Pavlidou, E. 1, 2 , Buis, B. . 2 , Springlfield, J. . 2 , Mencl, E. . 2, 3 & Pugh, K. 2, 3
1 University of York
2 Haskins Laboratories
3 Yale University
We acquired functional images (fMRI) for a set (N=30) of TD and RD children during an AGL task. In the passive learning phase, children were presented with twelve alternating blocks of Structured (strings of unfamiliar shapes obeying the grammar) and Unstructured stimuli (strings of randomly ordered Greek letters). At test (event-related), children were presented with novel stimuli (unfamiliar shapes) and they were asked to indicate the structured ones (i.e. those that obeyed the grammar) with a button press. Standard regression analysis was used to identify activations in response to the Structured and Unstructured items across subjects and between groups (i.e. TD vs RD & Learner vs Non-Learner). Our contrasts revealed higher activation bilaterally for structured items in a number of sites already implicated in visual learning and AGL, including the IFG, posterior occipital areas, and posterior MTG, extending into Precuneus and IPL. Overall, we observed increased activity throughout the occipital, parietal, and prefrontal cortical areas; reflecting the complex nature of the task and suggesting that such areas maybe recruited early to facilitate and optimize test performance. We discuss our findings in the light of open theoretical questions on implicit statistical learning and its relationship with typical and atypical reading.