Learning new actions from observing others across development

Learning from observing others actions and outcomes (observational learning) has a major evolutionary advantage as it allows a rapid acquisition of new actions and has been found in various intelligent species, including humans. The developmental mechanism underlying learning from observed actions are not understood. Recent studies in adults have started to investigate individual action learning from a neurocomputational perspective. Here, reward-based learning through prediction errors (the difference between expected and received action-outcomes) seems to be crucial for learning new actions. The current project aims to describe neural and behavioral associations related to developmental differences in observational action learning using a computational approach.