Cutting through the noise: reducing bias in motor adaptation analysis

During goal-directed movements, the magnitude of error correction by a person on a subsequent movement provides important insight into a person9s motor learning dynamics. Observed differences in trial-by-trial adaptation rates might indicate different relative weighting placed on the various sources of information that inform a movement, e.g. sensory feedback, control predictions, or internal model expectations. Measuring this trial-by-trial adaptation rate is not straightforward, however, since externally observed data are masked by noise from several sources and influenced by inaccessible internal processes. Adaptation to perturbation has been used to measure error adaptation as the introduced external disturbance is sufficiently large to overshadow other noise sources. However, perturbation analysis is difficult to implement in real-world scenarios, requires a large number of movement trials to accommodate infrequent perturbations, and the paradigm itself might affect the movement dynamics being observed. Here we focus on error adaptation during unperturbed and naturalistic movements. With increasing motor noise, the conventional estimation of trial-by-trial adaptation increases, a counterintuitive finding that is the consequence of systematic bias in the estimate due to noise masking the learner9s intention. We present an analytic solution relying on stochastic signal processing to reduce this effect of noise, producing an estimate of motor adaptation with reduced bias. The result is an improved estimate of trial-by-trial adaptation in a human learner compared to conventional methods. We demonstrate the effectiveness of the new method in analyzing simulated and empirical movement data under different noise conditions. The analytic approach is applicable across different types of movements in varied contexts and should replace the regression analysis method in future motor analysis studies.