Performance anxiety is associated with biases in learning from reward and punishment in skilled individuals

Many individuals experience performance anxiety (PA) in high-stakes situations, from public speaking to the performing arts. While debilitating PA is associated with physiological, cognitive, and affective alterations, its underlying mechanisms remain unclear. Using behavioural analysis, computational modelling, and electroencephalography, we investigated whether PA predisposes individuals to learn faster from punishment than reward, particularly under high task uncertainty. Across three experiments with 95 skilled pianists, participants learned hidden melody dynamics through reinforcement with graded reward or punishment feedback. Bayesian hierarchical modelling revealed that performers with greater PA levels learn faster from punishment in low-uncertainty environments but increasingly rely on reward as uncertainty escalates. These biases were mediated by reinforcement-driven modulation of motor variability— increasing following poor outcomes— and shifts in frontal theta (4–7 Hz) activity encoding feedback changes and signalling upcoming motor adjustments. The findings reveal that PA alters the weighting of reward and punishment signals based on task uncertainty.