Motor similarity shapes predictive processes in action perception: evidence from behavioral and neuroelectrical markers

Predictive processes during action observation are known to benefit from motor expertise, yet it remains unclear which factors drive successful prediction in everyday actions, where everyone can be considered an “expert”. We hypothesized that prediction accuracy depends on kinematic overlaps between observer and agent, referred to as motor similarity. In Study 1 (N=90, action execution) we introduced a kinematics-based metric - motor distance - to quantify inter-individual motor (dis)similarities. In Study 2 (N=20, behavioral) and Study 3 (N=21, behavioral and EEG) we tailored action observation tasks to each participant by presenting movements varying in similarity to their own motor style. Behavioral results showed that motor similarity enhances perceptual accuracy. EEG results revealed that beta-band desynchronization, a marker of prediction uncertainty, is stronger over sensorimotor regions when observing more dissimilar movements. These findings demonstrate that agent-observer motor style overlap is a key determinant of prediction processes, extending beyond traditional expertise-based accounts.