Limits to learning by observing: when learning by doing leads to differences in brain activity and behaviours in comparison to observation only

Practitioners often use “learning by observation” when teaching new skills; however, there is limited research on how the brain - behavioural responses compare with “learning by doing”. We compared the brain responses when performing and watching a novel joystick-tracing task in three groups: a physical practice group, an observational practice group and a group that had no practice. We then compared their behavioural performance during retention. Following physical practice when all groups watched a video of the task, there was activation in motor regions bilaterally. However, following observational practice there was only activation in the left motor regions. No activation was shown in the control (no-practice) group. These patterns of activation matched what we saw during physical and observational practice. The physical practice group also showed fewest errors when assessed in retention in comparison with the other two groups. These results suggest that when teaching a novel skill, observational learning may activate some areas of the brain but this does not necessarily translate to significant improvements in behavioural measures related to motor learning. We need to consider other ways to bring about functional change, including interspersing observational and physical practice or stimulating brain areas before observation.