Abstract
Gradual acquisition of knowledge of predictive relationships: learning curves in contingency judgements. The aim of this paper was studying how predictive relationships between events are learnt. Participants had to evaluate the relationship arranged between a cue and an outcome and it was studied how their judgements evolved across trials. Four different contingency conditions were programmed, positive, negative and two zero contingency conditions with a high and a low outcome probability. The results showed that terminal ratings agreed with the contingency programmed and that the knowledge about predictive relationships was gradually acquired. Regarding their theoretical implications, such results and the simulations performed showed the limitations of the ∆P statistical model and their compatibility with the predictions made by the Rescorla and Wagner associative model.
