Estimating the implicit component of visuomotor rotation learning by constraining movement preparation time.

Abstract

When sensory feedback is perturbed, accurate movement is restored by a combination of implicit processes and deliberate re-aiming to strategically compensate for errors. Here, we directly compare two methods used previously to dissociate implicit from explicit learning on a trial-by-trial basis: 1) asking participants to report the direction that they aim their movements, and contrasting this with the directions of the target and the movement that they actually produce, 2) manipulating movement preparation time. By instructing participants to re-aim without a sensory perturbation, we show that re-aiming is possible even with the shortest possible preparation times, particularly when targets are narrowly distributed. Nonetheless, re-aiming is effortful and comes at the cost of increased variability, so we tested whether constraining preparation time is sufficient to suppress strategic re-aiming during adaptation to visuomotor rotation with a broad target distribution. The rate and extent of error reduction under preparation time constraints were similar to estimates of implicit learning obtained from self-report without time pressure, suggesting that participants chose not to apply a re-aiming strategy to correct visual errors under time pressure. Surprisingly, participants who reported aiming directions showed less implicit learning according to an alternative measure, obtained during trials performed without visual feedback. This suggests that the process of reporting can affect the extent or persistence of implicit learning. The data extend existing evidence that restricting preparation time can suppress explicit re-aiming, and provide an estimate of implicit visuomotor rotation learning that does not require participants to report their aiming directions.\n\nNew and NoteworthyDuring sensorimotor adaptation, implicit, error-driven learning can be isolated from explicit strategy-driven re-aiming by subtracting self-reported aiming directions from movement directions, or by restricting movement preparation time. Here, we compared the two methods. Restricting preparation times did not eliminate re-aiming, but was sufficient to suppress reaiming during adaptation with widely-distributed targets. The self-report method produced a discrepancy in implicit learning estimated by subtracting aiming directions, and implicit learning measured in no-feedback trials.