[PS-1.18] Statistical learning as a basic tool for generating expectations, an individual differences approach

Cunillera, T. 1 , François, C. 1 & Morís, J. 2

1 University of Barcelona
2 University of Oviedo

The ability to discover regular patterns in the environment, often called statistical learning (SL), seems to depend on the brain's sensitivity to the statistical properties of the input. Infants and adults can learn to segment statistically organized sequences of auditory, visual and tactile stimuli probably via a domain-general SL mechanism. Crucially, the discovery of these regularities may allow the successful prediction of incoming information. In the present study, we explored the role of predictability in generating expectations using a set of tasks, typically used in different learning domains. Specifically, we exposed participants to two different tasks in which successful predictions of the incoming information was crucial to correctly solve the tasks and maximize performance. A working memory task was also used as discriminant validity control. Interestingly, preliminary data point towards a positive correlation between the different measures of learning. These results may support the idea that predictability is one important outcome of SL.