Lectures‎ > ‎

Statistical learning and L2 learning: Many questions and a few answers - Ram Frost, PhD

Like any cognitive ability, individuals seem to differ in their sensitivity to correlations in the environment. This is revealed in a large number of tasks that monitor statistical learning. In the context of language, our working hypothesis is that native speakers implicitly develop sensitivities to the statistical properties of their language and its writing system, and while learning an L2 they pick-up and assimilate a new set of statistical regularities. This predicts that individual differences in statistical learning will be correlated with individual differences in L2 learning. Although we have demonstrated this with one specific statistical learning task, our approach raises important questions regarding the theoretical construct of “statistical learning capacity”, its internal structure, validity and reliability. We will therefore focus on mapping the possible dimensions of statistical learning, their relations to general cognitive abilities, their relations to specific linguistic skills, their relations to other implicit learning tasks, as well as their test-retest reliability.