Motor Learning Mechanisms are not modified by Feedback Manipulations in a Real-World Task

Federico Nardi,A. Aldo Faisal,Shlomi Haar
DOI: https://doi.org/10.1101/2024.04.10.588812
2024-04-14
Abstract:This study examines the distinctiveness of error-based and reward-based mechanisms in motor learning, which are traditionally isolated in laboratory tasks but co-occur in real-world scenarios. Using Embodied Virtual Reality (EVR) of pool billiards - allowing for full proprioception via interaction with the physical pool table, cue stick, and balls - we introduced visual perturbations to a real-world task. Participants underwent sessions learning a visual rotation with either error or reward feedback. While participants corrected the entire rotation with error feedback, only partial correction was observed with reward feedback, highlighting the influence of the feedback regime. However, lag-1 autocorrelation and inter-trial variability decay, indicators of skill learning, showed no significant differences between sessions, suggesting that the provided visual feedback did not exclusively engage specific learning mechanisms. Analysis of post-movement beta rebound (PMBR), a marker of brain activity associated with learning mechanisms, revealed a decrease in PMBR with reward feedback but no consistent trend during error feedback sessions. These findings suggest that while reward feedback was absent in error conditions, participants still engaged in reward-based learning, indicating that feedback manipulation alone may not sufficiently challenge individual learning mechanisms in real-world settings. This study underscores the complexity of motor learning processes and highlights that visual feedback by itself can not elucidate the interplay between error-based and reward-based mechanisms in real-world contexts.
Neuroscience
What problem does this paper attempt to address?