning about a stimulus can be influenced by prior experience of a predictive relationship between that stimulus and an outcome. The current research will assess the ability of a recent model of such effects to account for the various ways in which the experienced predictive validity of a stimulus influences its processing power. This model proposes that overall processing power is controlled by two separate "associability" processes that are set in opposition. The aim of the research is to investigate the properties and mechanisms of each of these processes individually, and establish the way in which they interact, in order to gain a better understanding of how what we have learnt in the past might influence the way we learn information in the future.