Motion trajectory information and agency influence motor learning during observational practice

Abstract

Fundamental to performing actions is the acquisition of motor behaviours. We examined if motor learning, through observational practice, occurs by viewing an agent displaying naturalistic or constant velocity, and whether motion trajectory, as opposed to end-state, information is required. We also investigated if observational practice is sensitive to belief regarding the origin of an agent. Participants had to learn a novel movement sequence timing task, which required upper-limb movements to a series of targets within a pre-specified absolute and relative time goal. Experiment 1 showed learning after viewing naturalistic and constant velocity, but not end-state information. For Experiment 2, in addition to learning the movement sequence, participants observed a series of movement stimuli that were either the trained or new sequences and asked to rate their confidence on whether the observed sequence was the same or different to observational practice. The results indicated agency belief modulates how naturalistic and constant velocity is coded. This indicated the processes associated with belief are part of an interpretative predictive coding system where the association between belief and observed motion is determined. When motion is constant velocity, or believed to be computer-generated, coding occurs through top-down processes. When motion is naturalistic velocity, and believed to be human-generated, it is most likely coded by gaining access to bottom-up sensorimotor processes in the action-observation network

Atypical biological kinematics are represented during observational practice.

“…As well as facilitating socio-cognitive functioning during interactions between people (Cook, Blakemore, & Press, 2013;Press, Cook, Blakemore, & Kilner, 2011), biological tuning is important for the acquisition of novel motor actions during observational practice . We have confirmed biological tuning across a series of behavioural studies where participants observe a series of model stimuli that depict typical or atypical human biological kinematics (Hayes, Dutoy, Elliott, Gowen, & Bennett, 2016;Hayes, Roberts, Elliott, & Bennett, 2014;Hayes, Timmis, & Bennett, 2009;Roberts, Bennett, Elliott, & Hayes, 2015). Typical kinematics had a movement profile where peak velocity occurred at approximately 50% of the trajectory, which is consistent with goal-directed upper-limb aiming movements (Elliott et al, 2010).…”