Behavioral treatments for children with autism spectrum disorder (ASD) seek to improve how children learn about others. The variability in the effectiveness that these treatments have on children’s social skills remains poorly understood.
Gabriela Rosenblau plans to leverage state-of-the-art approaches in cognitive neuroscience to specify the cognitive and neural mechanisms governing learning in children with ASD. Using age-appropriate learning tasks, the current project aims to 1) characterize children’s learning strategies in social and nonsocial contexts with precise mathematical models, 2) identify how children’s learning strategies are implemented in the brain and 3) test whether these model-based strategies can forecast the extent to which a child benefits from an evidence-based treatment.
This proposal is based on a pilot study showing that typically developing (TD) adolescents integrate feedback from peers into their judgments about them. In contrast, adolescents with ASD rely only on their own preferences when judging peers, and this group difference in learning strategies scales with differences in neural activity during the task1.
The current project is intended to advance this initial study in several important ways. The project will 1) evaluate a younger group of children (TD and ASD) to test whether model-based differences in learning strategies are present at an earlier developmental stage; 2) determine whether learning strategies are specifically social or whether they generalize beyond social decision-making; 3) test whether children have general knowledge of peers and how this conceptual knowledge is encoded in neural activity; and 4) probe whether an individual’s learning profile, captured by a specific computational model, can predict their treatment outcomes (i.e., the extent to which a child learns social skills and achieves certain treatment goals).
Taken together, this proposal will rigorously test and tease apart learning strategies of children with ASD with the goal of improving diagnostic and treatment decisions.