Training on brain–computer interfaces is shown to be markedly improved by using real-time reinforcement feedback instead of relying only on sensory feedback, which can be impaired in users of these ...